interventions to prevent skin cancer by reducing exposure ......interventions to prevent skin cancer...

IEAMCABP

A

I

Sabac

FHiPCRNCBC

bng

CP3

4

nterventions to Prevent Skin Cancer by Reducingxposure to Ultraviolet RadiationSystematic Review

ona Saraiya, MD, MPH, Karen Glanz, PhD, MPH, Peter A. Briss, MD, MPH, Phyllis Nichols, MPH,ornelia White, PhD, MSPH, Debjani Das, MPH, S. Jay Smith, MHPA, Bernice Tannor, MPH,ngela B. Hutchinson, PhD, MPH, Katherine M. Wilson, PhD, MPH, Nisha Gandhi, MPH, Nancy C. Lee, MD,arbara Rimer, DrPH, Ralph C. Coates, PhD, Jon F. Kerner, PhD, Robert A. Hiatt, MD, PhD,atricia Buffler, PhD, MPH, Phyllis Rochester, PhD

bstract: The relationship between skin cancer and ultraviolet radiation is well established.Behaviors such as seeking shade, avoiding sun exposure during peak hours of radiation,wearing protective clothing, or some combination of these behaviors can provide protec-tion. Sunscreen use alone is not considered an adequate protection against ultravioletradiation. This report presents the results of systematic reviews of effectiveness, applicabil-ity, other harms or benefits, economic evaluations, and barriers to use of selectedinterventions to prevent skin cancer by reducing exposure to ultraviolet radiation. TheTask Force on Community Preventive Services found that education and policy approachesto increasing sun-protective behaviors were effective when implemented in primary schoolsand in recreational or tourism settings, but found insufficient evidence to determineeffectiveness when implemented in other settings, such as child care centers, secondaryschools and colleges, and occupational settings. They also found insufficient evidence todetermine the effectiveness of interventions oriented to healthcare settings and providers,media campaigns alone, interventions oriented to parents or caregivers of children, andcommunity-wide multicomponent interventions. The report also provides suggestions forareas for future research.(Am J Prev Med 2004;27(5):422–466) © 2004 American Journal of Preventive Medicine

cmiaaa

cbrpiiip1nc2ham

ntroductionkin cancer is the most common type of cancer inthe United States.1 Estimates for 2004 indicatethat more than 1 million people will be diagnosed

s having the two most common types of skin cancer—asal cell carcinoma and squamous cell carcinoma—nd approximately 2300 deaths from both cancersombined are predicted. Both basal cell and squamous

rom the National Center for Chronic Disease Prevention andealth Promotion, Division of Cancer Prevention and Control (Sara-

ya, Das, Tannor, Wilson, Lee, Coates, Rochester), and Epidemiologyrogram Office (Briss, Nichols, White, Smith, Hutchinson, Gandhi),enters for Disease Control and Prevention, Atlanta, Georgia; Canceresearch Center, University of Hawaii (Glanz), Honolulu, Hawaii;ational Cancer Institute, National Institutes of Health, Division ofancer Control and Population Sciences (Rimer, Kerner, Hiatt),ethesda, Maryland; and School of Public Health, University ofalifornia-Berkeley (Buffler), Berkeley, CaliforniaNotes: Affiliation reflects author’s location while this research was

eing conducted. Names and affiliations of Task Force on Commu-ity Preventive Services members are listed at www.thecommunityuide.org.

Address correspondence and reprint requests to: Mona Saraiya,enters for Disease Control and Prevention, Division of Cancer

arevention and Control, 4770 Buford Highway, MS K55, Atlanta GA0341. E-mail: [email protected].

22 Am J Prev Med 2004;27(5)© 2004 American Journal of Preventive Medicine • Publish

ell carcinoma respond well to treatment. However,elanoma, the third most common type of skin cancer,

s much more likely to be fatal. Diagnoses of melanomare anticipated in approximately 55,000 people and willccount for 7900 deaths, more than three quarters ofll skin cancer fatalities.2

In the United States, although the incidence of mostancers has been declining, melanoma incidence haseen on the rise.3 Since 1973, the annual incidenceate for melanoma (new cases diagnosed per 100,000eople) has more than doubled, from 6.8 per 100,000

n that year to 17.4 cases per 100,000 in 1999.4 Thencrease is likely a result of several factors, includingncreased exposure to ultraviolet (UV) radiation and,ossibly, earlier detection of melanoma.5 From 1973 to999, the number of melanoma deaths also rose: An-ual deaths per 100,000 people from melanoma in-reased by about 40% during this period, from 1.9 to.7 per 100,000 people. During the past 10 years,owever, melanoma mortality rates have remained rel-tively stable among women, but less so amongen.3,6,7 (Mortality rates for white males by economic

rea are presented in Figure 1.)

0749-3797/04/$–see front mattered by Elsevier Inc. doi:10.1016/j.amepre.2004.08.009

http://www.thecommunityguide.org

http://www.thecommunityguide.org

PE

HraUcpRbttdecibchetcs

C

Elmoosdi

cmmrSt

atol

Fr

reventable Risk Factors for Skin Cancerxcessive Exposure to UV Radiation

igh levels of exposure to UV radiation increase theisk of all three major forms of skin cancer, andpproximately 65% to 90% of melanomas are caused byV exposure.8 Studies have shown that the damage

aused by UV radiation, particularly damage to DNA,lays a central role in the development of melanoma.9

isk of melanoma and other skin cancers can thereforee reduced by limiting exposure to sunlight, which ishe primary source of UV radiation. (Sunlamps andanning beds are other sources.) Total UV exposureepends on the intensity of the light, duration of skinxposure, and whether the skin is protected by shade,lothing (including hats), or sunscreen. Severe blister-ng sunburns are associated with an increased risk ofoth melanoma and basal cell carcinoma. For theseancers, intermittent intense exposures seem to carry aigher risk than do lower-level, chronic, or cumulativexposures, even if the total amount of UV exposure ishe same. The risk of squamous cell carcinoma, inontrast, is strongly associated with chronic UV expo-

HAWAII

ALASKA

igure 1. Cancer mortality rates by state economic area, whitate for United States overall: 3.07/100,000.

ure but not with intermittent exposure.10 a

hildhood and Adolescent UV Exposure

xposure to UV radiation during childhood and ado-escence plays a role in the future development of both

elanoma and basal cell cancer.11–16 The risk of devel-ping melanoma is strongly related to a history of oner more sunburns (an indicator of intense UV expo-ure) in childhood or adolescence.12,17–19 Sunburnsuring these periods have also recently been found to

ncrease the risk of basal cell carcinoma.14,15

Nevi, or moles (lesions of pigment forming skinells), are an important risk factor for skin cancer, andost develop in childhood through early adulthood. Itay be that sun exposure in childhood heightens the

isk of melanoma by increasing the number of moles.18

un protection during childhood may therefore reducehe risk of melanoma in adulthood.20,21

Children and adolescents have more opportunitiesnd time than adults to be exposed to sunlight,22 andhus more opportunities to increase their risk of devel-ping skin cancer.9,23,24 At least 25% of a person’s

ifetime UV exposure occurs during childhood and

Deaths per 100,000

3.53–5.11 (highest 20%)

3.18–3.52

2.83–3.17

2.50–2.82

1.22–2.49

DC

es, 1970 to 1998 (age-adjusted 1970 U.S. population). Death
e mal
dolescence.25–27

Am J Prev Med 2004;27(5) 423

S

AcasaHnfw1EtapnfAAn

cwncdasrmoanf

anaemti

E

EeaassttpE

S

Bo(aa

stppbmCisctor

tUahrbsprpssbrwtpa

oc

P

Il

48

4

kin Color and Ethnicity

lthough anyone can get skin cancer, people withertain characteristics are particularly at risk. Whitesre 80 times more likely to develop basal cell andquamous cell carcinoma than African Americans,28

nd 20 times more likely to develop melanoma.29

ispanics appear to be at lower risk of melanoma thanon-Hispanic whites: a study conducted in Los Angeles

ound Hispanic incidence rates to be 2 to 3 per 100,000,hereas the rate for non-Hispanic whites is 11 per00,000.30 According to the data from the Surveillance,pidemiology, and End Results (SEER) cancer regis-

ry31 for the 1995–1999 period, average annual age-djusted incidence rates for melanoma per 100,000opulation were 23.5 for men and 15.7 for women foron-Hispanic whites; 3.8 for men and 3.7 for women

or Hispanics; 1.8 for men and 1.3 for women forsians; 1.5 for men and 0.9 for American Indian/laskan Natives; and 1.2 for men and 0.9 for women foron-Hispanic blacks.The racial and ethnic differences in skin

ancer rates are mostly due to skin color,hich is determined by the amount of mela-in produced by skin cells called melano-ytes. These cells protect the skin from theamage produced by UV radiation. However,lthough darkly pigmented people developkin cancer on sun-exposed sites at lowerates than lightly pigmented people, incre-ental UV exposure does increase their risk of devel-

ping skin cancer.32 The risk of skin cancer is greatermong those who sunburn readily and tan poorly,33

amely those with red or blond hair and fair skin thatreckles or burns easily.34–36

Other strong predictors of melanoma include havinglarge number of nevi, or moles, including atypical

evi; family history of melanoma; and increasingge.33,37–39 The incidence of skin cancer increasesxponentially with age because older people have hadore opportunities to be exposed to UV radiation and

heir capacity to repair the damage from UV radiations diminished.9,34,35

nvironmental Factors Affecting UV Radiation

nvironmental factors that increase the amount of UVxposure include proximity to the equator; higherltitude; lower levels of cloud coverage (which canllow up to 80% of UV rays to penetrate the atmo-phere); the presence of materials that reflect the sun,uch as pavement, water, snow, and sand; exposure tohe sun around midday; and spending time outside inhe spring or summer.30,40 Ozone depletion couldotentially increase levels of solar radiation at the

reCommon pa

and

arth’s surface.30,41 i

24 American Journal of Preventive Medicine, Volume 27, Num

un-Protective Behaviors

ehaviors that reduce skin cancer risk include limitingr minimizing exposure to the sun during peak hours10 A.M. to 4 P.M.) because UV rays are more intenseround midday, wearing protective clothing, or usingppropriate sunscreen protection.

Scientific knowledge about sunscreen has undergoneome recent evolution. Although sunscreen is thoughto be an important adjunct to other types of UVrotection, it should not be counted on to provide UVrotection by itself. Sunscreen clearly prevents sun-urn, and using sunscreen is one of the most com-only practiced behaviors for preventing skin cancer.linical trials have found sunscreens effective in reduc-

ng the incidence of actinic keratoses, the precursors toquamous cell carcinoma,42,43 and one randomizedlinical trial showed sunscreens to be moderately effec-ive in reducing squamous cell carcinoma itself.44 An-ther randomized trial found sunscreens effective ineducing the number of moles—the precursors and

strongest risk factor for melanoma45—amonghigh-risk children.

However, recent research suggests thatsunscreen, by itself, is not an adequatestrategy for UV protection. Many peopleuse sunscreens if they intend to stay out inthe sun for a long period of time, and theyreduce the use of other forms of sun pro-tection (e.g., clothing or hats). They

hereby receive the same or even a higher amount ofV exposure than they would have obtained duringshorter stay with no sunscreen.46,47 Some studies

ave shown a high incidence of sunburning despiteelatively high rates of sunscreen use,48,49 which maye the result of weakened sun-protection qualities ofunscreen when inadequately or infrequently reap-lied. For these reasons, although an expert groupecently concluded that topical use of sunscreenrobably prevents squamous cell carcinoma of thekin, the panel drew no conclusions about sun-creen’s contribution to reducing the incidence ofasal cell carcinoma or melanoma.50 The panelecommended avoiding the sun, seeking shade, orearing protective clothing that reduces exposure to

he full spectrum of UV radiation as the first line ofrotection against skin cancer, with sunscreen as andjunct form of protection only.In some instances, sunscreen may be the only viable

ption. However, to be effective, it must be appliedorrectly.51–53

revalence of Sun-Protective Behaviors

n 1992, a total of 53% of U.S. adults were “veryikely” to protect themselves from the sun by practic-

daries4824.

Seelateentges

ng at least one protective behavior (sunscreen,

ber 5

wtsAcswawrbIT2tw

habag(p(msap

T

TomFSUcCpmttRlm

HP

Ti

1

F1 ion).

earing clothing, or seeking shade).54 Less than onehird of white adults used sunscreen (32%), soughthade (30%), or wore protective clothing (28%).mong African Americans, 28% wore sun-protectivelothing and 45% sought shade, but only 9% usedunscreen.55 The sun-protective behaviors of bothhites and African Americans were more commonmong those who were more sensitive to the sun,ere female, and were older. When the survey wasepeated in 2000, the prevalence of sun-protectiveehaviors was similar (A. Hartmann, National Cancernstitute, personal communication, January 2003).wo independent surveys, conducted in 1999 and in000, showed that sunburn rates over the past year inhe U.S. population were between 35% and 40% andere highly variable by state56 (Figure 2).Among youth aged 11 to 18 years, 72% reported

aving had at least one summer sunburn, 30% reportedt least three, and 12% reported at least five sun-urns.49 Routinely practiced sun-protective behaviorsmong these youth on sunny days were wearing sun-lasses (32%) or long pants (21%), staying in the shade22%), and applying sunscreen (31%). At the beach orool, 58% used a sunscreen with sun-protective factorSPF) of �15.57 Among U.S. white children aged 6onths to 11 years, 43% experienced one or more

unburns within the past year, with sunscreen (62%)nd shade (27%) being the most frequently reported

igure 2. Sunburn rates among white people by state in the U999 (age and ethnicity adjusted to the 1999 BRFSS populat

rotection methods.58,59

he Guide to Community Preventive Services

he systematic reviews in this report represent the workf the independent, nonfederal Task Force on Com-unity Preventive Services (the Task Force). The Task

orce is developing the Guide to Community Preventiveervices (the Community Guide) with the support of the.S. Department of Health and Human Services in

ollaboration with public and private partners. Theenters for Disease Control and Prevention (CDC)rovides staff support to the Task Force for develop-ent of the Community Guide. A special supplement to

he American Journal of Preventive Medicine, “Introducinghe Guide to Community Preventive Services: Methods, Firstecommendations, and Expert Commentary,” pub-

ished in January 2000, presents the background andethods used in developing the Community Guide.

ealthy People 2010 Goals and Objectives forreventing Skin Cancer

he interventions reviewed in this article may be usefuln reaching the objectives set in Healthy People 2010 60:

. Increase to 75% the proportion of people who use atleast one of the following protective measures thatmay reduce the risk of skin cancer: avoid the sunbetween 10 A.M. and 4 P.M., wear sun-protective

States, Behavioral Risk Factor Surveillance System (BRFSS),
nited
clothing when exposed to the sun, use sunscreen

Am J Prev Med 2004;27(5) 425

2

RI

ICpwm

●

●

●

●

h

●

●

●

●

C

ItttnUvsrtAS

U

IUf

●

●

●

●

●

●

●

●

●

M

Ttwisb

4

with a sun-protection factor (SPF) of �15, and avoidartificial sources of ultraviolet light.

. Reduce melanoma deaths to �2.5 per 100,000people.

ecommendations from Other Advisory Groupsnternational Agency for Research on Cancer

n 2001, the International Agency for Research onancer (IARC), an independent organization sup-orted by the World Health Organization, convened aork group to address sunscreen use. The work groupade the following recommendations:50

Protect the skin from solar damage by wearing tightlywoven protective clothing that adequately covers thearms, trunk, and legs, and a hat that provides ade-quate shade to the entire head; seeking shade when-ever possible; and avoiding outdoor activities duringperiods of peak UV radiation.Avoid the use of sunscreens as the first choice or thesole agent for protection against the sun or forextending the duration of solar exposure, such asprolonging sunbathing.Residents of areas of high UV radiation who workoutdoors or engage in regular outdoor recreationshould daily use sunscreen with a high SPF (�15) onexposed skin.Pay particular attention to adequate solar protectionfor children. The first two recommendations above(protecting the skin against sun damage and avoid-ing reliance on sunscreen as the primary or solesun-protection agent) are more important duringchildhood than at any other time in life, and shouldbe rigorously applied by parents and schoolpersonnel.

The IARC also recommended the following publicealth strategies50:

Design health promotion interventions to increasethe appropriate and effective use of sunscreens bythe general public, as well as those subgroups atparticular risk for skin cancer because of their skintype or a tendency to seek solar exposure.Stringently evaluate the safety of sunscreens, partic-ularly with regard to long-term effects, and makesuch data available in the public domain, to allowindependent scientific evaluation.Subject sunscreens to the same regulatory safetyrequirements as pharmaceuticals.Require sunscreen advertising to promote a globalsun-protection strategy rather than portraying sun-screen use for intentional exposure to the sun orpromoting a false sense of security for people using
sunscreen. a

enters for Disease Control and Prevention

n 2002, the CDC published guidelines recommendinghat schools engage in skin cancer prevention activi-ies.61 Specific recommendations include implementa-ion of policies; creation of physical, social, and orga-izational environments that facilitate protection fromV rays; education of young people; professional de-

elopment of staff; involvement of families; healthervices; and program evaluation. These guidelineseceived the support of the National Cancer Institute,he American Academy of Dermatology, the Americancademy of Pediatrics, and the American Cancerociety.

.S. Preventive Services Task Force

n 2003, in the Guide to Clinical Preventive Services,62 the.S. Preventive Services Task Force (USPSTF) drew the

ollowing conclusions:

The benefits of sun-protective measures exceed anypotential harms.Evidence is poor to determine the effects of sunlampuse or skin self-examination on melanoma risk.There is fair to good evidence that increased sunexposure increases the risk of nonmelanoma skincancer.The relationships between sun exposure and mela-noma risk are complex, and observational studiessuggest that intermittent or intense sun exposure is agreater risk factor for melanoma than chronicexposure.Light-skinned people are at much higher risk forskin cancer than are those with darker skin.There is good evidence that sunscreens can reducethe risk of squamous cell cancer.There is insufficient evidence to determine the effectof sunscreen use on risk of melanoma.There is insufficient evidence to determine whetherclinician counseling is effective in changing patientbehaviors to reduce skin cancer risk.Counseling parents may increase their use of sun-screen for children, but there is little evidence todetermine effects of counseling parents on otherprotective behaviors (use of protective clothing, re-ducing sun exposure, avoiding sun lamps, or practic-ing skin self-examination).

ethods for Conducting the Review

he general methods used to conduct systematic reviews forhe Community Guide have been described in detail else-here.63,64 The specific methods for conducting this review,

ncluding selection of interventions and outcomes and theearch strategy for interventions to increase sun-protectiveehaviors, are presented in Appendix A. The conceptual
pproach to the review, critical both for describing the
ber 5

md

C

FlasarokbT3os

isbaasicsdcss

gpapat(heui

caIe

SO

MssiWbpts

FMcoctiw ectan

ethods and for understanding the results of the review, isescribed below.

onceptual Model

irst, the team (i.e., the authors of this article) developed aogic framework as a model to depict its overall conceptualpproach to preventing skin cancer by reducing UV expo-ure. Then for each intervention we reviewed, we developedn analytic framework, a conceptual model that shows theelationship of the intervention to relevant intermediateutcomes (e.g., knowledge, attitudes, beliefs, and intentions),ey sun-protective behaviors, and the assumed relationshipsetween sun-protective behaviors and skin cancer prevention.he analytic framework for mass media interventions (Figure) is a representative example. Analytic frameworks for thether interventions are similar to this example, althoughome include environmental and policy components.

The analytic frameworks focused on sun-protective behav-ors (e.g., avoiding peak sun, covering up, or using sun-creen) and intermediate outcomes that were postulated toe associated with sun-protective behaviors (e.g., knowledge,ttitudes, intentions, and environmental characteristics). Welso sought information on selected health outcomes (e.g.,unburn or nevi). Recommendations were based on eithermproved health outcomes (rare in this subject matter be-ause cancer outcomes occur long after the intervention) orun-protective behaviors thought by the systematic reviewevelopment team to be established proxies for cancer out-omes (in this case, avoiding sun or covering up and notunscreen use alone). In general, although we think sun-

igure 3. Analytic framework for media interventions to reajor stratification variables were type of media (e.g., sm

haracteristics of target population (e.g., age, sex, skin color,ccupation); intervention intensity (i.e., comparison; some inharacteristics (e.g., urban, rural, climate of location [e.g., suo media). The analytic frameworks for the other interventionnclude environmental and policy components. Key to shapeith rounded corners indicate intermediate outcomes; and r

creen use is an important outcome of sun-protection pro- c

rams, we did not consider it, by itself, to be an establishedroxy for better health (see Sun-Protective Behaviors sectionbove). Increased sunscreen use had to be a part of arogram that also had improvements in other behaviors suchs avoiding the sun or covering up. Similarly, if an interven-ion could not discern the individual sun-protective behaviorsespecially sunscreen use) and just reported composite be-aviors, the results of the intervention could not be consid-red as evidence for improving behaviors as it would benclear what particular behavior was contributing to the

mproved overall behavior.The relationship between UV radiation and risk of skin

ancer was assumed by the review team to be well establishednd, subsequently, was not the focus of the systematic review.nstead the team focused on interventions to decrease UVxposure.

electing and Summarizing Information onutcomes

any of the studies included in the body of evidence targetedeveral sun-protective behaviors: seeking shade; avoiding theun; wearing protective clothing; using sunscreen; or compos-te behaviors, a combination of at least two of these behaviors.

e abstracted one measure per study for each of theseehavioral constructs (avoiding sun [includes seeking shade],rotective clothing, using sunscreen, and composite) whenhey were available. Although the measures in the individualtudies were diverse, within these four categories the team

ultraviolet exposure and increase sun-protective behaviors.media [posters, brochures] vs large media [TV, radio]);type, baseline risk, socioeconomic status, sunburn incidence,tion; high level of intervention); geographic/environmentals cloudy]); and intervention characteristics (e.g., size, accessre similar to this example; however, other frameworks mightnalytic framework: circles indicate interventions; rectanglesgles with square corners indicate health outcomes.

duceall

skintervennny vs we

s in a

onsidered the available measures to be sufficiently similar

Am J Prev Med 2004;27(5) 427

ca

S

Itwogpiiam(tnoplotc

Imiustkfmtaotos

Eaes(hprcsssltbosvpbes

Mncdathvtohoo

MwttdTimdmatpaa

CT

T(svp

●

●

●

●

●

●

wm

b(

pvg

4

onceptually to transform the measures to a common scalend summarize the results when it was otherwise appropriate.

election of Interventions for Review

nterventions to reduce UV exposure are diverse and difficulto classify. Although some studies can be easily identifiedithin a single category, many more involve several methodsr communication strategies. Intervention studies often tar-et multiple audiences, such as parents and children orhysicians and patients. Bearing in mind these complexities,

t is useful to provide a broad typology of four types ofnterventions that readers may use to group various strategiesnd studies65: (1) individual-directed strategies, (2) environ-ental and policy interventions, (3) media campaigns, and

4) community-wide multicomponent interventions. Eachype of intervention is briefly characterized here. Because theature of interventions is also strongly influenced by theirrganizational context or setting (i.e., the setting is often aroxy for important characteristics of both the target popu-

ation and relevant providers), the team has tended torganize both individual-directed strategies and environmen-al or policy changes by the setting in which they areonducted.

ndividual-directed strategies. These strategies include infor-ational and behavioral interventions aimed primarily at

ndividuals or relatively small groups. These interventionssually occur within an organizational context, such aschool, recreation program, or healthcare settings. Theyypically aim to educate and motivate individuals by providingnowledge, teaching attitudes, and teaching behavioral skillsor skin cancer prevention. They include the use of small

edia (e.g., brochures, pamphlets, printed materials); didac-ic programs (e.g., classroom lessons, lectures); interactivectivities (e.g., games, multimedia programs); and skill devel-pment (e.g., role playing, instruction in sunscreen applica-ion). These strategies can be directed toward any age,ccupation, or risk group and are often combined with othertrategies.

nvironmental and policy interventions. These interventionsim to provide or maintain a physical, social, or informationalnvironment that supports sun protection and promotesun-safety practices for all people in a defined populatione.g., school, community setting), not just those who areighly motivated. The interventions reach populations byassively reducing UV exposure, providing sun-protectionesources, and broadening the accessibility and reach of skinancer prevention information. Examples include increasinghade areas, supplying sunscreen, providing environmentalources of information and prompts, and many other possibletrategies. Policies establish formal rules or standards thatead to organizational actions, legal requirements, or restric-ions related to skin cancer prevention measures. Policies maye developed by a school, school board, or communityrganization, or by other legal entities, such as municipal,tate, and federal governments. Environmental strategies pro-ide supportive resources for skin cancer prevention in thehysical, social, or informational environment. They may beased on, and restricted or assisted by, policies. However,nvironmental supports can also be undertaken in the ab-
ence of a formal policy. g

edia campaigns. These campaigns use mass media chan-els such as print media (e.g., newspaper, magazines), broad-ast media (e.g., radio, television), and the Internet toisseminate information and behavioral guidance to a wideudience. They may be aimed at specific target audiences, butypically use broad distribution channels. Media campaignsave some of the characteristics of individual-directed inter-entions, but they lack the face-to-face interpersonal interac-ion and “captive audience” that is possible in a definedrganizational setting. Media campaigns tend to have a publicealth orientation and often seek to raise levels of awarenessr concern, and to help shape the policy agenda that drivesther interventions.

ulticomponent programs and comprehensive community-ide interventions. These interventions, often called popula-

ion-wide programs or campaigns, combine elements of thehree other types of strategies into an integrated effort in aefined geographic area (city, state, province, or country).hey often include individual-directed strategies carried out

n a range of settings, environmental and policy changes,edia campaigns, and a variety of setting-specific strategies

elivered with a defined theme, name or logo, and set ofessages. The team included studies in this review if they

ddressed a defined geographic area and included at leastwo components or at least two settings. Additionally, com-rehensive community-wide interventions are further defineds interventions that may include more than two componentsnd two settings to drive the campaign.

lassification of Strategies to Create a Practicalaxonomy for Reviews

he team organized most individual-directed strategiesmostly informational and behavioral interventions) by theetting in which they were conducted. Some of these inter-entions also incorporated the following environmental andolicy interventions:

Educational and policy interventions in child care settingsEducational and policy interventions in primary schoolsEducational and policy interventions in secondary schoolsand collegesEducational and policy interventions in recreational andtourism settingsPrograms in outdoor occupational settingsHealthcare system and provider settings

The team defined the category of mass media campaignsithout other activities to include either mass media alone orass media in combination with small media.The team organized one individual-directed intervention

y the target population of interest: programs for caregiverse.g., parents or teachers).

A final category was community-wide multicomponentrograms, including comprehensive community-wide inter-entions, which combine two or more of the other strate-ies into an integrated effort for an entire defined geo-
raphic area.
ber 5

RESEC

Ip�i(ctb(tc(o

cschttstho

etoctfbrUchomc

Epcrqdig

wa

abwwccsccebsaiequto

Ale

Ocieoas

Epw

Bib

Clade

Ta

RRRQD

a

b

aa

esults. Part I: Interventions to Decrease UVxposure and Promote UV Protection in Specificettingsducational and Policy Interventions in Childare Centers

nterventions in child care settings involve efforts toromote sun-protective behaviors among children aged5 years who attend programs in these settings. These

nterventions include at least one of the following:1) provision of information directly or indirectly to thehildren (through instruction or small media educa-ion); (2) additional activities to influence children’sehavior (modeling, demonstration, or role playing);3) activities intended to change the knowledge, atti-udes, or behavior of teachers, parents, and otheraregivers; and (4) environmental or policy approachessuch as providing sunscreen and shade or schedulingutdoor activities to avoid hours of peak sunlight).A large proportion of lifetime sun exposure occurs in

hildhood.25,66 Sun exposure among infants and pre-chool-aged children is largely dependent on the dis-retion of parents and adult care providers, and isighly variable. Studies have found that parental pro-

ective behaviors often depend on whether the childends to sunburn, and that parents often rely onunscreen as the most common method of protec-ion.59,67–71 As children progress from infancy to child-ood, increased mobility and a greater tendency to playutdoors often lead to increased UV exposure.72

Additionally, the responsibility for limiting the sunxposure of young children is shifting from the parento alternate care providers. In 1995, approximately 31%f preschool-aged children were being cared for inhild care centers; this number is expected to grow withhe increasing number of women entering the work-orce (projected to include 64% of all eligible womeny the year 2005).73 Child care centers therefore rep-esent an important opportunity to reduce children’sV exposure. However, a recent study found that of 25

hild care centers surveyed in Connecticut, nearly alleld outdoor activities during peak UV hours, and onlyne third had 50% shade in the play area. Further-ore, the observed use of sunscreen and protective

lothing was limited.74

ffectiveness. The team’s search identified nine re-orts75–83 on the effectiveness of interventions in childare centers. Information about the disposition of theeports is provided in Table 1. After exclusions foruality and redundancy, two reports of least suitableesign remained for review. Details of the two qualify-

ng reports are available at www.thecommunityuide.org/cancer.

One report75 evaluated the “Be Sunsafe” curriculum,hich includes interactive classroom and take-home
ctivities that promote covering up, finding shade, and t
sking for sunscreen. The report did not evaluateehavioral or policy outcomes. Another study78 used aorkshop for staff, an activity packet for parents, and aorking session to develop skin protection plans forenters. The intervention focused on increasing appli-ation of sunscreen, scheduling activities to avoid peakun, increasing availability of shade, and encouraginghildren to play in shady areas and to wear protectivelothing. This study did not show statistically significantffects on policy outcomes or measures of children’sehavior. Both studies showed generally consistent andtatistically significant improvements in the intermedi-te outcome of knowledge. Evidence of effectiveness isnsufficient because of (1) limitations in the design andxecution of available reports; (2) small numbers ofualifying reports; (3) variability in interventions eval-ated; (4) short follow-up times; and (5) little substan-ial or statistically significant improvement in outcomesther than knowledge and attitudes.

pplicability. Evidence about applicability was not col-ected because effectiveness of the intervention was notstablished.

ther positive or negative effects. The reports in-luded in the team’s search for effectiveness did notnclude information on other potential additional ben-fits of these interventions, such as reduction in the riskf overexposure to heat, nor on potential harms, suchs reductions in outdoor physical activity or transmis-ion of lice via hats or other clothing.84

conomic efficiency. Economic evaluations were noterformed because effectiveness of the interventionas not established.

arriers to implementation. Evidence about barriers tomplementation of this intervention was not collectedecause effectiveness was not established.

onclusion. Because of the small number of studies,ack of relevant outcome, and inconsistent results,ccording to Community Guide rules of evidence,64 evi-ence is insufficient to determine the effectiveness ofducational and policy interventions in child care cen-

able 1. Interventions in child care centers: informationbout reportsa

Number

eports meeting inclusion criteria 975–83

eports excluded, limited quality of executionb 276,79

eports on the same study 577,80–83

ualifying reports 275,78

esign suitabilityLeast 275,78

Reports may include more than one intervention arm.Grant-Petersson79 had two designs, one on school policies andnother on behavioral outcomes, which were independently evalu-ted as separate studies.

ers in reducing children’s adverse health effects or

Am J Prev Med 2004;27(5) 429

http://www.thecommunityguide.org/cancer


csecep

ES

Iptomo((b(ius

oUcmtssFcscs

E3eptimt

opd

itaedgs

tiikait

tptrcrebr

bhs(ahamswpbm

T

RRRQD

a

b e outo

4

hanging children’s behavior related to sun expo-ure, changing caregivers’ behavior related to sunxposure, changing policies and practices in child careenters, or changing children’s or caregivers’ knowl-dge or attitudes related to sun exposure and sunrotection.

ducational and Policy Interventions in Primarychools

nterventions in primary schools promote sun-rotective behaviors among children in kindergartenhrough 8th grade. These interventions include at leastne of the following activities: (1) provision of infor-ation to children (through instruction, small media,

r both); (2) activities to influence children’s behaviore.g., modeling, demonstration, or role playing);3) activities to change the knowledge, attitudes, orehavior of caregivers (e.g., teachers or parents); and4) environmental and policy approaches (e.g., provid-ng sunscreen, increasing availability of shade, or sched-ling outdoor activities to avoid hours of peakunlight).

Students spend a large amount of their day in schoolr in school-related activities that occur during peakV hours. Children are more receptive than adoles-

ents to practicing sun-protective behaviors and areore amenable to parents’ or other adults’ instruc-

ion.85 Primary schools are also more likely to haveuccess with sun-protection programs than highchools, as is the case with programs in Australia.86

ormal education settings facilitate integration of skinancer education into existing learning situations, andupport policy and environmental interventions. Re-ent review articles have addressed many of thesetudies in detail.22,87

ffectiveness. The team’s search identified a total of3 reports (30 intervention arms)79–82,88–116 on theffectiveness of educational and policy interventions inrimary schools. Information about the disposition ofhe reports is provided in Table 2. Many of the reportsncluded several intervention arms and multiple inter-

ediate and behavioral outcomes. Details of a subset of

able 2. Interventions in primary school: information about

eports meeting inclusion criteriaeports excluded, limited quality of executioneports on an already included studyualifying reportsb

esign suitabilityLeastGreatest

Reports may include more than one intervention arm.Grant-Petersson79 report was included for evaluation of knowledgutcomes.

he 20 qualifying studies that evaluated behavioral c


utcomes (increased sun-protective behaviors or im-rovement in health outcome) are provided in Appen-ix B and at www.thecommunityguide.org/cancer.A wide range of intervention activities was used,

ncluding didactic classroom teaching,103,108,116 didac-ic teaching using sunscreen samples,97 interactive classnd home-based activities,89,91,92,107 health fairs,91 anducational picture book,115 teaching by medical stu-ents,98 interactive CD-ROM multimedia pro-rams,90,101 and peer education.110,111 Relatively fewtudies included environmental or policy approaches.

The team did not conduct quantitative analyses ofhe intermediate outcomes of knowledge, attitude, andntentions. However, the overwhelming majority ofntervention arms showed a significant increase innowledge (22 out of 25) and a significant change inttitude (13 out of 17). Only four reports evaluatedntentions and their findings were inconsistent in direc-ion and generally not statistically significant.

Only one study evaluated the effect of an interven-ion on policies. This study reported significant im-rovements in adoption of a comprehensive sun-pro-ection policy in primary schools, but showed littleelationship between adoption of the policy and asso-iated sun-protective behavior changes that might haveesulted from the change in policy.112 Only one studyvaluated the effect of an intervention to reduce sun-urns; the intervention led to a 43% reduction ineported sunburns.88

Table 3 shows summary changes in sun-protectiveehaviors by study design. Specific sun-protective be-aviors included (1) covering up (wearing hats, long-leeved clothing, or pants); (2) using sunscreen;3) avoiding the sun (seeking shade, reschedulingctivities, not going out in the sun during peak UVours); and (4) composite behaviors (a combination oft least two of the above behaviors). Study designarkedly affected the effect size in these data. For

un-avoidance behaviors, the median relative changeas 4% for those studies that had a concurrent com-arison group and 16% for those studies that had aefore-and-after design. For covering-up behaviors, theedian relative change ranged from 25% (concurrent

rtsa

Number

3379–82,88–116

595

880–82,94,99,104–106

2079,88–93,97,98,100–103,107,108,110–112,115,116

779,88,93,98,103,115,116

1389–92,97,100–102,107,108,110–112

comes, but had limited quality of execution for behavioral/policy

repo

omparison) to 70% (before-and-after). For sunscreen

ber 5


u(Cw2

eiio(gdoaneaaid

Amoctscaae

Ocitoro

Etr

BtpragnbdftsnOh

Ceoifiscfibd

ES

Ieacs(ayiap(o

p

Tp

Ob

ACUC

CUAa

ob

scc

fudoN

se, the median relative change ranged from 17%concurrent comparison) to 34% (before-and-after).omposite behaviors were measured only in studiesith controls, where the median relative change was%.In general, evidence is sufficient to determine the

ffectiveness of interventions in primary schools inmproving the covering-up behavior; however, evidences insufficient to determine effectiveness in improvingther sun-protective behaviors, such as avoiding the sunminimizing UV exposure by seeking shade or notoing outside), because of inconsistent results. Evi-ence is also insufficient to determine the effectivenessf interventions in primary schools in changing policiesnd practices or health outcomes, based on smallumbers of studies and limitations in their design andxecution. These interventions improved knowledgend attitudes related to skin cancer prevention. Theylso increased sunscreen use (although this is not, bytself, a recommendation outcome, as previouslyiscussed).

pplicability. Studies that examined the team’s recom-endation outcomes of policies, behaviors, or health

utcomes were conducted in diverse geographical lo-ations, including Arizona,89–92 North Carolina,101 Aus-ralia,97,107,112 Canada,98 and France.88 Most of thetudies that reported race and ethnicity89–91,98 wereonducted among a predominantly white population,nd one study included only children of Europeanncestry.107 Four studies did not report race or

able 3. Median and interquartile relative changes in sun-rotective behaviors from interventions in primary schoolsa

utcomeehaviorsb

Interventionarms

Relative change

25th Median 75th

Studies with concurrent comparison groups

void sun 7 0.92 1.04 1.16over up 13 1.01 1.25 1.04se sunscreen 6 1.02 1.17 1.32ompositec 15 0.94 1.02 1.72

Before-and-after studies

over up 5 1.42 1.70 2.00se sunscreen 2 NA 1.34 NAvoid sun 1 NA 1.16 NA

Based on six reports.88,89,92,98,100,107 (A report may have more thanne intervention arm and more than one outcome measure.)Outcome behaviors: avoiding the sun includes behaviors such astaying in the shade, and avoiding the sun during peak UV hours; andover up includes wearing long-sleeved clothing or wearing a hat.Composite behavior was a combination of at least two of theollowing behaviors: avoiding sun, using protective clothing, andsing sunscreen. Often, presentation of the results did not allowifferentiation of the effect of one outcome from another behavioralutcome.A, not available.

thnicity.88,92,97,112 a

ther positive or negative effects. The studies in-luded in the team’s search for effectiveness did notnclude information on other potential benefits ofhese interventions, such as reduction in the risk ofverexposure to heat, or on potential harms, such aseductions in outdoor physical activity or transmissionf lice via hats or other clothing.

conomic efficiency. No studies were found that methe requirements for inclusion in a Community Guideeview.63

arriers to implementation. A potential barrier to in-ervention implementation might be the concerns ofarents or teachers that these interventions will lead toeductions in physical activity. Parents or teachers maylso be concerned that covering up will lead to wearingang insignia. (Until recently, California schools didot permit children to wear clothing such as hatsecause of concerns about gang affiliation. Now, stu-ents in California can wear hats to protect themselvesrom the UV rays and, ultimately, skin cancer.) Possibleransmission of head lice among younger children whohare hats has also been raised as a concern; however,one of the intervention studies cited this as a barrier.ne paper has examined the issue and reported thatats were not a major factor in transmission.84

onclusion. According to Community Guide rules ofvidence,64 available studies provide sufficient evidencef the effectiveness of interventions in primary schools

n improving covering-up behavior. Evidence was insuf-cient to determine effectiveness in improving otherun-protective behaviors (e.g., avoiding the sun) be-ause of inconsistent evidence; evidence was also insuf-cient to determine effectiveness in decreasing sun-urns because only a single study, with limitations inesign and execution, reported on this outcome.

ducational and Policy Interventions inecondary Schools and Colleges

nterventions in secondary schools and colleges involvefforts to promote sun-protective behaviors amongdolescents and young adults. These interventions in-lude at least one of the following activities: (1) provi-ion of information to adolescents and young adultse.g., instruction, small media, or both); (2) additionalctivities to influence the behavior of adolescents andoung adults (e.g., modeling, demonstration, role play-ng); (3) activities intended to change the knowledge,ttitudes, or behavior of caregivers (i.e., teachers orarents); and (4) environmental and policy approachese.g., providing sunscreen and shade, or schedulingutdoor activities to avoid hours of peak sunlight).Interventions in secondary schools and colleges are

otentially important because adolescents and young
dults are more likely to be exposed to UV radiation
Am J Prev Med 2004;27(5) 431

tslssdetausieattt

Eatatiot

cstdpup

ccsasieitm

ice

ytsimstti

nai((so

Ale

Ocitoro

Ele

Bib

Ceest

ER

Ivai((b

Ti

RR

QD

a

4

han younger children and are less likely to adoptun-protective behaviors, parents and caretakers haveess influence in promoting sun protection, and highchools and colleges can provide an infrastructure toupport intervention activities. Some data indicate thatespite high levels of knowledge about the healthffects of unprotected sun exposure, changes in atti-udes and social norms during adolescence are associ-ted with increases in high-risk behaviors and present anique challenge to health educators.117–119 Overall,un-protection programs have reported more success inmproving sun-protective practices for infants (by par-nts) and among younger children, but less successmong adolescents.118 Efforts in sun-protection educa-ion, supportive environments, and policies are difficulto sustain effectively as primary school children transi-ion to secondary schools.120

ffectiveness. The team’s search identified a total of 17rticles (25 intervention arms)110–112,121–134 on the effec-iveness of education and policy interventions in second-ry schools and colleges. Information about the disposi-ion of the reports is provided in Table 4. Mostntervention arms reported on multiple outcomes. Detailsf the qualifying reports are provided at www.

hecommunityguide.org/cancer.The intervention activities used in these studies in-

luded didactic classroom teaching combined withome interactive classroom and home-based activi-ies,126–128 Internet-based activities,121 small me-ia,123,126,129,131–133,135 and provision of sunscreen sam-les, extra class credit, or money.129,132,134 One studysed a strategy of information dissemination and sup-ort of school staff to facilitate policy implementation.112

Only four reports (six intervention arms) examinedhanges in sun-protective behavior or poli-y,112,121,128,129 and each report measured differentun-protective behaviors. One report focused on themount of time spent in the sun129 and another onunscreen use.121 Another report measured a compos-te behavior, which did not allow us to determine theffect of the intervention on specific protective behav-ors.128 The fourth paper presented self-reported prac-ices, but the presentation did not allow us to deter-

able 4. Interventions in secondary schools and colleges:nformation about reportsa

Number

eports meeting inclusion criteria 17110–112,121–134

eports excluded, limited executionquality

4110,111,122,124

ualifying reports 13112,121,123,125–134

esign suitabilityLeast 1126

Greatest 12112,121,123,125,127–134

Reports may include more than one intervention arm.

ine the relative effect of the intervention.112 The a


nconsistency of interventions undertaken and out-omes measured did not allow us to determine theffectiveness of the interventions.The team did not conduct formal quantitative anal-

ses of the intermediate outcomes of knowledge, atti-udes, and intentions. Nine intervention arms generallyhowed an increase in knowledge as a result of thentervention.123,126–128,130,131 Seven intervention arms

easured attitudes and beliefs, with inconsistent re-ults,121,125,130,133,134 and seven measured inten-ion125,129,132–134 (the majority of which looked only athe intention to use sunscreen.); these results were alsonconsistent.

Evidence is insufficient to determine the effective-ess of educational and policy interventions in second-ry schools and colleges because of (1) limitationsn the design and execution of available reports;2) variability in interventions and evaluated outcomes;3) short follow-up times; and (4) the small number oftudies that examined relevant outcomes such as healthutcomes and sun-protective behavioral changes.


ther positive or negative effects. The studies in-luded in the team’s search for effectiveness did notnclude information on other potential benefits ofhese interventions, such as reduction in the risk ofverexposure to heat, or on potential harms, such aseductions in outdoor physical activity or transmissionf lice via hats or other clothing.

conomic efficiency. Economic evidence was not col-ected because effectiveness of the intervention was notstablished.

arriers to implementation. Information on barriers tomplementation of this intervention was not collectedecause effectiveness was not established.

onclusion. According to Community Guide rules ofvidence,64 evidence is insufficient to determine theffectiveness of these interventions in secondarychools or colleges to reduce adverse health effects oro change behavior related to UV exposure.

ducational and Policy Interventions inecreational and Tourism Settings

nterventions in recreational and tourism settings in-olve efforts to promote sun-protective behaviorsmong adults, children, and their parents. Thesenterventions include at least one of the following:1) provision of information to children and adultsi.e., through instruction, small media education, oroth); (2) activities intended to change the knowledge,
ttitudes, beliefs, or intentions of children and adults;
ber 5



(coaop

aapDab2Fton

isosiseii

E(nsTebvo

nd(F

tsmuctbasssgoges

iiwcicphaitppaiaeo

sotd1ptssohdOfs

e(p1

Ti

R

R

R

QD

a

3) additional activities to influence the behavior ofhildren and adults (such as modeling, demonstration,r role playing); and (4) environmental or policypproaches, including provision of sunscreen or shade,r scheduling of outdoor activities to avoid hours ofeak sunlight.UV exposure often occurs during outdoor recre-

tional activity. The tourism industry has experiencedn increase in overseas vacationers traveling from tem-erate climates to regions where the UV level is high.omestic U.S. travel increased 8% from 1994 to 2001,

nd travel with children increased 17%.136 Of the 1illion domestic U.S. person-trips that took place in001, the top three destination states were California,lorida, and Texas; a high percentage of these vaca-ions involved outdoor activities, including generalutdoor activities (17%), visits to beaches (11%), orational or state park visits (10%).137

Participation in outdoor leisure activities has alsoncreased, thus increasing exposure to sunlight. Mosttudies show convincing trends towards increasing riskf melanoma with increasing recreational sun expo-ure.34 Recreational and tourism settings are thereforemportant sites for sun-protection programs. In suchettings, skin cancer education can be integrated intoxisting recreational or tourism activities, and support-ve policy and environmental interventions can also bemplemented.

ffectiveness. The team’s search identified 18 reports24 intervention arms)138–155 that evaluated the effective-ess of interventions in outdoor recreational or tourismettings. Information about the reports is provided inable 5. Details of a subset of the 11 qualifying studies thatvaluated behavioral outcomes (increased sun-protectiveehaviors or improvement in health outcome) are pro-ided in Appendix C and at www.thecommunityguide.rg/cancer.

The reports in the team’s body of evidence evaluatedumerous intervention activities aimed at both chil-ren (9 arms in 5 reports)141,144,146,149,150 and adults21 arms in 10 reports).138,139,141,144,146–148,150,153,155

able 5. Interventions in recreational or tourism settings:nformation about reportsa

Number

eports meeting inclusioncriteria

18138–155

eports excluded, limitedquality of execution

2151,152

eports on an alreadyincluded study

5140,142,143,145,154

ualifying reports 11138,139,141,144,146–150,153,155


Greatest 9138,139,144,146,147,149,150,153,155

Reports may involve more than one intervention arm.

our studies141,144,146,150 involved interactive sun-pro- t

ection education activities (stories, games, puzzles,tamps, arts and crafts) and sun-safe environment pro-otions; one study149 presented a UV-reduction curric-

lum at poolside, provided home-based activities forhildren and their parents, and measured degree ofanness using a colorimeter; three studies138,139,147 usedrochures to help educate participants about the prev-lence and severity of skin cancer,138 the effects of theun on the skin,139,147 or sun-protection mea-ures153,155; one study155 involved a sun sensitivity as-essment, sun damage imaging photographs, and sug-estions on reducing unprotected UV exposure; andne study148 relied on peer-leader modeling by life-uards, a poster listing goals, other informational post-rs and fliers, and a “commitment raffle” to influenceun-protective behaviors of children and adults.

Eleven arms from six reports141,144,146,148–150 exam-ned changes in parent-reported, sun-protective behav-ors among children (using sunscreen; seeking shade;earing a hat, shirt, or other protective clothing; andomposite behaviors). One report146 examined thencidence of children’s sunburn, and one examined149

hildren’s degree of tanness. Six arms from five re-orts141,146,148,153,155 examined adult sun-protective be-aviors, and two arms139,153 looked at incidence ofdult sunburn. Four arms from one report147 examinednformation-seeking behavior and follow-up study par-icipation of adults; six arms from three re-orts141,146,147 examined adult knowledge; two re-orts141,147 examined adult attitudes or beliefs; and 13rms from four reports138,147,149,153 examined adultntentions. Four arms from three reports141,144,146 ex-mined parent-reported sun-protection measures andnvironmental supports at outdoor recreational centersr swimming pools.Five arms from three reports146,148,155 demonstrated

ufficient evidence of effectiveness of the interventionn the adult sun-protective behavior of wearing protec-ive clothing (hat or shirt), showing a median relativeifference of 11.2% (interquartile range 5.1% to2.9%). Available studies that measured children’s sun-rotective behavior demonstrate inconsistent effects ofhe intervention on wearing a hat, wearing a shirt, andeeking shade. Two arms from one report146 demon-trated a desirable effect (a decrease) in the incidencef children’s sunburn (relative difference, �41.2%);owever, this evidence is insufficient for a recommen-ation because of an inadequate number of studies.verall, the evidence of effectiveness was inconsistent

or adult incidence of sunburn139,153 and children’sun-protective behaviors.141,144,146,148–150

Five arms from four reports141,144,146,149 providedvidence of effectiveness on children’s sunscreen usemedian relative difference, 9.8%), and composite sun-rotective behaviors (median relative difference,5.4%). Available reports found inconsistent effects on
he adult outcomes of information-seeking behavior or
Am J Prev Med 2004;27(5) 433



faAtms

AstagHir3eapIrfisfh$ew

Oretodttr

Ete

Btblnmasae

Cedob

itcs

dcst

P

Iptotiidp

atowfegraaoctcaernos2osswsb

E(epvr

4

ollow-up study participation,147 knowledge,141,146,147

ttitudes or beliefs,141,147 and intentions.138,147,148,150

vailable reports141,144,146 also demonstrated inconsis-ent effects on sun-protection measures and environ-

ental supports at outdoor recreational centers orwimming pools.

pplicability. These interventions from the evidenceubset took place in diverse outdoor recreational andourism settings such as recreational pools, beaches, zoosnd wild animal parks, and airplanes and in diverseeographical settings, including Australia,153 England,139

awaii,141,144,146 southern California,147,149,150 Virgin-a,148 and New England.138,146,155 Study participants’ agesanged from 6.5 years to 79 years, with a median age of1.5 years. Most of the reports that identified race orthnicity were based on studies conducted with

predominantly white population. Three re-orts141,144,146 involved Hawaiian or Asian/Pacificslander populations, and five138,139,147,150,153 did noteport race or ethnicity. Of the reports that identi-ed gender,138,139,141,144,146,147,149,153,155 most were oftudies conducted among predominantly (�50%)emale populations.138,139,141,144,147,149,153,155 Annualousehold income of study participants ranged from20,000 to �$90,000. Only one study141 reportedducation level; in this study, 88% of the participantsere high school graduates.

ther positive or negative effects. The studies in thiseview did not include information on other potentialffects of interventions in recreational or tourism set-ings. These may include reaching populations nottherwise exposed to skin cancer prevention and re-ucing the risk of overexposure to heat and UV radia-ion resulting from over-reliance on sunscreen. Poten-ial negative effects of the intervention includeeductions in outdoor physical activity.

conomic efficiency. No reports were found that methe requirements for inclusion in a Community Guideconomic review.63

arriers to implementation. Three potential barrierso implementation were identified in the literature ory experts. Recreational center staff may have only

imited time to implement the special activity compo-ent of an intervention144; swimming class schedulesay limit intervention activities at swimming pools141;

nd some in the tourism trade might worry that sun-afety concerns might adversely affect their business,nd hence the trade might be unwilling to partner infforts that may involve more than sunscreen use.

onclusion. According to Community Guide rules ofvidence,64 available reports provide sufficient evi-ence of effectiveness of interventions in recreationalr tourism settings to increase adult sun-protective
ehavior of covering up. Available reports provide t

nsufficient evidence to determine effectiveness of in-erventions in reducing sunburn in adults139,153 andhildren,146 because results were inconsistent (adultunburn) or too few in number (children’s sunburn).

Available reports141,144,146,149 also demonstrate evi-ence of effectiveness of the intervention based onhildren’s sun-protective behaviors, including sun-creen use and composite sun-protective behaviors;hese, however, are not recommendation outcomes.

rograms in Outdoor Occupational Settings

nterventions in occupational settings promote sun-rotective behaviors among workers. These interven-ions include at least one of the following: (1) provisionf information to the workers (instruction, educationhrough small media, or both); (2) additional activitiesntended to change the knowledge, attitudes, beliefs,ntentions, or behaviors of workers (i.e., modeling oremonstration); and (3) environmental or policy ap-roaches, including provision of sunscreen and shade.Outdoor workers in the United States are a crucial

udience for receiving skin cancer prevention informa-ion. According to the U.S. Census Bureau,156 in 1991,ver 8% of the U.S. national workforce (�9 millionorkers) primarily worked outdoors in one of the

ollowing occupational groups: construction, farm, for-stry, fishing, land surveying and mapping, gardeners,roundskeepers, mail carriers, and amusement park orecreational center attendants. From both scientificnd programmatic perspectives, occupational settingsre ideal sites for sun-protection programs. High ratesf nonmelanoma (basal cell and squamous cell) skinancer have been found among occupational groupshat work outdoors, and rates for nonmelanoma skinancer among outdoor workers are significantly associ-ted with cumulative UV exposure.28,157 Outdoor work-rs may receive up to six to eight times the dose of UVadiation that indoor workers receive.157 A recent Ca-adian survey found low levels of sun protection amongutdoor workers158: 44% seek shade, 38% avoid theun, 58% wear a hat or protective clothing, and 18% to3% reported using sunscreen while at work. Becauseutdoor workers receive intense and prolonged expo-ure to the sun, and are at increased risk of developingquamous cell cancer, interventions that educate theseorkers and modify their work environment are well

uited to the workplace and could provide substantialenefit.

ffectiveness. The team’s search identified 14 reports16 intervention arms)140 –142,145,148,151,159 –166 thatvaluated the effectiveness of interventions in occu-ational settings. More detailed information is pro-ided in Table 6. Details of the eight qualifyingeports140,142,145,159,161–163,165 are available at www.
hecommunityguide.org/cancer.
ber 5



neowccimpgeaeuab

eeSaptm

dissn

pvsbmdfii

a

msedimste

Asn

Odoceoit

Eln

Bib

Cdocr

H

Isschatpu(citsaptic

Ta

R

R

R

QD

a

The reports in the team’s body of evidence involvedumerous intervention activities, and evaluated a vari-ty of outcomes. Five reports141,145,160,162,163 consistedf interventions that provided sun safety training toorkers; two160,164 involved sun-protection and skinancer education sessions and skin exams by a physi-ian; six studies141,145,148,160,162,163 promoted cover-ng-up behaviors; five141,145,148,162,163 involved role-

odeling by lifeguards or aquatics instructors; one160

rovided sun protection to outdoor workers (sun-lasses, brimmed hat, and sunscreen); one165 usedducational brochures designed for men aged �45 andbody chart for self-assessment of pigmented lesions toducate male workers about skin cancer; and two145,163

sed environmental supports (sunscreen dispensersnd shade structures) to promote sun-protectiveehavior.Eleven arms from seven reports141,145,148,160,162–164

xamined changes in sun-protective behaviors and UVxposure, and one163 examined incidence of sunburn.ix arms from five141,145,163–165 reports141,145,163–165 ex-mined knowledge; five arms from four re-orts141,145,163,164 examined attitudes or beliefs; and

hree arms from two reports145,163 examined environ-ental pool policies.Available reports provided insufficient evidence to

etermine effectiveness of the intervention in increas-ng the sun-protective behaviors of covering up162,163 oreeking shade,163 or in decreasing the incidence ofunburn163 and UV exposure,160 because of the smallumber of reports and inconsistent results.162,163

Three arms from two reports145,163 examining sunrotection demonstrated desirable effects of the inter-ention on sun safety measures and environmentalupports (provision of sunscreen dispensers and porta-le shade structures) at recreational centers and swim-ing pools. Six arms from five reports141,145,163–165

emonstrated inconsistent effects on knowledge, andve arms from four reports141,145,163,164 demonstrated

nconsistent effects on attitudes or beliefs.According to Community Guide rules of evidence,64

able 6. Interventions in occupational settings: informationbout reportsa

Number


14140–142,145,148,151,159–166

eports excluded, limitedquality of execution

3151,161,166

eports on an alreadyincluded study

3140,142,159

ualifying reports 8141,145,148,160,162–165


Greatest 6145,160,162–165

Reports may involve more than one intervention arm.

vailable reports provide insufficient evidence to deter- a

ine the effectiveness of interventions in occupationalettings because of too few reports and inconsistentvidence. Although available reports demonstrate evi-ence of effectiveness of the intervention, based on

mprovements in sun-protection measures and environ-ental supports at outdoor recreational centers and

wimming pools, these policies are not, by themselves,hought to be adequate proxies for decreased UVxposure or better health.

pplicability. Evidence about applicability was not as-essed for this intervention because effectiveness wasot established.

ther positive or negative effects. Reviewed studiesid not include information on other potential effectsf these interventions. Other positive effects may in-lude reaching populations that might not otherwise bexposed to skin cancer prevention and reducing risk ofverexposure to heat. Potential negative effects of the

ntervention may include worker requests for reduc-ions in time spent working outdoors.

conomic efficiency. Economic evidence was not col-ected for this intervention because effectiveness wasot established.


onclusion. Available reports provide insufficient evi-ence to determine the effectiveness of interventions inccupational settings to reduce UV exposure and in-rease sun-protection behavior because of too feweports and inconsistent evidence.

ealthcare System and Provider Settings

nterventions to reduce UV exposure and promoteun-protective behaviors can take place in healthcareettings (e.g., pharmacies, drugstores, clinics, physi-ian’s offices, and medical schools) or can targetealthcare providers (e.g., physicians, nurses, physicianssistants, medical students, and pharmacists). Theeam included studies of interventions oriented towardrimary prevention and average-risk populations. Theysually include at least one activity aimed at providersto increase knowledge, attitudes, and intentions; in-rease positive role modeling for patients and clients;ncrease counseling behaviors or information provisiono patients and clients) or placed within a healthcareetting (to promote increased knowledge, attitudes,nd intentions about sun-protective behaviors amongatients and clients; to promote provision of informa-ion about skin cancer to patients and clients; and toncrease sun-protective behaviors among patients andlients).

Healthcare settings and primary care providers are in
unique position to provide advice and preventive
Am J Prev Med 2004;27(5) 435

sPtctFhesitpcp

tpsptplic

Epodiit

vmsTstca

cpttpssisutc

spStsmb

tsmiNopscactas

ore

Ale

Ocnse

Ele

Bi

Tsr

R

R

R

QD

a

4

ervices to the majority of the general population.eople in the United States make an average of 1.7 visitso a primary care provider annually,167 and surveysonsistently show that healthcare providers are arusted and important source of health information.or these reasons, healthcare providers and systemsave a unique opportunity to affect population knowl-dge, attitudes, and beliefs about reducing UV expo-ure and increasing sun-protective behaviors. Throughncreasing knowledge, changing attitudes and inten-ions, role modeling behaviors, and even establishingolicies, healthcare providers and healthcare settingsan greatly influence the behavior of the clients andatients who use their services.According to the USPSTF,62 evidence is insufficient

o recommend for or against regular counseling byrimary care clinicians to decrease sun exposure, avoidun lamps, use sunscreen or protective clothing, orractice skin self-examination. This review expands onhe USPSTF review by evaluating a broader range ofroviders and by including system approaches not

imited to providers. This review did not evaluatenterventions that focused on early detection of skinancers.

ffectiveness. The team’s search identified 21 re-orts80,94,99,168–186 on the effectiveness of interventionsriented to providers and healthcare systems. Moreetailed information about the reports is provided

n Table 7. Details of the 11 qualifying stud-es99,171,173–176,178,179,181–183 can be found at www.hecommunityguide.org/cancer.

The targets of the reviewed interventions were di-erse, as were the content of the interventions and theedia by which they were delivered. Several of the

tudies were targeted to diverse healthcare providers.wo interventions171 evaluated brief educational ses-

ions for physicians and house staff in a large, urbaneaching hospital; another181 evaluated a didactic skinancer prevention module aimed specifically at nurses;

able 7. Interventions in healthcare systems (healthcareettings and healthcare providers): information abouteportsa

Number


2194,99,168–186

eports excluded, limitedexecution quality

994,168–170,173,177,184–186

eports on an already includedstudy

1180

ualifying reports 1199,171,173–176,178,179,181–183

esign suitabilityLeast 599,174–176,182

Moderate 1178

Greatest design suitability 5171,173,179,181,183


nd another178 conducted a skin cancer prevention b


urriculum for medical students. One study182 taughthysicians through a curriculum on how to accuratelyriage skin lesions and counsel patients on skin cancer;wo studies used the Internet to train physicians174 (orhysicians, medical students, and house staff175) aboutkin cancer; one study taught medical students aboutkin cancer control, and then used the students to gonto elementary school classrooms and teach youngertudents about skin cancer control; and a final study179

sed videotapes and role modeling training procedureso teach and encourage pharmacists to engage theirlients in more skin cancer control behaviors.

Other studies were oriented to clients in healthcareettings. One study183 used the community drugstore toromote the message of appropriate sunscreen use andPF. Another study176 used a physician’s waiting roomo recruit and educate people about the importance ofunscreen. A final study173 tested the effects of differentessage content and sources of message on client

ehaviors.Two of the qualifying studies assessed recommenda-

ion outcomes,171,178 but results were generally incon-istent in direction and statistical significance. Measure-ents of provider behaviors were diverse in type and

nconsistent in direction and statistical significance.one of the studies in the review reported on behaviorsr exposures among clients, only on behaviors ofroviders toward clients. Several, but not all, qualifyingtudies showed improvements in intermediate out-omes, such as changes in knowledge, attitudes, beliefs,nd intentions of providers.99,178,181 Studies measuringlient knowledge, attitudes, beliefs, or intentionsended to show results in the desirable direction,lthough they did not consistently reach a level oftatistical significance.173,176,183

Lack of measurement of key behaviors and healthutcomes among clients and lack of consistency inesults provided insufficient evidence to determine theffectiveness of the intervention.


ther positive or negative effects. The studies in-luded in this body of evidence did not measureegative effects of reducing UV exposure. None of thetudies evaluated effects of these interventions on clinicfficiency or delivery of other preventive or clinical care.


arriers to implementation. Evidence about barriers tomplementation of this intervention was not collected
ecause effectiveness was not established.
ber 5



Ceefiitr

RRCM

Mbtpsmwbroiei

sbbtAAWwAmhc

E1ea

Dw

sdCcasartSppv

barilrrentptmptt

Ale

Orehpbs

Ele

Bib

Ceec

Tr

RRRQD

a

onclusion. According to Community Guide rules ofvidence,64 evidence is insufficient to determine theffectiveness of interventions in healthcare settings oror healthcare providers in reducing UV exposure orncreasing sun-protective behaviors. Not enough stud-es of sufficient design and execution quality evaluatedhe effectiveness of these interventions in changingecommendation outcomes.

esults. Part II: Interventions to Decrease UVadiation and Increase UV Protective Behaviors inross-Cutting Settingsass Media Campaigns Alone

ass media campaigns can promote sun-protectiveehaviors in a community. They provide informationhrough mass media (e.g., television, radio, newspa-ers, magazines, and billboards), and can also includemall media (e.g., brochures, flyers, newsletters, infor-ational letters, and posters). Mass media has beenidely used in public health programs to addressehavioral risk factors, and is a recognized vehicle foreaching wide audiences, particularly for the purposef raising awareness and concern about an issue. These

nterventions include activities to change the knowl-dge, attitude, beliefs, intentions, sun-protective behav-or, or health outcomes of children or adults.

In the area of skin cancer prevention and control,everal mass media campaigns in the United States haveeen initiated in the past decade, including campaignsy the Skin Cancer Foundation (www.skincancer.org);he federal government (www.chooseyourcover.gov); themerican Academy of Dermatology (www.aad.org); themerican Cancer Society (www.cancer.org); and theeather Channel (www.weather.com). These campaignsere launched because of the reported success ofustralia’s regional programs, which rely heavily onass media,187 but, to date no systematic reviewsave been conducted on the effectiveness of suchampaigns.188

ffectiveness. The team’s search identified a total of2 reports (eight intervention arms)189–200 on theffectiveness of mass media campaigns without other

able 8. Interventions in mass media: information abouteportsa

Number

eports meeting inclusion criteria 12189–200

eports excluded, limited execution quality 8191,194–200

eports on an already included study 1190

ualifying reports 3189,192,193

esign suitabilityLeast 1193

Greatest 2189,192


ctivities. More information is provided in Table 8. c

etails of the qualifying studies can be found atww.thecommunityguide.org/cancer.The interventions included a three-segment televi-

ion program that emphasized early detection and theangers of sun exposure and sunburn193; the use of aD-ROM–based information kiosk housed at sites ac-essible to the public192; media reporting of skin cancerdvisories in the form of a UV index189; and a rating ofunlight intensity coupled with recommendations forppropriate sun-protective measures. The team did noteview interventions that did not allow for evaluation ofhe effect of mass media alone on behavior change.tudies that included mass media as part of multicom-onent programs are evaluated below (see Multicom-onent and Comprehensive Community-Wide Inter-entions section).

Only two of the three included reports measuredehavior change, and neither of those studies allowedssessment of specific sun-safe behaviors.189,192 Oneeport examined the outcome of self-reported changen nonspecific sun-safe behavior among a select popu-ation (those who were aware of the UV index) as aesult of media dissemination of the UV index. Anothereport measured change by using a composite score ofight behavioral questions, which included a compo-ent of early detection or self-detection behaviors, and

hus did not allow the team to disentangle primaryrevention from secondary prevention behaviors. Thus,hese studies provide insufficient evidence to deter-

ine the effectiveness of mass media approaches toromote sun-safe behaviors or reduce UV exposure. Allhree reports found that mass media campaigns tendedo show increases in some aspects of knowledge.


ther positive or negative effects. The studies in thiseview did not include information on other potentialffects of the interventions. Some authors, however,ave cited a concern that a primary prevention cam-aign may result in increased unnecessary excisions ofenign skin lesions; this should be addressed in futurekin cancer public awareness campaigns.201


arriers to implementation. Evidence about barriers tomplementation of this intervention was not collectedecause effectiveness has not been established.

onclusion. According to Community Guide rules ofvidence,64 evidence is insufficient to determine theffectiveness of mass media interventions alone inhanging sun exposure behaviors. Evidence is insuffi-
ient because of (1) limitations in the design and
Am J Prev Med 2004;27(5) 437

http://www.skincancer.org

http://www.chooseyourcover.gov

http://www.aad.org

http://www.cancer.org

http://www.weather.com


eqa

P

ItcnpcbmuoirtS

hopfci(tcmvoi

pdccatdcesa

tcoptstpsstpatp

E(tUcpro

e(crTpavosstcgsocct

T

RRRQD

4

xecution of available studies; (2) the small number ofualifying studies; and (3) variability in interventionsnd outcomes evaluated.

rograms for Caregivers

nterventions for parents or caregivers involve activitieshat primarily promote sun-protective behaviors forhildren under their care. A caregiver is defined as aonparental adult who assumes responsibility, at leastart-time on a regular basis, for the care of at least onehild (e.g., professional nannies, mother’s helpers,abysitters, grandparents, or other family or householdembers, or daycare providers). Occupational or vol-

nteer caregivers such as lifeguards, teachers, coaches,r scout leaders were not included in this category of

nterventions, but were instead included in the team’seview of interventions in outdoor recreational andourism or occupational settings (see “Occupationalettings” and “Recreational and Tourism Settings”).Recommendation outcomes included changes in

ealth and behavioral outcomes of adults (i.e., parentsr other caregivers) or children. Interventions forarents and caregivers include at least one of theollowing: (1) provision of information to parents oraregivers and the children under their care (throughnstruction, education through small media, or both);2) activities intended to change the knowledge, atti-udes, beliefs, intentions, or behavior of parents oraregivers and the children under their care (i.e.,odeling, demonstration, or role playing); or (3) en-

ironmental or policy approaches, including provisionf sunscreen or shade, or scheduling of outdoor activ-

ties to avoid hours of peak UV radiation.Parents and caregivers play an important role in

rotecting children from UV radiation. In addition toirectly reducing children’s UV exposure, parents andaregivers can support sun-protective behaviors by in-orporating preventive behaviors into family routinesnd by serving as role models for the children underheir care. Parental beliefs about and involvement inisease prevention are important components of suc-essful skin cancer prevention programs for children,specially young children. Parents control family re-ources and activities and the availability of sunscreen

able 9. Interventions for parents or caregivers: information

eports meeting inclusion criteriaeports excluded, limited quality of executioneports on an already included studyualifying reportsesign suitabilityLeastGreatest

nd protective clothing.202 b


The literature reports significant correlations be-ween parental use of sunscreen and use by theirhildren,71,203 but no such relationship has been dem-nstrated with other sun-protective behaviors. Somearents know the risks of skin cancer, but do not realizehat children are at risk.67,68 Some parents believe auntan is a sign of good health; others use sunscreen onheir children as their only or preferred skin cancerrevention measure,59 even when other methods (e.g.,hade, clothing) are available. Sometimes parents applyunscreen on their children incorrectly and inconsis-ently50,204,205 (e.g., only after a child has experienced aainful sunburn). Reports of high sunburn ratesmong youth49,58 highlight the need for better educa-ion of parents and caregivers about appropriate sun-rotective behaviors.

ffectiveness. The team’s search identified 16 reports23 intervention arms)87,95,140–142,144,146,149–151,200,206–210

hat evaluated the effectiveness of interventions to preventV exposure or skin cancer, directed to parents or

aregivers. Information about the disposition of the re-orts is provided in Table 9. Details of the nine qualifyingeports are available at www.thecommunityguide.rg/cancer.

The reviewed reports used numerous activities andvaluated a variety of outcomes in one or both parentsall nine reports)141,144,146,149,150,206,207,209,210 and theirhildren (eight studies).141,144,146,149,150,207,209,210 (Noeports evaluated outcomes among other caregivers.)hree reports141,144,146 involved interactive sun-rotection activities (stories, games, puzzles, stamps,rts and crafts) and environmental supports (e.g., pro-iding sunscreen, shade, and signage); two207,210 reliedn educational materials or presentations; one149 pre-ented a UV exposure-reduction curriculum at pool-ide, provided home-based activities for parents andheir children, and measured degree of tanness using aolorimeter; one206 targeted new mothers, who wereiven sun-protective guidelines, postcard reminders,unscreen samples, baby sun hats, and sun umbrellas;ne150 used point-of-purchase prompts and discountoupons for children’s hats and sunscreen; and one209

ombined focused behavioral strategies with communi-y-wide publicity campaigns to change attitudes and

t reports

Number

1687,95,140–142,144,146,149–151,200,206–210

395,151,200

487,140,142,208

9141,144,146,149,150,206,207,209,210

2141,209

7144,146,149,150,206,207,210

abou

ehaviors of parents and their children.

ber 5



pUpcscNceFcrtieacaac

dssisstspbe

eaaaa

Asn

Ovfalepsitiand

Eln

Bib

Cedfsf

tbtmno(

CII

TparacsuataTtmmpl

adtbwacrIiph

Five intervention arms from four re-orts141,146,206,207,209 examined changes in parentalV exposure. Fourteen arms from eight re-orts141,144,146,149,150,207,209,210 examined changes inhildren’s sun-protective behaviors (using sunscreen;eeking shade; wearing a hat, shirt, or other protectivelothing; and composite sun-protective behaviors).ine arms from four reports150,206,207,209 examined

hanges in children’s UV exposure, and one report146

xamined changes in incidence of children’s sunburn.ive arms from three reports141,144,210 examinedhanges in parental knowledge; five arms from threeeports141,207,210 examined changes in parental atti-udes or beliefs; and one report150 examined changesn parental intentions. Three arms from two reports211

xamined changes in children’s attitudes or beliefs,nd four arms from three reports141,144,146 examinedhanges in parent-reported, sun-protection measuresnd environmental supports (provision of sunscreennd portable shade structures) at outdoor recreationalenters or swimming pools.

The reviewed reports demonstrated insufficient evi-ence of effectiveness of the intervention on parentalun-protective behavior,141,146,207 parental UV expo-ure,141,206,207,209 children’s sun-protective behav-or,141,144,146,149,150,207,209,210 children’s UV expo-ure,150,206,207,209 and incidence of children’sunburn146 because there were too few reports (paren-al UV exposure206,207,209 and incidence of children’sunburn146), or results were inconsistent (parental sun-rotective behavior,141,146,207 children’s sun-protectiveehavior,141,144,146,149,150,207,209,210 and children’s UVxposure150,206,207,209).Effects of the intervention on parental knowl-

dge,141,144,210 parental attitudes or beliefs,141,207,210

nd parental intentions were inconsistent.150 Threerms from two reports209,210 demonstrated desirablend consistent effects of the intervention on children’sttitudes or beliefs (median relative difference, 67.6%).

pplicability. Evidence about applicability was not as-essed for this intervention because effectiveness wasot established.

ther positive or negative effects. The systematic re-iew development team identified other potential ef-ects of interventions aimed at parents or caregivers,nd further evaluation is needed to determine theikelihood of their occurrence. Interventions for par-nts or caregivers may not only reduce risk of overex-osure to UV radiation based on over-reliance onunscreen use by the parent or caregiver, but thentervention may also reduce reliance on sunscreen forhe children under their care. Additionally, a reductionn UV exposure among this population may be associ-ted with reductions in cataract formation. Potentialegative effects of the intervention include vitamin D
eficiency and reductions in outdoor physical activity. l
conomic efficiency. Economic evidence was not col-ected for this intervention because effectiveness wasot established.


onclusion. According to Community Guide rules ofvidence,64 available reports provide insufficient evi-ence to determine the effectiveness of interventionsor parents or caregivers in preventing UV exposure orkin cancer. Evidence was insufficient because of tooew reports or inconsistent evidence.

The reviewed reports demonstrate that the interven-ion led to improvements in children’s attitudes oreliefs, as well as sun-safety measures and environmen-al supports at outdoor recreational centers and swim-

ing pools. Because changes in reported policies didot necessarily translate into changes in UV exposurer behavior, these are not recommendation outcomessee methods in Appendix A).

ommunity-Wide Multicomponent Programs,ncluding Comprehensive Community-Widenterventions

he team defined community-wide multicomponentrograms as those that used combinations of individu-l-directed strategies, mass media campaigns, and envi-onmental and policy changes in an integrated effort in

defined geographic area (city, state, province, orountry). Such programs may also incorporate setting-pecific strategies within the larger program. They aresually delivered with a defined theme, name, logo,nd set of messages.80,187 The team included studies inhis review if they occurred in a defined geographicrea, and had at least two components and two settings.he team defined comprehensive community-wide in-

erventions as being multilevel (i.e., those that includeultiple individual-directed, setting-specific, and com-unity-wide components), addressing a substantial

roportion of the population in a defined area, andasting longer than 1 year.

Multicomponent sun-protection programs aim tochieve behavioral changes among the population of aefined geographic area (e.g., counties, states, coun-ries). Some are relatively modest efforts, such as com-ining a setting-specific program with a community-ide mass media or small media effort, whereas othersre multilevel and comprehensive, involving entireommunities, schools, workplaces, healthcare and rec-eation settings, mass media, and other organizations.n addition to education, these programs may alsonclude significant efforts to institute sun-protectionolicies and structural supports. Programs like theseave been in place for 2 decades in Australia, with the

ongest-standing and most commonly cited ones being

Am J Prev Med 2004;27(5) 439

ttNPo

taprwctd

Etcfva

mcserc

gsaapnsocnAnc

hc

hsmoisapti

AtetaAoUiwheUmfmrS

Octii

Ele

Bib

Ce

Tr

RRRQD

4

he Slip! Slap! Slop! and SunSmart campaigns in Vic-oria.187 Two U.S. programs, the SafeSun Project inew Hampshire80,81 and the Falmouth Safe Skinroject in Massachusetts,209 have used similar strategiesn a smaller scale.The team conducted this review to assess the effec-

iveness of these programs and the extent to whichvailable evidence (which comes mostly from sustainedrograms in Australia, a country with high skin cancerates) may or may not generalize to the United States,here skin cancer is a less prominent public healthoncern and the population includes a higher propor-ion of dark-skinned individuals who are at lower risk ofeveloping skin cancer.

ffectiveness. The team’s search identified 35 reports ofhe effectiveness of multicomponent and comprehensiveommunity-wide interventions.79–82,94,187,188,209,212–238 In-ormation about the disposition of the reports is pro-ided in Table 10. Details of the ten qualifying reportsre available at www.thecommunityguide.org/cancer.Of the seven studies81,209,218,221,228,231,233 in the com-unity-wide multicomponent programs that measured

overing-up or sun-avoidance behaviors, four218,228,231,233

howed generally positive outcomes, and three oth-rs81,209,221 showed essentially no significant change in theecommended behaviors of interest (i.e., sun avoidance orovering up).

Results of comprehensive, community-wide pro-rams were generally more positive. All three suchtudies showed changes in covering-up or sun-avoid-nce behaviors in the desired direction. These results,ll from Australia, are promising, but by themselves stillrovide insufficient evidence to determine effective-ess due to small numbers of studies and limitations intudy design and execution (two time-series studies andne before-and-after study, all with fair quality of exe-ution). Insufficient evidence to determine effective-ess is not the same as evidence of ineffectiveness.dditional studies are needed to confirm the effective-ess of these programs and to determine whether theyan be successfully applied to the U.S. population.

Studies that evaluated self-reported sunscreen usead generally demonstrated desirable effects (in-

able 10. Interventions in community-wide multicomponenteports

eports meeting inclusion criteriaeports excluded, limited quality of executioneports on the same studyualifying reportsesign suitabilityLeastModerateGreatest

reased sunscreen use).231,233,237 Many of the compre- e


ensive community-wide studies evaluated changes inchool and government policy, and changes in environ-ent; they generally showed positive process measure

utcomes (e.g., increase in number of sun-safe policiesn schools or governments, increase in number oftores with available low-cost sunscreen or increase inmount of information and number of postersrovided).217,237 The effects on knowledge, atti-

udes, and beliefs of adults and children werenconsistent.209,231,234

pplicability. Because evidence was insufficient to de-ermine effectiveness, the team has not included a fullvaluation of applicability. However, it should be notedhat the most promising results of the available studiesre from three long-term, intensive interventions inustralia.217,231,237 The context in which those studiesccurred may differ in some important ways from the.S. context: For example, the incidence of skin cancer

s higher and thus more of a health priority in Australia,here there is less heterogeneity in skin color and thusigher skin cancer risk than in the United States, UVxposure is probably on average higher than in thenited States, and in some of those studies, the massedia component was heavily subsidized. Any of these

actors might affect the extent to which these resultsay or may not translate to the U.S. context. Additional

eplications in other contexts, especially in the Unitedtates, would be useful.

ther positive or negative effects. The studies in-luded in this body of evidence did not address poten-ial harms of reducing UV exposure, such as increasesn the incidence of vitamin D deficiencies or reductionsn physical activity.



onclusion. According to Community Guide rules ofvidence,64 evidence is insufficient to determine the

comprehensive community-wide settings: information about

Number

3579–82,94,187,188,209,212–238

5224,228,229,232,234

2279,80,82,94,187,188,212–217,220,222,223,225–227,230,235,236,238

881,209,218,219,221,231,233,237

3209,221,233

3218,231,237

281,219

and

ffectiveness of multicomponent programs to reduce

ber 5


Ucccina

G

AvivioriitrTeq

D

Oedfdvvo(dwthisentsostsehatdtA

aetsa

D

Sbtysetei(ssnsit

D

Ssdmsvtinetetptitp

Du

A�spiay

V exposure or increase sun-protective behaviors be-ause of inconsistent results. Evidence was also insuffi-ient to determine the effectiveness of comprehensiveommunity-wide programs to reduce UV exposure orncrease sun-protective behaviors because of smallumbers of studies and limitations in their study designnd execution.

eneral Research Issues

lthough the effectiveness of two recommended inter-entions (i.e., interventions in primary schools formproving children’s covering-up behaviors and inter-entions in recreational or tourism setting for improv-ng adult covering-up behaviors) for reducing the riskf skin cancer has been established, important generalesearch issues remain, some of which have beendentified in previous literature.22,65,239–242 The mostmportant and surprising conclusion is that, despite allhe issues, settings, and populations examined, littleesearch measures key behavioral and health outcomes.he research issues are organized below by the wayach individual study is scored from a design anduality perspective.243

esign and Analysis Considerations

f the 85 studies reviewed for all categories, 55 usedxperimental designs, and many involved group-ran-omized trials (note that overlap may exist for reportsound in more than one category). Most of the ran-omized controlled trials involved setting-specific inter-entions, whereas many of the community-wide inter-entions and one healthcare setting intervention usedther appropriate designs, such as time-series designsn �4). The remaining 26 studies used before-and-afteresigns. All of the designs have important strengths andeaknesses. Studies that incorporate concurrent con-

rol groups help to control for changes over time (e.g.,istory, maturation) that are not attributable to the

ntervention. Several of the included studies demon-trated the importance of this control and showedither desirable80 or undesirable changes (includingo change)146,192 in the control groups over time. The

eam’s review of school-based interventions foundmaller effects in those studies with greatest suitabilityf design, suggesting that simpler before-and-after de-igns may have overestimated intervention effects inhat review. On the other hand, before-and-after, time-eries, and similar designs have strengths as well. Forxample, before-and-after and time-series designs mayave fewer problems with contamination and may havedvantages with respect to external validity, becauseheir populations may be less highly selected (e.g., lessependent on volunteers, less highly motivated) thanhe populations included in tightly controlled trials.
dditional diverse approaches, in terms of study design i
nd execution, and with attention both to internal andxternal validity, are certainly worth pursuing. Consis-ently rigorous analytic methods are needed, and futuretudies should control for relevant confounders, suchs risk levels and weather conditions.

escription of Target Population and Context

everal reports in this area of research did not containasic descriptions about the intervention or popula-ion. For example, many studies did not report theear(s) in which the study took place. In many in-tances, the distribution of the population by race andthnicity or sun sensitivity was not described. Many ofhe settings could have been better described. Forxample, many reports did not describe the character-stics of the schools in which interventions took placee.g., whether schools were private or public, how manytudents they served, and the characteristics of thetudents and faculty). Finally, better descriptions areeeded of annual UV exposure in the places in whichtudies were conducted. Better descriptions of thesemportant issues will help to assess likely applicability ofhe findings or to explain any variability of effects.

escription of Intervention

everal reports mentioned that their programs empha-ized skin cancer prevention, but it was difficult toisentangle what the specific components were or howuch emphasis there was on primary prevention (ver-

us early detection) or on promoting use of sunscreenersus on covering-up or sun-avoidance behaviors. Fur-her information is needed on which attributes of thenterventions contribute most to intervention effective-ess or ineffectiveness (e.g., do policy components orducation components contribute more to interven-ion effectiveness; what are the central “active ingredi-nts” in complex interventions). Describing interven-ion characteristics in greater detail might also helpractitioners replicate successes. If journal space limi-ations are a barrier to more complete and usefulntervention descriptions, supplemental communica-ion strategies might be explored (e.g., Internetublication).48,244

uration of Interventions and Length of Follow-p

bout two thirds of the interventions had a duration of6 weeks, and more than half the evaluations followed

ubjects for �3 months. Given the seasonality of sun-rotective behaviors and the importance of encourag-

ng habitual as opposed to short-term behavior change,longer follow-up is crucial. The trend toward multi-

ear interventions and longer follow-up periods is an
mportant improvement, although limitations in re-
Am J Prev Med 2004;27(5) 441

st

I

Tpvmo

ob

ttatrbdpptciia

M

Fhiha

M

Meitbssssmcmmasih

n

pto

iboiiwantsuoisbiesshpt

artripsbrSriH“metap

cstbavsmo

4

earch resources can be an important barrier to longer-erm programs and evaluations.

ntervention Quality

his is one of the most difficult attributes to assess fromublications describing intervention studies.245 Inter-ention quality is also very hard to conceptualize andeasure. However, the following observations are

ffered.An encouraging trend can be seen in increasing use

f formative research and pre-testing of interventionsefore they are implemented.142,150,163

Mediating factors deserve greater attention and needo be correlated with behavior changes. For example, ifhe intervention is thought to work through changes inttitude, did these changes in attitude occur along withhe behavior change? To date, few studies in this area ofesearch have reported on both mediating factors andehavioral or health outcomes. A need also exists toevelop measures of the effects of environmental andolicy change strategies. Few interventions addressedolicy or environmental changes and in those that did,he effects of the policy or environmental componentsould not be disentangled from other aspects of thentervention. Often, the reports measured the increasen number of policies, but did not measure changes inctual practice.

easurement of Exposure

ew studies reported process evaluation data, which canelp to assess how much of the intervention was actually

mplemented. Improvement in this area would beelpful, especially for interventions of longer durationnd increased complexity.

easurement of Outcomes

ore behavioral and health outcomes need to bexamined. The majority of intervention studies exam-ned intermediate outcomes, such as knowledge, atti-udes, and intentions, rather than actual sun-protectiveehaviors or health outcomes. For example, in primarychool settings, only one study examined incidence ofunburn. In the secondary school setting, very fewtudies examined sun-protective behaviors other thanunscreen use. The pattern of examining many inter-ediate outcomes and few behavioral and health out-

omes was generally consistent across reviews. Further-ore, many studies that did measure behaviorseasured only sunscreen use. Given recent concerns

bout the adequacy of sunscreen as a sole protectivetrategy (see Sun-Protective Behaviors section for morenformation on sunscreens), additional behavioral andealth outcomes should also be measured.Outcomes need to be similar to evaluate effective-

ess. For example, measuring protective behaviors c


rior to intervention, but measuring behavioral inten-ions after the intervention does not allow calculationf an effect.Measurement of specific sun-protective behaviors is

mportant. Many of the studies reported compositeehaviors and did not allow measurement of the effectf the intervention on a specific sun-protective behav-

or. Other studies concentrated on only one behavior:mprovement in sunscreen use. More interventionsith a greater focus on covering up and sun avoidance,nd a decreased emphasis on sunscreen use areeeded. Given recent research findings on the effec-

iveness of sunscreens, more detailed research on sun-creen use is needed. Are higher SPF sunscreens beingsed? Does sunscreen use extend the amount of timeut in the sun? How do different sun-protective behav-

ors interact (e.g., does seeking shade make wearingunscreen or a hat unnecessary)? Also, the distinctionetween intentional sun exposure (to achieve a biolog-

cal response, such as a tan) and unintentional sunxposure (no specific intention to acquire a tan or totay out in the sun, but rather a result of daily activitiesuch as work or sports) has not been well studied. Itas been reported that intention may affect sun-rotective behavior and thus might affect interven-

ion effectiveness.50

The adequacy of the sun-protective behaviors is usu-lly not accounted for in self-report measures or parent-eported measures (for their children), although newools may enable us to improve the validity of self-eports.168 Most studies rely on self-reporting of behav-ors and their presumed determinants. Self-report isarticularly vulnerable to social-desirability bias. A fewtudies have used multiple self-report measures (e.g.,oth surveys and diaries), and have examined theelative merits of each for assessing behavior.140,141,246

tudies in nonschool settings have more often (but stillarely) used third-party observations of behaviors, visualnspection, and occasionally, physical measures.246

owever, observations may not reflect adoption of ahabit” but rather only a moment in time.150 Physicaleasures using erythema meters, spectrophotom-

ters,246 colorimeters,247 and polysulphone dosime-ers248 are useful, but may be impractical in large trials,nd measures of tanness may not be valid in nonwhiteopulations.None of these measurement strategies is without

hallenges, and additional triangulation using multipletrategies might be useful. Also, further work is neededo increase consistency between at least a core set ofehavior change measures that can be used to comparend contrast study results. Lack of consistency makes itery difficult to compare and contrast results acrosstudies. In addition, measures should be developed thatake the public health importance (or lack thereof) of

bserved changes easier to discern (e.g., small average
hanges in Likert-like scales are very hard to interpret).
ber 5

mocepthf

●

●

●

●

Ars

●

●

●

Ttama

TP

Lcwtcegtosg

tmtbcpodhsH

rmmTmbtttaiwstmtt

iritgswtittltmfd

mtbitiaa

Addressing this concern of a lack of consistency ineasurement, in 1998 a Canadian National Workshop

n measurements of sun-related behaviors developed aonsensus on a standard set of measures for programvaluation and for monitoring of sun exposure androtective behaviors.249 The workshop summarized

hat sun-related behaviors are different from otherealth behaviors such as nutrition and tobacco in the

ollowing ways:

Sun exposure varies substantially seasonally and geo-graphically, and collection measures must be sensi-tive to such external influences.Sun-related behaviors require individuals to inter-pret and respond to their risk according to a com-plex set of environmental and physiological clues.Comprehensive sun protection requires individualsto undertake a set of behaviors. Therefore, researchmust account for a set of behavioral outcomes ratherthan a single indicator.A number of nonbehavioral factors influence indi-vidual risk, including phenotype, occupation, andage.

dditionally, the workshop identified several issueselated to data collection and measurement unique toun-related behaviors:

Most research relied on self-reported behaviors byyouth and adults and proxy reports for children.More objective data collection tools (e.g., daylightexposure monitors) might be useful to validate self-reported measures.Operational definitions of sun-related variables havevaried substantially across published studies.Recall periods also vary substantially. Some studiesassess behavior over a long period of time and someover a short period of time (such as last weekend).

he workshop concluded with six essential core itemso include for future behavioral surveillance surveysnd smaller evaluation projects. These items includedeasurement of sunburn, sun exposure, phenotype,

nd five separate sun-protective behaviors.

heory, Conceptual Models, and Evidence forlanning Interventions

ike other types of health promotion efforts, skinancer prevention programs are most likely to succeedhen they are based on a clear understanding of the

argeted health behaviors and their environmentalontext. Theories about why people do or do notngage in sun-protective behaviors, and data about aiven target audience (e.g., about a particular popula-ion’s barriers to and facilitators of sun protection) areften helpful in guiding the search for promising anduitable interventions.250 Conceptual models of pro-
ram characteristics and outcomes of interest, such as t
he analytic frameworks developed for these reviews,ight also be helpful for program planning. Popula-

ion-wide surveys have been used to examine the distri-ution of behaviors and their determinants, and canontribute to the design of strategies to increase sun-rotective behaviors. Other formative research meth-ds, such as focus groups, have also been used toevelop targeted skin cancer prevention programs, andave been especially useful when working with under-tudied audiences in new locations, such as multiethnicawaii.142

Although few evaluation reports describe the theo-etical bases of interventions to prevent skin cancer,any of the more recent publications specify one orore of the theories that have guided their programs.hese have included learning theory,83 theories ofessage framing and fear arousal,87,133,138,208 applied

ehavioral analysis,148 stages of change,155 social cogni-ive theory,145,146 and ecologic approaches.187 Theorieshat suggest both individual-directed and environmen-al strategies are most compatible with multicomponentnd community-wide interventions. People who engagen the planning of skin cancer prevention programs—hether for community health improvement or for

cientific evaluation—should identify and examineheir assumptions and, in turn, build on the theories,

odels, and local context underlying their approacheso improve sun-protective behaviors in their communi-ies and target audiences.

An important area of inquiry in the arena of behav-oral research to reduce UV exposure concerns theelative effectiveness of various messages. Three stud-es123,132,134 in a secondary school setting reported aheory base for message development, but only one134

ave sufficient detail to replicate message testing. Thesetudies, which for the most part were not conducted asidely distributed interventions, were useful to guide

he development of messages used in setting-specificnterventions to reduce UV exposure. Some persuasionheories directed researchers to build messages thatested cognitive features, such as inductive or deductiveogic about sun-protective behaviors or benefits ofanning.132 Other theories directed researchers to craft

essages with emotional features, such as humor orear.134 The small number of available studies and theiriversity preclude overall conclusions.Research to improve the use of message-develop-ent theories of message development, especially in

erms of instruments and message templates, couldenefit interventions to increase sun-protective behav-

ors. In all studies, messages were assumed, rather thanested, to possess the characteristics that they werentended to have. Although differences in opinion exists to what constitutes a message development study andstudy in which the message is the intervention, either

ype of study could benefit by including a control group

Am J Prev Med 2004;27(5) 443

to

R

Ttiobbititapr

S

Awsts

I

Mbdta

I

Stlwwa

I

AsMotaiibasi

I

Stino

IS

A(tav

SP

TicatrUfotiutm

tuBoirwmrttpae

tkpcs

4

o rule out chance having as much to do with anutcome as the message.

esearch Needs and Work in Progress

he field of behavior change for skin cancer preven-ion has progressed significantly in the past decade, butmportant areas for further advancement exist. Asutlined above, these include design, measurement,etter description of interventions, development of aetter understanding of how environmental and policy

nterventions work, and studies in multiethnic popula-ions. The use of new communication technology andnternational collaborations can make significant con-ributions in these areas. The team hopes that thevailability of systematic reviews that identify bothrogress to date and the remaining gaps will help toeduce the gaps in available research.

pecific Research Issues

lthough most of the research gaps described aboveere general, and could explain why most setting-

pecific categories did not produce sufficient evidenceo determine effectiveness, a few research issues werepecific to the setting or target group.

nterventions for Secondary Schools

ore studies are needed to examine sun-protectiveehaviors of adolescents and young adults, and toetermine what kind of approach might work best inhis population, especially given the low baseline prev-lence of sun-protective behaviors.

nterventions in Occupational Settings

tudies that target the most common outdoor occupa-ional workers—mail carriers, agricultural workers,andscapers, horticulturists, foresters, constructionorkers, telephone line workers, commercial fisheryorkers, land surveyors and mappers, oil field workers,musement park attendants, and athletes—are needed.

nterventions in Healthcare Settings

lmost all studies in this category examined the coun-eling behavior of the provider and not the patient.

ore studies that examine the behavioral or healthutcome of the end user—the patient—are needed. Inhis small subset of studies, the provider was most oftenphysician or a physician-in-training, but studies exam-

ning the role of the nonphysician provider would helpdentify if counseling skills to change behavior might beetter suited for providers with the time and skills, suchs a nurse or a health educator. Additionally, moretudies are needed to examine healthcare system stud-
es oriented directly to patients. b

nterventions for Parents and Caregivers

tudies are needed to examine the effect of interven-ions on nonparental caregivers, as it is becomingncreasingly common for children to be cared for byonparental caregivers while both parents are at workutside the home.

nterventions in Multicomponent Communityettings

pproaches to better define the “active ingredients”i.e., the most important components that contributeo the success of these interventions) would be helpful,s would determining the applicability of these inter-entions to the U.S. population.

ummary: Findings of the Task Force on Communityreventive Services

he Task Force recommends two interventions tomprove sun-avoidance or covering-up behaviors: edu-ational and policy interventions in primary schools,nd programs for adults in outdoor recreational orourism settings. These Task Force recommendationsepresent tested interventions that promote decreasedV exposure at the community level. They can be used

or planning interventions to promote UV protectionr to evaluate existing programs. The other interven-ions that were reviewed, but for which evidence wasnsufficient to determine effectiveness, may also proveseful in providing a broader taxonomy of interven-ions that might be tried in communities and in pro-

oting additional testing and evaluation.The Task Force reviews and recommendations iden-

ify promising strategies for reducing UV exposuresing the interventions that have been proven effective.ecause sun-protective behaviors are not practicedften enough, and because the incidence of skin cancer

s increasing, interventions for which evidence is cur-ently insufficient deserve more research attention,hile interventions that have been proven effectiveerit increased attention to diffusion, dissemination,

eplication, and implementation. These reviews byhemselves do not provide advice about implementa-ion of effective programs, but the referenced articlesrovide additional detail for those with an interest, anddditional implementation advice is also availablelsewhere.61,65,244,251–253

Many of the recommended interventions had smallo moderate behavior change scores. Readers shouldeep in mind that the interventions were targeted toopulations rather than single individuals. Smallhanges in behavior in large populations can result inubstantial public health benefits.

The reviews on which the recommendations are
ased also provide a starting point for improving the
ber 5

qhmaen

U

CnipitGitUso

dpttamchwytsoassmtgppUtspca

TrPWDDNS

UMLACPPoNHVRJMtSSlHpJBM

PC

R

uality and usefulness of existing research. As mightave been predicted in this emerging area of research,any questions about these interventions remain to be

nswered. The team hopes that the documentation ofvidence gaps in these reviews will help to improve theext generation of research.

sing These Recommendations

hoosing interventions that are well matched to localeeds and capabilities, and then carefully implement-

ng those interventions, are vital steps for increasing UVrotection. In setting priorities for the selection of

nterventions to meet local objectives, recommenda-ions and other evidence provided in the Communityuide should be considered along with such local

nformation as skin cancer incidence, skin cancer mor-ality, prevalence of sun-protective behaviors, latitude,V index averages, resource availability, administrative

tructures, and economic and social environments ofrganizations and practitioners.An assessment of the community is essential to

eciding how, when, and where to focus skin cancerrevention efforts. Some of the most important issueso consider are priorities, place, population, and prac-ices. Establishing skin cancer prevention as a priority ischallenge in various settings. Skin cancer is but one ofany health topics that merit attention, and for certain

ommunities, skin cancer prevention may not be asigh priority as other cancers or other diseases. Inarmer climates, children spend more time outdoorsear-round, whereas in colder climates, outdoor activi-ies are much more limited by the weather. Snow-belttates and rainy regions may afford less opportunity forutdoor recreation, but may want to invest in educationbout UV exposure during winter sports seasons, whenunshine is prevalent, and for residents who travel tounnier regions. Population characteristics in a com-unity, such as relevant age group, gender distribu-

ion, skin phototype, and socioeconomic status canuide program planning. Although light-skinned peo-le are at a higher risk for skin cancer, darker-skinnedeople also need to take precautions when exposed toV radiation. All of these factors can affect the atti-

udes, ability, and behaviors of the community to takeun-protective precautions. Awareness of the existingractices for sun exposure and sun protection amongommunity members helps define the challenge inchieving optimal sun-safety practices.

he team thanks its consultants for their contributions to theeview: Ross Brownson, PhD, St. Louis University School ofublic Health, St. Louis MO; Robert Burack, MD, MPH,ayne State University, Detroit MI; Linda Burhansstipanov,rPH, Native American Cancer Research, Pine CO; Allenietrich, MD, MPH, Dartmouth Medical School, HanoverH; Russell Harris, MD, MPH, University of North Carolina
chool of Medicine, Chapel Hill; Tomas Koepsell, MD, MPH,
niversity of Washington, Seattle; Howard Koh, MD, MPH,assachusetts Department of Public Health, Boston; Peter

ayde, MD, MSc, Medical College of Wisconsin, Milwaukee;l Marcus, PhD, AMC Cancer Center, Denver CO; Margaret. Mendez, MPA, Texas Department of Health, Austin;atricia D. Mullen, DrPH, University of Texas School ofublic Health, Houston; Amilie Ramirez, PhD, Baylor Collegef Medicine, San Antonio TX; Linda Randolph, MD, MPH,ational Center for Education on Maternal and Childealth, Arlington VA; Lisa Schwartz, MD, Department ofeterans Affairs Medical Center, White River Junction VT;obert Smith, PhD, American Cancer Society, Atlanta GA;

onathan Slater, PhD, Minnesota State Health Department,inneapolis; Stephen Taplin, MD, Group Health Coopera-

ive, Seattle WA; Sally Vernon, PhD, University of Texaschool of Public Health, Houston; Fran Wheeler, PhD,chool of Public Health, University of South Carolina, Co-umbia; Daniel B. Wolfson, MHSA, Alliance of Community

ealth Plans, New Brunswick NJ; Steve Woloshin, MD, De-artment of Veterans Affairs Medical Center, White Riverunction VT; John K. Worden, PhD, University of Vermont,urlington; Jane Zapka, PhD, University of Massachusettsedical Center, Worchester.Points of view are those of the Task Force on Community

reventive Services and do not necessarily reflect those of theenters for Disease Control and Prevention.

eferences1. Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer statistics, 2000. CA

Cancer J Clin 2000;50:7–33.2. American Cancer Society. Cancer facts and figures—2003. Atlanta GA:

American Cancer Society, 2003.3. Ries LA, Wingo PA, Miller DS, et al. The annual report to the nation on

the status of cancer, 1973–1997, with a special section on colorectalcancer. Cancer 2000;88:2398–424.

4. National Cancer Institute. SEER Cancer Statistics Review, 1973–1999.Available at: http://seer.cancer.gov/csr/1973_1999/melanoma.pdf. Ac-cessed October 4, 2004.

5. Jemal A, Devesa SS, Hartge P, Tucker MA. Recent trends in cutaneousmelanoma incidence among whites in the United States. J Natl CancerInst 2001;93:678–83.

6. Hall HI, Miller DH, Rogers JD, Bewerse B. Update on the incidence andmortality from melanoma in United States. J Am Acad Dermatol1999;40:35–42.

7. Jemal A, Devesa SS, Fears TR, Hartge P. Cancer surveillance series:changing patterns of cutaneous malignant melanoma mortality ratesamong whites in the United States. J Natl Cancer Inst 2000;92:811–8.

8. Armstrong B, Kricker A. How much melanoma is caused by sun exposure?Melanoma Res 1993;3:395–401.

9. Gilchrest BA, Eller MS, Geller AC, Yaar M. The pathogenesis of melanomainduced by ultraviolet radiation. N Engl J Med 1999;340:1341–8.

10. Armstrong BK, Kricker A. The epidemiology of UV induced skin cancer.J Photochem Photobiol B 2001;63(1–3):8–18.

11. Whiteman DC, Whiteman CA, Green AC. Childhood sun exposure as arisk factor for melanoma: a systematic review of epidemiologic studies.Cancer Causes Control 2001;12:69–82.

12. Westerdahl J, Olsson H, Ingvar C. At what age do sunburn episodes playa crucial role for the development of malignant melanoma? Eur J Cancer1994;30A:1647–54.

13. Elwood JM. Melanoma and sun exposure: contrasts between intermittentand chronic exposure. World J Surg 1992;16:157–65.

14. Kricker A, Armstrong BK, English DR, Heenan PJ. Does intermittent sunexposure cause basal cell carcinoma? A case-control study in WesternAustralia. Int J Cancer 1995;60:489–94.

15. Gallagher R, Hill G, Bajdik CD, et al. Sunlight exposure, pigmentary
factors, and risk of nonmelanocytic skin cancer. I. Basal cell carcinoma.Arch Dermatol 1995;131:157–63.
Am J Prev Med 2004;27(5) 445

4

16. Gallagher RP. Sun exposure and non-melanocytic skin cancer. In: Grob JJ,Stern RS, MAckie RM, Weinstock WA, eds. Epidemiology, causes, andprevention of skin diseases. London: Blackwell Science, 1997:72–7.

17. Elwood JM, Jopson J. Melanoma and sun exposure: an overview ofpublished studies. Int J Cancer 1997;73:198–203.

18. Armstrong BK. Melanoma: childhood or lifelong sun exposure. In: GrobJJ, Stern RS, MAckie RM, Weinstock WA, eds. Epidemiology, causes, andprevention of skin diseases. London: Blackwell Science, 1997:63–6.

19. Whiteman D, Green A. Melanoma and sunburn. Cancer Causes Control1994;5:564–72.

20. Autier P, Dore JF, Cattaruzza MS, et al. Sunscreen use, wearing clothes,and number of nevi in 6- to 7-year-old European children. J Nat CancerInst 1998;90:1873–80.

21. Autier P, Dore JF, Lejeune F, et al. Recreational exposure to sunlight andlack of information as risk factors for cutaneous malignant melanoma.Results of an European Organization for Research and Treatment ofCancer (EORTC) case-control study in Belgium, France and Germany.Melanoma Res 1994;4:79–85.

22. Buller DB, Borland R. Skin cancer prevention for children: a criticalreview. Health Educ Behav 1999;26:317–43.

23. Autier P, Dore J-F. Influence of sun exposures during childhood andduring adulthood on melanoma risk. Int J Cancer 1998;77:533–7.

24. Taylor CR, Stern RS, Leyden JJ, Gilchrest BA. Photoaging/photodamageand photoprotection. J Am Acad Dermatol 1990;22:1–15.

25. Stern RS, Weinstein MC, Baker SG. Risk reduction for nonmelanoma skincancer with childhood sunscreen use. Arch Dermatol 1986;122:537–45.

26. Godar DE, Urbach F, Gasparro FP, Van Der Leun JC. UV doses of youngadults. Photochem Photobiol 2004;77:453–7.

27. Williams ML, Pennella R. Melanoma, melanocytic nevi, and other mela-noma risk factors in children. J Pediatr 1994;124:833–45.

28. Scotto J, Fears TR, Fraumeni JF Jr. Incidence of nonmelanoma skin cancerin the United States. Bethesda MD: National Cancer Institute, NationalInstitutes of Health, 1983 (83-2433).

29. National Cancer Institute. SEER Cancer Statistics Review, 1973–1998.Available at: http://seer.cancer.gov/Publications/CSR1973_1998/mela-noma.pdf. Accessed January 17, 2003.

30. Diffey BL. Solar ultraviolet radiation effects on biological systems. PhysMed Biol 1991;36:299–328.

31. National Cancer Institute. Incidence: melanoma of the skin. Available at:http://seer.cancer.gov/faststats/html/inc_melan.html. Accessed January17, 2003.

32. Pennello G, Devesa S, Gail M. Association of surface ultraviolet B radiationlevels with melanoma and nonmelanoma skin cancer in United Statesblacks. Cancer Epidemiol Biomarkers Prev 2000;9:291–7.

33. Rhodes AR, Weinstock MA, Fitzpatrick TB, Mihm Jr, MC Sober AJ. Riskfactors for cutaneous melanoma: a practical method of recognizingpredisposed individuals. JAMA 1987;258:3146–54.

34. Armstrong BK, English DR. Cutaneous malignant melanoma. In: Schot-tenfeld D, Fraumeni JF, eds. Cancer epidemiology and prevention. 2nded. New York: Oxford University Press, 1996:1282–312.

35. Scotto J, Fears TR, Kraemer KH, Fraumeni JF. Nonmelanoma skin cancer.In: Schottenfeld D, Fraumeni JF, eds. Cancer epidemiology and preven-tion. 2nd ed. New York: Oxford University Press, 1996:1313–30.

36. Kricker A, Armstrong BK, English DR, Heenan PJ. Pigmentary andcutaneous risk factors for non-melanocytic skin cancer—a case-controlstudy. Int J Cancer 1991;48:650–62.

37. Holly EA, Kelly JW, Ahn DK, Shpall SN, Rosen JI. Risk of cutaneousmelanoma by number of malnocytic nevi and correlation of nevi byanatomic site. In: Gallagher RP, Elwood JM, eds. Epidemiological aspectsof cutaneous malignant melanoma. Boston: Kluwer, 1994:159–72.

38. Holly EA, Kelly JW, Shpall SN, Chiu SH. Number of melanocytic nevi as amajor risk factor for malignant melanoma. J Am Acad Dermatol1987;17:459–68.

39. Goldstein AM, Tucker MA. Genetic epidemiology of familial melanoma.Dermatol Clin 1995;13:605–12.

40. International Agency for Research on Cancer. Solar and ultravioletradioation. Monographs on the evaluation of carcinogenic risks to hu-mans. Lyon, France: International Agency for Research on Cancer, 1992.

41. Koh H, Sinks T, Geller A, Miller D, Lew R. Etiology of melanoma. In:Nathanson L, ed. Current research and clinical management of mela-noma. Boston: Kluwer, 1993:1–28.

42. Thompson SC, Jolley D, Marks R. Reduction of solar keratoses by regular
sunscreen use. N Engl J Med 1993;329:1147–51.

43. Naylor MF, Boyd A, Smith DW, Cameron GS, Hubbard D, Neldner KH.High sun protection factor sunscreens in the suppression of actinicneoplasia. Arch Dermatol 1995;131:170–5.

44. Green A, Williams G, Neale R. Does daily use of sunscreen or beta-carotene supplements prevent skin cancer in healthy adults? West J Med2000;173:332.

45. Gallagher R, Rivers JK, Lee T, Bajdik CD, McLean DI, Coldman AJ.Broad-spectrum sunscreen use and the development of new nevi in whitechildren. A randomized controlled trial. JAMA 2000;283:2955–60.

46. Vainio H, Bianchini F. Cancer-preventive effects of sunscreens are uncer-tain. Scand J Work Environ Health 2000;26:529–31.

47. Weinstock MA. Do sunscreens increase or decrease melanoma risk: anepidemiologic evaluation. J Investig Dermatol Symp Proc 1999;4:97–100.

48. Geller AC, Colditz G, Oliveria S, et al. Sunscreen use, sunburning rates,and tanning bed use among more than 10,000 US children and adoles-cents. Pediatrics 2002;109:1009–14.

49. Davis KJ, Cokkinides VE, Weinstock MA, O’Connell MC, Wingo PA.Summer sunburn and sun exposure among US youths ages 11 to 18:national prevalence and associated factors. Pediatrics 2002;110:27–35.

50. Vainio H, Bianchini F, eds. Sunscreens. Vol. 5 of IARC handbooks ofcancer prevention. Lyon, France: International Agency for Research onCancer, 2001.

51. Odio MR, Veres DA, Goodman JJ, et al. Comparative efficacy of sunscreenreapplication regimens in children exposed to ambient sunlight. Photo-dermatol Photoimmunol Photomed 1994;10:118–25.

52. Bech-Thomsen N, Wulf HC. Sunbathers’ application of sunscreen isprobably inadequate to obtain the sun protection factor assigned to thepreparation. Photodermatol Photoimmunol Photomed 1992;9:242–4.

53. McLean DI, Gallagher R. Sunscreens. Use and misuse. Dermatol Clin1998;16:219–26.

54. Hall HI, May DS, Lew RA, Koh HK, Nadel M. Sun protection behaviors ofthe U.S. white population. Prev Med 1997;26:401–7.

55. Hall HI. Sun protection behaviors among African-Americans. Ethn Dis1999;9:126–31.

56. Saraiya M, Hall HI, Uhler RJ. Sunburn prevalence among adults in theUnited States, 1999. Am J Prev Med 2002;23:91–7.

57. Cokkinides VE, Johnston-Davis K, Weinstock M, et al. Sun exposure andsun-protection behaviors and attitudes among U.S. youth, 11 to 18 years ofage. Prev Med 2001;33:141–51.

58. Hall HI, McDavid K, Jorgensen CM, Kraft JM. Factors associated withsunburn in white children aged 6 months to 11 years. Am J Prev Med2001;20:9–14.

59. Hall HI, Jorgensen CM, McDavid K, Kraft JM, Breslow R. Protection fromsun exposure in U.S. white children ages 6 months to 11 years. PublicHealth Rep 2001;116:353–61.

60. U.S.Department of Health and Human Services. Healthy people 2010.2nd ed. Washington DC: U.S. Government Printing Office, 2000.

61. Glanz K, Saraiya M, Wechsler H. Guidelines for school programs toprevent skin cancer. MMWR Morb Mortal Wkly Rep 2002;51:1–18.

62. U.S. Preventive Services Task Force. Counseling to prevent skin cancer:recommendation statement. Available at: www.ahrq.gov/clinic/uspstf/uspsskco.htm. Accessed March 1, 2004.

63. Carande-Kulis VG, Maciosek MV, Briss PA, et al. Methods for systematicreviews of economic evaluations for the Guide to Community PreventiveServices. Task Force on Community Preventive Services Am J Prev Med2000;18(suppl 1):75–91.

64. Briss PA, Zaza S, Pappaioanou M, et al. Developing an evidence-basedGuide to Community Preventive Services—methods. The Task Force onCommunity Preventive Services Am J Prev Med 2000;18(suppl 1):35–43.

65. Glanz K, Saraiya M, Briss PA. Impact of intervention strategies to reduceUVR exposure. In: Hill D, Elwood JM, English D, eds. Prevention of skincancer. Melbourne, Australia: Kluwer, 2002:58–82.

66. Marks R. Role of childhood in the development of skin cancer. AustPaediatr J 1988;24:337–8.

67. Grob JJ, Guglielmina C, Gouvernet J, Zarour H, Noe C, Bonerandi JJ.Study of sunbathing habits in children and adolescents: application to theprevention of melanoma. Dermatology 1993;186:94–8.

68. Maducduc L, Wagner R, Wagner K. Parents’ use of sunscreen onbeach-going children. Arch Dermatol 1992;128:628–9.

69. Weinstein JM, Yarnold PR, Hornung RL. Parental knowledge and practiceof primary skin cancer prevention: gaps and solutions. Pediatr Dermatol2001;18:473–7.

70. Centers for Disease Control and Prevention. Sun-protection behaviors
used by adults for their children—United States, 1997. MMWR MorbMortal Wkly Rep 1998;47:480–1.
ber 5

http://www.ahrq.gov/clinic/uspstf/uspsskco.htm

http://www.ahrq.gov/clinic/uspstf/uspsskco.htm

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

71. Johnson K, Davy L, Boyett T, Weathers L, Roetzheim RG. Sun protectionpractices for children: knowledge, attitudes, and parent behaviors. ArchPediatr Adolesc Med 2001;155:891–6.

72. Moise AF, Harrison SL, Gies HP. Solar ultraviolet radiation exposure ofinfants and small children. Photodermatol Photoimmunol Photomed1999;15:109–14.

73. National Institute of Child Health and Human Development Early ChildCare Research Network. Characteristics and qualities of child care fortoddlers and preschoolers. Appl Dev Sci 2000;4:116–35.

74. Grin CM, Pennoyer JW, Lehrich DA, Grant-Kels JM. Sun exposure ofyoung children while at day care. Pediatr Dermatol 1994;11:304–9.

75. Loescher LJ, Emerson J, Taylor A, Christensen DH, McKinney M. Educat-ing preschoolers about sun safety. Am J Public Health 1995;85:939–43.

76. Boldeman C, Jansson B, Holm LE. Primary prevention of malignantmelanoma in a Swedish urban preschool sector. J Cancer Educ1991;6:247–53.

77. Boldeman C, Ullen H, Mansson-Brahme E, Holm LE. Primary preventionof malignant melanoma in the Stockholm Cancer Prevention Programme.Eur J Cancer Prev 1993;2:441–6.

78. Crane LA, Schneider LS, Yohn JJ, Morelli JG, Plomer KD. “Block the Sun,Not the Fun”: evaluation of a skin cancer prevention program for childcare centers. Am J Prev Med 1999;17:31–7.

79. Grant-Petersson J, Dietrich AJ, Sox CH, Winchell CW, Stevens MM.Promoting sun protection in elementary schools and child care settings:the Sun Safe Project. J Sch Health 1999;69:100–6.

80. Dietrich AJ, Olson AL, Sox CH, et al. A community-based randomized trialencouraging sun protection for children. Pediatrics 1998;102:E64.

81. Dietrich AJ, Olson AL, Sox CH, Tosteson TD, Grant-Petersson J. Persistentincrease in children’s sun protection in a randomized controlled commu-nity trial. Prev Med 2000;31:569–74.

82. Olson AL, Dietrich AJ, Sox CH, Stevens MM, Winchell CW, Ahles TA.Solar protection of children at the beach. Pediatrics 1997;99:E1.

83. Loescher LJ, Buller MK, Buller DB, Emerson J, Taylor AM. Publiceducation projects in skin cancer. The evolution of skin cancer preventioneducation for children at a comprehensive cancer center Cancer 1995;75(suppl 2):651–6.

84. Speare R, Buettner PG. Hard data needed on head lice transmission. IntJ Dermatol 2000;39:877–8.

85. Cockburn J, Hennrikus D, Scott R, Sanson-Fisher R. Adolescent use ofsun-protection measures. Med J Aust 1989;151:136–40.

86. Dobbinson S, Peipers A, Borland R, Nolan K. Are Victorian primaryschools Sunsmart? Health Promotion J Aust 2000;10:43–50.

87. Buller DB, Borland R, Burgoon M. Impact of behavioral intention oneffectiveness of message features: evidence from the family sun safetyproject. Hum Community Res 1998;24:433–53.

88. Bastuji-Garin S, Grob JJ, Grognard C, Grosjean F. Melanoma prevention:evaluation of a health education campaign for primary schools. ArchDermatol 1999;135:936–40.

89. Buller DB, Buller MK, Beach B, Ertl G. Sunny Days, Healthy Ways:evaluation of a skin cancer prevention curriculum for elementary school-aged children. J Am Acad Dermatol 1996;35:911–22.

90. Buller DB, Hall JR, Powers PJ, et al. Evaluation of the “Sunny Days,Healthy Ways” sun safety CD-ROM program for children in grades 4 and5. Cancer Prev Control 1999;3:188–95.

91. Buller MK, Goldberg G, Buller DB. SunSmart Day: a pilot program forphotoprotection education. Pediatr Dermatol 1997;14:257–63.

92. Buller MK, Loescher LJ, Buller DB. “Sunshine and Skin Health”: acurriculum for skin cancer prevention education. J Cancer Educ1994;9:155–62.

93. DeLong M, LaBat K, Gahring S, Nelson N, Leung L. Implications of aneducational intervention program designed to increase young adoles-cents’ awareness of hats for sun protection. Clothing Textiles Res J1999;17:73–83.

94. Dietrich AJ, Olson AL, Sox CH, Winchell CW, Grant-Petersson J, CollisonDW. Sun protection counseling for children: primary care practicepatterns and effect of an intervention on clinicians. Arch Fam Med2000;9:155–9.

95. Fleming C, Newell J, Turner S, Mackie R. A study of the impact of SunAwareness Week 1995. Br J Dermatol 1997;136:719–24.

96. Fork HE, Wagner RFJ, Wagner KD. The Texas peer education sunawareness project for children: primary prevention of malignant mela-noma and nonmelanocytic skin cancers. Cutis 1992;50:363–4.

97. Girgis A, Sanson-Fisher RW, Tripodi DA, Golding T. Evaluation of
interventions to improve solar protection in primary schools. Health EducQ 1993;20:275–87.
98. Gooderham MJ, Guenther L. Sun and the skin: evaluation of a sunawareness program for elementary school students. J Cutan Med Surg1999;3:230–5.

99. Gooderham MJ, Guenther L. Impact of a sun awareness curriculum onmedical students’ knowledge, attitudes, and behaviour. J Cutan Med Surg1999;3:182–7.

00. Hoffmann III, RG Rodrigue JR, Johnson JH. Effectiveness of a school-based program to enhance knowledge of sun exposure: attitudes towardsun exposure and sunscreen use among children. Child Health Care1999;28:69–86.

01. Hornung RL, Lennon PA, Garrett JM, DeVellis RF, Weinberg PD,Strecher VJ. Interactive computer technology for skin cancer preventiontargeting children. Am J Prev Med 2000;18:69–76.

02. Hughes AS. Sun protection and younger children: lessons from the Livingwith Sunshine program. J Sch Health 1994;64:201–4.

03. Labat KL, DeLong MR, Gahring S. Evaluation of a skin cancer interven-tion program for youth. J Fam Consumer Sci 1996;88:3–10.

04. Milne E, Corti B, English DR, Cross D, Costa C, Johnston R. The use ofobservational methods for monitoring sun-protection activities in schools.Health Educ Res 1999;14:167–75.

05. Milne E, English DR, Corti B, et al. Direct measurement of sun protectionin primary schools. Prev Med 1999;29:45–52.

06. Milne E, English DR, Cross D, Corti B, Costa C, Johnston R. Evaluation ofan intervention to reduce sun exposure in children: design and baselineresults. Am J Epidemiol 1999;150:164–73.

07. Milne E, English DR, Johnston R, et al. Improved sun protectionbehaviour in children after two years of the Kidskin intervention. Aust NZ J Public Health 2000;24:481–7.

08. McWhirter JM, Collins M, Bryant I, Wetton NM, Bishop JN. Evaluating“Safe in the Sun”, a curriculum programme for primary schools. HealthEduc Res 2000;15:203–17.

09. Ramstack JL, White SE, Hazelkorn KS, Meyskens FL. Sunshine and SkinCancer: a school-based skin cancer prevention project. J Cancer Educ1986;1:169–76.

10. Reding DJ, Fischer V, Gunderson P, Lappe K. Skin cancer prevention: apeer education model. Wis Med J 1995;94:77–81.

11. Reding DJ, Fischer V, Gunderson P, Lappe K, Anderson H, Calvert G.Teens teach skin cancer prevention. J Rural Health 1996;12(4suppl):265–72.

12. Schofield MJ, Edwards K, Pearce R. Effectiveness of two strategies fordissemination of sun-protection policy in New South Wales primary andsecondary schools. Aust N Z J Public Health 1997;21:743–50.

13. Syson-Nibbs L. Measuring effectiveness of Sun Safety. Health Visit1996;69:274–7.

14. Syson-Nibbs L. Sun safety education in schools. Educ Health1996;14:37–41.

15. Thornton C, Piacquadio DJ. Promoting sun awareness: evaluation of aneducational children’s book. Pediatrics 1996;98:52–5.

16. Vitols P, Oates RK. Teaching children about skin cancer prevention: whywait for adolescence? Aust N Z J Public Health 1997;21:602–5.

17. Marks R, Hill D. Behavioural change in adolescence: a major challenge forskin-cancer control in Australia. Med J Aust 1988;149:514–5.

18. Hill D, Dixon H. Promoting sun protection in children: rationale andchallenges. Health Educ Behav 1999;26:409–17.

19. Arthey S, Clarke VA. Suntanning and sun protection: a review of thepsychological literature. Soc Sci Med 1995;40:265–74.

20. Dobbinson S, Peipers A, Reading D, Sinclair C. A national approach toskin cancer prevention: the National SunSmart Schools Program. Med JAust 1998;169:513–4.

21. Bernharndt J. Tailoring messages and design in a Web-based skin cancerprevention intervention. Int Electron J Health Educ 2001;4:290–7.

22. Castle CM, Skinner TC, Hampson SE. Young women and suntanning: anevaluation of a health education leaflet. Psychol Health 1999;14:517–527.

23. Cody R, Lee C. Behaviors, beliefs, and intentions in skin cancer preven-tion. J Behav Med 1990;13:373–89.

24. Hughes BR, Altman DG, Newton JA. Melanoma and skin cancer: evalua-tion of a health education programme for secondary schools. Br JDermatol 1993;128:412–7.

25. Jones JL, Leary MR. Effect of appearance-based admonitions against sunexposure on tanning intentions in young adults. Health Psychol1994;13:86–90.

26. Kamin CS, O’Neill PN, Ahearn MJ. Developing and evaluating a cancerprevention teaching module for secondary education: Project SAFETY(Sun Awareness for Educating Today’s Youth). J Cancer Educ
1993;8:313–8.
Am J Prev Med 2004;27(5) 447

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

4

27. Katz RC, Jernigan S. Brief report: an empirically derived educational programfor detecting and preventing skin cancer. J Behav Med 1991;14:421–8.

28. Lowe JB, Balanda KP, Stanton WR, Gillespie AM. Evaluation of athree-year school-based intervention to increase adolescent sun protec-tion. Health Educ Behav 1999;26:396–408.

29. Mahler HI, Fitzpatrick B, Parker P, Lapin A. The relative effects of ahealth-based versus an appearance-based intervention designed to in-crease sunscreen use. Am J Health Promotion 1997;11:426–9.

30. Mermelstein RJ, Riesenberg LA. Changing knowledge and attitudes aboutskin cancer risk factors in adolescents. Health Psychol 1992;11:371–6.

31. Mickler TJ, Rodrigue JR, Lescano CM. A comparison of three methods ofteaching skin self-examinations. J Clin Psychol Med Settings1999;6:273–86.

32. Prentice-Dunn S, Jones JL, Floyd DL. Persuasive appeals and the reduc-tion of skin cancer risk: the roles of appearance concern, perceivedbenefits of a tan, and efficacy information. J Appl Soc Psychol1997;27:1041–7.

33. Rothman AJ, Salovey P, Antone C, Keough K. The influence of messageframing on intentions to perform health behaviors. J Exp Soc Psychol1993;29:408–33.

34. Stephenson MT, Witte K. Fear, threat, and perceptions of efficacy fromfrightening skin cancer messages. Public Health Rev 1998;26:147–74.

35. Stephenson A, From L, Cohen A, Tipping J. Family physician’s knowledgeof malignant melanoma. J Am Acad Dermatol 1997;37:953–7.

36. Travel Industry Association of America. Domestic travel in the United States.Available at: www.tia.org/Travel/tripChar.asp. Accessed March 1, 2004.

37. Travel Industry Association of America. State tourism facts. Available at:www.tia.org/Travel/StateTourism.asp. Accessed December 11, 2002.

38. Detweiler JB, Bedell BT, Salovey P, Pronin E, Rothman AJ. Messageframing and sunscreen use: gain-framed messages motivate beach-goers.Health Psychol 1999;18:189–96.

39. Dey P, Collins S, Will S, Woodman CB. Randomised controlled trialassessing effectiveness of health education leaflets in reducing incidenceof sunburn. BMJ 1995;311:1062–3.

40. Glanz K, Silverio R, Farmer A. Diary reveals sun protective practices. SkinCancer Fdn J 1996;14:27–8.

41. Glanz K, Chang L, Song V, Silverio R, Muneoka L. Skin cancer preventionfor children, parents, and caregivers: a field test of Hawaii’s SunSmartprogram. J Am Acad Dermatol 1998;38:413–7.

42. Glanz K, Carbone E, Song V. Formative research for developing targetedskin cancer prevention programs for children in multiethnic Hawaii.Health Educ Res 1999;14:155–66.

43. Glanz K, Lew RA, Song V, Murakami-Akatsuka L. Effects of skin cancerprevention in outdoor recreation settings: the Hawaii SunSmart Program.Eff Clin Pract 2000;3:1–5.

44. Glanz K, Lew RA, Song V, Murakami-Akatsuka L. Skin cancer preventionin outdoor recreation settings: effects of the Hawaii SunSmart Program.Eff Clin Pract 2000;3:53–61.

45. Glanz K, Maddock JE, Lew RA, Murakami-Akatsuka L. A randomized trialof the Hawaii SunSmart program’s impact on outdoor recreation staff.J Am Acad Dermatol 2001;44:973–8.

46. Glanz K, Geller AC, Shigaki D, Maddock JE, Isnec MR. A randomized trialof skin cancer prevention in aquatic settings: the Pool Cool Program.Health Psychol 2002;21:579–87.

47. Keesling B, Friedman HS. Interventions to prevent skin cancer: experi-mental evaluation of informational and fear appeals. Psychol Health1995;10:477–90.

48. Lombard D, Neubauer TE, Canfield D, Winett RA. Behavioral communityintervention to reduce the risk of skin cancer. J Appl Behav Anal1991;24:677–86.

49. Mayer JA, Slymen DJ, Eckhardt L, et al. Reducing ultraviolet radiationexposure in children. Prev Med 1997;26:516–22.

50. Mayer JA, Lewis E.C, Eckhardt L, et al. Promoting sun safety among zoovisitors. Prev Med 2001;33:162–9.

51. Parrott R, Duggan A, Cremo J, Eckles A, Jones K, Steiner C. Communi-cating about youth’s sun exposure risk to soccer coaches and parents: apilot study in Georgia. Health Educ Behav 1999;26:385–95.

52. Presant CA, Presant SL, Mack J, Atterbury GB, French RA, Schroeder T.Youth cancer education through a combined American Cancer Society–Boy Scouts of America pilot program. J Cancer Educ 1987;2:229–31.

53. Segan C, Borland R, Hill D. Development and evaluation of a brochure onsun protection and sun exposure for tourists. Health Educ J1999;58:177–91.

54. Weinstock MA, Rossi JS, Redding CA, Maddock JE, Cottrill S. Sunprotection behaviors and stages of change for the primary prevention of


skin cancers among beachgoers in southeastern New England. Ann BehavMed 2000;22:286–93.

55. Weinstock MA, Rossi JS, Redding CA, Maddock JE. Randomized trial of amulti-component stage-matched intervention to increase sun protectionin at-risk beach bathers. Prev Med 2002;35:584–92.

56. U.S. Census Bureau. Statistical abstract of the United States: 2000.Available at: www.census.gov/prod/2001pubs/statab/sec13.pdf. AccessedMarch 1, 2004.

57. Holman CD, Gibson IM, Stephenson M, Armstrong BK. Ultravioletirradiation of human body sites in relation to occupation and outdooractivity: field studies using personal UVR dosimeters. Clin Exp Dermatol1983;8:269–77.

58. Shoveller JA, Lovato CY, Peters L, Rivers JK. Canadian National Survey onSun Exposure & Protective Behaviours: outdoor workers. Can J PublicHealth 2000;91:34–5.

59. Azizi E, Modan M, Fuchs Z, Kushelevsky AP. Skin cancer risk of Israeliworkers exposed to sunlight. Harefuah 1990;118:508–11.

60. Azizi E, Flint P, Sadetzki S, et al. A graded work site intervention programto improve sun protection and skin cancer awareness in outdoor workersin Israel. Cancer Causes Control 2000;11:513–21.

61. Borland RM, Hocking B, Godking GA, Gibbs AF, Hill DJ. The impact ofa skin cancer educational package for outdoor workers. Med J Aust1991;154:686–8.

62. Dobbinson S, Borland R, Anderson M. Sponsorship and sun protectionpractices in lifesavers. Health Promotion Int 1999;14:167–76.

63. Geller AC, Glanz K, Shigaki D, Isnec MR, Sun T, Maddock J. Impact ofskin cancer prevention on outdoor aquatics staff: the Pool Cool programin Hawaii and Massachusetts. Prev Med 2001;33:155–61.

64. Girgis A, Sanson-Fisher RW, Watson A. A workplace intervention forincreasing outdoor workers’ use of solar protection. Am J Public Health1994;84:77–81.

65. Hanrahan PF, Hersey P, Watson AB, Callaghan TM. The effect of aneducational brochure on knowledge and early detection of melanoma.Aust J Public Health 1995;19:270–4.

66. Novick M. To burn or not to burn: use of computer-enhanced stimuli toencourage application of sunscreens. Cutis 1997;60:105–8.

67. Nelson C, Woodwell D. National Ambulatory Medical Care Survey: 1993summary. Vital Health Stat 1998;13:136.

68. Azurdia RM, Pagliaro JA, Rhodes LE. Sunscreen application technique inphotosensitive patients: a quantitative assessment of the effect of educa-tion. Photodermatol Photoimmunol Photomed 2000;16:53–6.

69. Brandberg Y, Bergenmar M, Michelson H, Mansson-Brahme E, Sjoden P.Six-month follow-up of effects of an information programme for patientswith malignant melanoma. Patient Educ Couns 1996;28:201–8.

70. Brodkin RH, Altman EM. Controlling malignant melanoma: a focus onpediatricians. Am J Dis Child 1993;147:875–81.

71. Dolan NC, Ng JS, Martin GJ, Robinson JK, Rademaker AW. Effectivenessof a skin cancer control educational intervention for internal medicinehousestaff and attending physicians. J Gen Intern Med 1997;12:531–6.

72. Geller AC, Prout MN, Sun T, et al. Cancer skills laboratories for medicalstudents: a promising approach for cancer education. J Cancer Educ2000;15:196–9.

73. Gerbert B, Wolff M, Tschann JM, et al. Activating patients to practice skincancer prevention: response to mailed materials from physicians versusHMOs. Am J Prev Med 1997;13:214–20.

74. Harris JM, Salasche SJ, Harris RB. Can internet-based continuing medicaleducation improve physicians’ skin cancer knowledge and skills? J GenIntern Med 2001;16:50–6.

75. Harris JM Jr., Salasche SJ, Harris RB. Using the Internet to teachmelanoma management guidelines to primary care physicians. J Eval ClinPract 1999;5:199–211.

76. Johnson EY, Lookingbill DP. Sunscreen use and sun exposure. Trends ina white population. Arch Dermatol 1984;120:727–31.

77. Leinweber CE, Campbell HS, Trottier DL. Is a health promotion cam-paign successful in retail pharmacies? Can J Public Health 1995;86:380–3.

78. Liu KE, Barankin B, Howard J, Guenther LC. One-year followup on theimpact of a sun awareness curriculum on medical students’ knowledge,attitudes, and behavior. J Cutan Med Surg 2001;5:193–200.

79. Mayer JA, Slymen DJ, Eckhardt L, et al. Skin cancer prevention counselingby pharmacists: specific outcomes of an intervention trial. Cancer DetectPrev 1998;22:367–75.

80. Mayer JA, Eckhardt L, Stepanski BM, et al. Promoting skin cancerprevention counseling by pharmacists. Am J Public Health
1998;88:1096–9.
ber 5

http://www.tia.org/Travel/tripChar.asp

http://www.tia.org/Travel/StateTourism.asp

http://www.census.gov/prod/2001pubs/statab/sec13.pdf

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

81. McCormick LK, Masse L, Cummings SS, Burke C. Evaluation of skincancer prevention module for nurses: change in knowledge, self-efficacy,and attitudes. Am J Health Promotion 1999;13:282–9.

82. Mikkilineni R, Weinstock MA, Goldstein MG, Dube CE, Rossi JS. Theimpact of the basic skin cancer triage curriculum on provider’s skincancer control practices. J Gen Intern Med 2001;16:302–7.

83. Palmer RC, Mayer JA, Eckhardt L, Sallis JF. Promoting sunscreen in acommunity drugstore. Am J Public Health 1998;88:681.

84. Robinson JK. Behavior modification obtained by sun protection educa-tion coupled with removal of a skin cancer. Arch Dermatol1990;126:477–81.

85. Robinson JK. Compensation strategies in sun protection behaviors by apopulation with nonmelanoma skin cancer. Prev Med 1992;21:754–65.

86. Robinson JK, Rademaker AW. Skin cancer risk and sun protectionlearning by helpers of patients with nonmelanoma skin cancer. Prev Med1995;24:333–41.

87. Montague M, Borland R, Sinclair C. Slip! Slop! Slap! and SunSmart,1980–2000: skin cancer control and 20 years of population-based cam-paigning. Health Educ Behav 2001;28:290–305.

88. Smith BJ, Ferguson C, McKenzie J, Bauman A, Vita P. Impacts fromrepeated mass media campaigns to promote sun protection in Australia.Health Promotion Int 2002;17:51–60.

89. Geller AC, Hufford D, Miller DR, et al. Evaluation of the UltravioletIndex: media reactions and public response. J Am Acad Dermatol1997;37:935–41.

90. Centers for Disease Control and Prevention. Media dissemination of andpublic response to the Ultraviolet Index—United States, 1994–1995.MMWR Morb Mortal Wkly Rpt 1997;46:370–3.

91. Gelb BD, Boutwell WB, Cummings S. Using mass media communicationfor health promotion: results from a cancer center effort. Hosp HealthServ Adm 1994;39:283–93.

92. Kiekbusch S, Hannich HJ, Isacsson A, et al. Impact of a cancer educationmultimedia device on public knowledge, attitudes, and behaviors: acontrolled intervention study in southern Sweden. J Cancer Educ2000;15:232–6.

93. Theobald T, Marks R, Hill D, Dorevitch A. “Goodbye Sunshine”: effects ofa television program about melanoma on beliefs, behavior, and mela-noma thickness. J Am Acad Dermatol 1991;25:717–23.

94. Boutwell WB. Under Cover: a community-based skin cancer preventioninitiative. Cancer Bull 1993;45:279–81.

95. Boutwell WB. The Under Cover Skin Cancer Prevention Project. Acommunity-based program in four Texas cities Cancer 1995;75(suppl2):657–60.

96. Cameron IH, McGuire C. “Are you dying to get a suntan?” The pre- andpost-campaign survey results. Health Educ J 1990;49:166–70.

97. King PH, Murfin GD, Yanagisako KL, et al. Skin cancer/melanomaknowledge and behavior in Hawaii: changes during a community-basedcancer control program. Prog Clin Biol Res 1982;130:135–44.

98. Putnam GL, Yanagisako KL. Skin cancer comic book: evaluation of apublic educational vehicle. J Audiovisual Media Med 1985;8:22–5.

99. McGee R, Williams S. Adolescence and sun protection. N Z Med J1992;105:401–3.

00. Pfahlberg A, Gefeller O, Kolmel KF. Public awareness of malignantmelanoma risk factors in Germany. J Epidemiol Community Health1997;51:698–700.

01. Del Mar CB, Green AC, Battistutta D. Do public media campaignsdesigned to increase skin cancer awareness result in increased skinexcision rates? Aust N Z J Public Health 1997;21:751–4.

02. Buller DB, Callister MA, Reichert T. Skin cancer prevention by parents ofyoung children: health information sources, skin cancer knowledge, andsun-protection practices. Oncol Nurs Forum 1995;22:1559–66.

03. Foltz AT. Parental knowledge and practices of skin cancer prevention: apilot study. J Pediatr Health Care 1993;7:220–5.

04. Vail-Smith K, Watson CL, Felts WM, Parrillo AV, Knight SM, Hughes JL.Childhood sun exposure: parental knowledge, attitudes, and behaviors.J Health Educ 1997;28:149–55.

05. Robinson JK, Rigel DS, Amonette RA. Summertime sun protection usedby adults for their children. J Am Acad Dermatol 2000;42:746–53.

06. Bolognia JL, Berwick M, Fine JA, Simpson P, Jasmin M. Sun protection innewborns. A comparison of educational methods. Am J Dis Child1991;145:1125–9.

07. Buller DB, Burgoon M, Hall JR, et al. Using language intensity to increasethe success of a family intervention to protect children from ultraviolet
radiation: predictions from language expectancy theory. Prev Med2000;30:103–13.
08. Buller DB, Burgoon M, Hall JR, et al. Long-term effects of languageintensity in preventive messages on planned family solar protection.Health Community 2000;12:2000–275.

09. Miller DR, Geller AC, Wood MC, Lew RA, Koh HK. The Falmouth SafeSkin Project: evaluation of a community program to promote sunprotection in youth. Health Educ Behav 1999;26:369–84.

10. Rodrigue JR. Promoting healthier behaviors, attitudes, and beliefs towardsun exposure in parents of young children. J Consult Clin Psychol1996;64:1431–6.

11. Eskelin S, Pyrhonen S, Summanen P, Prause JU, Kivela T. Screening formetastatic malignant melanoma of the uvea revisited. Cancer1999;85:1151–9.

12. Anti-Cancer Council of Victoria. Sunsmart evaluation studies 1. Carlton,Victoria, Australia: Anti-Cancer Council of Victoria, 1994.




16. Anti-Cancer Council of Victoria. SunSmart evaluation studies 5. Carlton,Victoria, Australia: Anti-Cancer Council of Victoria, 1998.

17. Anti-Cancer Council of Victoria. SunSmart evaluation studies 6. Carlton,Victoria, Australia: Anti-Cancer Council of Victoria, 1999.

18. Anti-Cancer Council of Victoria. Sunsmart Evaluation Studies, 2000–2003.Carlton, Victoria, Australia: Anti-Cancer Council of Victoria, 2003.

19. Biger C, Epstein LM, Hagoel L, Tamir A, Robinson E. An evaluation of aneducation programme, for prevention and early diagnosis of malignancyin Israel. Eur J Cancer Prev 1994;3:305–12.

20. Borland R, Hill D, Noy S. Being SunSmart: changes in communityawareness and reported behaviour following a primary prevention pro-gram for skin cancer control. Behav Change 1990;7:126–35.

21. Carmel S, Shani E, Rosenberg L. The role of age and an expanded HealthBelief Model in predicting skin cancer protective behavior. Health EducRes 1994;9:433–47.

22. Carmel S, Shani E, Rosenberg L. Skin cancer protective behaviors amongthe elderly: explaining their response to a health education programusing the Health Belief Model. Educ Gerontol 1996;22:651–8.

23. Chapman S, Marks R, King M. Trends and tans and skin protection inAustralian fashion magazines, 1982 through 1991. Am J Public Health1992;82:1677–82.

24. Fielder H, Lo SV, Shorney S, Roberts DL. Skin, sun and sense: anevaluation of a skin cancer prevention campaign. Health Educ J1996;55:431–8.

25. Geller AC, Sayers L, Koh HK, Miller DR, Steinberg BL, Crosier-Wood M.The New Moms Project: educating mothers about sun protection innewborn nurseries. Pediatr Dermatol 1999;16:198–200.

26. Hill D, White V, Marks R, Theobald T, Borland R, Roy C. Melanomaprevention: behavioral and nonbehavioral factors in sunburn among anAustralian urban population. Prev Med 1992;21:654–69.

27. Hill D, White V, Marks R, Borland R. Changes in sun-related attitudes andbehaviours, and reduced sunburn prevalence in a population at high riskof melanoma. Eur J Cancer Prev 1993;2:447–56.

28. Holtrop JS. Sticking to it: a multifactor cancer risk-reduction program forlow-income clients. J Health Educ 2000;31:122–7.

29. Kelly PP. Skin cancer and melanoma awareness campaign. Oncol NursForum 1991;18:927–31.

30. Marks R. Melanoma prevention: is it possible to change a population’sbehavior in the sun? Pigment Cell Res 1994;7:104–6.

31. New South Wales Cancer Council. Report on the Seymour Snowman SunProtection Campaign (1997–1998). North Sydney, New South Wales,Australia: New South Wales Cancer Council, 1998.

32. Ramsdell WM, Kelly P, Coody D, Dany M. The Texas Skin Cancer/Melanoma Project. Tex Med 1991;87:70–3.

33. Rassaby J, Larcombe I, Hill D, Wake R. Slip Slop Slap: health educationabout skin cancer. Cancer Forum 1983;7:69.

34. Richard MA, Martin S, Gouvernet J, Folchetti G, Bonerandi JJ, Grob JJ.Humour and alarmism in melanoma prevention: a randomized con-trolled study of 3 types of information leaflet. Br J Dermatol1999;140:909–14.

35. Sanson-Fisher R. Me No Fry 1992/93 summer campaign evaluationreport. North Sydney, New South Wales, Australia: NSW Department of
Health, 1993.
Am J Prev Med 2004;27(5) 449

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

A

M

Ittdnfinveiip

ar

●●

●●●

●

●

●

●

dUrst

S

MhrailTT1T

●

●●

●

●●

tteci

*AIGIN

4

36. Sanson-Fisher R. Me No Fry 1993/1994 summer campaign evaluationreport. North Sydney, New South Wales, Australia: NSW Department ofHealth, 1994.

37. Sanson-Fisher R. Me No Fry 1994/1995 summer campaign evaluationreport. North Sydney, New South Wales, Australia: NSW Department ofHealth, 1995.

38. Staples M, Marks R, Giles G. Trends in the incidence of non-melanocyticskin cancer (NMSC) treated in Australia 1985–1995: are primary preven-tion programs starting to have an effect? Int J Cancer 1998;78:144–8.

39. Baum A, Cohen L. Successful behavioral interventions to prevent cancer:the example of skin cancer. Annu Rev Public Health 1998;19:319–33.

40. Morris J, Elwood M. Sun exposure modification programmes and theirevaluation: a review of the literature. Health Promotion Int 1996;11:321–32.

41. Melia J, Pendry L, Eiser JR, Harland C, Moss S. Evaluation of primaryprevention initiatives for skin cancer: a review from a U.K. perspective.Br J Dermatol 2000;143:701–8.

42. Peters L, Paulussen T. School health promotion and cancer prevention: areview of international effectiveness research on skin cancer prevention.NIGZ Neth Inst Health Promotion Dis Prev 1997.

43. Zaza S, Wright-De Aguero LK, Briss PA, et al. Data collection instrumentand procedure for systematic reviews in the Guide to Community Preven-tive Services. Am J Prev Med 2000;18(suppl 1):44–74.

44. National Cancer Institute, Centers for Disease Control and Prevention,American Cancer Society. Cancer Control PLANET (Plan, Link, Act, Net-work with Evidenced-Based Tools). Available at: http://cancercontrolplanet.cancer.gov. Accessed June 8, 2004.

45. Rimer B, Glanz K, Rasband G. Searching for evidence about healtheducation and health behavior interventions. Health Educ Behav2000;28:231–48.

46. Creech LL, Mayer JA. Ultraviolet radiation exposure in children: a reviewof measurement strategies. Ann Behav Med 1998;19:399–407.

47. Eckhardt L, Mayer JA, Creech L, et al. Assessing children’s ultravioletradiation exposure: the potential usefulness of a colorimeter. Am J PublicHealth 1996;86:1802–4.

48. Diffey BL, Gibson CJ, Haylock R, McKinlay AF. Outdoor ultravioletexposure of children and adolescents. Br J Dermatol 1996;134:1030–4.

49. Lovato C, Shoveller J, Mills C. Canadian national workshop on measure-ment of sun-related behaviours. Chronic Dis Can 1999;20:96–100.

50. Glanz K, Lewis FM, Rimer BK. Theory, research, and practice in healthbehavior and health education. In: Glanz K, Lewis FM, Rimer BK, eds.Health behavior and health education: theory, research and practice. 2nded. San Francisco: Jossey-Bass, 1997:22–40.

51. World Health Organization. Evaluating school programmes to promotesun protection. Geneva: World Health Organization, 2002.

52. World Health Organization. Sun protection: a primary teaching resource.Geneva: World Health Organization, 2002.

53. World Health Organization. Sun protection and schools: how to make adifference. Geneva: World Health Organization, 2002.

ppendix A

ethods

n the Community Guide, evidence is summarized on (1) effec-iveness of interventions; (2) applicability of evidence data (i.e.,he extent to which available effectiveness data might apply toiverse population segments and settings); (3) positive oregative effects of the intervention other than those assessed

or the purpose of determining effectiveness, including pos-tive or negative health and nonhealth outcomes; (4) eco-omic impact; and (5) barriers to implementation of inter-entions. When evidence is insufficient to determine theffectiveness of the intervention on a specific outcome,nformation about applicability, economics, or barriers tomplementation is not included, unless there is an issue ofarticular interest.The following process was used to review evidence system-

tically, and translate that evidence into the conclusions
eached in this article involved the following steps:
MU


Forming a systematic review development teamDeveloping a conceptual approach to organizing, group-ing, and selecting interventionsSelecting interventions to evaluateSearching for and retrieving evidenceAssessing the quality of and abstracting information fromeach studyAssessing the quality of and drawing conclusions about thebody of evidence of effectivenessTranslating the evidence of effectiveness intorecommendationsConsidering data on applicability, other effects, economicimpact, and barriers to implementationIdentifying and summarizing research gaps

This section summarizes how these methods were used ineveloping the reviews of interventions to reduce exposure toV radiation. The reviews were produced by the systematic

eview development team and a multidisciplinary team ofpecialists and consultants representing a variety of perspec-ives on cancer prevention.

earch for Evidence

Electronic searches for literature were conducted inEDLINE, PsycINFO, and CINAHL (nursing and allied

ealth). The team also reviewed the references listed in alletrieved articles, and consulted with experts on the system-tic review development team and elsewhere, including seek-ng published and unpublished articles in a sun protectionistserve sponsored by the Environmental Protection Agency.he team included journal articles and governmental reports.he initial literature search on the topic was conducted in999, and the search was updated monthly until June 2000.he MEDLINE search strategy is shown in Table A1.To be included, identified studies had to:

Evaluate a specified population-based intervention for theprevention of skin cancerBe published in English from 1966 to June 2000Involve primary prevention of skin cancer (i.e., studiespromoting screening were excluded because the effective-ness of screening is uncertain according to the USPSTF(see www.ahrq.gov/clinic/uspstf/uspsskca.htm)Evaluate effectiveness and assess at least one of the out-comes specified on the team’s analytic frameworks and/orprovide information on one or more of the followingdomains: applicability, other effects (i.e., harms or sideeffects), economic evaluation, or barriers to interventionimplementationBe conducted in an established market economy*Be a primary study rather than, for example, a guideline orreview

Studies of effectiveness or applicability also required thathe study compare a group of people who had been exposedo the intervention with a group of people who had not beenxposed or who had been less exposed. (The comparisonsould be concurrent or in the same group over time. Studiesn the other domains could be with or without a comparison.)

Established market economies as defined by the World Bank arendorra, Australia, Austria, Belgium, Bermuda, Canada, Channel

slands, Denmark, Faeroe Islands, Finland, France, Germany,ibraltar, Greece, Greenland, Holy See, Iceland, Ireland, Isle of Man,

taly, Japan, Liechtenstein, Luxembourg, Monaco, the Netherlands,ew Zealand, Norway, Portugal, San Marino, Spain, St. Pierre and
iquelon, Sweden, Switzerland, the United Kingdom, and thenited States.
ber 5

http://www.ahrq.gov/clinic/uspstf/uspsskca.htm

A

ri

rt

A

TccSdt

Ao

Csdatp

nGbi

●

●

●

tercvabbs

fiosu

fratzir

psmae

Tr

1

11

1

1

1

11

1122

22

2

2

2

22233

3333

bstraction and Evaluation of Studies

Each study that met the inclusion criteria was read byeviewers who used a standardized abstraction form to record

able A1. Search strategy for Community Guide skin cancereview

1 skin neoplasms/or skin cancer.tw.2 melanoma/3 carcinoma, basal cell/or carcinoma, squamous

cell/4 nevus/or nevi.tw.5 keratosis/6 actinic keratoses.tw.7 (sun damage or photodamage).tw. or skin aging/8 solar keratoses.tw.9 1 or 2 or 3 or 4 or 5 or 6 or 7 or 80 primary prevention/or prevention.mp. [mp �

title, abstract, registry number word, meshsubject heading]

1 pc.fs.2 knowledge/or knowledge, attitudes, practice/or

knowledge.mp. [mp � title, abstract, registrynumber word, mesh subject heading]

3 awareness/or awareness.mp. [mp � title, abstract,registry number word, mesh subject heading]

4 (attitude or attitude or attitudes).mp. [mp � title,abstract, registry number word, mesh subjectheading]

5 public policy/or policy.mp. [mp � title, abstract,registry number word, mesh subject heading]

6 health promotion/or health education/7 behavior/or behavior:.mp. [mp � title, abstract,

registry number word, mesh subject heading]8 10 or 11 or 12 or 13 or 14 or 15 or 16 or 179 9 and 180 19 not screen:.tw,hw.1 19 and screen:.tw,hw. and primary prevention.mp.

[mp � title; abstract, registry number word,mesh subject heading]

2 20 or 213 (sunburn: or suntan: or tanning).mp. [mp � title,

abstract, registry number word, mesh subjectheading]

4 ultraviolet rays/or ultraviolet radiation.mp. [mp �title, abstract, registry number word, meshsubject heading]

5 (sun exposur: or sun protect: or sun safety or solarprotect: or solar exposur:).mp. [mp � title,abstract, registry number word, mesh subjectheading]

6 sunlight/or protective clothing/or protectiveclothing.mp. [mp � title, abstract, registrynumber word, mesh subject heading]

7 sunscreening agents/8 23 or 24 or 25 or 26 or 279 18 and 280 29 not screen:.tw,hw.1 29 and screen:.tw,hw. and primary prevention.mp.

[mp � title, abstract, registry number word,mesh subject heading]

2 30 or 313 22 or 324 limit 33 to English language5 limit 34 to human

nformation from the study.1 Any disagreements between the i

eviewers were reconciled by consensus among the revieweam members.

ssessing the Suitability of Study Design

Design suitability was assessed for every identified study.2

he team’s study design classifications, chosen to ensureonsistency in the review process, sometimes differ from thelassification or nomenclature used in the original studies.tudies with good or fair quality of execution, and any level ofesign suitability, were included in the body of evidence forhe purpose of assessing effectiveness.

ssessing the Quality and Summarizing the Bodyf Evidence of Effectiveness

Quality of study execution was systematically assessed usingommunity Guide methods and its abstraction form.1,2 Sometudies had more than one separate intervention arm, (i.e.,istinct interventions that were compared with each othernd/or a control). Under these circumstances the teamreated distinct “arms” as independent interventions for theurposes of this review.Criteria for assessing the strength of evidence on effective-

ess in the Community Guide have been published elsewhere.2

enerally, by Community Guide standards, the minimal num-er of studies sufficient to draw a conclusion about a given

ntervention outcome is as follows:

One study of greatest design suitability and good executionand sufficient effect size, orThree studies of greatest or moderate design suitability andgood or fair execution with sufficient and consistent effectsize, orFive studies of greatest, moderate or lowest design suitabil-ity and good or fair execution with sufficient and consistenteffect size.

The team abstracted information from the studies abouthe outcomes of interest specific to the intervention undervaluation. Unless otherwise noted, the team represented theesults of each study as a point estimate for the relativehange in the outcome of interest associated with the inter-ention. The team calculated absolute and relative effect sizess shown in Table A2. When the team had to make choicesetween effect measures, it used the last available measureoth before and after the intervention in calculating effectizes.

The team reported the effect of the intervention as bene-cial, when the intervention is associated with a change in anutcome in the desired direction (i.e., an increase in sun-creen use or a decrease in exposure to peak sun) and asndesirable when an effect went in the opposite direction.When the team decided that a summary effect measure was

easible and useful, it reported the median and interquartileange to show effect sizes from multiple studies. The teamlso noted whether zero was included within the upper andhe lower interquartile range. Interquartile ranges includingero suggest that the results are inconsistent in direction;nterquartile ranges not including zero suggest that theesults are generally consistent in direction.

In some cases, the team had to select among severalossible effect measures for inclusion in its summary mea-ures of effectiveness. When available, the team includedeasures adjusted for potential confounders in multivariate

nalysis rather than crude effect measures. No studies werexcluded from the evaluation strictly on the basis of an
nsufficient follow-up period. If the intervention program had
Am J Prev Med 2004;27(5) 451

mc

attttbdpnwcb

eikoo

O

tobtt

E

cMwae

SI

st

S

iioigenqneofogitertcaddodnhem

R

T

AR

C

4

ultiple evaluations at different follow-up points, the teamhose the evaluation at the longest follow-up period.

In evaluating the body of evidence, the team assessed sizend consistency of reported effects and attempted, if possible,o explain any inconsistency. The team also assessed whetherhere were common threats to validity in the body of evidencehat either weakened or strengthened the conclusions. Theeam summarized the strength of the body of evidence on theasis of the number of available studies, the strength of theiresign and execution, and the size and consistency of re-orted effects, as described in detail elsewhere. When theumber of studies and their design and execution qualityere sufficient by Community Guide standards to draw aonclusion on effectiveness, the results were summarizedoth graphically and statistically.It is critical to note that when the team concludes that

vidence is insufficient to determine the effectiveness of thentervention on a given outcome, it means that it is not yetnown what effect, if any, the intervention has on thatutcome. It does not mean that the intervention has no effectn the outcome.

ther Effects

The Community Guide reviews of UV protection interven-ions systematically assessed the effects of the intervention onther outcomes that were identified as other harms orenefits of the intervention in the analytic frameworks. Theeam also notes other harms or benefits if they were men-ioned in the studies reviewed.

conomic Evaluations

Economic evaluations were conducted only there is suffi-ient or strong evidence of effectiveness of the interventions.ethods used in economic evaluations are described else-here.3 The standard abstraction form used for economicbstraction is available at www.thecommunityguide.org/methods/con-abs-form.pdf.

ummarizing Barriers to Implementation ofnterventions

Barriers to implementation are summarized only if there isufficient or strong evidence of effectiveness of the interven-

able A2. Summary effect measures

Before-and-aft

bsolute effect measure Post–preelative effect measure (Post–pre)/pre

, control; I, intervention; RCT, randomized controlled trial.

ion.


ummarizing Research Gaps

Systematic reviews in the Community Guide identify existingnformation on which to base public health decisions aboutmplementing interventions. An important additional benefitf these reviews is identification of areas in which information

s lacking or of poor quality. To summarize these researchaps, remaining research questions for each interventionvaluated were first identified. Where evidence of effective-ess of an intervention was sufficient or strong, remaininguestions about effectiveness, applicability, other effects, eco-omic consequences, and barriers were summarized. Wherevidence of effectiveness of an intervention was insufficient,nly remaining questions about effectiveness and other ef-ects were summarized. Applicability issues were summarizednly if they affected the assessment of effectiveness. Ineneral, the Community Guide has made the argument that its premature to identify research gaps in economic evalua-ions or barriers before effectiveness is demonstrated. Forach category of evidence, issues that emerged from theeview were identified, based on the informed judgment ofhe team. Several factors influenced that judgment. When aonclusion was drawn about evidence, the team decided ifdditional issues remained. Specifically, if effectiveness wasemonstrated by using some but not all outcomes, the teamid not necessarily identify gaps in all of the other possibleutcomes as important evidence gaps. If the available evi-ence was thought to be generalizable, the team did notecessarily list all subpopulations or settings where studiesad not been done as research gaps. Within each body ofvidence, the team considered whether there were generalethods issues that would improve future studies in that area.

eferences for Appendix

1. Zaza S, Wright-De Aguero LK, Briss PA, et al. Datacollection instrument and procedure for systematic re-views in the Guide to Community Preventive Services.Am J Prev Med 2000;18(suppl 1):44–74.

2. Briss PA, Zaza S, Pappaioanou M, et al. Developing anevidence-based Guide to Community Preventive Servic-es—methods. The Task Force on Community PreventiveServices Am J Prev Med 2000;18(suppl 1):35–43.

3. Carande-Kulis VG, Maciosek MV, Briss PA, et al. Meth-ods for systematic reviews of economic evaluations forthe Guide to Community Preventive Services. Task Forceon Community Preventive Services Am J Prev Med

ly design

Study with comparison group(RCT, cohort design,nonrandomized trial)

�I–�C00 �I/Ipre–�C/Cpre

er on

� 1

2000;18(suppl 1):75–91.

ber 5

http://www.thecommunityguide.org/methods/econ-abs-form.pdf

http://www.thecommunityguide.org/methods/econ-abs-form.pdf

Appendix BTable B1. Summary evidence table: interventions in primary schools

Author, year, location,study design, designsuitability, studyquality Follow-up interval; n; limitations

Demographics: age,gender, race/ethnicity, SES Intervention

Results: summary effect measures(behavioral and health outcomes only),p value, within or between groups

Bastuji-Garin (1999)88

FranceBefore-and-after studyLeast suitableGood quality

F/U: 3 months from end of In�203Limitations: Neither the content nor

the intensity of the education weredescribed in detail; no descriptionof the sampling frame for theschools or eligibility criteria; volun-teer samples of schools may haveselected for staff with an interest inthe subject and biased results awayfrom the null; single interviewer notblinded; intervention may havechanged self-reports rather than ac-tual behavior; unrelated time trendsmay have contributed to apparentpositive effects

Mean age: 9.2yFemale: 50.5%Race/ethnicity: NRSES: Upper middle class

Interactive sun-awarenessprogram consisting ofindividual journalsprepared by the children,and a skit and poster, orinteractive game tointegrate material of eachsun awareness topic;weekly packet provided adifferent topic each weekfor 4 weeks

Child health outcomes and sun-protec-tive behaviors (single items)

Children reporting that their skinnever sunburns:

Absolute change: �20.0Relative change: �42.7%

(p�0.001, within)Children reporting that they always

wear a hatAbsolute change: �9.9Relative change: �41.6%


wear a t-shirt:Absolute change: �2.5Relative change: �5.3%

(p�0.01, within)Children reporting that they avoid

sunny hoursAbsolute change: �10.8Relative change: �16.4%


wear sunscreen:Absolute change: �9.4Relative change: �37%

(p�0.03, within)Buller (1994)92

ArizonaRCTGreatest suitabilityGood quality

F/U: Immediately following end of In�139Limitations: Limited demographic in-

formation on study population; nodescription of schools; conveniencesample of schools; self-reported out-come measures; group design butindividual analyses may have re-sulted in overestimated effect

Mean age: NR(4th–6th graders)

Female: NRRace/ethnicity: NRSES: NR

I: Classroom curriculum for4th–6th graders, consistingof skin cancer and UVprotection education, les-son plans, inclass activities,take-home activities, news-letter, and disseminationsuggestions for involvingentire school

C: No intervention

Children wear protective clothing insummer (single item; range 1–3):

Absolute change: 0.37Relative change: �27.6%

(p�0.05, between)Children wear sunscreen in winter

(single item; range 1–3):Absolute change: 0.18Relative change: �13.55%

(p�0.05, between)

(continued on next page)

Am

JPrev

Med

2004;27(5)453

Table B1. Continued




Buller (1999)90


F/U: 2 months from B/Ln�159Limitations: No descriptions of

schools; convenience sample of vol-unteer schools and classrooms; chil-dren smart self-reported outcomemeasures; no comparison of groupsat B/L

Mean age: NR(range: 9–11y)�

Female: 53.7%Race/ethnicity: 57% white;

36% other (mainlyHispanic); 7% NR

SES: NR

I-1: CD-ROM sun safety gameand interactive activities,modified for grades 4 and5, with children earningpoints on each activity

I-2: Sun safety curriculumonly, sun safety game andactivities

I-3: Sun-safety curriculum �CD-ROM

C: No intervention

Child composite sun-protective behav-iors (mean score, 13 items, 3-pointscale):

I-1:Absolute change: �0.07Relative change: �3.5%

I-2:Absolute change: �0.14Relative change: 7.3%

I-3:Absolute change: �14Relative change: 7.2%

Overall p�0.074Buller (1996)89


F/U: 2 months from B/Ln�447Limitations: Low participation rates;

childrens’ reports of own and ofparents’ behavior; group design andindividual analyses may have over-estimated significance

Mean age: range 8–10yFemale: 49%Race/ethnicity: 63.7% white;

10.4% Hispanic; 2.8%African American; 15.6%other; 7.4% NR

SES: NR

I: Multidisciplinary curricu-lum on sun properties,composition of the skin,historic attitudes to tan-ning, and strategies toreduce sun exposure; in-cluded lesson materialsin-class and take-homeactivities, workbook, keyterm glossary, quick re-view, and newsletter

C: No intervention

Children wear protective clothing insummer (mean score, 3-point-scale):


(p�0.43, between)Children lay in sun to tan (mean

score, 3-point scale):Absolute change: �0.11Relative change: �4.4%

(p�0.11, between)Children play early or late when out-

side (mean score, 3-point scale)Absolute change: �0.12Relative change: �6.3%

(p � 0.27, between)Child composite sun-protective behav-

ior (13 items, 3-point scale):Absolute change: �0.02Relative change: �1.0%

(p�0.51, between)Children wear sunscreen in summer

(mean score, 2-item scale, 3-pointscale):


(p � 0.05, between)


454A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e27,

Num

ber5

Table B1. Continued




Girgis (1993)97

AustraliaRCTGreatest suitabilityGood quality

F/U: 8 months from end of In�648Limitations: Secular time

Mean age: 10yFemale: 53%Race/ethnicity: NRSES: father-NR (53%);

32.6% low income;15.3% high income

I-1: Standard: 30-minute di-dactic lecture focusing ondissemination of informa-tion; included posters andsunscreen samples

I-2: Intensive: Skin Safe skin-protection program incor-porated into teachers’curriculum, consisting ofcooperative learning tech-niques, student participa-tion, problem-based strat-egies to promoteawareness of problemsand potential solutionsassociated with solar ex-posure, encouragementof students to developresponsibility for theirown welfare by criticallyexamining and improvingtheir own environment

C: No intervention

Child composite sun-protective behav-iors (odds ratio using interventiongroups as variable):



Overall p value � 0.001

Gooderham (1999)98

CanadaBefore-and-after studyLeast suitableGood quality

F/U: 1 month from B/Ln�216Limitations: Nonvalidated self-reported

outcome measures; before-and-afterstudy design with questionable useof analysis of variance statistical test-ing; potential for test–retest biasaway from the null

Mean age: NRRange: (9–10y)Female: 47%Race/ethnicity: 90% white;

3% American Indianor Alaskan Native; 2%Asian; 2% East Indian;1% African American;2% other

Sun-awareness educationprogram consisting of two1-hour presentations, sun-awareness activity booklet,sun-safety workbook, take-home educationalmaterials, and incentives

Child sun-protective behaviors (3-pointscale):

Always wear a long-sleeved shirt:Absolute change: �3.0Relative change: �100.0%

(p�0.001, within)Always wear long pants:


(p�0.001, within)Always wear hat when outside:


(p�0.001, within)Always wear sunscreen when outdoors

in summer:Absolute change: �13.0Relative change: �31.7%

(p�0.001, within)


Am

JPrev

Med

2004;27(5)455

Table B1. Continued




Hoffman (1999)100

FloridaNonrandomized trialGreatest suitabilityGood quality

F/U: 2 weeks from end of In�181Limitations: School size not described;

convenience sample of schools; novalidation of assessment tool; self-reported outcome measures

Mean age: NR(5th graders)

Female: 52%Race/ethnicity: 81% white;

9% African American; 2%Hispanic; 8% other

SES: based on Hollingshead1975—0% strata 1; 6.5%strata 2; 17% strata 3; 38%strata 4; 37% strata 5

I: Lecture and interactiveintervention given to 5th-grade science classes torelay information onsun’s effect on the skin,evaluate students’ risksinherent in sun exposure,and promote change to-ward sunscreen use; in-cluded 10-minute ACSvideo, proper sunscreenapplication, student-pro-duced videotaped com-mercials emphasizing UVexposure dangers andmethods to reduce,homework assignments,and brochures forparents

C: No intervention

Sunscreen use (visual analog scale,mean score):


Homung (2000)101

North CarolinaRCTGreatest suitabilityGood quality

F/U: 7 months from end of In�192Limitations: Extremely limited descrip-

tion of standard didactic presenta-tion; no description of actual provi-sion of the intervention; grouprather than individual interactionwith CD-ROM may have limited in-dividual participation; no validationof behavioral measures

Mean age: 8.5 yFemale: 44%Race/ethnicity: NRSES: NR

I-1: CD-ROM sun-safety gameand interactive activities,modified for grades 3 and4, supplemented by AADpamphlets and informa-tion sheet

I-2: Standard didactic sunsafety curriculum

Child composite behavior (shade andsunscreen use) (100 point):


(p-value NR, between)I-2:


(p-value NR, between)


456A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e27,

Num

ber5

Table B1. Continued




Milne (2000)107

AustraliaNonrandomized trialGreatest suitabilityGood quality

F/U: 1.5y from B/Ln � 1386Limitations: Reliance on self-reported

outcome measures

Mean age: 6Female: 48.2%Race/ethnicity: 9.7%

Southern Europeanancestry: 90.3% other

SES: parents—59.3% highschool or less; 40.6%tertiary education

I-1: Moderate group received“Kidskin” curricula com-prised of developmentallyappropriate, learner-cen-tered skill and outcomebased materials, class-room and home-basedactivities, and guidelinesfor providing a sun-pro-tective schoolenvironment

I-2: High intervention groupreceived same interven-tion as the Moderategroup and were alsomailed program materialsover the summer holi-days, offered low-cost sunprotective swimwear, andwere actively assisted tointroduce and formalizepolicies to provide a sun-protective schoolenvironment

C: Regional standard West-ern Australian Health Ed-ucation curricula

Child sun-protective behaviors orassessments:

Time spent outdoors during peak UVhours (adjusted mean):



(p � 0.01, between)Covered back entire time:I-1:



(p � 0.001; between)Wore hat entire time:I-1:



(p � 0.6, between)Wore protective swimwear:I-1:



(p � 0.0005, between)Use shade more than half the time:I-1:



Am

JPrev

Med

2004;27(5)457

Table B1. Continued





(p�0.09, between)Composite observed sun exposure to

face:I-1:




forearm:I-1:




back:I-1:



(p�0.002, between)

AAD, American Academy of Dermatology; ACS, American Cancer Society; B/L, baseline; C, comparison; F/U, follow-up; I, intervention; n, sample size; NR, not reported; RCT, randomized controltrial; SES, socioeconomic status; UV, ultraviolet; y, year(s).

458A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e27,

Num

ber5

Appendix CTable C1. Summary evidence table: interventions in recreational/tourism settings

Author, year; location;study design; designsuitability; studyquality Follow-up interval; n; limitations



Dey (1995)139

EnglandRCTGreatest suitabilityFair quality

F/U: 0–21 days from end of In�2385Limitations: Limited description of

intervention (# questions and scor-ing); self-reported outcome mea-sures; no stratification of UV expo-sure by site; no assessment ofconfounding

Mean age: 32.5 yFemale: 52.5%Race/ethnicity: NRSES: NR

I: Leaflet “If you worshipthe sun, don’t sacrificeyour skin” by the HealthEducation Authorityplaced in seat-back pock-ets for airline passengers

C: No leaflet

Percentage of adults with self-reportedincidence of severe sunburn (flightlength assumed to correspond withduration of vacation):Absolute change:All flights: �0.9%

(p�0.38, between)Short flight: �0.8%

(p�0.60, between)Glanz (2000)143,144

HawaiiRCTGreatest suitabilityGood quality

F/U: 3 months from end of In�285Limitations: Self-reported outcome

measures

Mean age: parents: 38y;children: 7 y

Female: 86%Race/ethnicity: 79%

Hawaiian or otherSES: Most parents were

married, had somecollege education, and�$20K annual householdincome

I-1: Education arm: staff train-ing, on-site activities, take-home booklets, behavior-monitoring boards, andincentives

I-2: Education/environmentarm: same as educationarm plus provision of sun-screen and promotion ofsun-safe environments

C: Condensed educationalprogram after secondsurvey

Child sun-protection habits (scorerange 1–4, “rarely or never” to“always”):

Sunscreen use:I-1:


(p�0.05, within)I-2:


(p�0.05), within)No significant difference between

groupsSeek shadeI-1:

Absolute change �0.15I-2:

Absolute change �0.0(NR, between)

Relative changes not applicableWear a hat:I-1:


(NR, within)I-2:


(NR, within)


Am

JPrev

Med

2004;27(5)459

Table C1. Continued




Wear a shirt:I-1:

Absolute change �0.03Relative change: �9.0%

(NR, within)I-2:


(NR, within)Composite sun-protective habits index

(average score; required responsesfor at least three behaviors):


(p�0.01, within)I-2:


(p�0.01, within)No significant differences between

groupsGlanz (2002)146

Massachusetts andHawaii

RCTGreatest suitabilityGood quality

F/U: 8 weeks from B/Ln�842Limitations: Reliance on brief self-re-

port measures

Mean age: parents: NR;children: 6.5y

Female: 47%Race/ethnicity: Hawaii: 27%

white; Massachusetts: 90%white

SES: Hawaii: 55% withhousehold income �$50K

Massachusetts: 85% withhousehold income �$50K

I: Sun-protection: staff-train-ing; parent and childrensun-safety lessons, interac-tive activities, providingsunscreen, shade, and sig-nage, and sun-safe envi-ronment promotion

C: Injury prevention: lessonsand activities on bicycleand rollerblading safety,fire safety, traffic and walk-ing safety, poisoning andchoking prevention, andplayground safety

The following were measured frompossible scores ranging from 1 to 4,“rarely or never” to “always”:

Child sun-protection habitsSunscreen use:

Absolute change: �0.14Relative change: �4.53

(p�0.05, between)Wear a hat:


(p�0.28, between)Seek shade:


(p�0.01, between)Composite sun-protection habits index

(required responses for at leastthree behaviors):


460A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e27,

Num

ber5

Table C1. Continued





(p�0.03, between)Children’s incidence of sunburn

(moderate/high-risk children):Absolute change: �0.22Relative change: �41.16

(p�0.04, between)Adult sun-protection habitsUse sunscreen:


(p�0.01, between)Wear hat:


(p�0.01, between)Seek shade:


(p�0.24, between)Composite sun-protection habits index

(required responses for at leastthree behaviors):Absolute change: �0.11Relative change: �4.38

(p�0.95, between)


Am

JPrev

Med

2004;27(5)461

Table C1. Continued




Glanz (1998)141

HawaiiBefore-and-after

studyLeast suitableFair quality

F/U: 4 weeks from B/Ln�154Limitations: Limited demographic in-

formation on parents; no descrip-tion of sample selection; self-re-ported outcome measures

Mean age: parents: NR;children: 7 y; staff: 20 y

Female: 66.7%Race/ethnicity: Parents—

white or Asian/PacificIslanders % NR; Staff—4% white; 42% Hawaiian;27% Asian/Pacific Islander;27% mixed

SES: parents—well-educated;middle or upper income

Staff: 56% attended orgraduated college; 81%never married

Staff training, groupactivities, take-homebooklets, incentives forchildren and staff,providing sunscreen, andpromotion of sun-safeenvironments and policies;children receivedinteractive booklets, withstories, games, and puzzlesto be completed withparents; parents receivedadditional educationalbrochures and surveys tocomplete; delivered byrecreation leaders for c.u.4 weeks

Adult (parents) sun-protection behav-iors (stratified from compositescores; range of scores: 5–20)

Sunscreen use:Absolute change: �0.03Relative change: �4.1%

(NR, within)Seek shade:


(NR, within)Composite sun protection habits index:


(p�0.05, within)Child sun-protection behaviors (strati-

fied from composite scores; range ofscores: 5–20)

Sunscreen use:Absolute change: �0.07Relative change: �9.8%

(NR, within)Seek shade:


(NR, within)Composite sun-protection habits index:


(p�0.01, within)Lombard (1991)148

VirginiaBefore-and-after

studyLeast suitableFair quality

F/U: Average 1 month from B/Ln�NR (total 600 members)Limitations: Year study performed not

reported; no description of age,race/ethnicity, or gender of studypopulation; convenience sample oftwo swimming pools; no reportingof screening criteria; no statisticaltesting; no assessment ofconfounding

Mean age: NRFemale: NRRace/ethnicity: NRSES: primarily middle to

upper class

Peer leader modeling bylifeguards, informationalposters and fliers, postedfeedback, posted goals,free sunscreen andcommitment raffle; at twoswimming pools withapproximately 300members each; for averageof 25 days

Percentage of adults with the follow-ing sun-protection behaviors:

Seek shade:Absolute change:

Pool A: �20.1Pool B: �3.8

Relative change:Pool A: �187.9% (NR, within)Pool B: �62.3% (NR, within)

Wear a hat:


462A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e27,

Num

ber5

Table C1. Continued




Absolute change:Pool A: �9.5Pool B: �1.3


Wear a shirt:Absolute change:



Percentage of children with the follow-ing sun-protection behaviors:

Seek shade:Absolute change:

Pool A: �35.3Pool B: �25.6


Wear a hat:Absolute change:



Wear a shirt:Absolute change:

Pool A: �10.6Pool B: �13.7



Am

JPrev

Med

2004;27(5)463

Table C1. Continued




Mayer (2001)150

San Diego CANonrandomized trialGreatest suitabilityFair quality

F/U: Immediately following end of In�17,245Limitations: No description of study

population or setting; no reportingof percentage of daily visitors in-cluded in sampling frame; exposureto intervention based on self-reports;observed hat use at exit may not re-flect hat use during visit; and non-equivalent comparison group designmay have compromised internal va-lidity (intervention site was closer tothe coast than comparison site andhad somewhat cooler temperatures);no control for potential for partici-pants to visit both sites and thuscontaminate groups

Mean age: NRFemale: NRRace/ethnicity: NRSES: NR

I: Parent visitors of San Di-ego Zoo provided with tipsheets, stamp activity sheetfor children, discount cou-pons for hats and sun-screen in zoo gift shops,point-of-purchase prompts,sun-safety signs in re-strooms; at exhibits, and atstroller rental, and aerialtram loading areas, andthematically relevant chil-dren’s arts and craftsactivities

C: Evaluation only

Observed hat use at site exit:I vs C � site:

Winter: OR�1.84(p�0.01, between)

Summer: OR�0.90(p�0.46, between)

Mayer (1997)149

San Diego CARCTGreatest suitabilityFair quality

F/U: Average 2.5 weeks from B/Ln�169Limitations: All measures except color-

imeter were self-reports by parents;no comparison of respondents andnonrespondents

Mean age: 7.6 yFemale: 49.7%Race/ethnicity: 79.8% whiteSES: (annual income)�$30K 15%$30–49K 18%$50–69K 26%$70–90K 22%�$90K 20%reported by parents

I: UV reduction curriculumpresented at poolside byYMCA aquatics instructorsand home-based activitiesfor children and theirparents; tanness-associated skin colordimensions assessed witha colorimeter; generaland specific daily solar-protection behaviors ofchildren

Use sunscreen: (all values adjusted forage and gender)

Child colorimeter values (higher val-ues indicate more tan):Absolute change: �0.4Relative change: �2.8%

(p�0.084, between)Childrens’ sun-protection habits (5-

point scale ranging from “never” to“always”):Absolute change: �0.03Relative change: �0.82%



(p�0.049, between)Composite sun-protection habits (sum

of scores; possible scores, 0 to 16;higher score indicates more protec-tion):Absolute change: �0.16Relative change: �1.11%

(p�0.15, between)


464A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e27,

Num

ber5

Table C1. Continued




Segan (1999)153

AustraliaRCTGreatest suitabilityFair quality

F/U: Immediately after end of In�373Limitations: Intervention exposure

based on self-reports; no analysis ofrespondents vs nonrespondents

Mean age: 33yFemale: 64%Race/ethnicity: NRSES: NR

I: Full-color, culturally rele-vant, six-page fold-out edu-cational brochure “TheSunSmart Holiday Guide:How to enjoy your holidayin the sun without gettingburnt,” pre-holiday ques-tionnaire that assessedlength and destination ofholiday, reasons for holi-day, skin type, demograph-ics, sun tanning aspira-tions, dichotomousmeasures of whether a hatand sunscreen werepacked, and sun protec-tion intentions, and a post-holiday questionnaire thatassessed frequency andlocation, and extent andseverity of sunburn, rea-sons for sunburn, compos-ite sunburn and sun-pro-tection measures, measureof suntan acquired, andnumber of days outsideand frequency of sun pro-tection behaviors between10 A.M. and 2 P.M.

C: Pre- and post-holiday ques-tionnaires only

Adult sun-protection behavior scores(5-point scale; 1 � never, 5 �always):

Use sunscreen:Absolute change: �0.03

(p�0.72, between)Relative change: NR

Seek shade:Absolute change: �0.09


Wear clothes covering most of body:Absolute change: �0.13


Composite sun-protection behaviors:Absolute change: �0.04


Adult incidence of self-reportedsunburn

63% reported sunburn duringintervention

37% reported severe sunburn


Am

JPrev

Med

2004;27(5)465

Table C1. Continued




Weinstock (2002)155

Rhode IslandRCTGreatest suitabilityFair quality

F/U: 24 months from B/Ln�1449Limitations: self-reported outcomes

measured by questionnaires thatwere not validated; no analysis ofrespondents

Mean age: 33yFemale: 61%Race/ethnicity:94% whiteSES: 88% with high school

education

I: Initial assessment usingnine-item sun behavior–protection index and stageof change questionnaire;educational pamphlet, sunsensitivity assessment andfeedback, SPF-15 sun-screen, instant sun damageimaging photographs, fol-lowed up by three- to four-page feedback reports,mailed 2, 12, and 24months from B/L,matched to individualstage of change (reportsincluded suggestions onreducing unprotected UVexposure)

C: Initial assessment usingsun-protection index andstage of change

Adult sun-protection behavior scores(5-point scale; 1 � never, 5 �always):

Use sunscreen:Absolute change: �0.17Relative change: �5.6%



(p�0.016, between)Avoid the sun:


(p�0.002, between)Composite sun-protection behaviors:


(p�0.001, between)

ACS, American Cancer Society; B/L, baseline; C, comparison; F/U, follow-up; I, intervention; K, thousand; n, sample size; NR, not reported; RCT, randomized control trial; SES, socioeconomicstatus; SPF, sun-protection factor; UV, ultraviolet; y, year(s).

466A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e27,

Num

ber5

interventions to prevent skin cancer by reducing exposure ......interventions to prevent skin cancer...

Documents