randomised trial of asthma self-management programme for ... · randomised trial ofan...

Thorax 1995;50:731-738

Randomised trial of an asthma self-managementprogramme for adults

Rae M Allen, Michael P Jones, Brian Oldenburg

731

AbstractBackground - A hospital based, com-munity service asthma education pro-gramme for adults to improve asthmaknowledge, promote compliance withmedication, and reduce morbidity wasevaluated.Methods - The programme was evaluatedusing a randomised experimental andcontrol group design with repeated meas-urements over 12 months. A volunteercommunity sample of 192 respondents wasrecruited of whom 116 satisfied the in-clusion criteria. At the 12 month follow upsome data were obtained for all subjects.Intervention subjects attended four 2 5hour education sessions at weekly inter-vals. An asthma knowledge question-naire was administered and compliancewas assessed from diary records. Mor-bidity was assessed retrospectively byquestionnaire, prospectively by diary, andobjectively by spirometry and serial peakexpiratory flow rate monitoring. Theadequacy of medical treatment was alsoassessed. Data were collected at baseline,immediately after the intervention, andat three, six, nine, and 12 months afterintervention.Results - Improvements occurred inknowledge and compliance in the inter-vention group but the impact on mor-bidity was modest; this was due, at leastin part, to the inadequacy ofmedical treat-ment.Conclusions - Treatment ofasthma shouldbe reviewed and optimised in conjunctionwith self-management programmes inorder to improve health outcomes.(Thorax 1995;50:731-738)

Keywords: asthma, patient education, evaluationstudies.

Good control of asthma symptoms is a realisticgoal for most asthmatic patients. Good asthmacontrol requires both good medical treatmentand good self-management. If either of theseconditions is not met, reduction is unlikelyin any of the indices typically used to assessmorbidity. Until recently asthma treatment hasemphasised symptomatic relief rather than pro-phylaxis. As a result of this approach, asthmaticindividuals have needed considerable know-ledge and skills to assess their symptomsaccurately and to make the appropriatetherapeutic and behavioural responses. In ad-dition, medical regimens are often complex andrequire skilled use of devices such as metered

dose inhalers, nebulisers, and peak flow meters.Given the complexity of the task, it is notsurprising that inadequacies in self-man-agement have been implicated as contributingto asthma mortality' and morbidity.2

In recent years a number of group educationprogrammes for adults with asthma have beendeveloped3`10 that focus on improving patientself-management. Randomised controlled trialsof the effectiveness of such interventions8-10have shown a significant but modest impact onmorbidity. Acquisition of knowledge, com-pliance with medication, and the adequacy ofmedical treatment are factors that may in-fluence the health outcome of these strategies.However, data for these factors are not widelyreported. Examination of these factors mayprovide clues as to how to optimise the potentialof adult self-management programmes.We report the randomised controlled evalu-

ation of a community service self-managementeducation programme for adults with asthma.It was hypothesised that the programme wouldbe effective in improving asthma knowledgeand compliance with treatment and, wheretreatment was adequate, improved compliancewould result in reduced morbidity. To identifywhen any treatment effects occurred and howlong any such effect lasted, assessments weremade before and immediately after the inter-vention, and three, six, nine, and 12 monthslater.

MethodsADULT ASTHMA EDUCATION PROGRAMMEThe programme was developed and is con-ducted by the Department ofHealth Promotionand Patient Education and Department ofThoracic Medicine at Royal North Shore Hos-pital, Sydney, Australia, a 750 bed universityteaching hospital. The programme was initiatedin 1983 and an evaluation of the impact of theprogramme on asthma management knowledgeand skills was reported in 1989.5A detailed account of the content and theo-

retical background of the programme can befound in that report. In brief, groups of 10-12adult asthmatic patients attended four 2 5 houreducational sessions held at weekly intervals.A family member or friend was encouraged to-accompany each participant in the programme.The programme was led by two asthma edu-cators and a respiratory physician attended onesession. Small group discussion was used toassess the understanding, priorities, and needsof the participants. Emphasis was placed ondeveloping a supportive setting in which asthmamanagement skills and behaviour could be

Departmentof Health Promotionand PatientEducation,Royal North ShoreHospital,Australia 2065R M Allen

Departmentof BehaviouralSciences,Macquarie University,Australia 2113M P Jones

School ofPublic Health,Queensland Universityof Technology,Australia 4059B Oldenburg

Reprint requests to:Mr R M Allen.Received 7 June 1994Retumed to authors21 September 1994Revised version received3 January 1995Accepted for publication9 March 1995

on Novem

ber 6, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.50.7.731 on 1 July 1995. Dow

nloaded from

http://thorax.bmj.com/

Allen, Jones, Oldenburg

practised and reinforced. Information was pre-sented by means of lectures, audiovisual, andwritten materials.5 Each participant received aprogramme manual."The programme content covered areas of:

asthma physiology and aetiology; identificationand control of triggers; recognition of symp-toms and assessment of symptom severity;medications; lifestyle factors; psychosocial fac-tors; inhaler and peak flow meter skills; andmonitoring and recording of peak flow rate(PFR).Although hospital-based, the programme is

offered as a community service for which nomedical referral is required. The programme isintended to supplement that of the medicalcare provider and programme participants wereadvised to return to their usual asthma careprovider.

STUDY DESIGNA randomised controlled trial design was usedand the knowledge, skills, compliance withmedication, and morbidity of participants weremeasured before intervention, immediatelyafter intervention, and 12 months later. In orderto monitor implementation, compliance andmorbidity were also measured three, six, andnine months after intervention.

RECRUITMENT AND ELIGIBILITYA volunteer community sample of 192 asth-matic adults responded to newspaper ad-vertisements in February and May 1989 callingfor adult asthmatics with moderate to severeasthma to volunteer for asthma research. Therewere no significant differences in response orrecruitment rates or group assignment betweenthe recruitment periods. All the respondentswere contacted by telephone. The purpose ofthe study and obligations were explained andeligibility for inclusion in the study was assessedusing a standardised interview protocol.

Respondents were included in the study ifthey were aged 18-65 years, their diagnosis ofasthma had been made by a doctor, and theyhad experienced asthma symptoms rated asmoderate to severe according to publishedconsensus guidelines.'2 Respondents wereexcluded from the study if they had anypotentially confounding conditions or were cur-rent tobacco smokers, had quit in the threemonths prior to the study, or had previouslyattended any asthma education programme.Adequate English literacy skills were also re-quired. Of the 192 respondents, 116 (60%)were eligible to participate in the study. Themain reasons for exclusion were age (50%) andprevious asthma education (25%).

STRATIFICATION AND RANDOM ASSIGNMENTAs the sample was drawn from the community,some variation in the standard of medical carewas anticipated. Before group assignment thesample was stratified according to peak flowmeter ownership as this was considered likelyto indicate better asthma management. The

eligible respondents were then randomly as-signed - 58 to the intervention group and 58to the control group.

PROTOCOLThe study protocol was approved by the RoyalNorth Shore Hospital medical research andethics committee and all participants gave theirinformed consent to take part in the study.Control group participants were offered placesin the education programme at the end of thestudy period.

Baseline assessmentAfter recruitment and random assignment toa group all participants attended a baselineassessment session. Separate sessions were heldfor intervention group and control group mem-bers. Participants completed questionnaires toobtain more detailed demographic data and toassess asthma knowledge, beliefs, attitudes, andquality of life. Inhaler and peak flow meter skillswere also assessed. Spirometric measurementswere performed to obtain objective measuresof lung function. The percentage predictedpeak flow rate was determined for each subject.Those members of the intervention group

who did not have a peak flow meter were givenone and trained in its use. To assess morbidityand compliance, members of both groups keptdaily records of symptoms and medicationusage for two weeks. All intervention groupmembers and those 26 control group memberswho had peak flow meters also recorded morn-ing and evening PFR before and after bron-chodilator use.

Intervention periodFollowing the two week baseline assessmentphase the intervention group participants at-tended the four weekly education sessions oftheasthma education programme. Opportunitieswere available for them to make up any sessionmissed. Following the baseline assessmentphase the control group members kept dailydiary records for four weeks, corresponding tothe period of the education programme for theintervention group. The four week diary recordwas returned by mail.

Post intervention assessmentImmediately after the intervention question-naires were readministered and a furtherfour weeks ofdiary records were kept. To assessongoing compliance and morbidity asthma di-aries were also kept by members ofboth groupsfor four consecutive weeks at three, six, nine,and 12 months following the intervention. Allstudy participants received a reminder letterbefore each diary keeping period. Diaries werereturned by post and a photocopy sent to theparticipant for their own records.To assess maintenance of changes in know-

ledge, attitudes, and subjectively reportedmorbidity 12 months after the interventionthe appropriate questionnaires were readmin-istered, participants were reinterviewed,

732

on Novem



/T


nloaded from


Randomised trial of an asthma self-management programme for adults

inhaler and peak flow meter skills were re-assessed, and spirometric measurements wererepeated.

MEASURESAsthma general knowledge questionnaireA 31 item questionnaire was developed to as-sess knowledge of asthma concepts covered inthe education programme. Knowledge ques-tions were marked as either "true", "false", or"not sure". The knowledge score was the totalof correct answers. Content and face validitywere assessed to be adequate by the respiratoryphysicians and asthma educators involved inthe development of the programme. To assessthe discriminant validity of the questionnaireit was completed by 35 adults who did nothave asthma and had never been involved inasthma management. The difference betweenthe mean (SD) total score for the "no asthma"group (17-6 (4 5)) and the mean baseline scorefor study participants (25-1 (3-1)) was stat-istically significant (p<00001).As an index of reliability the internal con-

sistency of the items was estimated using theKuder-Richardson 20 coefficient (KR-20) fordichotomous data as 0 56 for the questionnairecompleted at baseline and as 075 for the ques-tionnaire completed at the 12 month followup. A separate sample of 20 subjects completedthe questionnaire on two occasions two weeksapart. The reproducibility of scores on thequestionnaire was examined using the paired ttest. The mean score of 24-4 (2 4) for the firstcompletion and 27-9 (2-3) for the second werenot significantly different.

Compliance with medicationParticipants recorded on the asthma diary rec-ord form the name and dosage of prescribedasthma medications. They also recorded anyother medications they took for asthma. Along-side each recorded medication they kept a dailyrecord ofhow much medication was taken andthe reasons for taking more or less than theprescribed dose. To be classified as "com-pliant" participants had to have taken at least90% of each of their prescribed medications,using the formula of prescribed - taken/pre-scribed x 100. The proportion of subjectscategorised as compliant was calculated foreach group.

Strategies to enhance reliability and validityof asthma diary records suggested by Creer andWinder'3 were employed in diary design anddata collection. Data collected were kept to aminimum and were limited to periods of onemonth during key phases of the study. Duringthe explanation of the diary keeping procedureit was acknowledged that there were manyreasons why asthmatic patients do not take alldoses of all medications prescribed for theirasthma. Study participants were asked to re-cord the reasons why they took more or lessthan the prescribed dose of any medicationto allow difficulties encountered in complyingwith medication regimens to be identified andstrategies developed to overcome them.

Morbidity(1) Assessment by interview: information wascollected concerning the number of scheduledand unscheduled doctor and hospital visitsmade and the number of days when asthmahad disrupted daily activities in the previous12 months. "Disruption" was defined as "beingconfined to bed or a chair because of asthmasymptoms". Ratings were made of the fre-quency of morning wheeze, nocturnal asthmasymptoms, and poor response to broncho-dilator medications as follows: 0, never; 1,less than monthly; 2, 1-3 times a month; 3,1-3 times a week; 4, 4-7 times a week. Thereference periods were the 12 months beforethe study and the 12 month study period.

(2) Spirometric measurements: two Vitalo-graph compact spirometers (Vitalograph,Buckingham, UK) were used and recalibratedbefore use. The two asthma educators whoperformed the spirometric measurements weretrained by an experienced respiratory physicianwho assessed the values obtained. The best oftwo measures was recorded for each par-ticipant.

Participants were asked not to use bron-chodilator medications in the four hours beforespirometry unless necessary to control symp-toms. The FEV1/FVC before bronchodilatormedication was measured for each participant.Values obtained were categorised as: > 80%,normal; 50-79%, mild to moderate airwaysnarrowing; <50%, severe airways narrowing;and the proportion of subjects in each categorywas calculated for the two groups.

(3) Assessment ofmorbidity by diary record:all participants recorded the presence or ab-sence ofasthma symptoms during each 24 hourperiod of diary keeping and the frequency (asa percentage) of symptom occurrence was cal-culated. Daily symptom severity was rated ac-cording to defined rating categories: 0, nosymptoms; 1, symptoms controlled by med-ication taken more than three hourly; 2,symptoms controlled by an extra dose ofbronchodilator medication; 3, bronchodilatormedication required 2-3 hourly; 4, bron-chodilator medication required two hourly orless. The frequency ofnocturnal symptoms wasalso calculated, rated as follows: 0, not woken;1, woken once; 2, woken twice; 3, woken morethan twice.

Daily records were kept of morning andevening PFR before and after administrationof bronchodilator using mini-Wright peak flowmeters (Boots, Sydney, Australia). The per-centage of predicted PFR was calculated forthe recorded values. The proportion of eachgroup who recorded a PFR of <70% was cal-culated. Variability in PFR was also calculatedusing the formula: highest - lowest/highest x100, using the highest and lowest PFR valuesfor each collection period. The median vari-ability in PFR was calculated for both groups.

Adequacy of asthma control and adequacy ofmedical therapyTo assess the adequacy of asthma control andmedical therapy we used the assessment and

733

on Novem



/T


nloaded from


734 Allen, Jones, OldenburgTable 1 Cross sectional and longitudinal effects of education on knowledge scores

Intervention group Control group Group difference compared Cross sectional groupwith baseline differencesn Median Interquartile range n Median Interquartile range p value p value

Baseline 56 26-0 23-27 57 25-0 23-27 - NSImmediately after intervention 55 29-0 28-30 55 26-0 24-28 <0 0001 <0-000112 months after intervention 55 29-0 27-30 57 27-0 25-28 <0-0001 0-0002

Table 2 Mean (SD) cross sectional and longitudinal effects of education on complianceIntervention group Control group

Group difference 12 months - baseline Cross sectional group differencesn % compliant n % compliant Odds ratio (95% CI) Odds ratio (95% CI)

Baseline 55 709 (39) 52 59-6 (31) - 1-73 (0-78 to 3 84) (NS)Immediately after intervention 55 83-6 (46) 49 61-2 (30) 1-95 (0-85 to 4 44) (NS) 3-41 (1-37 to 8 47) (p<0-01)3 months after intervention 46 78-2 (36) 45 64-4 (29) 1-41 (0-67 to 2.97) (NS) 2-11 (0-84 to 5 3) (NS)6 months after intervention 42 83-3 (35) 35 62-8 (22) 2-05 (0-79 to 5 28) (NS) 2-95 (1-02 to 8 55) (p<0 05)9 months after intervention 43 88-3 (38) 37 56-7 (21) 4 03 (1-27 to 12-78) (p<0 02) 6-15 (1-99 to 19-06) (p<0 002)12 months after intervention 47 85-1 (40) 45 53-3 (24) 3-4 (1-22 to 9-46) (p<0 02) 5-24 (1-95 to 14-09) (p=0-001)

therapy criteria suggested in the report of aninternational panel of asthma specialists.'4 Foreach participant in the study asthma severitywas rated according to the report criteria. Theseverity ratings are based on FEV1, and thesymptoms and inhaled 12 agonist use that wasrecorded in the asthma diaries. Prescribed med-ical treatment was then compared with thelevel ofmedical treatment advised for the ratedasthma severity. Medication was rated as "ad-equate" if the prescribed medication was atleast that recommended for the level of symp-toms recorded. The occurrence of symptomswould indicate non-compliance with adequatetreatment. Medication was rated as "in-adequate" if the recorded treatment was lessthan that recommended for the severity ofsymptoms recorded. The use of these severityand treatment criteria allowed a rating algo-rithm to be objectively applied.

STATISTICAL ANALYSISTo test whether participation in the programmeresulted in significant sustained improvementsin compliance and morbidity required repeatedmeasurements on each subject to be analysed.The generalised estimating equations method"5using a binomial link function was used in theanalysis of dichotomous variables. A subjecteffect was included in the ordinal regressionmodel for ordered categorical outcomes. Ineach case the model included main effects oftreatment and time, as well as their interaction.The latter provided a test of the homogeneityof treatment effect during the follow up periodas well as a comparison of the study groupswith respect to longitudinal changes in a givenmeasure. When appropriate, post hoc analyseswere performed to compare the group differ-ence at the relevant time points with that atbaseline - that is, to compare the treatmenteffect (for example, odds ratio at follow up)with that at baseline.

Similarly, differences between the groups ateach time point were examined to identifywhere significant changes had occurred. Forthe cross sectional comparisons of groups di-chotomous outcome measures were analysedby unconditional logistic regression andordered categorical outcome measures by or-

dinal regression. None ofthe quantitative meas-ures was normally distributed and none couldbe normalised. Consequently, for quantitativeoutcome measures the Wilcoxon rank sum testwas used to test for differences between groupsand medians and interquartile ranges are re-ported. Proportions are expressed as per-centages and the odds ratios and 95%confidence intervals (CI) are reported.Because of the number of analyses being

performed the statistical significance level wasdefined as p<001. Analyses were performedusing the programme SPIDA.'6

ResultsPARTICIPATION AND RESPONSE RATESThree subjects (two in the intervention groupand one in the control group) were withdrawnfrom the study before the immediate post-intervention assessment. One member of theintervention group developed a confoundingmedical condition and the other moved over-seas before completing the education pro-gramme. The control group member withdrewbecause of family health reasons. All remaining113 subjects were followed up. Re-interviewand return rates for questionnaires were closeto 100% (table 1). Return rates for diaries weremore variable (table 2). Furthermore, somediary data were not included in the analysesbecause they were incomplete or, in the caseofmorbidity data, oral steroid medications werebeing taken for acute asthma or some othercondition. The sample size for each calculationis shown.

DEMOGRAPHIC DETAILS OF PARTICIPANTSThere was an approximately equal proportionofmen (46%) and women (54%) in the sample.The average age was 40 years (range 19-63).Most (93%) were Australian bom or had Eng-lish-speaking backgrounds; 28% had nevermarried and 65% were married or in de factorelationships. Nearly half (49%) of the subjectshad 12 or more years of education, only 4%had six years or less. There were no significantdifferences between the groups on any of thedemographic variables measured.

on Novem



/T


nloaded from



A: Intervention40 A: Nocturnal asthma C Control

30

0

0

10

0Never Less Monthly 1-3 4-7 Never Less Monthly 1-3 4-7

than times times than times timesmonthly weekly weekly monthly weekly weekly

Cl)40a)C.)

a)

-0

a)

U)

0-

I-Ro

-i

Bronchodilator medication needed 2-3 hourly

Never Less Monthly 1-3 4-7 Never Less Monthly 1-3 4-7than times times than times times

monthly weekly weekly monthly weekly weekly

Baseline 12 months afterintervention

Perceived morbidity at baseline and 12 months after intervention.(A) Nocturnal asthma: odds ratio (95% confidence interval) for group difference betweenbaseline values and 12 months after intervention 0 40 (0-85 to 1 11) (p= 0-08); crosssectional group difference 12 months after intervention 1-93 (0-99 to 3178) (p= 0-05).(B) Morning wheeze: group difference between baseline values and 12 months afterintervention 0-08 (0-03 to 0-24) (p<0-0001); cross sectional group difference 12 monthsafter intervention 4-02 (1-98 to 8-18) (p= 0-0001).(C) Bronchodilator medication needed 2-3 hourly: group difference between baselinevalues and 12 months after intervention 0-25 (0-09 to 0-71) (p<0-01); cross sectionalgroup difference 12 months after intervention 2-38 (1-15 to 4-94) (p<0-02).

KNOWLEDGEFrom table 1 it can be seen that general asthmaknowledge was good for both groups at entryto the study. Compared with the group differ-ence at baseline, the immediate post-inter-vention difference was statistically significantwith those in the intervention group sig-nificantly improving their asthma generalknowledge. This improvement was maintainedfor the members of the intervention group 12months after the intervention. Patients in thecontrol group became more knowledgeableover time but the change from baseline wassignificantly greater for patients in the inter-vention group at both assessment points.

735

COMPLIANCE WITH MEDICATIONAt baseline 65% of subjects were compliantwith prescribed medical treatment (table 2).The group x time interaction was not sig-nificant. However, the comparisons of oddsratios during follow up with those at baseline,although of marginal statistical significance,suggest that the significant differences betweenthe groups immediately after intervention andnine and 12 months later are not simply re-flections ofpre-existing baseline differences be-tween the groups. The odds ratios comparingthe two study groups at each assessment indi-cate that the odds of compliance in the inter-vention group are always greater than forthe control group, varying from twofold tosixfold. This is particularly pronounced to-wards the end of the follow up period whenthe odds ratios were 5-6 fold.

MORBIDITYThere were no significant differences betweenthe groups at baseline for any measure of mor-bidity, nor were there any significant baselinedifferences within either group for any mor-bidity measure for those groups members whohad a peak flow meter at entry to the studycompared with those who did not.

In the 12 months before the study littledisruption to daily activities or need to seekacute asthma care was reported (table 3). Smallreductions occurred in these measures over thestudy period for both groups and the groupdifferences were not significant. Participants inthe programme were encouraged to seek reg-ular review of asthma, even when asympto-matic. For both groups the median of fourconsultations for asthma review (interquartilerange 2-7) 12 months after the intervention wasnot significantly different from that at baseline.

Table 3 Morbidity in the 12 months before the study andadequacy of medical treatment at baseline

Intervention Controlgroup group

(a) Retrospective interviewNo. of days activities disrupted by asthma:Median 2 1Interquartile range 0-9 0-6No. of subjects 56 57

Unscheduled use of medical services for acute care:Median 2 1Interquartile range 1-3 0-2No. of subjects 56 57

(b) Prospective diary assessmentPercentage of days asthma symptoms recorded:Median 71-0Interquartile range 29-100No. of subjects 55

Percentage of nights asthma symptoms recorded:Median 14-0Interquartile range 0-57No. of subjects 55

Percentage reporting PFR <70% predicted:70 0

No. of subjects 55Peak flow lability:MedianInterquartile rangeNo. of subjects

(c) FEV,/FVCNormal (>80%)Mild/moderate (79-50%)Severe (<50%)No. of subjects

(d) Adequacy of medical treatmentPercent adequateNo. of subjects

50*014-8653

11-00-43

53

65-026

58-4 48-3350-770 370-75054 26

16%77%7%

56

12%56

10%78%12%57

11%53

on Novem



/T


nloaded from


Allen, Jones, Oldenburg

Although asthma does not appear to havecaused much disruption or need for acute care,considerable asthma morbidity was recordedat baseline (table 3, figure). Additionally, inthe baseline diaries 19% of subjects recordedthat they had required bronchodilator med-ication 2-3 hourly to control asthma symptomsand 29% reported waking more than oncenightly with asthma symptoms. Comparison ofthe odds ratio at 12 months after interventionwith that at baseline indicates that the inter-vention group experienced a greater reductionin the perception of morning wheeze and re-quirement for bronchodilator medication threehourly or more frequently than did the controlgroup. Otherwise, there were no significantreductions on any of the morbidity measuresshown in table 3 between the groups or betweenthe group differences at any time point andthat at baseline.

ADEQUACY OF ASTHMA TREATMENTAt baseline surprisingly few of the study par-ticipants had adequate medical treatment tocontrol their asthma as rated according to thepublished criteria.'7 At the 12 month follow upassessment point this situation was virtuallyunchanged, only eight subjects in each groupbeing rated as having adequate treatment.There were too few participants with adequatemedical treatment to conduct the planned com-parisons of morbidity indices for those com-pliant with adequate medical treatment andthose compliant with inadequate treatment.

DiscussionThe Royal North Shore Hospital asthma self-management education programme for adultshas been successful in achieving and main-taining its education objective of improvingknowledge about asthma and self-managementskills. In addition, the programme aim of sus-tained improvement in adherence to treatmentregimens was achieved. However, despite sig-nificant improvements in knowledge and com-pliance, the overall impact on morbidity wasmodest. There were significant improvementsin perceived morbidity but no delayed or short-lived improvements in objective morbiditymeasures. The continuing high levels of mor-bidity are more likely to reflect the inadequacyof medical treatment than deficiencies in studymethods.The study design and methods have a num-

ber of strengths. Sample size was clearly ade-quate in this study as statistically significantdifferences were found in central outcome vari-ables. The response rates were good and resultsare likely to be representative of the targetpopulation. The measure of asthma knowledgewas shown to be reliable and valid. Compliancedata were based on prospective diary recordsand published criteria were used to assess theadequacy of medical treatment. The clinicalsigns and physiological measures consideredcharacteristic of asthma are neither universalnor unique and the intensity of physiologicalmeasures to define asthma and to categorise

severity are arbitrary. This lack of precisionhampers research design and the interpretationand comparison of findings. We assessed mor-bidity retrospectively by interview and pro-spectively by diary record, as well as moreobjectively by spirometry. Terms and ratingswere clearly defined.The ultimate goal of asthma education strat-

egies is to reduce morbidity. The reduction inperceived morbidity, but not in physiologicalmeasures of asthma, found in this study isconsistent with other randomised controlledtrials.89 In this study utilisation of medical ser-vices for acute asthma care was low in the12 months before intervention (table 3) andconsequently the small reductions that oc-curred were non-significant.

Utilisation of acute asthma care services iscommonly used as an outcome indicator in theevaluation of asthma education interventions.Examination of the findings of this and similarstudies suggests that, for samples of asthmaticsubjects drawn from the community, recentuse of acute care for asthma may motivateparticipation in programmes but a more ex-tensive follow up period than 12 months maybe needed for significant group differences tobe detected. The sample of university clinicpatients recruited by Bailey and coworkers8 hadhigh rates of acute care utilisation at baseline.The group difference was non-significant 12months after intervention as a dramatic de-crease in hospital presentation or admissionhad occurred for both groups. Wilson and col-leagues9 found no significant difference in acutecare consultations 12 months after inter-vention, but significant reduction at 24 monthsin a sample drawn from members of a healthcare plan.

In contrast, Yoon and coworkersl' recruiteda sample from patients admitted to hospital foracute asthma care. They reported significantgroup differences in physiological measures fivemonths after intervention that were not sus-tained in the longer term. They also showed adramatic reduction in hospital presentationsand admissions. However, as the authors havedocumented,'7 participation rates were low andpatients of one of the authors were more likelyto participate.Asthma morbidity may be influenced by a

variety of factors within and outside the directcontrol of the programme. Our study allowedus to look at some of these factors - acquisitionof knowledge, compliance with treatment re-gimens, and adequacy of medical treatment.The study participants were knowledgeable atthe outset, but programme attendance resultedin a significant and sustained improvement inasthma knowledge. Similarly, for the morecommitted subjects who continued to par-ticipate in the study of Yoon et all' there wasa significant improvement in knowledge of as-pects of asthma self-management. However,knowledge is not invariably translated into be-haviour.

In the study reported here compliance withmedical treatment, assessed after group as-signment, was good at baseline for both groups(table 2). Following intervention the odds of

736

on Novem



/T


nloaded from



compliance in the intervention group are alwaysgreater than for the control group. It might beargued that compliance was overstated despitestrategies to encourage honest reporting. Par-ticipants in the study were actively encouragedto record reasons for departures from pre-scribed treatment and their responses were dis-armingly frank. Bailey and colleagues8 alsorealised their programme aim of improvedadherence to treatment regimens, reportingsubstantial improvements in inhaler and med-ication adherence in their intervention group.The adequacy of medical treatment is gen-

erally assumed in the evaluation of self-man-agement strategies. No matter how compliantthe participant, if medical treatment is in-adequate there is little likelihood of significantimprovement in any of the morbidity indicestypically used in assessment of programmeeffectiveness. A standard protocol was availableto guide the medical treatment of participantsin the study by Bailey and colleagues.8 How-ever, it is not recorded whether this treatmentprotocol was uniformly and consistently im-plemented. We adopted a set of criteria'4 torate the adequacy of medical treatment ir-respective ofpatient compliance. As the asthmaeducation programme is a community serviceprogramme, some variation in the standardof medical treatment was expected. However,nearly 90% of the study participants werejudged to have inadequate medical treatmentat baseline and there was virtually no changein this situation during the study period.The poor ratings of medical adequacy were

principally attributable to prescription ofregular daily 0, agonist therapy (84%) andinadequate anti-inflammatory medication(64%). The previously prevailing view of air-ways hyperresponsiveness as the primaryasthma mechanism had favoured the use ofbronchodilator medications for prophylaxis.The acceptance of inflammation as the primaryasthma mechanism, increased availability ofinhaled anti-inflammatory medications sincethe late 1980s, and evidence that long term P2agonist therapy may increase airways hyper-responsiveness"8 has led to the recom-mendation of anti-inflammatory medicationsas maintenance treatment for asthma.'4Lack of continuity of care because of geo-

graphical mobility or seeking treatment onlyfor acute attacks of asthma may have led toinadequate medical treatment. The par-ticipants in this study had considerable contactwith the medical profession in the 12 monthsbefore the study and the 12 months ofthe studyperiod. Nevertheless, their medical treatmentremained suboptimal. Several Australian com-munity based studies of adults with asthmahave shown that undertreatment of the con-dition is as common as 60%.19220Programme outcomes research can be

broadly categorised as efficacy or effectivenesstrials.2' Efficacy trials are designed to test whata programme achieves under optimum con-ditions and require committed, motivated sub-jects and controlled designs. Efficacy trials areconsidered to be necessary precursors to effect-iveness trials that test programme achievements

in more real life settings - for example, gen-eralisability trials.The randomised controlled study reported

here approximates the requirements for anefficacy trial. The participation rate and base-line levels of knowledge and compliance in-dicate a committed, motivated sample. Thesample was representative of the area which,according to 1986 census data,22 has higheremployment rates, average household income,and educational qualifications than either thestate or the country as a whole. Under thesesocioeconomically advantageous conditionsthe programme has demonstrated efficacy inachieving significant, sustained improvementsin knowledge and compliance. The efficacy ofthe programme in conjunction with optimalmedical treatment may need to be dem-onstrated before effectiveness trials of gen-eralisability are undertaken.Good asthma control requires not only good

self-management but also good medical treat-ment. To date the focus has been on improvingself-management and the adequacy of medicaltreatment has been neglected. The findings ofour study suggest that the adequacy of medicaltreatment needs to be taken into account whenevaluating asthma self-management inter-ventions. To maximise programme usefulness itis also desirable to review and optimise medicaltreatment in conjunction with the programme.

We thank Ms Gill Browne and Ms Wendy Starr who led theeducation groups and provided invaluable assistance with thedata collection. We gratefully acknowledge Dr David Allen forhis helpful comments on the manuscript. The study was partlyfunded by the NSW Department of Health. Peak flow meterswere supplied by Allen and Hanburys.

1 Robertson CF, Rubinfeld AR, Bowes G. Deaths from asthmain Victoria: a 12 month survey. Med JT Aust 1990;152:511-7.

2 Hom CR, Clark TJH, Cochrane GM. Compliance withinhaled therapy and morbidity from asthma. Respir Med1990;84:67-70.

3 Hindi-Alexander MC, Throm J, Middleton E. Collaborativeasthma self-management: evaluation designs. Clin RevAllergy 1987;5:249-58.

4 Snyder SE, Wmder JA, Creer TL. Development and evalu-ation of an adult asthma self-management program:Wheezers Anonymous. J Asthma 1987;24: 153-8.

5 Bauman AE, Craig AR, Dunsmore J, Browne G, AllenDH, Vandenberg R. Removing barriers to effective self-management. Patient Education and Counseling 1989;14:217-26.

6 Worth H. Patient education in asthmatic adults. Lung 1990;Suppl:463-8.

7 Ringsberg KC, Wiklund I, Wilhelmsen L. Education ofadult patients at an 'asthma school': effects on quality oflife, knowledge and need for nursing. Eur Respir Jf 1990;3:33-7.

8 Bailey WC, Richards JM, Brooks CM, Soong S, WindsorRA, Manzella BA. A randomised trial to improve self-management practices of adults with asthma. Arch InternMed 1990;150:1664-8.

9 Wilson SR, Scamagas P, German DF, Hughes FW, LullaS, Coss S, et al. A controlled trial of two forms of self-management education for adults with asthma. Am J. Med1993;94:564-76.

10 Yoon R, McKenzie DA, Bauman A, Miles DA. Controlledtrial evaluation of an asthma education programme foradults. Thorax 1993;48:1110-6.

11 Browne G. Asthma education handbook. Royal North ShoreHospital, Sydney, Australia, 1986.

12 Woolcock AJ, Rubinfeld AR, Seale JP, Landau LL, AnticR, Mitchell C, et al. Thoracic Society of Australia andNew Zealand asthma management plan, 1989. MedjAust1989;151:650-3.

13 Creer TL, Winder JA. Asthma. In: Holroyd KA, Creer TL.eds. Self-management ofchronic disease. Orlando: AcademicPress, 1986;269-303.

14 Hargeave FE, Dolovich J, Newhouse MT. The assessmentand treatment of asthma: a conference report. 7 AllergyClin Immunol 1990;85:1098-111.

15 Zeger SI, Liang K-Y Longitudinal data analysis for discreteand continuous outcomes. Biometrics 1986;42:121-30.

737

on Novem



/T


nloaded from


Alen, J'ones, Oldenburg16 Gebski V, Leung 0, McNeil D, Lunn D. SPIDA User'

Manual. Version 6. Sydney: Southwood, 1992.17 Yoon R, McKenzie DA, Miles DA, Bauman A. Char-

acteristics of attenders and non-attenders at an asthmaeducation programme. Thorax 1991;46:886-9.

18 Sears MR, Taylor DR, Print CG, Lake DC, Li Q, FlanneryE, et al. Regular inhaled beta-agonist treatment in bron-chial asthma. Lancet 1990;336:1391-6.

19 Henry DA, Sutherland D, Francis L and the NewcastleRetail Pharmacy Research Group. The use of non-pre-scription salbutamol inhalers by asthmatic patients in the

Hunter Valley, New South Wales. Med Aust 1989;150:445-9.

20 Gibson PG, Henry D, Francis L, Cruickshank D, DupenF, Higginbotham N, et al Association between availabilityof non-prescription ,2-agonist inhalers and under-treatment of asthma. BMJ 1993;306:1514-8.

21 Flay BR. Efficacy and effectiveness trials (and other phasesof research) in the development of health promotion pro-grammes. Prev Med 1991;15:451-74.

22 Australian Bureau of Statistics. Tables CX1148, CX1155;1986 Census.

738

on Novem



/T


nloaded from


randomised trial of asthma self-management programme for ... · randomised trial ofan...

Documents