PEDIATRICS Vol. 97 No. 4 April 1996 597

Committee Report: Population-to-Pediatrician Ratio Estimates:

A Subject Review

Committee on Careers and Opportunities

Many pediatricians evaluating practice opportuni-ties are interested in population-to-provider ratiosinsofar as they have obvious implications for de-mand for services and economic viabffity of practice.Managed-care systems are intensely interested in do-termining the optimal staffing ratio of children orenrollees to full-time equivalent (VFE) pediatrician.This statement summarizes available information re-lating to staffing ratios of children to FTE pediatri-cians and provides formulas that can be of assistancein evaluating an area’s demand for pediatricians.

PREVIOUS WORK FORCE STUDIES

In 1980, the Graduate Medical Education NationalAdvisory Committee (GMENAC) produced a reportdetailing estimates for supply of and requirementsfor health professionals for 1990. The GMENAC usedan adjusted, needs-based method to estimate ro-quirements for various specialties and also used itsown model to estimate the future supply of physi-cians by specialty. Although the GMENAC’s methodhas been criticized,’ the ideal ratio of child popula-tion to general pediatrician arrived at by theGMENAC was 2033:1.2

In 1991, the US Bureau of Health Professions con-traded with the consulting firm of Abt Associates toupdate the needs-based requirements model of theGMENAC for seven physician specialties, includingpediatrics. Abt estimated requirements based onideal levels and types of medical care (preventiveand curative) without regard to barriers posed byabifity to pay, access, availability, or lack of knowl-edge. Abt Associates recommended a ratio of childpopulation to general pediatrician of 2430:1.�

STAFFING RATIOS

Most of the available staffing ratio informationcomes from health maintenance organizations(HMOs) or from practices serving privately insuredpatients; therefore, it is based on children currentlyunder the care of physicians. Staffing ratios of chil-dren to VFE pediatricians from a variety of HMOsare summarized in the Table. These ratios vary fromthe ratio of 1400:1 found in the private setting of 91practices that belong to the American Academy ofPediatrics (AAP) PROS Network, a nationwide net-

The recommendations in this statement do not indicate an exclusive course

of treatment or serve as a standard of medical care. Variations, taking intoaccount individual circumstances, may be appropriate.PEDIATRIcS (ISSN 0031 4005). Copyright C 1996 by the American Acad-

emy of Pediatrics.

work of 700 clinicians engaged in office-based re-search. Definitions of the ratios in these estimatesvary somewhat according to the age criteria used.Pediatricians also must keep in mind that HMO staff-ing ratios are different from population ratios, whichencompass all children, including those who do nothave a usual source of care and those who receivecare from other physicians. Furthermore, the numberof active patients must be distinguished from thenumber of inactive patients when examining staffingratios.

Children enrolled in prepaid group practices havebeen found to use preventive health care services ata higher rate than children in fee-for-service arrange-ments. In general, children enrolled in prepaid grouppractices make four to five visits per year, comparedwith three visits per year made by children in fee-for-service practices.4’�

Ratios such as those presented in the Table shouldbe viewed with caution. As the Table demonstrates,ratios are variable among HMO practices. Greater orsmaller numbers of children per pediatrician are notnecessarily an indication of the quality of care. Suchdifferences are often a result of a number of factors,induding the level of automation in a practice, theproductivity of physicians, the abffity to delegatespecific tasks to nonphysidan providers, and thehealth status and/or psychosocial and behavioralstatus of patients.

METHODOLOGY FOR EVALUATING AN AREA’SDEMAND FOR A PEDIATRICIAN

Two models based on current use and productiv-ity of physicians are discussed in this section.68These models incorporate an area’s total or childpopulation and the number of local pediatricians andother physicians who see children. This informationoften can be obtained from city or county depart-ments of health, school districts, local libraries, andchambers of commerce.

Both of the following models estimate an area’sdemand for pediatricians but differ with regard tothe variables involved and the end product. Themodels do not suggest an ideal number of childrenper pediatrician based on need or an appropriatenumber of visits that should be made by childrenbut, instead, reflect the status quo and as such mayrepresent underuse according to some standards.9Furthermore, because pediatric subspecialty servicework force assessments are beyond the scope of thisanalysis, for the two models, we conceptualize pedi-

by guest on January 22, 2020www.aappublications.org/newsDownloaded from

Visits/wk

25th percentile = 85

50th percentile = 11075th percentile = 150

Weeks

Worked/y Visits/y

X47 =3995X47 =5170

X47 =7050

* From a staff model Midwestern health maintenance organization

(HMO).

t From an Independent Practice Association model New EnglandHMO.:$:The average number of HMO enrollees younger than 15 yearsper full-time pediatrician from 33 different HMOs, 1992. GroupHealth Association of America, Inc. HMO Industry Profile; 1993.Washington, DC: Group Health Association of America.

§ From a staff model Midwestern HMO.IIRatios of children per pediatrician for two different group modelHMO practices advertising for pediatricians in the Northeast.

I From a staff and group model New England HMO.

TABLE. Staffing Ratios for Children per Pediatrician

598 POPULATION-TO-PEDIATRICIAN RATIO ESTIMATES

Ages of

Children, yChildren per Pediatrician

in HMO

0-18 885�1105:1*0-17 1300:lt

0-14 1157:1�

0-11 1200:1�

Unknown 1125 :111

Unknown 1750:111Unknown 1200:11

atric services as only those provided by non-hospital-based general pediatricians.

Model of Hicks and Glenn

The model of Hicks and Glenn6’7 focuses on pro-viders considering entrance to a market and is there-fore designed to estimate the population necessary tosupport a pediatrician. The model requires knowl-edge of visit rates to pediatricians and the number ofweeks worked per year per pediatrician. The modelcontrols for visits by children to nonpediatrician pro-viders through the visit rates. Therefore, it is notnecessary to know the number of family physicians,for example, in the area.

The visit rates and productivity figures providedin the articles by Hicks and Glenn6’7 may not beconsistent with current estimates. Therefore, in thefollowing description of the model, more recent in-formation is used.

Step I

. . office visits/wk X weeks worked/yPopulation size = . .

visits per capita/y

This example will provide three population esti-mates, demonstrating the versatility of this model forincorporating different work styles and/or patientcare demands. The 25th, 50th, and 75th percentiles ofthe number of ambulatory visits general pediatri-cians have per week will be used. On average, thereare 0.31 visits per capita to a pediatrician in an am-bulatory setting per year (AAP Department of Re-search tabulations from the 1990 Health InterviewSurvey data tape). General pediatricians work anaverage of 47 wk/y.’#{176}The median number of ambu-latory visits general pediatricians have per week isI 10, the 25th percentile is 85 visits per week, and the75th percentile is 150, based on AAP Department ofResearch, Periodic Survey No. 21. The number ofweekly visits is multiplied by the number of weeksworked per year to calculate the number of visits peryear. This number is then divided by the number ofvisits per capita to derive the population size neces-

Visits/yDivided

by Visitsper Capita

= Population

3995/0.31 = 12 8875170/0.31 = 16677

7050/0.31 = 22742

sary to obtain the desired productivity, given theaverage number of visits to pediatricians by the pop-ulation.

Please note: These figures represent populationsizes, not practice sizes. Some children may not bereceiving care, or they may be receiving care fromphysicians in other specialties, eg, family physiciansor internists.

Step 2

Population

4.19

1.56

1.822.03

2 X 5170 visits8 X (0.212 x 5574) visits

COMMITFEE REPORT 599

that all pediatric care is being handled by the G/FPs.On average, G/FPs have 118.6 visits per week andwork an average of 47 wk/y, totaling 5574 visits peryear.’#{176}Using the same model, one G/FP needs a totalservice area population of 4160 to generate an aver-age of 5574 ambulatory visits per year at 1.33 visitsper capita, based on AAP Department of Researchtabulations and those of Gonzalez.’#{176} Obviously, if asimilar community, equally as small, already has apediatrician, then, by this model, there would not bea need for an additional one.

Model of Miller and Associates

The model of Miller et al8 is designed to assistpediatricians and community leaders in estimatingthe primary-care needs of patients in the communityand the economic feasibffity of establishing a pediat-ric practice. The model requires information on thevisit rates of children to physicians, as well as agebreakdowns of the area’s child population.

Step 1

Physician visits = visit rate

x No. of children in each age group

The visit rates provided in the article8 are dated;more current approximations of ambulatory visits toall physicians for 1990 are as follows (AAP Depart-ment of Research tabulations from the 1990 HealthInterview Survey data tape).

Age, y Average Visit Rate/y

0-56-10

11-1415-21

One must calculate how many physician visits bychildren and adolescents will be made in a commu-nity by multiplying these visit rates by the number ofchildren in each age group. If such age distributionsare unavailable, then an age-adjusted overall visitrate of 2.52 for children 0 to 21 years of age willsuffice. As stated previously, about 40% of all chil-dren’s and adolescents’ visits are to pediatricians.

Step 2

No. of pediatricians in the community

plus x average visits/y to pediatricians

No. of G/FPs in the community

x % of visits made by children and adolescents

x average visits/y to G/FPs

Next, one must determine the number of visits thatthe area’s child health physicians can accommodate.For pediatricians, the median productivity of 5170 vis-its per year is used (AAP Department of Research,Periodic Survey No. 21). An estimate of the number ofvisits to G/FPs from children must be induded. Asstated above, on average, G/FPs have 118.6 visits perweek and work an average of 47 wk/y, totaling 5574visits per year.’#{176}Approximately one fifth (21.2%) or1182 of these visits come from children 0 to 21 years of

age (AAP Department of Research tabulations from the1990 Health Interview Survey data tape). One mustfactor in this number, multiplied by the number offamily and general practitioners in the area.

Step 3

Subtract from the results of step I (the expectednumber of total visits to be made by the community’schildren) the results of step 2 (number of visits thatcan be managed by the area’s physicians) to get thenumber of visits that currently are not being accom-modated by local physicians. Divide this value bythe average number of visits per year per pediatri-cian to derive the number of additional pediatriciansan area can support.

Example

An area has 11 655 children in the following agegroups, generating 29 000 total visits:

Step 1: Age,y No. of children = Total Visitsx Visit Rate

0-5 3391 x 4.19 visits 14 2086-10 2681 x 1.56 visits = 4182

11-14 2016 x 1.82 visits = 366915-21 3567 x 2.03 visits = 7241

A. 29 300total visits

Step 2: This community has 2 pediatricianS and 8 general/familypractitioners

= 10340= 9453

B. 19793accommodated

visits

Step 3: The remaining potential ambulatory visitsin this area equal A - B, or 9507 visits. Dividing thisremainder by the average number of visits per year apediatrician can accommodate (9/5170) results in thenumber of additional pediatricians the area couldpotentially support (1 .8), given that there are noother child health care providers that have not beenaccounted for.

Obviously, the same factors (productivity, visitrate, and competition) that can influence the modelof Hicks and Glenn6� are apparent here.

Insurance Effects

DISCUSSION

Current estimates of the use of pediatric servicesreflect national averages across a broad array of typesof insurance coverage, ranging from no coverage tofirst-dollar coverage, private insurance. Approximately86% of US children are covered by private or publichealth insurance plans.’3 Insurance coverage plays asignificant role in the use of services. For example, ithas been shown that insured children are more likely tosee physicians for acute illnesses than are those withoutinsurance.’4 In fact, it is estimated that implementationof universal health insurance would increase the num-ber of pediatric contacts by 5% to 15%.15,16

Managed-care Implications

In using the models of Hicks and Glenn6’7 and Mifieret al,8 it is important to keep in mind the differences in

by guest on January 22, 2020www.aappublications.org/newsDownloaded from

600 POPULATION-TO-PEDIATRICIAN RATIO ESTIMATES

use among fee-for-service patients and those enrolledin managed-care organizations. Data on use fromHMOs and other managed-care systems have certaininherent limitations. To the extent that managed-careorganizations have healthier than average populationsenrolled in their plans, use would be expected to belower than what is observed for the general population.Additionally, most managed-care plans have imple-mented elaborate utilization review programs to con-trol the demand for and provision of services. On theother hand, evidence indicates that, from the stand-point of patients, prepaid coverage (especially whenout-of-pocket expense is minimal or absent) results in alower threshold for seeking care.’7”8 One study foundthat overall visit rates for children enrolled in an FIMOwere about 20% higher and preventive visits were 40%higher than for those with the cost-sharing� fee-for-service plan, whereas hospital admission rates wereabout 17% lower for the FIMO group.5

Approximately 34% of all individuals with healthinsurance coverage are enrolled in managed-careplans.’9 Visit rates to pediatricians may increase withan increase in enrollment of children in managed-careplans. Although adjustment factors to extrapolate fromHMOs to the overall US population have been estimat-ed,�#{176}the features unique to managed-care organiza-tions and their enrollees make it difficult to generalizefrom their use patterns to the general population. How-ever, because the nation seems increasingly headedtoward managed-care systems as the predominantmode of health services delivery, the experience ofliMOs to date is of obvious importance.

CONCLUSION

This statement is intended to assist pediatricians,health care administrators, and policymakers in evalu-ating an area’s or a population’s demand for pediatri-cians; it is not meant to recommend an ideal number ofchildren per pediatrician based on need. This statementprovides a variety of numbers to work with, rangingfrom low ratios, reflecting the number of children cur-rently in the health care system that can be seen real-istically by pediatricians (see Table), to high ratios,reflecting the current use of pediatricians by the entirepopulation younger than 21 years of age (see the mod-els of Mifier et al8 and Hicks and Glenn6’�D.

In applying the models, it is important to keep inmind the several previously discussed factors that af-fed the ratio of children per pediatrician. Other factorsthat also must be considered when examining ratiosare differences in physician productivity that may varyaccording to gender, personal characteristics or life-styles, and geographic location. In addition, local dif-ferences in demand for pediatric services will varyaccording to child health status and use of services.Finally, universal access to health care and/or man-aged care would increase the demand for pediatricservices. Limitations in available data prevent the abil-ity of the models to be used for subspecialty care or foroutpatient care provided in hospitals.

There is currently a lack of information examiningratios, quality of care, and health outcomes. Studieslinking these three issues are greatly needed. Alsoneeded are studies that focus on the appropriatenumber of children per pediatric subspecialist.

COMMITTEE ON CAREERS AND OPPORTUNITIES,1994 TO 1995

Robert L. Johnson, MD, ChairmanElena Fuentes-Afflick, MDMarion J. Balsam, MDM. Rosario Gonzalez-De-Rivas, MDStephen N. Keith, MDMary Antenen Mdllroy, MDKathleen C. Nelson, MDWanessa P. Risko, MDJeffrey J. Stoddard, MDDoris Wethers, MD

NATIONAL MEDICAL ASSOCIATION LIAISON

Frances L. Dunston, MD, MPH

STAFF CONSULTANTS

Sarah Brotherton, PhDSuzanne Connaughton, MPH

REFERENCES

1. Budetti PP. The impending pediatric “surplus”: causes, implications,

and alternatives. Pediatrics. 1981;67:597-606

2. DepartmentofHealth and HumanServices. Summary Re�rtofthe Graduate

Malice! Education NationalMvisory Committee to the Secretary, Dqxzrtment of

Health and Human Services. Washington, DC: Government Printing Office;

1980. US Dept of Health and Human Services publication (HRA) 81-6513. Abt Associates. Reexamination ofthe Adequacy ofPhysician Supply Made in

1980 by the Graduate Medical Education National Advisory Committee

(GMENAC) for Selected Specialties. Springfield, VA: National Technical

information Service; 1991. Health Resources and Services Administra-

tion publication 240-89-0041 PB 91-18813

4. Smoller M. Telephone calls and appointment requests: predictability in

an unpredictable world. HMO Pract. 19926:25-29

5. Valdez RB, Ware JE Jr. Manning WG, et al. Prepaid group practiceeffects on the utilization of medical services and health outcomes for

children: results from a controlled trial. Pediatrics. 1989;83:168-1806. Hicks LL, Glenn JK. Too many physicians in the wrong places and

specialties? Populations and physicians from a market perspective. IHealth Care Manage. 19899:18-26

7. Hicks LL, Glenn JK. Rural populations and rural physicians: estimates

of critical mass ratios, by specialty. I Rural Health. 1991;7:357-3728. Miller KA, Miller DA, Doeksen GA,Jacobs P. A model to determine the

feasibility of a pediatric practice. Am I Dis Child. 1989;143:919-9239. Green M, ed. Bright Futures. Guidelines for Health Supervision of Infants,

Children, and Adolescents. Arlington, VA: National Center for Education

in Maternal and Children Health; 1994

10. Gonzalez ML, ed. Socioeconomic Characteristics of Medical Practice.

Chicago, IL: American Medical Association; 1993

11. National Center for Health Statistics. Health, United States, 1992 and

Healthy PeopIe2OfYJ Review. Hyattsville, MIY. Public Health Service; 1993.

US Dept of Health and Human Services publication PHS 93-1232

12. Ing CD, Tewey BP. Summary of Data on Children and Youth With Disabil-ities. Washington, DC: National Institute on Disabifity and Rehabilita-

tion Research; 1994. US Dept of Education publication HN 93027001

13. Cartland JD, Yudkowsky BK. State estimates of uninsured children.

Health Aff(Millwood). 1993;12:144-151

14. Stoddard ll� St Peter RE,Newacheck PW. Health insurance status andambulatory care for children. N EngI I Med. 1994,330:1421-1425

15. Lawn ICF. American Academy ofPediatrics Proposalfor UniversalAccess to

Health CareforChildren and Pregnant Women. Washington, DC: American

Academy of Pediatrics; 1991

16. WeinerJP. The demand for physician services in a changing health caresystem: a synthesis. Med Care Rev. 199350:411-449

17. Leibowitz A, Manning WGJr, Keeler ES, Duan N, Lohr KN, NewhouseJP.

Effect of cost-sharing on the use of medical services by children: interimresults from a randomized controlled trial. Pediatrics. 1985;75:942-951

18. Lohr KN, Brook RH, Kamberg CJ, et al. Use of medical care in theRAND Health Insurance Experiment: diagnosis- and service-specific

analyses in a randomized controlled trial. Med Care. 198624:S1-S87

19. American Academy of Pediatrics, Committee on Child Health Financ-

ing. Strategies for managed care. AAP News. 1994;10

20. Weiner JP. Forecasting the effects of health reform on US physicianworkforce requirement: evidence from HMO staffing patterns. JAMA.1994;272:222-230

by guest on January 22, 2020www.aappublications.org/newsDownloaded from

1996;97;597Pediatrics Committee on Careers and Opportunities

Committee Report: Population-to-Pediatrician Ratio Estimates: A Subject Review

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/97/4/597including high resolution figures, can be found at:

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlentirety can be found online at: Information about reproducing this article in parts (figures, tables) or in its

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on January 22, 2020www.aappublications.org/newsDownloaded from

http://http://pediatrics.aappublications.org/content/97/4/597http://www.aappublications.org/site/misc/Permissions.xhtmlhttp://www.aappublications.org/site/misc/reprints.xhtml

1996;97;597Pediatrics Committee on Careers and Opportunities

Committee Report: Population-to-Pediatrician Ratio Estimates: A Subject Review

http://pediatrics.aappublications.org/content/97/4/597the World Wide Web at:

The online version of this article, along with updated information and services, is located on

Copyright © 1996 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397. American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007.been published continuously since 1948. Pediatrics is owned, published, and trademarked by the Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has

by guest on January 22, 2020www.aappublications.org/newsDownloaded from

http://pediatrics.aappublications.org/content/97/4/597