RESEARCH 9*0--0*-0--- Preventive care and barriers to effective prevention How do_famil physicians see it? BRIAN G. HUTCHISON, MD, MSC, CCFP JULIA ABELSON, MSC CHRISTEL A. WOODWARD, PHD GEOFFREY NORMAN, PHD Dr Hutchison is an Associate Professor in the Department ofFamily Medicine, the Centrefor Health Economics and Policy Analysis, and the Department of Clinical Epidemiology and Biostatistics; Ms Abelson is Health Policy Researcher; Dr Woodward is a Professor in the Centrefor Health Economics and Policy Analysis and the Department of Clinical Epidemiology and Biostatistics; and Dr Norman is a Professor in the Department of Clinical Epidemiology and Biostatistics, all at McMaster Universiy in Hamilton, Ont. OBJECTIVES To assess how adequately family physicians think they arc delivering preventive care and to examine barriers to providing preventive care. DESIGN Cross-sectional survey. SETTING Primary care medical practices in south-central Ontario. PARTICIPANTS Four hundred eighty family physicians and general practitioners who graduated from medical school between 1972 and 1988. MAIN OUTCOME MEASURES Satisfactory preventive care delivery versus self-assessed coverage of patients for 15 preventive maneuvers. Perceived reasons for lack of success in providing recommended preventive care. RESULTS For 10 of the 15 maneuvcrs, the proportion of physicians who regarded 90% or higher as satisfactory coverage was twicc as great as the proportion who thought they provided that level of coverage. For 11 of the 15 mancuvers, most respondents reported coverage lower than the level they regarded as satisfactory. For six maneuvers, more than two thirds thought they provided less than satisfactory coverage. More than two thirds of rcspondents suggested these barriers to providing recommended preventive care: patient is healthy and does not visit; patient refuses, is not interested, or does not comply; no effective systems to remind patients to come in for preventive care; and priority given to presenting problem. CONCLUSION Many family physicians and general practitioners in south-central Ontario provide preventive care to their patients at lower levels than they consider satisfactory. TI'hey identified barriers to providing preventive services successfully; these barriers suggest approaches for improving care. OBJECTIFS Evaluer la perception des medecins conccrnan-t la qualite des soins preventifs qu'ils dispensent et analyser les obstacles A la prestation des soins preventifs. CONCEPTION Enqu&te transversale. \ CONTEXTE Cliniques de soins medicaux de premiere ligne du Centre et du Sud de l'Ontario. PARTICIPANTS Quatre cent huit medecins de famille et omnipraticiens qui ont recu leur dipl6me de medecin entre 1972 et 1988. PRINCIPALES MESURES DES RESULTATS Prestation satisfaisante des soins preventifs comparativement A l'auto-appr6ciation de leur efficacite A appliquer 15 interventions preventives. Raisons pernues pour l'insucces A dispenser les soins preventifs recommand6s. RESULTATS Peu de medecins avaient l'impression que leur niveau de couverture etait satisfaisant. Pour 10 des 15 interventions, la proportion des medecins, qui avaient etabli que le taux de couverture etait satisfaisant lorsque l'intervention etait appliquee dans 90% des cas, fut deux fois plus 6levee que la proportion de ceux qui pensaient offrir ce niveau de couverture. Pour 11 des 15 intcrventions, la plupart des r6pondants ont mentionne que leur niveau de couverture ctait inferieur au niveau qu'ils jugeaient satisfaisant. Pour six interventions, plus des deux tiers avaient l'impression d'offrir un niveau de couverture non satisfaisant. Plus des deux tiers des repondants ont mentionne les obstacles suivants A la prestation des soins preventifs recommandes: le patient est en bonnc santc et ne consulte pas; le patient refuse, n'est pas interessc ou ne respecte pas les recommandations; absence de systerme efficace pour rappeler aux patients de consulter pour des soins preventifs ; et priorite accordee A la raison de consultation. CONCLUSION De nombreux mcdecins de famille et omnipraticiens du Centre et du Sud de l'Ontario dispensent des soins preventifs A leurs patients A un niveau inferieur A celui qu'ils consid&rent satisfaisant. Ils ont identifie les obstacles qui les emp&chent d'offrir des soins preventifs adequats ; ces obstacles suggerent des approches pour ameliorer la qualite des soiims. Can Fam Physician 1996;42:1693-1700. VOL 42: SEPTEMBER * SEPTEMBRE 1996 + Canadian Family Plhysician * Le Aledecin defamille canadien 1693
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
RESEARCH9*0--0*-0---
Preventive care and barriers
to effective preventionHow do_famil physicians see it?
BRIAN G. HUTCHISON, MD, MSC, CCFPJULIA ABELSON, MSCCHRISTEL A. WOODWARD, PHDGEOFFREY NORMAN, PHD
Dr Hutchison is an
Associate Professor in theDepartment ofFamilyMedicine, the CentreforHealth Economics andPolicy Analysis, and theDepartment of ClinicalEpidemiology andBiostatistics;Ms Abelson is HealthPolicy Researcher;Dr Woodward is a
Professor in the CentreforHealth Economics andPolicy Analysis and theDepartment of ClinicalEpidemiology andBiostatistics; andDr Norman is a
Professor in the Departmentof Clinical Epidemiology andBiostatistics, all at
McMaster Universiy in
Hamilton, Ont.
OBJECTIVES To assess how adequately family physicians think they arc delivering preventive careand to examine barriers to providing preventive care.
DESIGN Cross-sectional survey.SETTING Primary care medical practices in south-central Ontario.PARTICIPANTS Four hundred eighty family physicians and general practitioners who graduatedfrom medical school between 1972 and 1988.MAIN OUTCOME MEASURES Satisfactory preventive care delivery versus self-assessed coverage ofpatients for 15 preventive maneuvers. Perceived reasons for lack of success in providingrecommended preventive care.RESULTS For 10 of the 15 maneuvcrs, the proportion of physicians who regarded 90% or higheras satisfactory coverage was twicc as great as the proportion who thought they provided that levelof coverage. For 11 of the 15 mancuvers, most respondents reported coverage lower than thelevel they regarded as satisfactory. For six maneuvers, more than two thirds thought theyprovided less than satisfactory coverage. More than two thirds of rcspondents suggested thesebarriers to providing recommended preventive care: patient is healthy and does not visit; patientrefuses, is not interested, or does not comply; no effective systems to remind patients to come infor preventive care; and priority given to presenting problem.CONCLUSION Many family physicians and general practitioners in south-central Ontario providepreventive care to their patients at lower levels than they consider satisfactory. TI'hey identified barriersto providing preventive services successfully; these barriers suggest approaches for improving care.
OBJECTIFS Evaluer la perception des medecins conccrnan-t la qualite des soins preventifs qu'ilsdispensent et analyser les obstacles A la prestation des soins preventifs.CONCEPTION Enqu&te transversale. \
CONTEXTE Cliniques de soins medicaux de premiere ligne du Centre et du Sud de l'Ontario.PARTICIPANTS Quatre cent huit medecins de famille et omnipraticiens qui ont recu leur dipl6mede medecin entre 1972 et 1988.PRINCIPALES MESURES DES RESULTATS Prestation satisfaisante des soins preventifs comparativement Al'auto-appr6ciation de leur efficacite A appliquer 15 interventions preventives. Raisons pernuespour l'insucces A dispenser les soins preventifs recommand6s.RESULTATS Peu de medecins avaient l'impression que leur niveau de couverture etait satisfaisant.Pour 10 des 15 interventions, la proportion des medecins, qui avaient etabli que le taux decouverture etait satisfaisant lorsque l'intervention etait appliquee dans 90% des cas, fut deux foisplus 6levee que la proportion de ceux qui pensaient offrir ce niveau de couverture. Pour 11 des15 intcrventions, la plupart des r6pondants ont mentionne que leur niveau de couverture ctaitinferieur au niveau qu'ils jugeaient satisfaisant. Pour six interventions, plus des deux tiers avaientl'impression d'offrir un niveau de couverture non satisfaisant. Plus des deux tiers des repondantsont mentionne les obstacles suivants A la prestation des soins preventifs recommandes: le patientest en bonnc santc et ne consulte pas; le patient refuse, n'est pas interessc ou ne respecte pas lesrecommandations; absence de systerme efficace pour rappeler aux patients de consulter pour dessoins preventifs ; et priorite accordee A la raison de consultation.CONCLUSION De nombreux mcdecins de famille et omnipraticiens du Centre et du Sud del'Ontario dispensent des soins preventifs A leurs patients A un niveau inferieur A celui qu'ilsconsid&rent satisfaisant. Ils ont identifie les obstacles qui les emp&chent d'offrir des soinspreventifs adequats ; ces obstacles suggerent des approches pour ameliorer la qualite des soiims.
Can Fam Physician 1996;42:1693-1700.
VOL 42: SEPTEMBER *SEPTEMBRE 1996 + Canadian Family Plhysician * Le Aledecin defamille canadien 1693
RESEARCH
Preventive care and barriers to effective prevention
TUDIES OF A RANGE OF PREVENTIVE MANEU-
vers have demonstrated in various pri-mary care settings that many, oftenmost, eligible patients do not receive
recommended preventive services."5 Even whenefforts are made to improve preventive care, cov-erage often falls far short of target levels.7'5 Whatare physicians' perceptions regarding preventivecare? Do physicians consider preventive interven-tions unimportant? What levels of preventive caredo they regard as satisfactory? What levels of cov-erage do they think they are achieving, and howdo these compare with those they consider desir-able? What do physicians see as the main barriersto providing satisfactory preventive care?
Information on the relationship between physi-cians' perceptions of the importance of preventivemaneuvers and their performance of thosemaneuvers is scanty. Dietrich and Goldberg'6showed a correlation between importance ratingand performance of seven preventive maneuvers(r = 0.45). However, several interventions (tetanusimmunization, mammography, and influenzaimmunization) were offered to less than half ofthe eligible patients. Overall, only 58% of indicat-ed procedures considered important were offeredor performed.We have found no studies assessing physi-
cians' perceptions of satisfactory levels of pre-ventive coverage in relation to their self-assessedor measured levels of performance. This issuehas important implications for the developmentof strategies to improve preventive care. If physi-cians believe they are achieving satisfactory lev-els of coverage, efforts to improve performancemust be directed toward encouraging them toaim higher or toward demonstrating that theyare overestimating current coverage. If, on theother hand, physicians think that current levelsof coverage are not satisfactory, they might bereceptive to strategies to enhance preventivecare performance.
Few studies have explored physicians' percep-tions of barriers to providing preventive care.Attarian and colleagues'7 have reported thatNorth Carolina family physicians and generalpractitioners most commonly cite lack of time,
patients' lack of motivation, patients' expecta-tions, and the reimbursement system as barriersto health promotion couinseling. McPhee and col-leagues'8 assessed perceived barriers to cancerscreening among 52 physicians in a universitygeneral medical practice. The main reasonsphysicians offered for not doing recommendedscreening tests were physicians' objections to thetests, physicians' forgetfulness, lack of time, andpatients' dislike or refusal.
In this paper, we present results from a surveyof family physicians in which we examined thegap between the level of preventive care physi-cians thought was satisfactory and the level ofcoverage they thought they achieved for each of15 preventive maneuvers, which have been eval-uated by the Canadian Task Force on thePeriodic Health Examination.'9 We also reporton physicians' perceptions of barriers to provid-ing preventive care.
METHODS
Between October 1993 and March 1994, a pre-ventive care survey was conducted among familyphysicians and general practitioners practising inrural areas and large, medium, and small cities insouth-central Ontario. The survey area was limit-ed to communities within 1 hour's drive ofMcMaster University in order to facilitate thesecond phase of the study, which involved intro-ducing unannounced standardized patients intothe practices of consenting physicians.
Questionnaires were mailed to all physicianswith addresses in the study area who were listedin the Canadian Medical Association's PhysicianResource Databank as family physicians or gener-al practitioners and whose recorded graduationfrom medical school was between 1972 and 1988.The first mailing was sent to 1236 physicians. Areminder postcard was sent 10 days later.Nonrespondents received follow-up mailings1 month and 2 '/2 months after the initial mailing.Shortly after the survey was begun, we realizedthe sample included many ineligible physicians.After the second mailing, we used the 1993Canadian Medical Directory to identify and eliminate
1694 Canadian Family Physician * Le Midecin defamille canadien , VOL 42: SEPTEMBER * SEPTEMBRE 1996
RESEARCH
Preventive care and barriers to effective prevention
24 ineligible doctors.After the final mailing,337 randomly selectednonrespondents (of atotal of 750) werecontacted by telephoneto check their eligibilityand encourage themto return the question-naire.The questionnaire
solicited information onphysician and practicecharacteristics, attitudestoward preventive care,perceptions of theimportance of a range ofpreventive maneuvers,self-assessed perfor-mance of preventiveinterventions, and barri-
Table 1. Questions on preventive care coverageand barriers to providing preventive care
What would you consider to be a satisfactory level ofperformance (in terms ofthe proportion of eligible patientscovered) for the following preventive maneuvers?
Considering the "real-world" limitations ofyour actualpractice, for what proportion ofyour eligible patients doyou think you are ordering or performing the followingpreventive maneuvers?
For those cases in which a preventive service isrecommended but you do not succeed in getting it done,which ofthe following best describe the reason(s)?
0.56 to 0.74 for self-assessed performance,and 0.44 to 0.51 for bar-riers to effective preven-tive care performance.
Data were enteredinto an SPSS-PC(Statistical Package forthe Social Sciences)database and auditedfor accuracy. We useddescriptive statistics andlog linear analysis tocompare respondentsand nonrespondents asto sex, certificationstatus, and decade ofgraduation, based oninformation about non-respondents from theCanadian Medical Directogy.
ers to providing preventive care. The wording ofthe items addressing issues of preventive care per-formance and barriers to providing preventivecare is presented in Table 1.Two preventive maneuvers on the question-
naire concerned mammography for women 50 to59 years of age, one reflecting the recommenda-tion of the Canadian Task Force on the PeriodicHealth Examination'9 (annual mammography)and the other the recommendation of theOntario Breast Screening Program2"' (mammog-raphy at 2-year intervals). The remaining13 maneuvers were separate interventions. Thequestion regarding perceived barriers to provid-ing preventive care was modified from an itemused in a survey of attitudes, knowledge, andpractice of disease prevention and health promo-tion developed by investigators at the JohnsHopkins Health Institution.2'
The questionnaire was pretested on a conveniencesample of family physicians. Fifty physicians includedin the survey participated in a study assessing test-retest reliability. For each question type, a randomsample of items was selected. For items relevant tothis paper, intraclass correlation coefficients rangedfrom 0.53 to 0.67 for satisfactory performance,
RESULTS
Of the 1236 physicians surveyed, 272 were foundto be ineligible: 180 because they were not familyphysicians or general practitioners; 34 becausetheir year of graduation was before 1972 or after1988; and the rest because they moved and couldnot be located, were not practising in Ontario,were on maternity leave, were not in practice,were out of town for an extended period, or hadparticipated in the pretest. Usable responses were
obtained from 480 (50%) of the 964 eligiblephysicians, of whom 41 % were women, 7 1 %were in group practice, 63% were certificants ofthe College, and 870% were in fee-for-servicepractice. Respondents were substantially more
likely than nonrespondents to be certificants ofthe College (63.3% vs 43.7%; P<0.0001), butboth groups were similar in sex distribution anddecade of graduation.
Respondents' ratings of the importance ofmaneuvers and their perceptions of satisfactoryand self-assessed levels of performance are pre-sented in Table 2. In general, the level of cover-
age respondents considered satisfactory reflected
VOL. 42: SEPTEMBER * SEPTEMBRE 1996 *:. Canadian Familv Plasician * LeAidecin defamnille canadien 1695
RESEARCH
Preventive care and barriers to effective prevention
their perceptions of the importance of the inclusion) from the Canadian Task Force on themaneuver. With few exceptions, importance rat- Periodic Health Examination than for thoseings and perceived satisfactory levels of coverage with a Class C (poor evidence for inclusion) or
were higher for maneuvers with a Class A or B Class D (fair evidence for exclusion). The mostrecommendation (good or fair evidence for striking exception was annual digital rectal
Table 2. Perceptions of satisfactory versus self-assessed performance ofpreventive interventions
LEVEL OF % OF PATIENTS COVERED IMPORTANCE* RECOMMENDATIONMANEUVER COVERAGE <10 25 50 75 >90 N MEAN SD N OF CTFPH'
Pap smears at regular intervals for Satisfactory .4 .4 1.5women who are sexually active Self-assessed .4 1.8 13.1
..........................................................I..........................................................I........Annual mammography for women Satisfactory 35.5 10.6 18.050 to 59 years Self-assessed 41.9 16.6 20.6
..................................................I. .........................................................................Measurement of prostate-specific Satisfactory 31.8 13.4 19.3antigen at regular intervals in Self-assessed 50.4 16.1 17.3middle-aged and elderly men
*1 - not important, 5 - veiy important.tCanadian Task Force on the Periodic Health Examination. Assessment ofevidence supporting the recommendation that the condition be specifically considered in apenrodic health examination: A good evidence; B -fair evidence; C poor evidence, but recommendations may be made on other grounds; D -fair evidenceforexclusionfrom consideration in a penrodic health examination.+Cforgeneralpopulation, Dforpostmenopausal women.
1696 Canadian Family Physician Le AMidecin defamille canadien VOL 42: SEPTEMBER * SEPTEMBRE 1996
A
C
N/A
A
469 C
450
466
A
D
465 C
469 C
D
RESEARCH
Preventive care and barriers to effective prevention
examination (DRE) for men older than 50,which, despite only a Class C recommendation,received a mean importance rating of 4.4 on afive-point scale. Almost two thirds of respon-dents considered satisfactory coverage for DREto be 90% or more of men older than 50.
For every maneuver, fewer physicians believedthey were providing the intervention to at least90% of their patients than saw this level of cov-erage as satisfactory. For 10 of the 15 preventivemaneuvers, twice as many physicians thought90% or higher coverage was satisfactory thanthought they provided that level of coverage.The gap in the proportion of physicians report-ing satisfactory versus self-assessed coverage of90% or more exceeded 30% for six maneuvers(annual breast examination for women 50 to 59years, mammography every 2 years for women50 to 59 years, annual DRE for men over50 years, Pap smears at regular intervals for sex-ually active women, tetanus booster immuniza-tion every 10 years, and smoking cessationcounseling). The gap was evident both for inter-ventions supported by strong evidence and forthose that are not.
Table 3 presents the proportion of respondentswho thought their delivery of preventive maneu-vers was less than satisfactory. For 11 of the15 maneuvers, most respondents reported cover-age lower than the level they thought satisfactory.For six interventions (tetanus booster immuniza-tion, DRE, annual breast examination, pneumo-coccal vaccination for community-dwellingelderly, mammography at 2-year intervals, andtesticular examination), more than two thirdsreported less than satisfactory coverage.
Respondents' perceived reasons for lack ofsuccess in providing recommended preventivecare are presented in Table 4. Reasons werepatient-related, physician-related, patient- andphysician-related, and systems-related. Morethan two thirds agreed or strongly agreed withthe following statements: patient is healthy anddoes not visit; patient refuses, not interested, ordoes not comply; no effective systems to remindpatients to come in for preventive services; andpriority given to presenting problem. More than
one third endorsed the following statements: noeffective systems to remind physicians to providepreventive services, intervention not clearlyeffective, not enough time during patient visits,intervention causes patient discomfort or
Table 3. Proportion of physicians whothought their delivery of preventivemaneuvers was less than satisfactory
PERCEIVEDPERFORMANCELESS THAN
MANEUVER SATISFACTORY (%)
Tetanus booster immunization every10 years (n = 428)
Annual digital rectal examination formen older than 50 years (n = 429)
Annual breast examination for women50 to 59 years (n = 43 1)
Pneumococcal vaccination for community-dwelling patients older than 65 years (n = 416)
Mammography every 2 years forwomen 50 to 59 years (n = 408)
Testicular examination at regularintervals (n = 426)
75.5
71.3
71.0
69.0
68.4
68.1
Smoking cessation counseling (n = 432) 63.0
Pap smears at regular intervals for women 61.8who are sexually active (n = 434)
Measurement of prostate-specificantigen at regular intervals in middle-aged and elderly men (n = 421)
Testing stools for occult blood atreg-ular intervals for middle-aged andelderly adults (n = 421)
Blood pressure measurement at regularintervals for adults (n = 431)
Annual influenza immunization forpatients older than 65 years (n = 428)
Annual mammography for women50 to 59 years (n = 394)
Thyroid-stimulating hormonemeasurement at regular intervals (n = 422)
Chest x-ray examination at regularintervals (n = 415)
58.2
52.3
51.7
48.6
48.2
34.8
22.9
VOL 42. SEPTEMBER * SEPTEMBRE 1996 *+ Canadian Fanily Phlysician Le elidecin defamille canadien 1697
RESEARCH
Preventive care and barriers to effective prevention
inconvenience. Less than one third agreed withthe following: intervention is too expensive,intervention is not reimbursed, do not remem-ber to offer the service, and preventive careguidelines too complex to apply.
Table 4. Reasons for not providingrecommended preventive care
% OF RESPONDENTS
STRONGLYREASONS AGREE AGREE
PATIENT-RELATED
Patient is healthy and does notvisit (n = 453)
Patient refuses, is not interested,or does not comply (n = 455)
PHYSICIAN-RELATED
Do not remember to offerthe service (n = 450)
Preventive care guidelines toocomplex to apply (n = 438)
PATIENT- AND PHYSICIAN-RELATED
Priority given to presentingproblem (n = 447)
44.8 39.1
46.2
21.6
16.0
23.7
3.6
2.5
53.2 20.8
SYSTEMS-RELATED
Lack of effective systems toremind patienats to come in forpreventive services (n = 452)
Lack of effective systems toremind physicians to providepreventive setvices (n = 452)
Not eniough time duringpatient visit (n = 449)
INTERVENTION-RELATED
Intervention is not clearlyeffective (n = 448)
Intenrvention causes paticnt dis-comfort or inconvenience (n = 449)
Intervention is tooexpensive (n = 450)
Intervention is not
reimbursed (n = 447)
43.4 23.7
32.6 12.5
27.8 9.4
37.9 10.5
33.9 6.7
..........................7....25.8 4.7
10.4 4.7
DISCUSSION
Our results indicate a substantial differencebetween the level of preventive care physicianswant to provide and the level they perceive theydo provide. This gap could provide motivation forchange, creating a receptive climate for strategiesto enhance preventive care performance.
Physicians' perceptions of barriers to providingpreventive care could affect the strategies consid-ered. Many respondents appeared to relate diffi-culties in providing preventive care to the
tendency to view medical care as illness care;
more than three quarters saw the failure ofhealthy patients to visit regularly and prioritygiven to presenting problems as reasons for theirfailure to provide care. Another widespread per-
ception was that patients' lack of interest in or
refusal of preventive care is a serious problem.Many respondents saw a need for effective sys-
tems to remind patients and physicians about pre-
ventive services. Almost half saw uncertaintyabout the effectiveness of preventive interventionsas an issue, and a few saw the complexity of pre-
ventive care guidelines as a problem. Among theleast endorsed reasons was inadequate reimburse-ment. (Although few respondents perceived inad-equate reimbursement as a barrier to providingpreventive services, more than 80% agreed else-where in the questionnaire that time spent on
preventive care is inadequately reimbursed.)Given these findings, many of our respondents
might support changes in practice organization,
clinical records, and information systems to give
greater prominence to preventive and anticipa-tory care; initiatives to inform the public aboutpreventive care; and development and imple-mentation of effective reminder systems. Ourfindings provide less support for fee schedulemanipulation.Our findings are consistent with those of
Dietrich and Goldberg that importance ratingcorresponds to performance for most but not allpreventive maneuvers 16; with the findings ofAttarian and colleagues that physicians most
commonly cite lack of time, patients' lack of moti-vation, and patients' expectations as barriers to
1698 Canadian Family Plysician * Le edecin defanmille ainadien :- VOL012: SEPTEMBER * SEPTEMBRE 1996
RESEARCH
Preventive care and barriers to effective prevention
performing health promotion counseling'7; andwith the findings of McPhee and colleagues thatlack of time and patient dislike or refusal wereserious barriers to cancer screening.'8 Our find-ings diverge in the importance of reimbursement,which was the fourth most commonly cited barri-er in the North Carolina study, and physicianobjection to the tests and physician forgetfulness,which were identified as important barriers in thestudy by McPhee and colleagues.We acknowledge several limitations of this
study. Using the Canadian Medical Association'sPhysician Resource Databank as the samplingframe for this study was somewhat unsatisfactory.We anticipated that some specialist physicians,but not so many, would be mislabeled as familyphysicians or general practitioners.The response rate (50% of eligible physi-
cians) was lower than expected, despite aggres-sive follow-up of nonrespondents. The relativelylow response rate warrants caution in generaliz-ing our study results to the entire physician pop-ulation of the study area. We suspect thatrespondents are more likely to be preventionoriented than nonrespondents. If this were true,our results might overestimate importanceratings and satisfactory and self-assessed levelsof coverage.
Certificants of the College were overrepresent-ed among respondents. However, none of thethree Canadian studies that have examined therelationship between residency training and pre-ventive care performance has shown better per-formance by residency-trained physicians aftercontrolling for confounding variables, such as ageand practice location.1"6,22
Data in this study were obtained by physicianself-report and should, therefore, be viewed withcaution. Studies comparing self-report withobjective measures of performance of preventivecare suggest that physicians tend to overestimatetheir performance.23125 The gap between the levelof performance physicians consider satisfactoryand their actual performance might, therefore, beunderstated in our study. In a second phase of thestudy, we introduced unannounced standardizedpatients into the practices of randomly selected
consenting physicians in order to assess actualperformance of preventive care maneuvers.
CONCLUSION
Many physicians in south-central Ontario pro-vide preventive care to their patients at lower lev-els than they consider satisfactory. Theirperception of a gap between self-assessed cover-age and the level of coverage they regard as satis-factory could create fertile ground for initiativesto improve performance. They identify barriersto the successful provision of recommended pre-ventive services that point to possible strategiesfor improving preventive care delivery.
AcknowledgmentDr Hutchison is supported as a National Health ResearchScholar by Health Canada. Ms Abelson is supported by aHealth Research Personnel Development Program Fellowshipfrom the Ontario Ministry ofHealth.
Correspondence to: Dr B. G. Hutchison, Centre forHealth Economics and Policy Analysis, Health Sciences Centre,Room 3HIE, McMaster University, 122 Main St WKHamilton, ON L8N3Z5
References1. Borgiel A, WilliamsJL, Bass MJ, Dunn EV, Evensen MK,Lamont CT, et al. Quality of care in family practice: doesresidency training make a difference? Can MedAssocj1989; 140:1035-43.
2. Lurie N, Manning WG, Peterson C, Goldberg GA, PhelpsCA, Lillard L. Preventive care: do we practice what wepreach? Am 7 Public Health 1987;77:801-4.
3. Lewis CE. Disease prevention and health promotion prac-tices ofprimary care physicians in the United States.AmJ Prev Med 1988;4(Suppl 4):9-16.
4. Battista RN. Adult cancer prevention in primary care: pat-terns ofpractice in Quebec. AmjPlOic Health 1983;73:1036-9.
5. Battista RN, Palmer CS, Marchand BM, Spitzer WO.Patterns of preventive practice in New Brunswick. Can MedAssocJ 1985;132:1013-5.
6. Smith HE, Herbert CP Preventive practice among prima-ry care physicians in British Columbia: relation to recom-mendations of the Canadian Task Force on the PeriodicHealth Examination. Can MedAssocJ 1993; 149:1795-800.
VOL 42: SEPTEMBER *SEPTEMBRE 1996 * Canadian Family Physician Le Medecin defamille canadien 1699
RESEARCH0 0*a-** v* * ---
7. Rosser WW, McDowell I, Newell C. Useof reminders for preventive procedures infamily medicine. Can Aled Assoc J 1991;145:807-14.
8. ShankJC, Powell T, LlewelyriJ.A five-year demonstration project associ-ated with improvement in physicianhealth maintenance behaviour. Fam Med1989;21:273-8.
9. Cheney C, RamsdellJW Effect of med-ical records' checklists on implementationof periodic health measures. AmJ Med1987;83: 129-36.
10. McDonald CJ, Hui SL, Smith DM,Tierney WM, Cohen SJ, Weinberger M,et al. Reminders to physicians from anintrospective computer medical record:a two-year randomized trial. Ann InternMed 1984; 100: 1 30-8.
11. Tierney WM, Hui SL, McDonald CJ.Delayed feedback of physician perfor-mance versus immediate reminders toperform preventive care: effects on physi-cian compliance. Aled Care 1986;24:659-66.
12. KornJE, Schlossberg LA, Rich EC.Improved preventive care following anintervention during an ambulatory carerotation: carryover to a second setting.J Gen Intern Med 1988;3: 156-60.
13. Cohen DI, Littenberg B, Wetzel C,Neuhauser DvB. Improving physiciancompliance with preventive guidelines.Med Care 1982;20:1040-5.
14. Schreiner DT, Petrusa ER, Rettie CS,Kluge RM. Improving compliance withpreventive medicine procedures in a housestaff training program. South NMedJ 1988;81:1553-7.
15. Ornstein SM, Garr DR, Jenkins RG,Rust PF, Arnon A. Computer-generatedphysician and patient reminders: tools toimprove population adherence to selectedpreventive services. J Fam Pract 199 1;32:82-90.
16. Dietrich AJ, Goldberg H. Preventivecontent of adult primary care: do general-ists and subspecialists differ? AmJ PublicHealth 1984;74:223-7.
17. Attarian L, Fleming M, Barron P,Strecher V A comparison of health
promotion practices of general practitionersand residency trained physicians.J C(ommunioy Healthl 1987; 12:31-9.
18. McPhee SJ, Richard R, BirdJ,Solkowitz S,Jenkins C. Reasons physi-cians do not perform cancer screening.Clin Res 1985;33:727A.
19. T'he Canadian Task Force on thePeriodic Health Examination. TheCanadian guide to clinical preventive health care.Ottawa, Ont: Health Canada, 1994.
20. Ontario Cancer Treatment andResearch Foundation. Proposalfor a province-wide program of breast cancer screening.Toronto, Ont: Ontario Cancer Treatmentand Research Foundation, 1989.
21.Johnson KC, Ford DE, Smith GS.The current practices of internists inprevention of residential fire injury.Am J Prev Med 1993;9:39-44.
22. Maheux B, Beaudoin C,Jacques A,LambertJ, Levesque A. Effects of residen-cy training in family medicine v. intern-ship training on professional attitudes andpractice patterns. Can Med Assoc§J 1992;146:901-7.
23. Woo B, Woo B, Cook EF, Weisberg M,Goldman L. Screening procedures in theasymptomatic adult: comparison of physi-cians' recommendations, patient desires,published guidelines and actual practice.JAAIIA 1985;254: 1480-4.
24. McPhee SJ, Richard RJ, Solkowitz SN.Performance of cancer screening in a uni-versity general internal medicine practice:comparison with American CancerSociety guidelines. 7 Gen Intern Med 1986;1:275-81.
25. McPhee SJ, BirdJA,Jenkins NH,Fordham D. Promoting cancer screening:a randomized, controlled trial of threeinterventions. Arch Intern Aled 1989;149: 1866-72.
Canadian Libraryof Family Medicine
The Library Serviceof the College of FamilyPhysicians of Canada
Services of the LibraryFor Practising Physicians* Literature on
- advances in diagnosis and therapy- pharmaceutical information- practice management topics
* Medical literature searches* Copies of medical articles* Information on ordering books
and journalsFor Researchers inFamily Medicine* Literature searches and
bibliographies* Location of documents and studies* Literature on research methodologyFor Teachers ofFamily Medicine* Preparation of booklists on family
medicine topics* Advice on collecting literature
to support programs* Searches of medical literature
on topics related to trainingfamily doctors
Residents inFamily Medicine* Literature searches in support
of research projects* Advice on using libraries and
bibliographic resources
Charges forBibliographic Serces
Literature searches for CFPC Members:One free search per year, $8/search thereafter.Articles copied ..... ......... 20c/pageArticles faxed .... .......... 50c/pageBooks/Av loans ..... ......... No ChargeCLFM bibliographies.............. No ChargeGeneral reference .............. No ChargeLiterature searches for non-members
$25/search. Searches limited to family medicine topics.
To use services ofCLFM, contact the librarian at:Canadian Ubrary of Family MedicineCollege of Family Physicians of Canada2630 Skymark Avenue, Mississauga ONL4W5A4, Telephone: (905) 629-0900,Fax (905) 629-0893
OR
Canadian Library of Family medicine,Natural Sciences Centre, Rm. 1 70CUniversity of Western Ontario, London, ONN6A 5B7, Telephone: (519) 661-31 70,Fax: (519) 661-3880
1700 Canadian Famil Physician . L Ae'Pdecin defamille canadien * VolI 42: SEPTEMBER * SEPTEMBRE 1996