Does the Content of Primary Care Visits Differ by the Racial Composition of Physicians’ Practices?

The American Journal of Medicine (2006) 119, 348-353

CLINICAL RESEARCH STUDY

Does the Content of Primary Care Visits Differ by the Racial Composition of Physicians’ Practices? Kevin Fiscella, MD, MPH,a Peter Franks, MDb a

Departments of Family Medicine and Community and Preventive Medicine, University of Rochester School of Medicine and Dentistry, Rochester, NY; bDepartment of Family and Community Medicine, Center for Health Services Research in Primary Care, University of California, Davis, Calif. ABSTRACT PURPOSE: The study compared the content of primary care visits between physicians with larger and smaller African American practices. METHODS: We compared the content of primary care adult visits between physicians with larger and smaller African American practices using data from the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey 1997 to 2002. Larger African American practice (⬎14% of primary care visits by African Americans) was empirically defined to conform with previous studies. RESULTS: Larger African American practices comprised 24% of all practices but accounted for 80% of all visits by African American patients. In adjusted analyses, physicians with larger African American practices were less likely to refer patients to specialists (adjusted odds ratio 0.77; 95% confidence internal 0.61-0.98) and marginally less likely to perform rectal examinations (adjusted odds ratio 0.84; 95% confidence interval 0.56-1.00) but were more likely to schedule a return appointment (adjusted odds ratio 1.25; 95% confidence interval, 1.02-1.52). However, there were no other significant differences in the reported content of the office visits. Results were not altered when different cutoffs for larger African American practices were used or when obstetricians-gynecologists were excluded from the analyses. CONCLUSION: There are few differences in the content of office visits between physicians with larger and smaller African American practices. © 2006 Elsevier Inc. All rights reserved. KEYWORDS: Primary Health Care; African Americans; Delivery of Health Care; Physician Practice Patterns

Provider quality may contribute to racial disparities in health care. African Americans are more likely to be hospitalized in facilities that provide lower quality medical treatment for acute myocardial infarction.1 Some studies show they are more likely to undergo cardiac surgery by surgeons and in hospitals with higher mortality rates2,3 and receive carotid endarterectomies in lower volume hospitals by less-experienced surgeons.4 Elderly African Americans are more likely to be enrolled in health maintenance orga-

This study was supported by the Agency for Healthcare Research and Quality R01 HS10910-02. Requests for reprints should be addressed to Kevin Fiscella, MD, MPH, 1381 South Ave, Rochester, NY 14620. E-mail address: [email protected]

0002-9343/$ -see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2005.08.040

nizations with lower quality ratings5 and to be admitted to lower quality nursing homes.6 Whether these findings extend to primary care is not clear. Some data show relatively few racial disparities measured at the visit level.7-9 Other data suggest that disparities may be driven by differences between physicians with larger and smaller African American practices.10 By using Medicare claims data, Bach et al10 showed that four-fifths of office-based visits of elderly African Americans are in the practices of one-fifth of physicians. Physicians with larger African American practices were less likely to be board certified, less likely to report that they are able to provide high-quality care to all their patients, and more likely to report more difficulty in obtaining access for their patients to high-quality spe-

Fiscella and Franks

Primary Care Visit Content

349

cialists, diagnostic imaging, and non-emergency admission to the hospital.10 In this study, we used primary care visit data to examine physician practice patterns, nationally, based on the racial composition of physicians’ practices. We compared the content of adult visits between physicians with larger and smaller African American practices.

METHODS Sample

racial disparities and those activities performed frequently enough in the surveys to generate stable estimates. These included advice-giving (exercise, diet, cholesterol, smoking cessation, breast health, and mental health counseling), prevention procedures (Pap test, mammogram in women aged more than 40 years, breast examination, pelvic examination, rectal examination, blood pressure check, CLINICAL SIGNIFICANCE vision examination, cholesterol test, and prostate specific antigen testing ● Based on nationally representative amin men aged more than 50 years), bulatory visits, we observed few differvisit duration, and follow-up plans ences in tasks and procedures commonly (specific follow-up and referral to performed by primary care physicians another physician).

Data for this study were derived from national samples of all priwith larger and smaller African American mary care visits from 1997 to practices. Independent Variables 2002. The National Ambulatory The primary independent variable ● These findings do not support the hyMedical Care Surveys (NAMCS) was the proportion of visits that is a nationally representative sampothesis that racial disparities in priinvolved African American paple of office visits, and the Namary care are largely driven by the tients. We selected a cutoff to tional Hospital Ambulatory Medsource of care. match as closely as possible that ical Care Surveys (NHAMCS) is a ● Our findings do not preclude the possireported by Bach et al,10 namely, nationally representative sample bility of unmeasured differences in qualthose physicians seeing 80% of all of hospital ambulatory clinic visAfrican American patients. On the ity not captured by these data. its. Data for these surveys were basis of empirical analysis, this derived from average of 20 enwas dichotomized as physician or counters for each physician or clinic practices with greater than clinic. The physician or his/her 14% or 14% or less of visits by African Americans. Sensidesignee completed a standardized form regarding content tivity analyses were conducted by using cutoffs at 10% and of the encounter immediately after each visit. Data included 18% of visits by African Americans and by excluding obage, sex, race, insurance status, up to 3 diagnoses assigned, stetricians/gynecologists from the sample. up to 6 medications prescribed, whether the patient had been seen before for the presenting problem or other problems, services provided, advice given, disposition, and visit duraCovariates tion. Not all questions were asked each year. Information on The following measures were used: age (years), gender, specialty (for NAMCS) or clinic specialty (NHAMCS), race (black or white), principal source of payment (private, rurality (standard metropolitan statistical area or not) and Medicaid, Medicare, worker’s compensation, self-pay, no geographic region was provided for each visit. Complete charge, other, or unknown), rurality (living in a metropolidetails about the surveys are available at http://www. tan statistical area or not), region of the country (Northeast, Midwest, South, or West), year of visit, whether the patient cdc.gov/nchs/about/major/ahcd/ahcd1.htm. had been seen before, case-mix (based on the number of The subset of encounters selected for this analysis were medications prescribed, and Ambulatory Diagnostic Groups visits by African American or white patients 18 years of age [ADGs] described below), and practice type-specialty (ofand older to primary care physicians and obstetrician-gynefice-based family physician or general practitioner, officecologists. For NAMCS, primary care physician encounters based internist, office-based obstetrician-gynecologist, genwere defined as visits to general internists, general practieral medicine clinic, and obstetrics-gynecology clinic). tioners, and family physicians who reported they were the primary care physician of the patient for the visit. For NHAMCS, primary care physician encounters were defined Case-Mix Adjustment as visits to the general medicine clinic (which includes Case-mix adjustment was based on the Ambulatory Care internal medicine and family practice clinics) where the Groups System11 using ADGs of the Ambulatory Care physicians reported they were the primary care physician Groups system and the number of medications prescribed. for the patient of the visit. Visits to obstetrician-gynecoloThe ADGs comprised 32 diagnostic and preventive clusters gists in NAMCS and obstetric-gynecologic clinics to which each International Classification of Diseases-9 (NHAMCS) were also included. Pregnancy-related visits, diagnostic and preventive code can be uniquely assigned. emergency department visits, and visits that were the result Each patient was assigned a series of dummy values (0, 1) of a referral were excluded. Elements were selected for for each ADG depending on the presence of up to 3 diagnoses assigned by the physician for the visit. ADGs have analysis on the basis of a review of the current literature on

350 previously been used to adjust for case-mix in NAMCS studies.12-15 Comorbidity measured in this way has been found to increase with patient age and distinguishes the type of primary care delivery site by morbidity burden.13 We also included an adjustment for the number of medications prescribed, both as an additional measure of morbidity16 and as a measure of competing demands;17 that is, we considered the number of medications prescribed as a proxy for the extent of issues competing for the attention of both physician and patient during the encounter. In this dataset, the number of medications prescribed correlated with both number of ADGs (r ⫽ 0.32) and age (r ⫽ 0.31).

Statistical Analyses NAMCS and NHAMCS use a complex survey design, involving the clustering of visits within physician or clinic, and the use of weights designed to yield population estimates for visits. Data were analyzed with STATA (Version 8.2, StataCorp, College Station, Tex) to adjust for the sampling design and weights. The weights on these public-use data adjust each physician-patient encounter according to its sampling probability and the probability of physician nonresponse to yield unbiased national estimates of annual total visits. The content of visits of physicians with larger African American practices were compared with those with smaller African American practices using bivariate statistics. The relationship between practice and visit characteristics was examined using logistic regression models, except visit duration, which was examined using linear regression. We adjusted for potential confounding by physician specialty and setting, whether the patient was seen before, patient sociodemographic characteristics (age, race, gender, insurance), rurality, region, year of visit, and case-mix (based on the number of medications prescribed and the ADGs coded). We also assessed for interactions between practice setting (NAMCS and NHAMCS) and race.

RESULTS The sample included 77 393 visits, 11.6% by African Americans, to primary care providers and clinics in 3838 settings. In this sample, with a cutting point of 14% or more of visits to the practice by African Americans, 24% of physician practices provided 80% of all the visits by African Americans. The demographic characteristics of patients and visit characteristics of the 2 groups are shown in Table 1. Visits to physicians serving more African American patients were more likely to involve female patients, patients with Medicaid insurance, and those living in nonrural areas and the South. There were few statistically significant differences by patient race in any of the unadjusted visit characteristics. Visits to physicians serving more African American patients were slightly less likely to involve a rectal examination or

The American Journal of Medicine, Vol 119, No 4, April 2006 mental health counseling and more likely to result in a return appointment. Comparisons of the content of visits between physicians with larger (⬎14%) and smaller (ⱕ14%) African American practices, after adjustment for patient age, race, insurance, previous visit and comorbidity, physician specialty-practice type, region of the country, community size, and year of the survey, are shown in Table 2. Physicians with larger African American practices were significantly less likely to refer patients to specialists and marginally less likely to perform rectal examinations, but were more likely to schedule a return appointment. Notably, there were no other statistically significant differences in the reported content of the office visit. Use of alternative cutoffs for the racial composition of physicians’ practices (Table 2) or exclusion of obstetricians/gynecologists from the sample (results not shown), did not alter the finding of little association between practice composition and visit content. In separate analyses of visit content, there were no significant interactions between the percentage of African American patients seen in practice and race (African American patient vs white) or setting type (hospital clinic vs office).

DISCUSSION These analyses of nationally representative data from adult primary care office and hospital clinic visits are notable in several respects. First, we extended previous findings for Medicare beneficiaries to all adult patients showing that a relatively small proportion of providers cared for most African American patients.10 Approximately 80% of all primary care visits for African Americans were made to approximately 24% of providers. These findings likely reflect a combination of continuing residential racial segregation, as well as racial differences in presence and type of insurance.18,19 We found that African Americans were more likely to be seen in hospital clinics than were whites, which is consistent with previous findings.13 Although we observed few differences in the content of visits between hospital clinics and office practices, provider continuity of care is typically less within hospital clinics.20 Provider continuity has been linked to improved patient outcomes.21-23 Conceivably, reduced continuity for African Americans (because of differences in site of care) could contribute to racial disparities in outcomes, but this question cannot be addressed with these data. Second, we found that physicians in larger African American practices were less likely to refer patients for specialty care. This finding was consistent with previous work showing that these physicians reported more difficulty accessing specialty care.10 Lower African American referral rates likely contributed to disparities in specialty procedures.24-26 We also found that these physicians were more likely to schedule patients for a return appointment. Quite possibly, this reflected the need for these physicians to follow up on problems that otherwise would have been

Fiscella and Franks Table 1

Primary Care Visit Content

351

Characteristics of Patients in the 2 Settings (Offices with ⱕ 14% Patients vs ⬎ 14% African American)

Characteristic

Age Male African American race Medicaid insurance Rural practice Region of practice Northeast Midwest South West Hospital clinic No. medications prescribed No. ADGs Pap screening Mammography screening PSA testing Cholesterol testing Blood pressure check Clinical breast examination Pelvic examination Rectal examination Vision screening Diet counseling Exercise counseling Breast self-examination teaching Smoking advice Mental health counseling Return appointment scheduled Patient referral made Time with MD

ⱕ14% African American

⬎14% African American

N

%/Mean (SD)

N

%/Mean (SD)

45469 45469 45469 45469 45469

52.5 (18.7) 34.2 3.0 4.8 22.8

31924 31924 31924 31924 31924

51.2 (18.8)y 29.8* 38.8* 12.3* 11.7*

10841 12341 13024 9293 45469 45469 45469 31574 19435 7908 45469 45469 18983 18983 28447 28447 45469 45469 18983 45469 45469 31302 14280 38308

22.2 26.5 29.3 21.8 5.7 1.8 (1.8) 1.6 (0.7) 14.4 8.6 0.0 8.7 75.7 18.0 21.3 8.2 2.1 1.9 1.4 7.1 4.3 3.6 5.8 6.1 17.2 (10.6)

9515 8712 11162 2535 31924 31924 31924 24319 12835 4027 31924 31924 13584 13584 18642 18642 31924 31924 13584 31924 31924 22392 9110 20770

22.0 20.2 50.8 7.0 14.0* 1.9 (1.8) 1.5 (0.7) 15.1 8.5 9.1 8.7 76.3 16.7 20.7 6.0* 1.7 1.9 1.1 5.6 3.5 2.4* 6.6* 7.0 18.2 (10.8)

SD ⫽ Standard deviation; ADG ⫽ Ambulatory Diagnostic Group; PSA ⫽ prostate specific antigen. N is number of observations; percentage and means are weighted to reflect population values. *Difference between 2 settings is significant P ⬍ .05.

managed by a specialist. It also may have reflected incomplete adjustment for greater severity of illness among African Americans in these settings, requiring closer follow-up or chance finding. Notably, we were not able to assess whether greater return appointments for African Americans translated into more return visits. Third and most importantly, we observed few differences in the content of visits based on the proportion of African Americans patients in the physician’s practice. Only 1 of 16 office procedures/interventions approached statistical significance. Physicians caring for more African Americans performed fewer rectal examinations. However, this finding is of dubious importance given its marginal statistical significance, the multiple tests conducted, and the unproven benefit of rectal examinations.27,28 The finding of few differences in the content of adult primary care visits based on the racial composition of the practice militates against the hypothesis that physicians caring for African American patients provide inferior care based on commonly performed procedures. Although these data did not allow for direct assessment of quality, the provision of Pap smears, mammography, smoking cessation, and other behavioral coun-

seling are increasingly recognized as indicators of primary care quality. These findings suggest that differential quality between primary care providers may contribute less than presumed to disparities in primary care measures. Note that our main findings apply to the average patient seen by physicians with larger or smaller African American practices; that is, the findings do not fully address the extent to which the care provided to African American and white patients within practices is different. Although we examined the interaction between race and percentage of African American patients seen and found no significant effect, our power to detect a difference was modest. A previous analysis of NAMCS data showed a few differences in content of care by race. African Americans had lower odds of receiving a Pap test, rectal examination, smoking cessation counseling, and mental health advice but had higher odds for visual screening, weight advice counseling, and receiving a follow-up appointment.7 Our findings are strengthened by use of a nationally representative sample of adult primary care visits whose content was reported by the physician immediately after the visit. Use of these data allowed us to avoid potential con-

352 Table 2

The American Journal of Medicine, Vol 119, No 4, April 2006 Adjusted Relationship between the Racial Composition of the Practice and Visit Content

Care Element

Pap smear screening Mammography screening PSA testing Cholesterol testing Blood pressure check Clinical breast examination Pelvic examination Rectal examination Vision screening Diet counseling Exercise counseling Breast self-examination teaching Smoking advice Mental health counseling Return appointment scheduled Patient referral made Time with MD

⬎14% vs ⱕ14%

⬎10% vs ⱕ10%

⬎18% vs ⱕ18%

Effect

95% Confidence Interval

Effect

95% Confidence Interval

Effect

95% Confidence Interval

1.03 1.03 1.00 0.97 1.05 0.85 0.87 0.75 0.79 0.94 0.81 0.80 0.86 0.75 1.25 0.77 0.69

0.84-1.27 0.78-1.37 0.76-1.33 0.78-1.20 0.85-1.30 0.67-1.09 0.67-1.13 0.56-1.00* 0.39-1.60 0.78-1.14 0.65-1.01 0.53-1.23 0.65-1.13 0.50-1.1 1.03-1.53* 0.61-0.98* ⫺0.10-1.49

1.14 1.02 0.96 1.11 0.96 0.96 1.01 1.05 0.72 0.91 0.79 1.05 0.88 0.75 1.23 0.89 0.054

0.94-1.37 0.80-1.32 0.74-1.24 0.93-1.34 0.78-1.19 0.77-1.19 0.81-1.28 0.58-1.00* 0.39-1.31 0.77-1.08 0.64-0.96* 0.71-1.57 0.67-1.14 0.70-1.31 1.02-1.48* 0.70-1.13 ⫺0.20-1.27

0.96 0.99 0.84 1.03 0.98 0.80 0.89 0.84 0.79 0.95 0.91 0.74 0.84 0.81 1.33 0.84 0.87

0.76-1.22 0.71-1.37 0.62-1.13 0.81-1.30 0.77-1.24 0.61-1.05 0.66-1.20 0.52-0.93* 0.37-1.88 0.77-1.18 0.71-1.17 0.47-1.18 0.62-1.13 0.51-1.27 1.05-1.68* 0.65-1.10 0.10-1.73*

PSA ⫽ prostate specific antigen. Effect⫹ is adjusted odds ratio, except for “Time with MD” where effect is parameter estimate in minutes. All analyses adjusted for age, gender, race, rural versus urban, region, insurance category, specialty, setting (NAMCS vs NHAMCS), number of medications prescribed, dummy variables for each Ambulatory Diagnostic Group (ADG), clustering of observations within physician practice, and population weights. *Significant, P ⬍ .05.

founding by patient adherence or selection bias associated with use of billing data. Our findings were further strengthened by control for multiple patient, physician, and practice factors. Limitations of our findings merit comment. Despite aggregation of NAMCS and NHAMCS across multiple years, our power to detect modest effects was limited as shown by the confidence intervals surrounding our estimates. Thus, we could not exclude the possibility of small effects. Our data were based entirely on physician report. However, comparison of NAMCS with direct observation of office visits showed high specificity (range 90%-99%) for office procedures/counseling but moderate to low sensitivity (range 0.12-0.84).29 More importantly, it seemed unlikely that the reliability of the physicians’ reports would differ based upon the racial composition of the physicians’ practices. The reliability of coding of patient race in NAMCS/ NHAMCS has not been formally assessed to our knowledge. Any measurement error would likely bias results toward the null. NAMCS/NHAMCS allowed for physician coding of up to 3 diagnoses, thus limiting control for comorbidity. For this reason, we also controlled for the number of medications prescribed. Nonetheless, our analyses likely failed to fully account for the higher age-adjusted morbidity of African Americans.30 When confronted with greater patient morbidity, physicians spend less time on prevention17 or smoking cessation counseling,31 conduct less depression screening,32 and are less likely to recommend preventive services such as mammography or hormone-replacement

therapy.33 Failure to fully control for unmeasured morbidity related to race would likely bias results against physicians with larger African American practices. NAMCS/NHAMCS consists exclusively of cross-sectional views of patient encounters. Visit-based analyses do not account for patients’ prior visits. However, African Americans make fewer primary care visits than whites.34 This means that physicians must attend to more tasks, including preventive services, at any one visit with African Americans. In theory, this could result in higher rates of preventive care for African Americans at any one visit; although with increased competing demands, prevention may receive a lower priority.17 Thus, the net effect of less frequent visits by African Americans on comparisons between physicians with larger and smaller African American practices is uncertain. We were able to control for only a few practice level variables. Lower proportion of practice revenue derived from Medicaid, medium practice size, physician graduation from a US or Canadian medical school, and access to information technology to generate preventive care reminders or treatment guidelines were associated with improved delivery of preventive services.35 However, control for proportion of visits by patients with Medicaid, geography, and physician specialty did not alter our findings of little difference in the content of visit based on race. Last, we excluded emergency department visits from the analyses because most of them do not provide primary or preventive care. African Americans have higher rates of emergency department visits than whites.36 Community

Fiscella and Franks

Primary Care Visit Content

health centers are theoretically within the scope of NAMCS, but it is not possible to ascertain to what extent these facilities are actually represented within the survey. In conclusion, we confirmed that African Americans are seen by a small proportion of primary care physicians and that physicians working in practices with more African Americans are less likely to refer patients to specialists but more likely to schedule follow-up visits. However, we found little evidence that the content of adult primary care visits differs between providers according to the racial composition of their practices.

References 1. Barnato AE, Lucas FL, Staiger D, Wennberg DE, Chandra A. Hospital-level racial disparities in acute myocardial infarction treatment and outcomes. Med Care. 2005;43:308-319. 2. Mukamel DB, Murthy AS, Weimer DL. Racial differences in access to high-quality cardiac surgeons. Am J Public Health. 2000;90:17741777. 3. Konety SH, Vaughan Sarrazin MS, Rosenthal GE. Patient and hospital differences underlying racial variation in outcomes after coronary artery bypass graft surgery. Circulation. 2005;111:1210-1216. 4. Dardik A, Bowman HM, Gordon TA, Hsieh G, Perler BA. Impact of race on the outcome of carotid endarterectomy: a population-based analysis of 9,842 recent elective procedures. Ann Surg. 2000;232:704709. 5. Schneider EC, Zaslavsky AM, Epstein AM. Racial disparities in the quality of care for enrollees in Medicare managed care. JAMA. 2002; 287:1288-1294. 6. Grabowski DC. The admission of blacks to high-deficiency nursing homes. Med Care. 2004;42:456-464. 7. Franks P, Fiscella K, Meldrum S. Racial disparities in the content of primary care office visits. J Gen Intern Med. 2005;20:599-603. 8. Ma J, Stafford RS. Quality of US outpatient care: temporal changes and racial/ethnic disparities. Arch Intern Med. 2005;165:1354-1361. 9. Williams RL, Flocke SA, Stange KC. Race and preventive services delivery among black patients and white patients seen in primary care. Med Care. 2001;39:1260-1267. 10. Bach PB, Pham HH, Schrag D, Tate RC, Hargraves JL. Primary care physicians who treat blacks and whites. N Engl J Med. 2004;351:575584. 11. Weiner JP, Starfield BH, Steinwachs DM, Mumford LM. Development and application of a population-oriented measure of ambulatory care case-mix. Med Care. 1991;29:452-472. 12. Blumenthal D, Causino N, Chang YC, et al. The duration of ambulatory visits to physicians. J Fam Pract. 1999;48:264-271. 13. Forrest CB, Whelan EM. Primary care safety-net delivery sites in the United States: a comparison of community health centers, hospital outpatient departments, and physicians’ offices. JAMA. 2000;284: 2077-2083. 14. Forrest CB, Reid RJ. Prevalence of health problems and primary care physicians’ specialty referral decisions. J Fam Pract. 2001;50:427432. 15. Franks P, Bertakis KD. Physician gender, patient gender, and primary care. J Womens Health (Larchmt). 2003;12:73-80. 16. Perkins AJ, Kroenke K, Unutzer J, et al. Common comorbidity scales were similar in their ability to predict health care costs and mortality. J Clin Epidemiol. 2004;57:1040-1048.

353 17. Chernof BA, Sherman SE, Lanto AB, Lee ML, Yano EM, Rubenstein LV. Health habit counseling amidst competing demands: effects of patient health habits and visit characteristics. Med Care. 1999;37:738747. 18. Williams DR, Collins C. Racial residential segregation: a fundamental cause of racial disparities in health. Public Health Rep. 2001;116:404416. 19. Shi L, Stevens GD. Vulnerable Populations in the United States. San Francisco, Calif: Jossey-Bass; 2005. 20. Doescher MP, Saver BG, Fiscella K, Franks P. Racial/ethnic inequities in continuity and site of care: location, location, location. Health Serv Res. 2001;36:78-89. 21. Gill JM, Mainous AG, III. The role of provider continuity in preventing hospitalizations. Arch Fam Med. 1998;7:352-357. 22. Saultz JW, Albedaiwi W. Interpersonal continuity of care and patient satisfaction: a critical review. Ann Fam Med. 2004;2:445-451. 23. Doescher MP, Saver BG, Fiscella K, Franks P. Preventive care. J Gen Intern Med. 2004;19:632-637. 24. Laveist TA, Arthur M, Morgan A, Plantholt S, Rubinstein M. Explaining racial differences in receipt of coronary angiography: the role of physician referral and physician specialty. Med Care Res Rev. 2003; 60:453-467. 25. Earle CC, Neumann PJ, Gelber RD, Weinstein MC, Weeks JC. Impact of referral patterns on the use of chemotherapy for lung cancer. J Clin Oncol. 2002;20:1786-1792. 26. Ayanian JZ, Landrum MB, Guadagnoli E, Gaccione P. Specialty of ambulatory care physicians and mortality among elderly patients after myocardial infarction. N Engl J Med. 2002;347:1678-1686. 27. Collins JF, Lieberman DA, Durbin TE, Weiss DG. Accuracy of screening for fecal occult blood on a single stool sample obtained by digital rectal examination: a comparison with recommended sampling practice. Ann Intern Med. 2005;142:81-85. 28. U.S.Preventive Services Task Force. Guide to Clinical Preventive Services, 3rd Edition: Periodic Editions. 3rd ed. Rockville, Md: Agency for Healthcare Research and Quality; 2003. 29. Gilchrist VJ, Stange KC, Flocke SA, McCord G, Bourguet CC. A comparison of the National Ambulatory Medical Care Survey (NAMCS) measurement approach with direct observation of outpatient visits. Med Care. 2004;42:276-280. 30. Lethbridge-Cejku M, Schiller JS, Bernadel L. Summary health statistics for U.S. adults: National Health Interview Survey, 2002. Vital & Health Statistics-Series 10: Data From the National Health Survey 2004;1-151. 31. Jaen CR, Stange KC, Tumiel LM, Nutting P. Missed opportunities for prevention: smoking cessation counseling and the competing demands of practice. J Fam Pract. 1997;45:348-354. 32. Nutting PA, Rost K, Smith J, Werner JJ, Elliot C. Competing demands from physical problems: effect on initiating and completing depression care over 6 months. Arch Fam Med. 2000;9:1059-1064. 33. Nutting PA, Baier M, Werner JJ, Cutter G, Conry C, Stewart L. Competing demands in the office visit: what influences mammography recommendations? J Am Board Fam Pract. 2001;14:352-361. 34. Greenblatt JGreenblatt J. Statistical Brief #14. Trends in Access to Routine Care and Experiences with Care: 2001. Rockville, Md: Agency for Healthcare Research and Quality; 2003. 35. Pham HH, Schrag D, Hargraves JL, Bach PB. Delivery of preventive services to older adults by primary care physicians. JAMA. 2005;294: 473-481. 36. Burt CW, Schappert SM. Ambulatory care visits to physician offices, hospital outpatient departments, and emergency departments: United States, 1999-2000. Vital & Health Statistics-Series 13: Data From the National Health Survey 2004;1-70.