Doescher etLal., Preventive Blackwell O R I G I N Publishing, A AR T I CLtd. L E Care: Does Continuity Count?
Preventive Care Does Continuity Count? Mark P. Doescher, MD, MSPH, Barry G. Saver, MD, MPH, Kevin Fiscella, MD, MPH, Peter Franks, MD
OBJECTIVE: To examine the impact of provider continuity on preventive care among adults who have a regular site of care. DESIGN: Logistic regression analyses were conducted to explore whether continuity, categorized as having no regular care, site continuity, or provider continuity, was associated with receipt of 3 preventive care services (influenza vaccination, receipt of a mammogram, and smoking cessation advice), independent of predisposing, need, and enabling factors. PARTICIPANTS: This study examined 42,664 persons with private, Medicaid, Medicare, or no health insurance coverage who reported either having no site of care or being seen in a physician’s office, HMO, hospital outpatient department, or other health center. SETTING: The 1996/1997 Community Tracking Study (CTS) household survey, a telephone-based survey providing a crosssectional sample of 60,446 U.S. adults aged 18 and older representing the U.S. housed, noninstitutionalized population. MEASUREMENTS AND MAIN RESULTS: After adjustment for differences in predisposing, enabling, and need factors, site continuity was associated with significant increases of 10.4% in influenza vaccinations (P = .006) and 12.6% in mammography (P = .001), and a nonsignificant increase of 5.6% in smoking cessation advice (P = .13) compared to having no regular site of care. After adjustment for these factors, provider continuity was associated with an additional improvement of 6.0% in influenza vaccinations (P = .01) and 6.2% in mammography (P = .04), and a nonsignificant increase of 2.5% in smoking cessation advice (P = .30) compared to site continuity. CONCLUSIONS: Provider continuity and site continuity are independently associated with receipt of preventive services. Compared to having no regular site of care, having site continuity was associated with increased receipt of influenza vaccination and mammography and, compared to having site continuity, having provider continuity was associated with further increases in the receipt of these two preventive services. KEY WORDS: patient-physician relationship; continuity of care; satisfaction with care. J GEN INTERN MED 2004;19:632–637.
H
aving a sustained relationship with a primary care provider is considered to be a key component of pri1 mary care. Yet trends in health care delivery, including involuntary disenrollment from health insurance coverage, the emergence of managed care and larger physician groups, and the use of physician extenders, diminish 2,3 opportunities for provider continuity. Thus, research is needed to examine the impact of provider continuity on outcomes, such as the receipt of evidence-based preventive services. 4–7 While studies have reported that identifying a regular site of care is associated with receipt of preventive services, particularly for women and children, very few studies have examined the relationship between continuity 4,8,9 of care with a regular provider and preventive care. In the one study that used a nationally representative U.S. sample (1987) to evaluate the effect of provider continuity 4 beyond having a regular site of care, Lambrew et al. reported that having a regular physician offered no statistically significant advantage over having a “mainstream” site of care (physician’s office, clinic, or health maintenance organization) in increasing receipt of preventive care by women (clinical breast examinations, Pap smears, and mammograms) and children (measles, mumps, rubella [MMR] and polio vaccinations). We sought to revisit this question using a more recent, nationally representative sample. We used data from the Community Tracking Study (CTS) household survey of 1996/1997 to examine the independent associations of self-report of having a regular site of care and having a regular provider on receipt of 3 types of preventive services: influenza vaccinations, mammography, and tobacco cessation advice. Specifically, we hypothesized that compared to having no regular site of care, having a regular site would be associated with increased receipt of these preventive services and, compared to having a regular site of care, having provider continuity would be associated with additional gains in the receipt of these services.
METHODS Received from the Department of Family Medicine (MPD, BGS), University of Washington School of Medicine, Seattle, Wash; Departments of Family Medicine and Community and Preventive Medicine (KF), University of Rochester School of Medicine, Rochester, NY; and Department of Family Medicine (PF), Center for Health Services Research in Primary Care, University of California, Davis School of Medicine, Davis, Calif. Address correspondence and requests for reprints to Dr. Doescher: University of Washington, Box 354696, Seattle, WA 98195-4696 (e-mail:
[email protected]). 632
Data Source The Community Tracking Study Household Survey, a telephone survey of 60,446 individuals, was conducted in 1996/1997, representing the U.S. housed, noninstitu10 tionalized population. Sixty communities were randomly selected using stratified sampling with probability in proportion to population size in order to ensure representation of the U.S. population. Random-digit dialing was used to select most households; a small subsample of households
JGIM
Volume 19, June 2004
without phones was included in the sample by providing these respondents with cellular phones for the interviews. The survey included information regarding access to care, health care use, preventive care, satisfaction and health insurance, health status, and sociodemographic attributes. The final response for the CTS was 65%. The study sample for these analyses consisted of 42,664 adults aged 18 and over with private, Medicaid, Medicare, or no health insurance coverage who reported either having no site of care or being seen in a physician’s office, HMO, hospital outpatient department, or other health center.
Variables Primary Independent Variables. Continuity was categorized into 3 levels, as follows: 1. Category 1 (no regular care): included all persons who reported having no regular site of care. 2. Category 2 (site continuity): included all those who reported having a regular site of care (doctor’s office, HMO, a hospital outpatient department, or other health center) that they used when sick or in need of medical advice, but who did not report having a regular provider. 3. Category 3 (provider continuity): included all subjects who reported having site continuity and who additionally identified a regular provider at their site of care (“Do you usually see the same provider each time you go there?”). Secondary Independent Variables. Covariates were selected 11 based on the Andersen model of health care access. This model describes the numerous factors that influence access to health care and classifies access into predisposing factors, need factors, and enabling factors. This conceptual framework was selected because it includes characteristics of the population at risk, consideration of health policy, and utilization of health services. Predisposing factors were: age (18– 29, 30– 44, or 45 – 64 years); gender; race/ethnicity (non-Hispanic white, African American, primarily Englishspeaking Hispanic, primarily Spanish-speaking Hispanic, or Other); marital status; household size; community size (large metropolitan region of >200,000 population, small metropolitan region of
