Both physicians and patients highly value continuity in healthcare 12. The Institute of Medicine (IOM) holds that continuity, defined as an ongoing partnership between patients and physicians, is a central and important component of primary care 3. Continuity fosters personal relationships and is believed to improve health outcomes, although the latter has been difficult to demonstrate formally.

Continuity has been measured in various ways with different associated outcomes. Concepts of continuity include having a usual place for healthcare, having a usual physician and actually having a greater proportion of one's visits with a particular physician 456. Continuity reportedly increases use rates of some services, including some preventive services, and is associated with lower health care costs for certain populations 789101112. However, continuity has shown no effect on the use of some particular preventive services, including Pap smears and mammograms 413. Evidence for the relationship between continuity with hospitalization rates is conflicting 4101415. Lack of continuity care is associated with fewer follow-up care visits and fewer medical prescriptions given for chronic illness 16. Few studies, however, have examined outcomes for continuity as defined as having a longer standing relationship with one's physician 911. The outcomes of this notion of continuity are less clearly known.

Elements of trust and satisfaction, important to the ongoing patient-physician relationship, are also important to examine in the context of continuity. Continuity over time allows physicians to get to know and better understand their patients 17 and is associated with greater trust 18. In older patients, longer relationships are associated with the perception that their provider is more knowledgeable and thorough 11. Considering these aspects of satisfaction may further the understanding of the benefits of a continuous health care relationship 1920.

The purpose of this study is to examine the characteristics of rural adults who have longer versus shorter relationships with their current physicians. We also examine if the length of this relationship is associated with aspects of patient satisfaction and with the likelihood that patients have received recommended preventive services.

Participants were sorted into four groups based on the number of years they had received care from the same physician: less than a year, one to two years, three to five years and more than five years. Chi-square tests were used to compare the demographic, satisfaction and preventive health care characteristics of each group of years of continuous care to the group with over five-years of care.

Logistic regression was performed to assess the relationship of continuity to the outcomes of satisfaction and preventive service use rates. Dummy variables were constructed for the years subjects had received care from their primary source of care, with more than five years of care as the omitted (comparison) category. Models for each outcome were adjusted for subject age, gender, race, income, health insurance status and self-reported general health status.

For all analyses STATA 8 (College Station, Texas) 27 was used. When respondent demographics were compared against the 2000 U.S. Census data, survey participation rates were found to be lower for males, persons 18–39 years of age, African-Americans, and those with household incomes below $15,000. Analyses were accordingly weighted to adjust for both the over-sampling in small counties and to correct differential response likelihood by demographic groups.

Among the 3176 persons identifying a particular physician from whom they received their health care, 10.8% (N = 319) had seen the same individual for the past year, 11.8% (N = 369) for the past one to two years, 20.7% (N = 669) for the past three to five years and 56.7% (1819) had seen the same physician for more than five years. Associations between sociodemographic characteristics and duration of care are shown in Table 1. Compared to persons with one year or less continuity with the same physician, respondents with more than five years continuity were more often Caucasian (66.9% versus 61.5%, p = 0.007), more often had insurance (77.3% versus 68.5%, p = 0.001), had more education (43.6% versus 37.5% had at least some college, p = 0.006), more often had an income $25,000 or more (62.4% versus 45.6%, p < 0.001) and more often reported good to excellent health (76.0% versus 66.2%, p < 0.001). Respondents with more than five years continuity also more often had an income higher then $25,000 (62.4%) than the groups reporting only one to two years of continuity (50.1%, p < 0.001) and three to five years continuity (57.6%, p = 0.035). Otherwise, the one to two-year, three to five-year and over five-year continuity groups did not differ in their characteristics.

In terms of putative outcomes of continuity, compared to those with more than five years continuity, more respondents with one year or less of continuity reported being neutral or dissatisfied with their overall health care (10.6% versus 4.5%, p < 0.001), with the quality of their health care (6.7% versus 3.3%, p = 0.022) with having their health questions answered (7.2% versus 3.8%, p = 0.021) and with the concerns shown to them by their physicians (7.0% versus 3.4%, p = 0.004) (Table 2). A higher proportion of persons with one year or less continuity were also neutral or not confident in the abilities of their physician to help them than those with more than five years of continuity (20.1% versus 14.5%, p = 0.035). There were no significant differences in rates of any of the preventive services, except those who reported seeing a physician for one year or less were more likely to report being counseled in nutrition over the past year (57.8% versus 48.1%, p = 0.008).

Controlling for age, gender, race, income, insurance and health status, participants with one year or less and one to two years of continuity with their physician remained more likely to be neutral or dissatisfied with their overall health care compared to those with more than 5 years of continuity (OR 2.34; 95% CI: 1.39–3.93 and OR 1.78; 95% CI: 1.04–3.06, respectively) (Table 3). Those with one to two years of continuity were also more likely to be neutral or dissatisfied with the quality of health care they usually received (OR 2.37; 95% CI: 1.35–4.15). Those with one year or less continuity were also more likely to be neutral or dissatisfied with the concern shown by their physician (OR 1.90; 95% CI: 1.12–3.23) and having their questions answered (OR 1.98; 95% CI: 1.10–3.57). There were no significant differences across groups with various lengths of relationships with their physicians and their confidence in their physician, feeling welcome by office staff, reported rates of counseling for smoking and physical activity and receipt of preventive services (Table 4). Persons with shorter periods of continuity were more likely, however, to report having received nutrition counseling in the past year (OR 1.47; 95% CI: 1.08–2.02).

Continuity can be measured in a number of ways. When measured as the length of the patient-physician relationship, we find continuity varies with patient demographics and with elements of satisfaction. Over half of adults of this study's 150 rural communities report seeing the same physician for more than five years. Those who saw the same physician for less than five years tended to be nonwhite, without health insurance, less educated, more often report income of less than $25,000 and more often report fair to poor health status. This group was also less likely to be satisfied with their overall health care and its quality and with components of the doctor-patient relationship.

Patients and their physicians value continuity 28293031; in this study of rural participants, satisfaction appears associated with length of the continuity relationship, as similarly noted in other studies 111832. The elements of greater satisfaction found in persons with longer relationships with their physicians lend credence to the importance of the continuity relationship to patient outcomes. However, in this study, there appears to be a threshold effect at one or two years of continuity beyond which satisfaction does not rise significantly further.

Even a relationship length of two years is becoming difficult to maintain in the current U.S. healthcare system, especially in urban areas, with the pressures of competitive managed care plans which encourage patients and their employers to change health plans, and the growth of urgent care centers 33. In the 1996–97 Community Tracking Study household survey, 17% of privately insured persons changed their health plan during the year prior to the survey 34. Of those changing health plans, a little over half cited changes in their insurance as the reason for also changing their source of care. Increases in insurance premiums could contribute to further health provider switching.

Satisfaction, as an indicator of quality of care 28, has been found to affect other outcomes, including patient adherence to their physicians' recommendations 35. Mothers are more likely to follow a physician's treatment recommendations for their child if she feels the physician is friendly and understands the complaint 36. Patients with hypertension are more likely to adhere to treatment and have their blood pressure under control when the physician considers the patient an active participant in treatment 37.

Previous studies note the importance of having a usual source of care versus no source to the timely receipt of preventive services for younger adults 3839. However no differences in preventive care services were observed for older Americans in long-term relationships 9. In our study preventive services outcomes did not differ significantly by length of continuity as well. For preventive services, it appears that having a usual source of care is important but no additional benefit comes with having a longer-term relationship with that source of care.

Respondents with one year or less continuity, interestingly, were more likely to report receiving nutrition counseling than those in the five-year or more continuity category. Possibly more patients are asking about nutrition issues in the first year or are specifically changing doctors to discuss nutrition. Another possibility is that the developing familiarity with the patient in a long term relationship may be associated with less vigilance by the physician or less counseling in areas in which the patient may have initially shown resistance 40.

The cross-sectional design of this study limits our ability to attribute causation to the statistical relationships demonstrated or know the directions of any causal connections. We do not know if continuity results in higher trust and satisfaction or if the opposite is true. Our data are limited in that the outcomes of having received counseling and other preventive services were self-reported and may not accurately reflect care patients received. Telephone surveys also limit the population to those persons with a working phone, although weighting upwards for households with low incomes and minorities partially adjusts for this. The survey response rate of 51% was moderate but similar to other U.S. national telephone surveys 41. Response bias is a possibility. This study addressed a rural population statistically representative of the U.S. rural South, thus its findings may not apply to urban and other regions of the U.S. or to other countries. However, we know of no reason to expect that the association between longer-term doctor-patient relationships and satisfaction differs elsewhere.

Over half of this rural population has seen the same physician for more than five years. Longer continuity was significantly related to aspects of the patient-physician relationship, specifically people's satisfaction with and confidence in their physicians, but not with one's likelihood of receiving recommended preventive services. Fostering long-term relationships between patients and their physicians may help promote the outcome of greater patient satisfaction with care.

The author(s) declare that they have no competing interests.

KD conceived and designed the study, performed the analyses, and drafted the manuscript. EA participated in the design, interpretation and helped draft the manuscript. DP participated in the design, analyses and critical review of the manuscript. All authors read and approved the final manuscript.