Logo of annfammedLink to Publisher's site
Ann Fam Med. 2004 Sep; 2(5): 455–461.
PMCID: PMC1466710

Physician-Patient Relationship and Medication Compliance: A Primary Care Investigation

Ngaire Kerse, PhD, MBChB,1 Stephen Buetow, PhD,1 Arch G. Mainous, III, PhD,2 Gregory Young,1 Gregor Coster, MSc, MBChB,1 and Bruce Arroll, PhD, MBChB1


PURPOSE We assessed the relationship between 4 attributes of the physician-patient relationship and medication compliance.

METHODS We conducted a waiting room survey of patients consulting 22 general practitioners in 14 randomly selected practices in Auckland, New Zealand (81% response rate). A total of 370 consecutive patients (75% response rate) completed survey instruments about 4 attributes of the physician-patient relationship. Continuity of care (assessed from use of a usual physician, length of continuity, and perceived importance of continuity) and trust in the physician were ascertained before the consultation. After the consultation the Patient Enablement Index measured the physician’s ability to enable patients in self-care, and concordance between the patient and physician was measured by a 6-item inventory of perceived agreement about the presenting problem and management, were ascertained immediately after the consultation. Compliance with prescribed medication therapy was ascertained by telephone follow-up 4 days after the consultation.

RESULTS Overall, 220 patients (61%) received a prescription, and 79% of these patients were taking the medication at follow-up. In a univariate analysis adjusted for clustering, only trust and physician-patient concordance were significantly related to compliance. In analysis further adjusted for health and demographic factors, physician-patient concordance was independently related to compliance (odds ratio = 1.34, 95% confidence interval, 1.04–1.72).

CONCLUSIONS Primary care consultations with higher levels of patient-reported physician-patient concordance were associated with one-third greater medication compliance. An emphasis on understanding and facilitating agreement between physician and patient may benefit outcomes in primary care.

Keywords: Physician-patient relations, primary health care, office visits, patient compliance, prescriptions, drug, drug therapy


The physician-patient relationship is integral to the successful delivery of primary health care.1 The quintessential character of the physician-patient relationship, however, has been difficult to define, especially in the face of mounting societal pressures for fiscal constraint and population-based health care.2

In any consultation several attributes of the physician-patient relationship may affect the outcome, including a longitudinal relationship between patient and physician (continuity of care)3; concordance, or agreement, between the patient and physician on the problem or need and its management4; patients’ trust in their physician to act in their best interest; and the ability of the physician to enable the patient toward effective self-care (patient enablement).5 Which of these aspects is most closely associated with outcomes of primary health care has not been established.

Continuity of care between patients and physicians has been associated with improved health services use and preventive care,3,6 the physician’s understanding of the psychosocial aspects of patient care,7 and satisfaction with care.8 As continuity appears to be a general construct under which a variety of concepts fit,9 several tools have been developed to measure different aspects of continuity. Which particular aspect confers most benefit is unknown.

The level of agreement or concordance between physician and patient about the nature of the patient’s problem is associated with resolution of nonspecific signs and symptoms.10 Less rigorous methods show an association between concordance and patient adherence to management and medication plans.4,11,12 Patients are less likely to return for visits when there is a lack of physician-patient concordance,13 although the relationship is inconsistent.14,15 Trust is said to be essential to the medical consultation and is related to continuity and patient satisfaction.16,17 The ability of the physician to enable the patient toward effective self-care is a marker of consultation quality.5

Medication prescribing is a core component of medical care, and patient compliance with recommendations to take medications varies.18 In this study, we sought to establish potential associations between attributes of the physician-patient relationship and medication compliance. We report an investigation measuring self-reported continuity and trust before the consultation, and physician-patient concordance and enablement immediately after the consultation, and we then relate these attributes to subsequent medication compliance.



We asked general practitioners, randomly selected from the Auckland metropolitan area of New Zealand, for permission to approach consecutive patients or those responsible for the patients in their practice waiting room. Patients and caretakers who were bringing a child or adult under their care were approached in the waiting room and then invited to give written informed consent to participate. Respondents completed survey instruments before and immediately after their consultation. They were telephoned 4 days later to establish compliance with medication therapy. We made 4 attempts at telephone contact at differing times of the day for a 2-day period. The Auckland Ethics Committee approved the study.

Survey Instrument

The survey instrument we used has been used in an international study of continuity and trust16,17 and was adapted for a New Zealand audience. The instrument asked about standard demographic and health information, and contained several validated tools and a physician-patient concordance measure.

Measures: Survey Part 1

Part 1 of the survey was completed before the consultation and included the following components.

Health and Demographic Information

For ethnicity, we used patients’ self-described ethnicity in response to the 1996 New Zealand census question. Education was based on high-school completion, and perceived economic status was measured with the following question: “Thinking of your money situation, would you say that your household (choose one): can’t make ends meet; has just enough to get along on; is comfortable.” In addition, we used eligibility for the community services (low-income assistance) card, as a marker of socioeconomic status. Health status was estimated from the presence of a chronic illness and from scores on the EuroQol 5-dimensional questionnaire (EuroQol-5D),19 a 5-item health-related quality-of-life measure.

Source of Care

We ascertained source of care by asking, “Is there one particular surgery that you/the patient usually go to if you are sick or need advice about your health? (For example, a general practice.)” and “If yes, is there a particular doctor you/the patient usually see at this surgery?” This approach has been validated with the Usual Provider Continuity Index20 and shows patients’ ability to identify with a single source of care.

Continuity of Care

We assessed continuity of care with 3 measures. Length of continuity was established by asking, “How long have you/the patient been seeing this doctor?” This approach has been validated with other continuity measures.21 The importance of continuity was assessed by asking, “How important is it to you to see the same GP every time you have a health problem?” with responses recorded on a Likert scale ranging from 1 (not important at all) to 5 (very important).22 Finally, we calculated the Usual Provider Continuity Index23—the proportion of consultations with the usual physician—one of the most widely used measures of continuity.9

Trust in the Physician

We measured trust with the Trust in Physician Scale,24 which yields a score ranging from 11 to 55, with higher scores indicating more trust. Because we measured trust before the consultation, and this measure was intended to assess established trust (and the capacity to develop it), only patients who reported having a usual general practitioner answered this section.

Measures: Survey Part 2

Part 2 of the survey was completed immediately after the consultation and included the following components.

Physician-Patient Concordance

We assessed agreement between physician and patient with the following questions:

  1. “To what extent do you think the doctor understands why you came in today?”
  2. “How well do you think the doctor understood you today?”
  3. “To what extent did you and the doctor agree about the main problem or need today?”
  4. “To what extent did you and the doctor agree about what to do about the problem or need?”
  5. “To what extent do you and the doctor agree on what part you play in making decisions about health?”
  6. “To what extent do you and the doctor agree on who is responsible for different aspects of care?” The response to each question was recorded on a Likert scale ranging from 1 (completely) to 4 (not at all). We dichotomized the results to 1 (completely) and 0 (all other responses). Results for the 6 questions were then summed to give a cumulative score between 0 and 6, with higher scores indicating greater concordance. We calculated the Cronbach α for internal consistence of the scale and report the distribution of scores.

Patient Enablement

We used the Enablement Index5 to assess whether the physician enabled the patient toward effective self-care. This index, validated in primary care against patient satisfaction, asks whether patients are more or less able to cope with life, understand and cope with their illness, and help themselves as a result of the consultation with the physician. The responses are scored from 1 to 4, with higher scores meaning more enablement.

Measures: Medication Compliance

To assess medication compliance, we first asked the patients whether they received a prescription on the day of the consultation. Four days later, we telephoned patients and asked whether they had filled the prescription, were taking the medication, and if not, why not. Patients were classified as compliant if they had received a prescription and were taking the medication at the 4-day follow-up or if they were not taking it at follow-up because it had been a repeated prescription and they already had enough medication. If they were not taking the prescribed medication at follow-up for any other reason, they were classified as noncompliant. Those who had not received a prescription during the consultation were not included in the compliance analysis.


We produced statistics describing the sample and summary responses to the survey with SPSS v11 (SPSS Inc, Chicago, Ill). Intraclass correlations were computed for compliance within practices from a 1-way analysis of variance using Stata 7.0 (Stata Corp, College Station, Tex).

To examine relationships between consultation characteristics and the dichotomous outcome of compliance, we performed logistic regression analysis. As observations on patients were not independent within groups (practices) but were independent across groups, we adjusted for clustering with Stata 7.0. Multivariable analyses included health status, age, ethnicity, and education as control variables, and the results are reported as odds ratios. We also assessed for interaction effects on medication compliance. Values were coded as missing if follow-up with the respondent had not been possible. We omitted from the logistic regression model cases that were missing values for any of the variables.


Twenty-two physicians of 27 approached in 14 surgeries were enrolled in the study. Of 490 consecutive patients, 370 were enrolled (75.5%). The number of patients attending on the recruitment days ranged from 5 to 40. On average, 16 patients were enrolled from each physician (range, 2–34 patients per physician; SD, 9.5). Figure 1 shows the disposition of patients and the number included in the multivariable analysis. The 59 patients who could not be followed up by telephone were indistinguishable from those that were by age (41.4 vs 42.1 years, P = .79), sex (62% vs 70% female, P = .25), and presence of a chronic disease (33% vs 38%, P = .52).

Figure 1.
Flow of primary care patients in a study about medication compliance.

Table 1 shows the demographic and health status of the sample. Three quarters of the patient sample consulted the physician for themselves. The remainder were bringing a child or adult under their care. About one third of respondents (30.9%) had a major illness, and the average number of consultations in the previous year was 5.3 (SD, 6.7 consultations).

Table 1.
Demographic Characteristics of a Sample of Primary Care Patients Gathered in Waiting Rooms (N = 370)

Table 2 shows that most participants had a regular source of care and that most patients perceived continuity of care to be quite or very important. Reported concordance between physician and patient was high when considering individual agreement items. The 6-item scale was internally consistent (α = 0.87), and scores ranged from low (1–3) to medium (3–4) to high (5), with 23%, 20%, and 57% of respondents in each range, respectively.

Table 2.
Consultation Characteristics of a Sample of Primary Care Patients Gathered in Waiting Rooms (N = 370)

Most respondents (61.3%) received a prescription for medication, and 78.6% of those who could be reached by telephone took this medication. Compliance with medications within practices yielded an intraclass correlation of .07 (SE, .06). The 150 patients not receiving a prescription were indistinguishable from those receiving a prescription in age (t = 0.94, P = .34) and sex (χ2 = 1.3, P = .71) but were less likely to have a chronic disease (26% vs 36%, respectively; χ2 = 4.3, df = 1, P = .037).

Table 3 shows univariate and multivariable associations between the outcome variable—medication compliance—and the consultation variables. All analyses control for the effect of the clustered design. The trust score was associated with medication compliance in univariate analysis, but this association lost statistical significance after further adjustment. In multivariable logistic regression analysis, physician-patient concordance was independently related to medication compliance. Source of care was also independently related to medication compliance, but the association was not consistent, as in univariate analysis there was no significant relationship identified. We found an interaction effect between patient’s sex and physician-patient concordance; specifically, the association of physician-patient concordance with compliance was stronger for women.

Table 3.
Relationship Between Consultation Attributes and Compliance With Medications (the Dependent Variable) in Logistic Regression Analysis (N = 172)


Principal Findings

Patients reporting high levels of concordance with the physician were one third more likely to be compliant in taking medications prescribed during that consultation. In contrast, continuity of care measures, trust in the physician, and enablement were not consistently or not independently related to compliance with medications.

Study Strengths and Weaknesses

This study was a preplanned evaluation of the impact of attributes of primary care, using validated measures where possible, on compliance with medication. The representative sample of primary care patients and high initial response rate add validity. The analytical techniques used took into account the impact of the clustered sampling frame used in patient selection and adjusted for health and demographic variables, whereas previous studies have reported results from unadjusted analyses.16,25

Our results should be interpreted with caution because medication compliance was based on self-report, a commonly used but crude measure that may be limited by patients’ desire to appear compliant. Pill counts and pharmacy audits were beyond the scope of this study. In addition, the measure of physician-patient concordance used in this study has not previously been validated, and responses showed a potential ceiling effect. Perceived concordance between physician and patient as measured in this study was high in contrast to that in some studies14,26 but comparable to that in others.27 Despite these limitations, the finding that physician-patient concordance is related to an important outcome is in agreement with findings of other studies10,12 and is worthy of further exploration. An important addition to further study would be the physician’s perception of physician-patient concordance.

This study may have been limited by response bias in that a sizable minority of the patients who were given a prescription during the consultation was unable to complete a follow-up telephone call. Reassuringly, these patients were indistinguishable from the main sample, and therefore their responses may have been similar.

Comparison With Published Findings

Compliance with medications is a key outcome of primary care consultations. The primary care physician is the key coordinator of prescribing, and correlates of this outcome will be useful in designing interventions to improve compliance.

Patterns in the source of care, usual provider continuity, and reported importance of continuity are comparable between our sample and a sample of US primary care patients.16 Medicaid populations have been found to have lower levels of continuity of care, but continuity was associated with fewer hospitalizations.6 Our finding that various continuity measures were not associated with compliance is surprising and contrary to the findings of previous research.25 We concur with Christakis28 that continuity may operate differently in different populations and that relatively high levels of continuity and quality of care might combine to create a ceiling effect, concealing the impact of continuity evident in other subpopulations.

As with continuity of care, levels of trust in one’s physician are similar among patients in the United States, the United Kingdom,16 and New Zealand. In this study, trust was not independently associated with medication compliance, nor was enablement, which questions their utility as markers of quality in primary care.5

Implications for Practice, Policy, and Future Research

This study is the first to add medication compliance to the outcomes positively related to physician-patient concordance,10,12 and our findings suggest that efforts to facilitate physician-patient concordance may improve primary care outcomes. Ensuring that the patient understands the physician, that there is agreement about the nature of the patient’s problem, and that management is acceptable will require excellent communication skills, especially listening skills. These skills could be a focus for educators, planners, and evaluators.

In conclusion, this study points to the importance of concordance in the physician-patient relationship during the delivery of primary health care and extends previous work by exploring various dimensions of the relationship simultaneously. Whether the physician-patient concordance is improved by using the patient-centered method29 may be a fruitful avenue for research.


We thank the general practitioners, staff, and patients for their willing participation. Roger Rosenblatt kindly made comments on a previous draft of the manuscript.


Conflicts of interest: none reported

Funding support: Support for this study was provided by the Health Research Council of New Zealand in the form of a University of Auckland Summer Studentship grant. In addition, the Department of General Practice and Primary Health Care provided project costs, and the manuscript was prepared while Dr. Kerse was supported by a Harkness Fellowship from the Commonwealth Fund.


1. Strasser R. The doctor-patient relationship in general practice. Med J Aust. 1992;156:334–338. [PubMed]
2. Balint J, Shelton W. Regaining the initiative. Forging a new model of the patient-physician relationship. JAMA. 1996;275:887–891. [PubMed]
3. Christakis DA, Mell L, Wright JA, Davis R, Connell FA. The association between greater continuity of care and timely measles-mumpsrubella vaccination. Am J Public Health. 2000;90:962–965. [PMC free article] [PubMed]
4. Vedsted P, Mainz J, Lauritzen T, Olesen F. Patient and GP agreement on aspects of general practice care. Fam Pract. 2002;19:339–343. [PubMed]
5. Howie JG, Heaney DJ, Maxwell M, Walker JJ. Quality at general practice consultations: cross sectional survey. BMJ. 1999;319:738–743. [PMC free article] [PubMed]
6. Gill J, Mainous AI. The role of provider continuity in preventing hospitalizations. Arch Fam Med. 1998;7:352–357. [PubMed]
7. Gulbrandsen P, Hjortdahl P, Fugeli P. General practitioners’ knowledge of their patients’ psychosocial problem: multipractice questionnaire survey. BMJ. 1997;314:1014–1018. [PMC free article] [PubMed]
8. Hjortdahl P, Laerum E. Continuity of care in general practice: effect on patient satisfaction. BMJ. 1992;304:1287–1290. [PMC free article] [PubMed]
9. Saultz JW. Defining and measuring interpersonal continuity of care. Ann Fam Med. 2003;1:134–143. [PMC free article] [PubMed]
10. Bass MJ, Buck C, Turner L, Dickie G, Pratt G, Robinson C. The physician’s actions and the outcome of illness in family practice. J Fam Pract. 1986;32:34–47. [PubMed]
11. Stanton AL. Determinants of adherence to medical regimes by hypertensive patients. J Behav Med. 1987;10:377–394. [PubMed]
12. Maly RC, Leake B, Frank JC, DiMatteo MR, Reuben DB. Implementation of consultative geriatric recommendations: the role of patient-primary care physician concordance. J Am Geriatr Soc. 2002;50:1372–1380. [PubMed]
13. Buetow S, Adair V, Coster G, Hight M, Gribben B, Mitchell E. GP care for moderate to severe asthma in children: what do infrequently attending mothers disagree with and why? Fam Pract. 2003;20:155–161. [PubMed]
14. Scheitel SM, Boland BJ, Wollan PC, Silverstein MD. Patient-physician agreement about medical diagnoses and cardiovascular risk factors in the ambulatory general medical examination. Mayo Clin Proc. 1996;71:1131–1137. [PubMed]
15. Britt H, Harris M, Driver B, Bridges-Webb C, O’Toole B, Neary S. Reasons for encounter and diagnosed health problems: convergence between doctors and patients. Fam Pract. 1992;9:191–194. [PubMed]
16. Mainous AG III, Baker R, Love M, Gray DP, Gill J. Continuity of care and trust in one’s physician: evidence from primary care in the United States and the United Kingdom. Fam Med. 2001;33:22–27. [PubMed]
17. Baker R, Mainous AG III, Gray DP, Love MM. Exploration of the relationship between continuity, trust in regular doctors and patient satisfaction with consultations with family doctors. Scand J Prim Health Care. 2003;21:27–32. [PubMed]
18. Cartwright A. Medicine taking by people aged 65 or more. Br Med Bull. 1990;46:63–76. [PubMed]
19. The EuroQol Group. EuroQol: a new facility for the measurement of health-related quality of life. Health Policy. 1990;16:199–208. [PubMed]
20. Reid RJ, Barer ML, McKendry R, et al. Patient-Focused Care Over Time: Issues Related to Measurement, Prevalence and Strategies for Improvement Among Patient Populations. Ottawa, Canada: Canadian Health Services Research Foundation. Final report, project #RC2-0276-10, 2003.
21. Reid R, Haggerty J, McKendry R. Defusing the Confusion: Concepts and Measures of Continuity of Health Care. Ottawa, Canada: Canadian Health Services Research Foundation; 2002.
22. Nutting PA, Goodwin MA, Flocke SA, Zyzanski SJ, Stange KC. Continuity of primary care: to whom does it matter and when? Ann Fam Med. 2003;1:149–155. [PMC free article] [PubMed]
23. Starfield B. Primary Care: Concept, Evaluation, and Policy. New York, NY: Oxford University Press; 1992.
24. Anderson L, Dedrick R. Development of the trust in physician scale: a measure to assess interpersonal trust in patient-physician relationships. Psychol Rep. 1990;67:1091–1100. [PubMed]
25. Charney E, Bynum R, Eldredge D, et al. How well do patients take oral penicillin? A collaborative study in private practice. Pediatrics. 1967;40:188–195. [PubMed]
26. Kwoh CK, O’Conner GT, Regan-Smith MG, et al. Concordance between clinicians and patient assessment of physical and mental health status. J Rheumatol. 1992;19:1031–1037. [PubMed]
27. Cramer JA. Consequences of intermittent treatment for hypertension: the case for medication compliance and persistence Am J Manag Care. 1998;4:1563–1568. [PubMed]
28. Christakis D. Continuity of care: process or outcome? Ann Fam Med. 2003;1:131–133. [PMC free article] [PubMed]
29. Stewart M, Roter D. Communicating with Medical Patients. Newbury Park, Calif: Sage Publications; 1989.

Articles from Annals of Family Medicine are provided here courtesy of American Academy of Family Physicians
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...