Logo of jgimedspringer.comThis journalToc AlertsSubmit OnlineOpen Choice
J Gen Intern Med. 2007 Dec; 22(12): 1641–1647.
Published online 2007 Oct 6. doi:  10.1007/s11606-007-0371-5
PMCID: PMC2219826

Primary Care Visit Length, Quality, and Satisfaction for Standardized Patients with Depression

Estella M. Geraghty, MD, MS, MPH,corresponding author1 Peter Franks, MD,2 and Richard L. Kravitz, MD, MSPH3



The contribution of physician and organizational factors to visit length, quality, and satisfaction remains uncertain, in part, because of confounding by patient presentation.


To determine associations among visit length, quality, and satisfaction when patient presentation is controlled.


A factorial experiment using standardized patients to make primary care visits presenting with either major depression or adjustment disorder, and a musculoskeletal complaint.


One hundred fifty-two primary care physicians, each seeing 2 standardized patients.


Visit length was determined from surreptitiously obtained audiorecordings. Other key measures were derived from physician and standardized patient report.


Mean visit length for 294 completed encounters was 22.3 minutes (range = 5.8–72.2, SD = 9.4). Key factors associated with visit length were: physician style (ρ = 0.68 and 0.54 after multivariate adjustment), nonprofessional experience with depression (11% longer, 95% CI = 0–23%), practicing within an HMO (26% shorter, 95% CI = 61–90%), and greater practice volume (those working >9 half-day clinic sessions/week had 15% shorter visits than those working fewer than 6, 95% CI = 0–27%, and those seeing >12 patients/half-day had 27% shorter visits than those seeing <10 patients/half-day, 95% CI = 13–39%). Suicidal inquiry (a process-based quality-of-care measure for depression) was not associated with adjusted visit length. Satisfaction was linearly associated with visit length but not with suicide inquiry or follow-up interval.


Despite experimental control for clinical presentation, wide variation in visit length persists, largely reflecting individual physician styles. Visit length is a significant determinant of standardized patient satisfaction.

KEY WORDS: visit length, quality, satisfaction, primary care, standardized patient


Time is a key commodity offered by primary care physicians to their patients. Patients often feel that they are not getting enough of it and physicians report feeling rushed in their efforts to deliver quality care, while facing mounting pressures to be more efficient and see more patients.1 As such, increased visit length has been related to patient satisfaction, enhanced delivery of evidence-based care, and improved patient outcomes.2,3 But studies examining visit length have been confounded by the effects of case-mix.

Previous studies have shown that there is wide variation in primary care visit length,46 and case-mix may account for up to 55% of that variance.7 Consequently, the relative contribution of physician and organizational factors to visit length is uncertain. In a 2002 systematic review, Wilson and Childs note that the average consultation length may represent the differences in processes and outcomes, but may also be a marker of other doctor attributes.8 There are reports of longer visits attributable to gender concordance,9 race concordance,10 primary care specialty,11 and practice type (e.g., HMO).12 But studies with robust case-mix controls are lacking.

Studies have also suggested that visit length is related to quality-of-care but the evidence is limited. Studies from England have shown a link between visit length and quality,2,13 but because of their typically brief visits, their results may not be generalizable to the United States.

Previous research indicates that both visit length1315 and quality-of-care are related to patient satisfaction.16,17 Patient satisfaction, in turn, has been shown to be associated with health outcomes.18 Again, these studies have not been adjusted for case-mix, the personal and biographical characteristics of providers, and the features of their practices. Furthermore, the range of satisfaction scores from real patients (as opposed to standardized patients) is often narrow and may be subject to selection bias (happy patients keep coming back) and cognitive dissonance reduction strategies (why would I be here if I am not happy with my doctor?).19

To address limitations in the literature, we examined the physician, practice setting, and contextual factors that affect visit length, using standardized patients (SP) presenting with depressive symptoms and a musculoskeletal complaint to control clinical presentation. We also examined whether quality-of-care was associated with visit length. To this end, suicidal inquiry was treated as a process-based measure for quality depression care. Notwithstanding guidelines that recommend screening for suicide risk in depressed patients, studies have found low rates of inquiry and detection of suicidal thoughts by primary care physicians.20,21 We considered visit length to be a function of physician, practice setting, and patient characteristics. Suicidal inquiry was included as an independent predictor of visit length.22 Satisfaction is viewed as an outcome in the context of visit length and quality-of-care. Our conceptual model and study questions are outlined in Figure 1.

Figure 1
Conceptual model and study questions. In this study, standardized patients were used to control for clinical presentation. The numbers correspond to the following study questions: (1) What physician personal and practice characteristics drive visit length? ...


General Study Design

The data used for this analysis were derived from a randomized controlled trial performed to examine the effects of patient prompting for antidepressant medications (details of the original study design are described elsewhere).23 In brief, SPs presented to participating primary care physician offices with moderately complex clinical conditions (a depressive disorder and musculoskeletal complaint). Informed consent was obtained from the participating physicians, and the study protocol was approved by the institutional review boards at each participating institution.23

Physician Sample and Data

Board certified and board eligible family physicians and Internists were recruited for the study from 4 physician networks: the University of California, Davis, Primary Care Network and Kaiser-Permanente in Sacramento, CA; Brown & Toland Medical Group in San Francisco, CA; and Excellus BlueCross BlueShield in Rochester, NY. Physicians were recruited via mail with telephone follow-up. Participation rates ranged from 53% to 61%.23

Each participating physician was asked to complete a Clinician Background Questionnaire detailing their personal and practice characteristics. From this survey, we used the independent variables gender, age, race, specialty, practice location, practice type (solo, HMO—as Kaiser, academic, or other), and practice volume expressed both as the number of half-days the physician is involved in patient care per week (herein called sessions per week) and as the number of patients the doctor sees per half-day (herein called ‘busyness’). Because previous research has shown that a physician’s orientation toward psychosocial problems is associated with visit length,7 we included several questions in the survey that focused on physician attitudes about the treatment of depression in primary care.24 Based on Bandura’s Social Cognitive Theory,25 these questions were grouped such that 8 questions assessed physician self-efficacy or confidence in providing depression treatment in primary care (4-point semantic differential scale) with a Cronbach’s α of 0.76, mean of 3.0, SD of 0.41, and range of 1.75–4.0; 9 questions evaluated the physician’s perceived barriers to recognition and treatment of depression (3-point semantic differential scale) with a Cronbach’s α of 0.75, mean of 2.0, SD of 0.45, and range of 1.0–2.88. One question asked whether or not the physician had personal or vicarious experience with depression, as prior research has suggested that such experience may inform clinical decision making (complete list of questions available in the Appendix).26 A total of 152 physicians completed the survey and were thus included in the study.

Standardized Patient Roles and Data Collection

Clinical presentation was controlled in this study using SPs. To create actor roles, two moderately complex clinical conditions (low back pain with symptoms of adjustment disorder or carpal tunnel syndrome with symptoms of major depression) were crossed with two medication request types, creating four possible roles. Patients either made an explicit antidepressant medication request or made no medication request and presented with symptoms only. All of the SPs were white, middle-aged, nonobese women. The SPs were trained and monitored to the point of achieving 95% accuracy in portrayal of their specified role.23

Each physician was randomly assigned 2 SP visits, one with each of the clinical conditions and one with each of the medication request types. All visits were conducted between May 2003 and May 2004 and were audiorecorded using minidisk recorders concealed in the SP’s handbags. Visit length was measured from the audiorecordings, beginning from the moment the physician entered the room, excluding physician breaks (leaving the room), and ending when the visit was completed. The number of physician breaks during the visit were also included as an independent variable. Irregularities in the recordings were noted by data entry analysts and were reviewed by study investigators for irrevocable problems. Four recordings (1%) were inaudible and were excluded from the analysis.

As soon as possible following each visit, SPs were instructed to complete a reporting form detailing specified visit attributes. Then, listening to the audiorecording, they were to check the accuracy of their responses and revise them as needed. An independent judge listened to 36 randomly selected audiorecordings and completed a report with an average of 92% agreement (mean κ = 0.82).23

We used suicidal inquiry and suggested follow-up time as surrogate measures of quality-of-care for depression care. Not only is suicidal inquiry recommended in depression screening,20,22 but is also a useful measure by which to distinguish major depressive disorder from adjustment disorder (the mental health components of our 2 SP roles).27 Therefore, from the SP reporting form, we use whether or not the physician asked about suicidal ideation during the visit as an independent variable. Guidelines for care in depressive disorders also advise early follow-up, particularly when medications are prescribed.28 Whereas early follow-up time is always considered to be less than 1 month, there is still uncertainty about the cost-effectiveness29 and efficacy30 of weekly versus biweekly or longer face-to-face follow-up visit schedules. Taking a more conservative approach, we created a categorical variable noting the physician’s suggestion for patient follow-up as: ≤2 weeks, >2 weeks and <1 month, or ≥1 month.

SP satisfaction was based on 2 questions asked in the SP reporting form: (1) overall satisfaction with the care provided (5-point Likert scale) and (2) whether the SP would want the doctor for their own personal physician (5-point Likert scale). These scores were combined to create a ‘satisfaction’ dependent variable (Cronbach’s α = 0.90).

Data Analysis

Data were analyzed using STATA, version 9.2 (Stata Corp, College Station, TX, USA). The key dependent variable in this study was visit length. To yield a normal distribution, visit length was logarithmically transformed (Fig. 2). Bivariate analyses used Student’s t tests, ANOVA, and linear regression as appropriate.

Figure 2
Histograms showing visit length variable before and after log transformation

Random effects linear regression models examined the influence of specific physician and organizational characteristics on log transformed visit length adjusting for SP characteristics. All regressions treat the physician as a random effect (accounting for the 2 visits to each physician). We report the physician random effect as an intraclass correlation coefficient (ρ) defined as the fraction of the total variance in visit length accounted for by the physician variance component.31

The initial analysis, to account for clinical presentation, included only SP characteristics regressed on visit length. Then, a regression model using all of the variables of interest (including suicidal inquiry) was performed. A third regression removed the potentially endogenous variable of physician ‘busyness’. Resulting beta coefficients and confidence intervals are reported as the percent effect of the variable category relative to the reference value (to avoid the retransformation bias).

To explore the relationship between SP satisfaction and visit length, we used another random effects regression model with satisfaction as the dependent variable and visit length (not log transformed, but divided into quintiles) as the independent variable.


Eighteen SPs made 298 primary care visits to 152 physicians. Visit length was recorded for 294 (99%) visits. The overall mean visit length was 22.3 minutes with a range of 5.8 to 72.2 minutes (SD = 9.4 minutes).

Table 1 shows the association between each of the studied variables and visit length (not log transformed or adjusting for nesting of observations within physician). In these bivariate analyses, significant physician characteristics included specialty and age. Internists had significantly longer visits with SPs than did family physicians. In addition, increasing physician age was associated with longer visits. All of the practice characteristics studied were significant including location, setting, practice volume, and number of physician breaks in a visit. Bivariate analysis also indicated a significant association between suicidal inquiry and visit length such that physicians who broached the topic of suicidality took almost 4 minutes longer than those who did not. The suggested amount of time to follow-up visit was not found to be associated with the visit length. And standardized patient characteristics (presenting condition and whether an explicit medication request was made) also did not appear to significantly affect visit length.

Table 1
Bivariate Analysis of Physician, Practice, and Standardized Patients’ Characteristics and Visit Length

Random effects regression provided adjusted estimates of the effects of physician, practice, and SP characteristics on visit length (Table 2). In the analysis adjusting only for SP effects, SPs complaining of low back pain with adjustment disorder symptoms had significantly shorter visits (91% as long) than those with carpal tunnel syndrome and major depression symptoms. However, medication requests made by the SPs had no significant effect on visit length. The ρ for this baseline model was 0.68.

Table 2
Regression Model Showing Adjusted Estimates of the Effects of Physician, Practice, and Standardized Patient Characteristics on Visit Length

In the analysis adjusting for SP, physician, and system effects, the only physician characteristic found to be associated with visit length was whether a physician had personal or vicarious experience with depression. Those who did spent 11% longer with their patients than those who did not (Table 2). Several practice characteristics had associations with visit length. HMO visits were shorter. Both of our measures of practice volume were related to visit length. Doctors who worked >9 half-day clinic sessions/week spent 15% less time per visit than those working <6 half-day sessions/week and physicians seeing >12 patients/half-day spent 27% less time per visit than those seeing <10 patients/half-day. In the full model, the physician ρ was 0.54, thus, 21% of the variance attributed to physicians in the baseline model (ρ = 0.68) could be accounted for by the physician’s personal characteristics and organizational factors.

In the full model, suicide inquiry was not related to visit length. Estimates based on an analysis that excluded physician ‘busyness’ were substantially similar to the primary analysis and are therefore not discussed further. The physician’s recommended follow-up time was not included in the visit length regression as it was not significant in bivariate analysis and was not considered, in our model, to be a predictor of visit length.

Finally, the results of regressing SP satisfaction on visit length, our quality measures, and the other physician and practice characteristics are shown in Table 3. SP satisfaction was linearly associated with longer visits and fewer clinic sessions per week, but not with suicidal inquiry or suggested time to follow-up visit.

Table 3
Regression Model Showing Adjusted Estimates of the Effects of Visit Length and Physician, Practice, and Standardized Patient Characteristics on Standardized Patient Satisfaction


We found that visit length for standardized patients with similar conditions varied more than 10-fold among primary care physicians. In our multivariate analysis, we found that doctors working in HMOs and those with the busiest schedules spent significantly less time with patients, whereas physicians with personal or vicarious experience with depression spent significantly more time with their patients.

Despite controlling for the clinical presentation with SPs, the range in visit length was wide, from 5.8 to 72.2 minutes. In our regression analysis, we made particular note of the intraclass correlation coefficient to determine how much of the physician variance component could be explained by the physician and practice variables we examined. The high ρ noted in our baseline model suggests that physicians exhibit a considerable style effect affecting visit length. Although the variables we examined explained 21% of this effect, much remained unexplained.

The association of practice characteristics (practice setting and patient volume) with visit length lends credence to the argument that managed care has had a substantial effect on the practice of medicine, a finding that contrasts with the conclusions drawn by Mechanic et al. in 2001.1 But does that effect have any association with the quality-of-care that patients receive? We examined suicidal inquiry and recommended follow-up time as process-based measures of quality in this study. In bivariate analysis, we observed an increase in visit length of almost 4 minutes when physicians asked about suicidality. But when added to the regression models (both with and without the potentially endogenous variable, ‘busyness’), no statistically significant association was revealed to indicate that this quality measure influenced visit length. Time to follow-up visit was not significant in any of our analyses. So it appears that the relationship between visit length and suicide inquiry observed in the bivariate analyses is mediated by other physician and practice characteristics.

Patient satisfaction may be jeopardized by shorter visits. Our analysis showed a linear relationship between SP satisfaction and visit length. Our results are similar to findings in other studies.1315 Patient satisfaction has been related to outcomes and thus may serve as a quality measure by itself.32,12 SP satisfaction was also inversely linked to the number of clinic sessions a physician worked per week. We ruled out collinearity of this variable with the ‘busyness’ variable, and speculate that physicians with fewer sessions per week may be less “burned out”, and thus able to take more interest in each patient, translating into greater patient satisfaction. Further examination of this phenomenon is needed. It is interesting to note that our SPs were not less satisfied with physicians who took one or more breaks, some of whom were potentially seeing other patients simultaneously. Despite the fact that SPs are considered connoisseurs of care and, in our study, were well trained to recognize key aspects of depression care, we found no association between suicidal inquiry, recommended follow-up time, and SP satisfaction.

Another remarkable observation in this study was the lack of effect of medication prompting on visit length. Contrary to prevailing opinion,33 a patient’s explicit request for antidepressant medication did not relate to longer visits in either bivariate or adjusted analyses.

The strength of this study is the experimentally controlled patient presentation but several limitations are acknowledged. In fact, the strength of this study is also a weakness: although we were able to control patient presentation experimentally, our results may not apply beyond the particular presentations we created. Indeed, the use of SPs, while having numerous methodological advantages, does create uncertainty as to whether our results would hold true with ‘real patients’ in other practice settings.23 This uncertainty may be particularly true in the assessment of patient satisfaction. As SPs are trained to monitor care processes, their rating of the encounter may be biased. We also make note that all of our SPs were women. The choice to limit the study to women was a convenient strategy (depression being more common in women) to limit the number of experimental factors. We acknowledge that this may limit generalizability. Second, the use of suicide inquiry and physician recommendations for follow-up may not be ideal measures of quality. It may be useful for future studies to relate visit length to quality measures with a stronger evidence base. Finally, the study was quite intrusive and thus our physician population may be highly selected.

In conclusion, despite experimental control of patient presentation, we found wide variation in visit length, some of which is related to the personal characteristics of the physician and practice setting. Suicidal inquiry (a marker for quality-of-care among depressed patients) was not associated with visit length. SP satisfaction was linearly associated with the length of the visit, but was not related to either of our quality measures. Further research is needed to examine the relationship of quality and visit length and to understand how physicians can maximize patient satisfaction within the time available.


This research was supported in part by a training award to Dr. Geraghty (Health Resources Services Administration grant no. D55 HP00232) and by grants from the National Institutes of Health (R01-MH064683 and K24-MH072756) to Dr. Kravitz.

Conflict of interest The sponsors played no role in the collection, management, analysis, or interpretation of the data. All of the authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The authors have no other financial interests, disclosures, or conflicts of interest to report.


Table 4
Questions Assessing Physicians Attitudes Toward Diagnosing and Treating Depression in Primary Care24


1. Mechanic D, McAlpine DD, Rosenthal M. Are patients’ office visits with physicians getting shorter? N Engl J Med. 2001;344:198–204. [PubMed]
2. Heany D, Howie J, Porter A. Factors influencing waiting times and consultation times in general practice. Br J Gen Pract. 1991;41:315–9. [PMC free article] [PubMed]
3. Kaplan SH, Greenfield S, Gandek B, Rogers WH, Ware JE. Characteristics of physicians with participatory decision-making styles. Ann Intern Med. 1996;124:497–504. [PubMed]
4. Hulka BS, Romm FJ, Parkerson GR Jr, Russell IT, Clapp NE, Johnson, FS. Peer review in ambulatory care: use of explicit criteria and implicit judgments. Med Care. 1979;17 (3 Suppl):i–vi, 1–73. [PubMed]
5. Hartzema A, Christensen D. Non-medical factors associated with the prescribing volume among family practitioners in an HMO. Med Care. 1983;21:990–1000. [PubMed]
6. Andersson S, Mattsson B. Features of a good consultation in general practice: is time important? Scand J Prim Health Care. 1994;12:227–32. [PubMed]
7. Deveugele M, Derese A, van den Brink-Muinen A, Bensing J, De Maeseneer J. Consultation length in general practice: cross sectional study in six European countries. BMJ. 2002;325:472–7. [PMC free article] [PubMed]
8. Wilson MA, Childs S. The relationship between consultation length, process and outcomes in general practice: a systematic review. Br J Gen Pract. 2002;52:1012–20. [PMC free article] [PubMed]
9. Franks P, Bertakis KD. Physician gender, patient gender, and primary care. J Womens Health. 2003;12(1):73–80. [PubMed]
10. Cooper L, Roter D, Johnson R, Ford D, Steinwachs D, Powe N. Patient-centered communication, ratings of care, and concordance of patient and physician race. Ann Intern Med. 2003;139(11):907–15. [PubMed]
11. Radecki SE, Kane RL, Solomon DH, Mendenhall RC, Beck JC. Do physicians spend less time with older patients? J Am Geriatr Soc. 1988;36:713–8. [PubMed]
12. Balkrishnan R, Hall MA, Mehrabi D, Chen GJ, Feldman SR, Fleischer AB. Capitation payment, length of visit, and preventive services: evidence from a national sample of outpatient visits. Am J Manag Care. 2002;8(4):332–40. [PubMed]
13. Howie J, Porter A, Heany D, Hopton J. Long to short consultation ratio: a proxy measure of quality of care in general practice. Br J Gen Pract. 1991;41:48–54. [PMC free article] [PubMed]
14. Gross D, Zyzanski S, Borawski E, Cebul R, Stange K. Patient satisfaction with time spent with their physician. J Fam Pract. 1998;47:133–7. [PubMed]
15. Hull F, Hull F. Time and the general practitioner: the patient’s view. J R Coll Gen Pract. 1984;34:71–5. [PMC free article] [PubMed]
16. Anderson R, Barbara A, Feldman S. What patients want: A content analysis of key qualities that influence patient satisfaction. J Med Pract Manage. 2007;22(5):255–61. [PubMed]
17. Otani K, Krurz RS, Harris LE. Managing primary care using patient satisfaction measures. J Healthc Manag. 2005;50(5):311–24. [PubMed]
18. Safran D, Taira D, Rogers W, Kosinski M, Ware J, Tarlov A. Linking primary care performance to outcomes of care. J Fam Pract. 1998;47:213–20. [PubMed]
19. Fiscella K, Franks P, Srinivasan M, Kravitz RL, Epstein R. Ratings of physician communication by real and standardized patients. Ann Fam Med. 2007;5:151–58. [PMC free article] [PubMed]
20. Schulberg HC, Bruce ML, Lee PW, Williams JW, Dietrich AJ. Preventing suicide in primary care patients: the primary care physician’s role. Gen Hosp Psych. 2004;26:337–45. [PubMed]
21. Williams J, Noel P, Cordes J, Ramirez G, Pignone M. Is this patient clinically depressed? JAMA. 2002;287:1160–70. [PubMed]
22. Wells KB, Schoenbaum M, Unutzer J, Lagomasino IT, Rubenstein LV. Quality of care for primary care patients with depression in managed care. Arch Fam Med. 1999;8:529–36. [PubMed]
23. Kravitz RL, Epstein RM, Feldman MD, et al. Influence of patients’ requests for direct-to-consumer advertised antidepressants: a randomized controlled trial. JAMA. 2005;293:1995–2002. [PMC free article] [PubMed]
24. Meredith LS, Rubenstein LV, Rost K, et al. Treating depression in staff-model versus network-model managed care organizations. J Gen Intern Med. 1999;14:39–48. [PMC free article] [PubMed]
25. Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall; 1986.
26. Funder D. Toward a social psychology of person judgments: implications for person perception accuracy and self-knowledge. In: Williams K, ed. Social Judgments: Implicit and Explicit Processes. New York: Cambridge University Press; 2003:115–33.
27. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Association; 1994.
28. Snow V, Lascher S, Mottur-Pilson C. Pharmacologic treatment of acute major depression and dysthymia. Ann Intern Med. 2000;132(9):738–42. [PubMed]
29. Stettin GD, Yao J, Verbrugge RR, Aubert RE. Frequency of follow-up care for adult and pediatric patients during initiation of antidepressant therapy. Am J Manag Care. 2006;12:453–61. [PubMed]
30. Schulberg HC, Katon W, Simon GE, Rush AJ. An update of the Agency for Health Care Policy and Research Practice Guidelines. Arch Gen Psychiatry. 1998;55:1121–7. [PubMed]
31. Snijders T, Bosker R. Multilevel Analysis: An Introduction to Basic and Multilevel Modeling. London: Sage; 1999.
32. Davidoff F. Time. Ann Intern Med. 1997;127:483–5. [PubMed]
33. Weissman JS, Blumenthal D, Silk AJ, et al. Physicians report on patient encounters involving direct-to-consumer advertising: doctors see both the positive and some negative effects on their patients and practices. Health Aff. 2004;W4:219–33. [PubMed]

Articles from Journal of General Internal Medicine are provided here courtesy of Society of General Internal Medicine
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • 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...