November 27, 2022


Objectives: To assess the relationship between self-rated mental health (SRMH) and infrequent routine care among Medicare beneficiaries and to investigate the roles of managed care and having a personal doctor.

Study Design: Cross-sectional analysis of data from the 2018 Medicare Consumer Assessment of Healthcare Providers and Systems survey.

Methods: Logistic regression was used to predict infrequent routine care (having not made an appointment for routine care in the last 6 months) from SRMH, Medicare coverage type (fee-for-service [FFS] vs Medicare Advantage [MA], the managed care version of Medicare), and the interaction of these variables. Models that did and did not include having a personal doctor were compared. All models controlled for demographics and physical health.

Results: Overall, 14.9% of beneficiaries did not make a routine care appointment in the last 6 months, with rates adjusted for demographics and physical health ranging from 14.5% for those with “excellent” SRMH to 19.2% for those with “poor” SRMH. Beneficiaries with poor SRMH were less likely to make a routine care appointment in FFS than in MA (20.1% vs 16.4%, respectively, had not done so in the last 6 months; P < .05). Accounting for having a personal doctor reduced the association between SRMH and infrequent routine care by about a third.

Conclusions: Extra efforts are needed to ensure receipt of routine care by beneficiaries with poor mental health—particularly in FFS, where more should be done to ensure that beneficiaries have a personal doctor.

Am J Manag Care. 2022;28(11):In Press


Takeaway Points

  • Medicare beneficiaries with poor self-rated mental health (SRMH) are less likely to have made a routine care appointment in the last 6 months than beneficiaries with excellent SRMH, even controlling for demographic variables and physical health.
  • The association between poor SRMH and infrequent routine care is stronger in fee-for-service (FFS) Medicare than in Medicare Advantage (MA), which may be because more beneficiaries with poor SRMH lack a personal doctor in FFS than in MA.
  • Efforts are needed to ensure receipt of routine care by beneficiaries with poor SRMH—particularly in FFS, where more should be done to ensure that beneficiaries have a personal doctor.


Annual or more frequent routine medical exams are associated with recognition and treatment of chronic disease, control of health risk factors, uptake of preventive services, and better patient-reported outcomes.1 Such exams may be especially valuable for Medicare beneficiaries, who generally have higher-than-average medical need.

Having a usual source of care,2,3 having a personal doctor,2-4 and living with a spouse5 facilitate getting routine care. Barriers include lack of time or transportation,6-9 low health literacy,10 and prior negative experiences with care.9,11 Poor mental health may also be a barrier. Two prior studies found psychological distress and clinical depression to be associated with less use of preventive care services.12,13

Our study uses data from a nationally representative sample of Medicare beneficiaries to investigate (1) the association of mental health status with infrequent routine care (not making an appointment for routine care in the last 6 months) among Medicare beneficiaries with otherwise similar health status, (2) whether the association differs by Medicare coverage type, and (3) the possible role of having a personal doctor in this association.

Those with poor self-rated mental health (SRMH) less often have a personal doctor,4 perhaps due to difficulty navigating the health care system or stigma from mental illness that impedes establishing care with a personal doctor.14 Those enrolled in fee-for-service (FFS) Medicare (vs Medicare managed care) also less often have a personal doctor or receive routine care.4,15,16 Although it does not uniformly outperform FFS,15,16 Medicare Advantage (MA), the managed care version of Medicare, focuses more heavily on preventive care to reduce potentially serious and costly illness.17,18 Thus, we hypothesized that (1) poorer mental health would be positively associated with infrequent routine care, (2) the association would be diminished after accounting for whether a beneficiary has a personal doctor, and (3) the association would be more evident in FFS than in MA.


Data Source

The RAND Institutional Review Board approved this study. Data came from the 2018 Medicare Consumer Assessment of Healthcare Providers and Systems (CAHPS) Survey, to which 208,560 community-dwelling MA beneficiaries and 94,982 community-dwelling FFS beneficiaries responded. Details such as response rates and factors associated with nonresponse are in the eAppendix (available at We excluded those who did not answer the question about whether they made a routine care appointment, leaving 201,886 MA and 93,463 FFS respondents.

Outcome Measure: Infrequent Routine Care

Survey respondents were asked, “In the last 6 months, did you make any appointments for a check-up or routine care at a doctor’s office or clinic?” We coded a “yes” response as 0 and a “no” response as 1, indicating infrequent routine care. Cognitive testing found this question to have a common, consistent interpretation across CAHPS survey participants. Visits that participants consider to be a check-up or routine care include annual exams, follow-up on a condition under current treatment, routine appointments related to a chronic condition, and treatment for something not severe enough to require urgent care.19


Respondents rated their overall mental and emotional health by selecting 1 of 5 options: excellent, very good, good, fair, or poor. This variable was analyzed categorically. This single-item measure was developed to provide a broad assessment of a person’s emotional well-being independent of the presence of a diagnosed mental illness. Scores on this measure are associated with scores on multi-item measures of mental health (eg, the Kessler Psychological Distress Scale), self-rated health, health problems, mental health service utilization, and service satisfaction.20

Not Having a Personal Doctor

The survey asked respondents whether they have a “personal doctor” (ie, the person you would see for a check-up, for advice about a health problem, or if you were sick or hurt).

Control Variables

In our multivariate statistical models, we controlled for demographic and health variables that could confound the association between SRMH and infrequent routine care (see eAppendix).

Statistical Analysis

First, we described characteristics of the study sample. Next, we used weighted logistic regression to test for bivariate associations between respondent characteristics and infrequent routine care and between SRMH and not having a personal doctor, stratified by Medicare coverage type. Finally, we estimated 2 pairs of multivariate logistic regression models to investigate the demographic- and physical health–adjusted association between SRMH and infrequent routine care and to assess the potential role of not having a personal doctor in this association. These models included coverage type as an additional control variable. The 2 initial models differed only in whether they included or excluded not having a personal doctor from the set of predictors. A second pair of models was designed to investigate whether the association between SRMH and infrequent routine care differed by Medicare coverage type. This pair of models was like the first pair but also included interactions between coverage type and each level of SRMH. We used covariate-adjusted proportions (“recycled predictions”)21 derived from these regression models to obtain rates of infrequent routine care adjusted for all control variables.

As a sensitivity test, we reran the first of the 2 unstratified logistic regression models adding beneficiaries’ 0-to-10 ratings of all care received in the last 6 months to see whether the association between SRMH and infrequent routine care remained after accounting for differences in patient experience.22 These results need cautious interpretation because poor SRMH may be both a risk factor for poor patient experience and a mechanism through which poor SRMH results in less utilization of routine care.

All analyses employed person-level poststratification weights that account for sample design and nonresponse (see eAppendix).23

Missing Data and Imputation

Missing data on control variables ranged from 1.2% to 7.1%. Missing values on all control variables except not having a personal doctor were imputed using mean values within contract for MA surveys and within state for FFS surveys, according to standard CAHPS procedures.24


Descriptive and Bivariate Analyses

Overall, 14.9% of respondents did not make a routine care appointment in the prior 6 months. In bivariate models, those who did not make a routine care appointment were more often male (P = .02), less educated (P < .001), Medicaid or low-income subsidy eligible (P < .001), and residing in the Pacific, Mountain, or West North Central Census divisions (all P < .001) (Table 1). There was a linear association between self-rated overall health status and infrequent routine care (P < .001), with the rate of infrequent routine care among those in excellent health being approximately twice that among those in poor health. The bivariate association between infrequent routine care and SRMH was curvilinear (P < .001): Those with fair, good, and very good SRMH were more likely to make a routine care appointment, whereas those with excellent and poor SRMH were less likely to do so.

Bivariate associations between SRMH and demographic and other health characteristics appear in eAppendix Table 1 stratified by Medicare coverage type.

Only 5.7% of respondents lacked a personal doctor, but fewer than half of these made a routine care appointment vs nearly 90% of those with a personal doctor (Table 1). Not having a personal doctor was also associated with SRMH. Those with poor SRMH were more likely to lack a personal doctor than were those with excellent SRMH (9.5% vs 5.3%; P < .001) (eAppendix Table 2). Not having a personal doctor was more common overall in FFS than in MA (6.4% vs 4.1%; P < .001), including for those with poor SRMH (10.8% vs 6.1%; P < .001) (eAppendix Table 2).

Multivariate Logistic Regression Analyses Predicting Infrequent Routine Care

The left side of Table 2 shows that when demographic- and physical health–related differences were accounted for, there was a nearly linear association between SRMH and infrequent routine care (in contrast to the bivariate association discussed previously). In this model, adjusted rates of infrequent routine care ranged from 14.5% for those with excellent SRMH to 19.2% for those with poor SRMH (full results in eAppendix Table 3).

The association between SRMH and infrequent routine care was also evident in a model that controlled for beneficiaries’ ratings of all care received in the past 6 months. This sensitivity test appears in eAppendix Table 4; adjusted rates of infrequent routine care by SRMH based on this model are in eAppendix Table 5.

In the model that added interactions between each level of SRMH and Medicare coverage type to the predictors in the initial multivariate model, there was a significant interaction (P = .01) between Medicare coverage type and poor SRMH in predicting infrequent routine care (all P values for interactions involving other levels of SRMH > .40) (eAppendix Table 6). The left side of Table 2 shows that whereas respondents with fair to excellent SRMH were about as likely to make a routine care appointment in FFS as in MA (all P ≥ .28), those with poor SRMH were significantly less likely to make a routine care appointment in FFS (20.1%) than in MA (16.4%; P < .05).

The right side of Table 2 shows that when not having a personal doctor was accounted for, the overall difference between those with excellent and poor SRMH in infrequent routine care decreased by about one-third, and the MA-FFS difference between the percentages of those with poor SRMH who did not make a routine care appointment was reduced by approximately half.


We found that Medicare beneficiaries with poor SRMH are approximately 30% less likely to have made a routine care appointment in the last 6 months than are those with excellent SRMH, even after controlling for demographic and physical health differences. Infrequent routine care may be especially problematic for those with poor SRMH given their likelihood of comorbid physical conditions.25 At minimum, the need for regular routine care for individuals with poor SRMH should be no less than for otherwise similar individuals with better SRMH. Thus, less frequent routine care among those with poor SRMH suggests that barriers exist for these beneficiaries.

We also found that beneficiaries with poor SRMH were 22% more likely to go without a routine care appointment in FFS Medicare than in MA. About half of this difference was associated with differences in not having a personal doctor.

Consistent with prior research,4 FFS beneficiaries were less likely than their MA counterparts to report not having a personal doctor, especially among those with poor SRMH. We found that a beneficiary with poor SRMH is about half as likely to report not having a personal physician in MA than in FFS (6.1% vs 10.8%).

Coverage type and not having a personal doctor did not account for all of the association between poor SRMH and infrequent routine care. Possibilities worth exploring are that those with poor SRMH are less motivated to adhere to medical recommendations,26 more isolated and withdrawn from individuals who could remind them about routine care,5 less tolerant of long wait times,27 or reluctant to seek care because they have internalized stigma related to poor mental health.14


Our study has limitations. First, results are based on self-reports of receiving a check-up or routine care in the last 6 months. Actual and self-reported receipt of routine care may differ. Second, the sample consisted of Medicare beneficiaries; results may not generalize to other segments of the US population, particularly uninsured individuals. Third, results could have been affected by nonresponse bias. However, response rate is a weak proxy for nonresponse bias,28 and the modest differences between responders and nonresponders on observed characteristics were accounted for by nonresponse weighting.29 Fourth, our cross-sectional study cannot establish causal directionality of the association between SRMH and routine care or attribute differences in routine care to mental health or differences between MA and FFS to the practices of MA plans.


Primary care is a first point of contact and a point of continuing care for individuals with mental health problems, many of whom have comorbid conditions that substantially increase their risk of mortality.25 Yet, our results suggest the need to facilitate access to routine care for adults with poor SRMH. They may need support to overcome psychological barriers that may prevent them from making and keeping appointments and prioritizing their health care.27,30 When individuals identified as having mental health problems do attend a visit, providers may want to discuss with them the importance of regular contact and routine check-ups. Our results also suggest that the need to further assist beneficiaries with poor SRMH may be greater in FFS Medicare than in MA and that more should be done to help FFS beneficiaries obtain and retain a personal doctor.

If the associations reported here were established to be causal, it would warrant efforts to identify the means through which MA facilitates having a doctor and scheduling routine care appointments among those with poor SRMH. Strategies that could work in either coverage context include sending targeted mailings and electronic reminders to schedule routine care appointments, covering the cost of routine physical exams, and offering members with poor SRMH additional assistance for finding a personal doctor and encouragement to get routine care.

Author Affiliations: RAND Corporation, Pittsburgh, PA (SCM, JB), and Santa Monica, CA (RDH, KH, NO, MNE); Carnegie Mellon University (AMH), Pittsburgh, PA; University of Alabama (RW-M), Birmingham, AL; CMS (SG), Baltimore, MD.

Source of Funding: This research was supported by a contract from CMS (HHSM-500-2017-00083G). The views expressed in this article are those of the authors and do not necessarily reflect the views of HHS or CMS.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (SCM, AMH, MNE); acquisition of data (RDH, SG, MNE); analysis and interpretation of data (SCM, RDH, KH, AMH, RW-M, JB, NO, MNE); drafting of the manuscript (SCM, KH, RW-M, JB, NO); critical revision of the manuscript for important intellectual content (RDH, AMH, RW-M, JB, NO, SG, MNE); statistical analysis (KH, AMH, JB); provision of patients or study materials (SG); obtaining funding (RDH, MNE); and supervision (MNE).

Address Correspondence to: Steven C. Martino, PhD, RAND Corporation, 4570 Fifth Ave, Ste 600, Pittsburgh, PA 15213-2665. Email:


1. Liss DT, Uchida T, Wilkes CL, Radakrishnan A, Linder JA. General health checks in adult primary care: a review. JAMA. 2021;325(22):2294-2306. doi:10.1001/jama.2021.6524

2. Okoro CA, Strine TW, Young SL, Balluz LS, Mokdad AH. Access to health care among older adults and receipt of preventive services. results from the Behavioral Risk Factor Surveillance System, 2002. Prev Med. 2005;40(3):337-343. doi:10.1016/j.ypmed.2004.06.009

3. Xu KT. Usual source of care in preventive service use: a regular doctor versus a regular site. Health Serv Res. 2002;37(6):1509-1529. doi:10.1111/1475-6773.10524

4. Martsolf GR, Elliott MN, Haviland AM, et al. Care experiences among Medicare beneficiaries with and without a personal physician. Med Care. 2018;56(4):329-336. doi:10.1097/MLR.0000000000000886

5. Lau DT, Kirby JB. The relationship between living arrangement and preventive care use among community-dwelling elderly persons. Am J Public Health. 2009;99(7):1315-1321. doi:10.2105/AJPH.2008.151142

6. DeVoe JE, Fryer GE, Phillips R, Green L. Receipt of preventive care among adults: insurance status and usual source of care. Am J Public Health. 2003;93(5):786-791. doi:10.2105/ajph.93.5.786

7. Green CA, Johnson KM, Yarborough BJH. Seeking, delaying, and avoiding routine health care services: patient perspectives. Am J Health Promot. 2014;28(5):286-293. doi:10.4278/ajhp.120702-QUAL-318

8. Kannan VD, Veazie PJ. Predictors of avoiding medical care and reasons for avoidance behavior. Med Care. 2014;52(4):336-345. doi:10.1097/MLR.0000000000000100

9. Taber JM, Leyva B, Persoskie A. Why do people avoid medical care? a qualitative study using national data. J Gen Intern Med. 2015;30(3):290-297. doi:10.1007/s11606-014-3089-1

10. Scott TL, Gazmararian JA, Williams MV, Baker DW. Health literacy and preventive health care use among Medicare enrollees in a managed care organization. Med Care. 2002;40(5):395-404. doi:10.1097/00005650-200205000-00005

11. Fiscella K, Franks P, Clancy CM. Skepticism toward medical care and health care utilization. Med Care. 1998;36(2):180-189. doi:10.1097/00005650-199802000-00007

12. Thorpe JM, Kalinowski CT, Patterson ME, Sleath BL. Psychological distress as a barrier to preventive care in community-dwelling elderly in the United States. Med Care. 2006;44(2):187-191. doi:10.1097/01.mlr.0000196965.54871.d5

13. Thorpe JM, Thorpe CT, Kennelty KA, Chewning BA. Depressive symptoms and reduced preventive care use in older adults: the mediating role of perceived access. Med Care. 2012;50(4):302-310. doi:10.1097/MLR.0b013e31821a933f

14. Knaak S, Mantler E, Szeto A. Mental illness-related stigma in healthcare: barriers to access and care and evidence-based solutions. Healthc Manage Forum. 2017;30(2):111-116. doi:10.1177/0840470416679413

15. Landon BE, Zaslavsky AM, Bernard SL, Cioffi MJ, Cleary PD. Comparison of performance of traditional Medicare vs Medicare managed care. JAMA. 2014;291(14):1744-1752. doi:10.1001/jama.291.14.1744

16. Agarwal R, Connolly J, Gupta S, Navathe AS. Comparing Medicare Advantage and traditional Medicare: a systematic review. Health Aff (Millwood). 2021;40(6):937-944. doi:10.1377/hlthaff.2020.02149

17. Dowd BE, Coulam RF, Feldman R, Pizer SD. Fee-for-service Medicare in a competitive market environment. Health Care Financ Rev. 2005;27(2):113-126.

18. Gold M, Casillas G. What do we know about health care access and quality in Medicare Advantage versus the traditional Medicare program? Kaiser Family Foundation. November 2014. Accessed June 10, 2021.

19. Harris-Kojetin LD, Fowler FJ Jr, Brown JA, Schnaier JA, Sweeny SF. The use of cognitive testing to develop and evaluate CAHPS 1.0 core survey items. Med Care. 1999;37(suppl 3):MS10-MS21. doi:10.1097/00005650-199903001-00002

20. Ahmad F, Jhajj AK, Stewart DE, Burghardt M, Bierman AS. Single item measures of self-rated mental health: a scoping review. BMC Health Serv Res. 2014;14:398. doi:10.1186/1472-6963-14-398

21. Graubard BI, Korn EL. Predictive margins with survey data. Biometrics. 1999;55(2):652-659. doi:10.1111/j.0006-341x.1999.00652.x

22. Martino SC, Elliott MN, Kanouse DE, Farley DO, Burkhart Q, Hays RD. Depression and the health care experiences of Medicare beneficiaries. Health Serv Res. 2011;46(6):1883-1904. doi:10.1111/j.1475-6773.2011.01293.x

23. Purcell NJ, Kish L. Postcensal estimates for local areas (or domains). Int Stat Rev. 1980;48(1):3-18. doi:10.2307/1402400

24. Medicare Advantage and Prescription Drug Plan CAHPS Survey. Updated September 23, 2022. Accessed February 3, 2022.

25. Firth J, Siddiqi N, Koyanagi A, et al. The Lancet Psychiatry Commission: a blueprint for protecting physical health in people with mental illness. Lancet Psychiatry. 2019;6(8):675-712. doi:10.1016/S2215-0366(19)30132-4

26. DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Arch Intern Med. 2000;160(14):2101-2107. doi:10.1001/archinte.160.14.2101

27. Ross LE, Vigod S, Wishart J, et al. Barriers and facilitators to primary care for people with mental health and/or substance use issues: a qualitative study. BMC Fam Pract. 2015;16:135. doi:10.1186/s12875-015-0353-3

28. Groves RM, Peytcheva E. The impact of nonresponse rates on nonresponse bias: a meta-analysis. Public Opin Q. 2008;72(2):167-189.

29. Elliott MN, Zaslavsky AM, Goldstein E, et al. Effects of survey mode, patient mix, and nonresponse on CAHPS hospital survey scores. Health Serv Res. 2009;44(2, part 1):501-518. doi:10.1111/j.1475-6773.2008.00914.x.

30. Dixon LB, Holoshitz Y, Nossel I. Treatment engagement of individuals experiencing mental illness: review and update. World Psychiatry. 2016;15(1):13-20. doi:10.1002/wps.20306

Source link