Communication about depression during rheumatoid arthritis patient visits




To examine whether rheumatologists and rheumatoid arthritis (RA) patients with depressive symptoms communicate about depression.


The data used in this study came from a randomized controlled trial to improve provider-patient communication about patients' agendas. The current secondary analysis used data collected at baseline, before any intervention occurred. A total of 200 RA patients from 4 rheumatology clinics participated in the study. Patient medical visits were audiotape recorded and patients were interviewed after their medical visits. Physicians recorded patients' American College of Rheumatology (ACR) functional status after their visits.


Twenty-one (11%) patients were scored as having moderately severe to severe symptoms of depression. Patients who were rated by their physicians as having worse ACR functional status were more than twice as likely to have moderately severe to severe symptoms of depression (odds ratio 2.23, 95% confidence interval 1.1–4.6). Only 4 (19%) of the 21 patients who were scored as having moderately severe to severe symptoms of depression discussed depression during their medical visits, and patients initiated the discussion each time.


Rheumatologists should consider assessing depressive symptoms among their patients, especially among those with worse functional status.


Rheumatoid arthritis (RA) is a chronic, debilitating disease characterized by inflammation of joint tissues, persistent pain, functional disability, stiffness, and fatigue (1, 2). RA is the most common chronic inflammatory arthritis, affecting 0.8% of the population (3). Depression is common among patients with RA, with a prevalence of 15–20% (4–6), and patients with RA are twice as likely as members of the general population to experience depression (7).

Depressive symptoms are more common among patients with RA who report higher levels of pain (8) and those who report more consequences from their RA (9). In contrast, patients who receive adequate emotional support report fewer depressive symptoms (10).

Physician-patient communication during RA visits is critical to help RA patients with depression (4, 8–10). Prior research has examined patient-physician communication about depression in primary care settings (11–13). Sleath and Rubin (12) found that depression was discussed during 25% of 383 primary care visits. Patients who rated their emotional health poorly were significantly more likely to discuss depression. Patients initiated the discussion of depression ∼55% of the time. Female patients and patients who were more educated were significantly more likely to raise the topic of depression before their physicians did. In a study of 69 primary care physicians and 308 adult patients with a score on the General Health Questionnaire that indicated psychological distress (score >5), patients were significantly more likely to disclose psychological distress if physicians asked about it, if there was greater physician-patient familiarity, and if there was greater severity of psychological distress; this patient disclosure was significantly related to physician recognition of a mental health problem (86% versus 28% for disclosure versus nondisclosure) (11). In a sample of 508 primary care patients, Callahan et al (13) found that physicians did not recognize depression among 72% of patients with a Beck Depression Inventory score indicating moderate to severe depression (score ≥9).

Although depression in primary care has been well studied, to our knowledge no prior study has examined the extent to which rheumatologists and patients with RA discuss depression during medical visits. Therefore, the purpose of this exploratory study was to examine whether rheumatologists and patients with RA with depressive symptoms communicate about depression and who initiates the discussion.



The data used in the present study came from a randomized controlled trial to improve provider-patient communication about patients' agendas. The current secondary analysis used data collected at baseline, before any intervention occurred. This study was approved by the University of North Carolina and Duke University Institutional Review Boards.


Rheumatologists and their patients with RA were recruited from 3 academically affiliated rheumatology clinics and 1 nonacademically affiliated rheumatology clinic. Rheumatologists were eligible if they saw patients age ≥45 years with RA and provided written informed consent to participate. Patients were eligible if they 1) were being seen by a participating rheumatologist, 2) were ≥45 years of age, 3) had a diagnosis of RA, and 4) were able to speak and read English. Clinic staff referred eligible patients to a research assistant in the clinic waiting area. The research assistant explained the study, obtained informed consent, and administered the 10-item Mental Status Questionnaire (14). The 10-item scale, which assesses patients' mental competence, has approximately the same sensitivity and specificity as the Folstein Mini-Mental Status Examination, but is easier and faster to administer (14). Patients who made ≥5 errors were considered ineligible. Patients who consented to participate had their medical visit audiotape recorded and were interviewed by a research assistant after their audiotaped visit.


Patient measures.

The 9-item validated Patient Health Questionnaire (PHQ) was used to measure depressive symptoms (15, 16). The diagnostic validity of the PHQ has been established in 2 studies involving 3,000 patients in 8 primary care clinics and 3,000 patients in 7 obstetrics-gynecology clinics (15, 16). Validity was established by comparing patient PHQ scores with diagnoses made by mental health professionals. Scores on the PHQ can range from 0 to 27; scores ≥15 indicate moderately severe to severe depressive symptoms (17), which can indicate that the patient has a possible major depressive disorder. We dichotomized moderately severe to severe depressive symptoms as present (PHQ score ≥15) or absent (PHQ score <15).

Global function was assessed by rheumatologists immediately after the audiotaped visit using the American College of Rheumatology (ACR) classification system: I = no restriction of ability to perform normal activities; II = moderate restriction but adequate for normal activities; III = marked restriction, inability to perform most duties of the patient's usual occupation or self-care; IV = incapacitation or confinement to a bed or wheelchair, permitting little or no self-care (18). In terms of demographic characteristics, age was analyzed as a continuous variable; race (based upon self-report) was coded as white or nonwhite; and education was categorized as less than high school, high school, or more than high school.

Physician measures.

Physician age was analyzed as a continuous variable and race (based upon self-report) was coded as white or nonwhite.


Using a detailed protocol, 3 trained coders recorded whether 1) depression was discussed during the encounter (depression, sadness, loss of interest in activities, down in dumps, low spirits, cries during the visit) and 2) the patient or physician initiated the discussion of depression. Because depression was one of many types of communication being coded, coders were functionally blinded to our specific focus on depression. To assess intercoder reliability, 24 randomly selected audiotapes were coded by each coder; these tapes were selected throughout the coding period to check against coder decay (19). Interrater correlation for discussion of depression was 0.93. Reliability coefficients for who initiated the discussion of depression could not be calculated because it occurred so infrequently. However, there was 100% agreement among the coders.

Statistical analysis.

All analyses were conducted using SPSS (SPSS, Chicago, IL). First, we used descriptive statistics for all variables. Second, we examined bivariate relationships using correlation coefficients or Pearson's chi-square statistics. Third, we conducted a multivariable logistic regression model to examine patient factors associated with having moderate to severe symptoms of depression, where we used odds ratios (ORs) and 95% confidence intervals (95% CIs). We intended to conduct multivariable logistic regression models to predict discussion of depression during the visit for all patients, as well as those with moderate to severe symptoms of depression. However, discussion of depression occurred so infrequently that this was not possible. We also used qualitative examples of depression discussions between patients with RA and their physicians.


Eight (72.7%) of 11 rheumatologists we approached participated in the study: 5 were men and the mean age was 43.1 years (range 32–59 years). In terms of race, 6 were white, 1 was African American, and 1 was Latino. After excluding 6 patients based on the mental status questionnaire, 200 (79%) of 254 eligible patients agreed to participate. Demographically (Table 1), 74% of patients were women, 21% were African American, and 45% completed more than high school. Their mean age was 62 years (range 45–88 years). A total of 25% of patients were rated by their physicians as ACR functional class I, and 16% were rated as class III or IV. Fifty-three percent of patients saw physicians who were the same sex as themselves, and 48% saw physicians who were the same race or ethnicity.

Table 1. Patient characteristics (n = 200)*
  • *

    Values are the number (percentage) unless otherwise indicated. ACR = American College of Rheumatology; PHQ = Patient Health Questionnaire.

 Male52 (26.0)
 Female148 (74.0)
Age, years 
 Mean ± SD62.35 ± 9.02
 African American43 (21.5)
 White137 (68.5)
 American Indian7 (3.5)
 Other3 (1.5)
 Missing10 (5.0)
Educational level 
 Less than high school40 (20.0)
 High school60 (30.0)
 More than high school90 (45.0)
ACR classification 
 I50 (25.0)
 II89 (44.5)
 III28 (14.0)
 IV4 (2.0)
Saw physician of same sex 
 Yes106 (53.0)
 No94 (47.0)
Saw physician of same race/ethnicity 
 Yes96 (48.0)
 No94 (47.0)
 Missing10 (5.0)
Depression severity (PHQ) 
 Not at all32 (16.0)
 Minimal55 (27.5)
 Mild61 (30.5)
 Moderate31 (15.5)
 Moderately severe13 (6.5)
 Severe8 (4.0)

In multivariable logistic regression (Table 2), patients who were rated by their physician as having worse ACR functional classification were more than twice as likely to have moderate to severe symptoms of depression than patients who were rated by their physician as having better functional status (OR 2.23, 95% CI 1.1–4.6).

Table 2. Multivariable logistic regression predicting whether patients had moderately severe to severe symptoms of depression (n = 161)*
Independent variablesOdds ratio (95% confidence interval)
  • *

    ACR = American College of Rheumatology.

  • P < 0.05.

Age0.98 (0.93–1.04)
Male sex0.36 (0.07–1.78)
ACR functional classification: worse2.23 (1.09–4.55)
Education level0.59 (0.30–1.15)
White race1.48 (0.43–5.07)

Twenty-one patients (10.5%) had moderate to severe depressive symptoms, only 4 of whom discussed depression during their visits. In fact, depression was only discussed during 7 (3.5%) of the 200 participating patients' visits. When depression was discussed, the patient initiated the discussion every time. Patients with moderate to severe depressive symptoms were significantly more likely to discuss depression than patients who did not have moderate to severe depressive symptoms (20% versus 1.7%; Pearson's χ2 = 17.11, 1 df, P < 0.000). No other patient or physician demographic characteristics were significantly related to discussion of depression during the visit, nor was whether patients and physicians were of the same sex or race.

Twenty-one patients (10.5%) were scored as having moderately severe to severe symptoms of depression. Only 4 of these patients (19%) discussed depression during their visits. If depression was discussed, the patients initiated the discussion every time. All patients who discussed depression were white, all were women, and all had a high school education or higher.

When depression was brought up during the medical visit, it was often not discussed at any length. Consider the following passage between a rheumatologist (D) and a patient (P):

P: “Somebody was saying what's so wrong with you that you aren't working and I said well I have rheumatoid arthritis and I have osteoporosis, so which one would you like to attack?”

D: “And it's not like, oh that's beautiful! It's not like you haven't tried.”

P: “Mmhmm, I would love to be working.”

D: “You have really tried. Oh, you know I've got plenty of papers showing that you have tried.”

P: “They don't know. I say you don't know some days I get so depressed and cry all day long because I think that I could be out working. I burnt my hand, I drop stuff.”

D: “I see that!”

P: “Dropped the grits on my hand cooking grits … I dropped 2 things out of the refrigerator this week. Thank God they weren't glass.”

D: “Yea, or else you would have had a mess to clean up.”

P: “Yea. And my husband said now when you start dropping stuff that means it's time for you to sit down because you are tired.”

The patient and the rheumatologist never discuss the patient's admission of depression and crying all day long. The rheumatologist examines the patient and discusses the patient's medication regimen for the RA and then tells the patient to come back in 3 months.

In another example the rheumatologist discusses the importance of treating the depression that a patient refers to and how it also might help the patient's pain:

D: “Ok… . Now when you were on the prednisone did it make you feel sick in any way?”

P: “No.”

D: “Did it increase your appetite?”

P: “I really didn't notice.”

D: “Ok, so, we're really on a small dose so—Alright.”

P: “I have been so depressed.”

D: “Really?”

P: “Just really depressed.”

D: “And is that something new for you? Have you had depression before?”

P: “Yes, I have always been really active and now I just don't feel like doing anything. I'm just really depressed.”

D: “Ok. And have you ever used any medicine for depression? Ok, would you be interested in trying something?”

P: “I guess—”

D: “Ok.”

P: “I guess??”

D: “It's overwhelming, isn't it? Ok.”

P: “Especially when you used to be so active and now all the sudden you can't do anything without hurting.”

D: “Mmmhmm. Are you sleeping at night?”

P: “I am at times. I have nights when I sleep good then there are other nights when I'm just up and down all night.”

D: “On the nights that you are up and down all night is it because you're depressed or because you're hurting, or why—what's keeping you up?”

In the above passage, the rheumatologist asks the patient if she wants to try something for the depression. The rheumatologist and patient continue to discuss the patient's sleep and pain, and the physician examines the patient. The physician then goes back to discussing the depression.

D: “No, it was fine … The encouraging thing I think here is that you don't have a lot of swelling and I'm not feeling in your joints a lot of restriction or inflammation so we've got pain to deal with. Some of the pain may be related to the depression and the sleep problems. In other words we have got multiple things going on here. You've got arthritis, your function has changed, you're not sleeping as well, you're not feeling as well, so it becomes kind of a vicious cycle. Sometimes actually treating the depression makes the pain better, which is kind of an interesting twist on this. I don't want to intensify your methotrexate right now because I think this is the right dose for you. We'll have to check the labs and make sure that dose is a dose you are tolerating. But I am going to start you on a medicine called—2 things. One is Zoloft, which is an antidepressant. We are going to start you on a low dose, 50 mg. You are just going to take one a day. Most people tolerate this extremely well. Occasionally people experience some side effects. If there is something that you are concerned about you should have a low threshold to call me if you are concerned. But, like I said, it's usually fairly well tolerated. I would suggest, um, that you try to take it around the same time every day. Some people like to take it at night. I heard reports, people sometimes say it gives them bad dreams, but that is pretty unusual. Um, but it's a pretty easy medicine to tolerate. It's going to take about 4 to 6 weeks to get this medicine into your system. It's not something that is going to work really fast.”

P: “Ok.”

D: “And then I am going to give you a prescription for prescription strength Motrin. We are not going to monkey around with the Advil doses, ok. And you are going to take 1 tablet every 8 hours if you need it and see if that doesn't help a lot with the pain.”

P: “Ok.”

In another example, the rheumatologist acknowledges the patient's depression but then immediately focuses on controlling the pain as a way of helping the depression:

D: “Is that sore here? … How about there? … Here?”

D: “Yea, alright. Take some deep breaths for me. Very good. … So how was your holiday?”

P: “It was alright.”

D: “It was alright? Were you too busy?”

P: “No.”

D: “Doing too much? Too much running around?”

P: “I was just depressed.”

D: “Depressed? … (mumbles) Well my guess is that over the next few days you are going to start to feel better because we are just kind of coming to the end of that Remicade and now that you have got another dose things should start to get better. If over the next week or so you don't feel like things are gradually improving, and it may not be that dramatic, just sort of gradually getting better where this discomfort isn't going away you need to call me and let me know. And do you need prescriptions for anything today?”

P: “I'll probably need the methotrexate.”

In the above example, the physician and patient never return to discussing the patient's depression.


Almost 11% of the sample were screened as having moderately severe to severe symptoms of depression. This is a slightly lower prevalence rate than other reported depression prevalence rates (15–20%) among patients with RA (4–6). However, the difference is most likely because our focus was on patients with RA with moderately severe to severe symptoms of depression.

Patients with RA who were rated by their physician as having worse functional status were significantly more likely to have moderately severe to severe symptoms of depression. Having a worse functional status means that patients are more restricted from participating in their normal activities. Our finding is similar to prior research that found that patients with RA who perceive more negative consequences from their disease are more likely to have worse depressive symptoms (9).

Only 1 of 5 patients who were scored as having moderately severe to severe symptoms of depression on the PHQ discussed depression with their rheumatologist. This finding is in contrast to Robinson and Roter (11) who found that 51% of patients with psychosocial distress disclosed this to their primary care physicians during their medical visits. Part of this difference may be because we examined communication about depression in rheumatology clinics, which is specialized care versus primary care.

When patients go to rheumatologists, the main focus of their visit is their RA. However, a chronic disease such as RA can greatly impact a patient's quality of life and psychosocial well-being. As discussed above, patients with RA are twice as likely as members of the general population to experience depression (7). If rheumatologists have limited time with patients, their main focus would most likely be on the RA, because they may be facing competing demands regarding which condition to focus on during an office visit (20). Yet many patients with RA see their rheumatologist more often than their primary care physician, on average, and some patients with RA do not have a primary care physician (21). For these patients, rheumatologists may become their primary care providers by default, thus stressing the importance of rheumatologists addressing patient issues beyond RA such as depression.

It is important for rheumatologists to consider addressing both the RA and the depression, because research in other chronic disease areas such as diabetes has shown that depression can impact adherence to diabetes medications and self-care regimens (22, 23). As illustrated by the transcripts of some of the medical visits in which depression was discussed, sometimes when the patient brought up depression the physician picked up on the discussion of depression, asked questions, and suggested treatment for the patient's depression.

Yet in other cases, the physician asked a few questions about the patient feeling depressed without resolving anything and then immediately began to focus on the patient's RA. In another example, the physician actually told the patient that he or she would feel better (the depression) if the pain was controlled. Although this may be true, this approach places the depression or psychosocial concern on the back burner to the RA. It also assumes that the depression is only due to the pain. What if the patient's pain was not controlled with the new medication regimen? An approach of only treating the pain and not co-occurring depression can potentially prolong the length of time the patient has untreated depression.

However, some rheumatologists may not feel comfortable discussing depression with their patients. In fact, in our sample, if depression was discussed, the patient initiated the discussion every time. This is in contrast to a study in primary care settings that found that patients brought up the discussion of depression ∼55% of the time whereas their physicians brought up the discussion 45% of the time (12). Rheumatologists might consider screening their RA patients for depression. They might consider having their office staff administer a brief depression screen immediately before the patient's visits so that problems are identified early and treatment can be initiated if appropriate. The 9-item PHQ that was used in this study is an easy-to-use depression screening instrument (17). If physicians do not have time to use a formal depression screener during the visit, they could ask about depression and the patient's loss of interest in usual activities.

In addition to screening for depression, it is also important that patients with RA who screen positive for depression have access to appropriate systems to receive treatment. Rheumatologists may want to treat the depression themselves, refer RA patients to mental health professionals in their area, or communicate with the patient's primary care physician about the patient's depression and coordinating a treatment plan. Rheumatology training programs should educate residents and fellows about the importance of screening for and treating depression.

Although our sample size was small and only 4 patients discussed depression with their physician, it is important to note that the 4 patients who brought up their depression were white, female patients who had a high school education or higher. None of the nonwhite patients, male patients, or patients with less than a high school education who screened as having moderately severe to severe symptoms of depression initiated the discussion of depression with their physicians, and therefore depression was not discussed.

Our findings are similar to those of a prior study that found that women and more educated patients were significantly more likely to bring up depression before their physicians (12). Also, prior research has found that female patients were more actively involved during their visits than male patients and more educated patients expressed a desire for more active participation during medical visits (24–26). Physicians should encourage all patients to discuss depression by asking about any feelings of depression or sadness or loss of interest in usual activities.

The study has limitations. We did not measure social support, which could have been related to whether patients had moderate to severe symptoms of depression. Patients did not have a physician diagnosis of depression, but the PHQ is a validated screener for depressive symptoms (15, 17). We did not obtain data relevant to the barriers and obstacles that prevented depression and psychological issues from being addressed. We did not have data on the physician-patient relationship (e.g., how long or how well the physician and patient knew one another). We also did not have data on how many years the physician had been in practice or prior training in communication. The study was conducted in only 4 practices, which may limit the external validity of the study findings.

Despite the limitations of our study, our results provide new information suggesting that rheumatologists rarely communicate about depression with their patients with RA who have moderately severe to severe depressive symptoms. Failure to detect and treat depression may compromise patients' adherence to regimens and, ultimately, their health outcomes. Rheumatologists may require additional training and resources to effectively detect and treat depression in their patients with RA. Future research should examine patient- and physician-reported barriers to communicating about depression in rheumatology practices and use these findings to design innovative interventions that can be delivered effectively in busy rheumatologist practices.


Dr. Sleath had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Sleath, Chewning, Brenda De Vellis, Weinberger, Robert De Vellis, Tudor.

Acquisition of data. Sleath, Brenda De Vellis, Beard.

Analysis and interpretation of data. Sleath, Chewning, Brenda De Vellis, Weinberger, Robert De Vellis, Tudor, Beard.

Manuscript preparation. Sleath, Chewning, Weinberger, Robert De Vellis, Tudor, Beard.

Statistical analysis. Sleath, Tudor.

Grant writing. Chewning.