There was variability in how depression was discussed and documented in the medical record. In combined scenarios A and B for cases where depression was diagnosed, it was both discussed with the patient and recorded in the medical record in 77.2% of encounters. It was discussed with the patient but not noted in the medical record in an additional 6.7% of encounters and appeared in the medical record but with no patient discussion in 5.4% of encounters. In 10.7% of encounters, patients received a prescription for an antidepressant or a recommendation for ongoing counseling with a psychologist or social worker, though the word “depression” was not used with the patient or noted in the medical record.
No statistical differences were found by region in the proportion of patients given a diagnosis of depression. Differences found in diagnosis and management according to physician and patient sex have been reported elsewhere.27 Briefly, male physicians explored symptoms and discussed a diagnosis with women significantly more often than with men. Both male and female physicians recommended counseling more often for women portraying scenario B standardized patients than for men portraying the same type of patient.
Discussion
Our study demonstrates that when mental distress is obvious, primary care physicians recognize it. When a patient presents with a somatic chief complaint, such as headaches or less obvious mental health distress, recognition is still common, although not universal. The claim that recognition of depression is only 50% in primary care settings should be regarded with reservation.
We identified some areas where recognition could be improved. All physicians who did not recognize depression in scenario B failed to ask the patient about depressed mood, and only 19% asked about anhedonia. When the physician asked about either, the recognition rate increased by approximately 50%, and if they asked about both, it approached 100%. We found that recognition of depression was 100% with patients presenting with major depressive disorder, which included a total of 7 DSM criteria. It was diagnosed in approximately 50% of patients portraying a headache presentation with minor depression without a prompt by the standardized patient. We found that the one-sentence prompt (“I’ve had a tough year”) activated an additional 30% of the physicians to consider depression, bringing the rate of diagnosis up to almost 80% for patients portraying scenario B.
This finding underscores the importance of physicians’ pattern of approach in determining the patient’s reason for the visit. We found in our focus group study25 that physicians describe 3 approaches in pursuing a diagnosis of depression. These include a biomedical approach, where physicians’ threshold for considering depression is high. In this case, providers pursue diagnostic testing of physical causes of the symptom first and address depression when physical explanations are lacking. In the psychosocial approach, physicians’ threshold for considering depression is low, and depression is considered as a possibility first; response to an SSRI determines if depression is present. The biopsychosocial approach integrates the first 2 approaches. For that approach the threshold for pursuing depression is moderate, and biomedical and psychosocial issues are simultaneously pursued in investigating patients’ problems. We found focus group physicians identified with all 3 approaches and often crossed over among the approaches, depending on patient cues. We suggest that those study physicians who did not diagnose depression, even with the standardized patient’s prompt, did not cross over among approaches despite 2 visits with the patient. Establishing a longer relationship and obtaining additional cues from the patient may have reduced their threshold for considering depression.
Certain diagnostic criteria were pursued more than others in diagnosing depression. Asking questions about sleep, for example, was strongly associated with such a diagnosis. Questions about sleep were pursued by more than 98% of the physicians diagnosing depression in patients enacting scenario B versus 44% who asked about sleep but did not diagnose depression. Anhedonia, weight changes, and fatigue were addressed in at least 56% to 76% of both patient encounters, while mood, suicide ideation, and family history of mental illness were addressed more often with patients enacting scenario A than B. This is consistent with a study done by Brody and colleagues,28 who found that asking about sleep, anhedonia, low self-esteem, and appetite led to recognition of the majority of cases of major depression in primary care. Substance abuse and previous history of mental health disorders were assessed in less than 40% of encounters. Asking about anhedonia and mood may increase recognition of depression. If depression is suspected, asking about previous mental health history and substance abuse, and exploring suicidal thoughts are important.