Dr. Chen is affiliated with the Charles B. Wang Community Health Center, 268 Canal Street, New York, New York 10013 (e-mail, tchen@cbwchc.org). Dr. Huang is with the Department of Psychiatry, University of California, San Francisco. Ms. Chang is with the University of Colorado Health Sciences Center, Denver. Dr. Chung is with the New York University Student Health Center and the New York University School of Medicine, New York.

The U.S. Preventive Services Task Force has recommended that screening for depression among adults occur in clinical practices that are capable of providing evidence-based depression management (1). This is particularly relevant for Asian Americans for a variety of reasons. Several studies have shown high levels of mental health needs in the Asian-American community, including high rates of suicidal ideation among older Asian-American primary care patients (2), high suicide rates among elderly and adolescent Asian-American females (3), and high levels of depressive symptom distress among Asian-American primary care patients (4,5). Despite these indicators of need, Asian Americans are known to underutilize and delay seeking mental health services, and those who eventually do receive specialty mental health services have severe psychiatric symptomatology (6). Stigma associated with mental disorders and treatment, somatization, and a lack of awareness of psychological distress are some of the reasons hypothesized for this underutilization by Asian Americans (6). Previous health service utilization research suggests that mental health services centered in the community can improve patterns of utilization and outcomes for Asian-American patients (6,7,8). Primary health care clinics in the community, in particular, have been identified as an important setting for the detection and treatment of mental illness for Asian-American populations and other racial or ethnic minority groups (6).

However, some studies have found that patients' race or ethnicity can increase the risk of nondetection of mental health problems in primary care (4,5,6,7,8,9). Specific to Asian Americans, Chung and colleagues (4) found that primary care providers significantly underdiagnosed depressive symptom distress among Asian Americans compared with Latinos, despite similar prevalence rates in these two groups. These studies, confirmed by our own clinical experience, indicate the need to develop systematic methods of improving the detection of depression among Asian-American primary care patients.

The nine-item depression module of the Patient Health Questionnaire (PHQ-9) is designed for depression screening in primary care and has cutoff severity scores that may be useful for monitoring care (10). The PHQ-9 has been widely used in research and clinical care and appears to work well at detecting clinical change (11), but study of its use in Asian-American populations has been limited. One recent validation study showed that the Chinese-translated PHQ-9 used in this analysis measures a concept of depression that is statistically similar to that in the English-language PHQ-9 and can also be used comparably to detect and measure severity of depression in racially and ethnically diverse populations (12). In the analysis presented here, we evaluated the clinical utility of using the PHQ-9 as part of a standard clinical practice in a community health center that primarily serves immigrant, Chinese-American patients with limited English proficiency.

Participants

The sample for this analysis consisted of 3,417 adults who received an initial or annual physical examination between January and October 2003 at the Chares B. Wang Community Health Center (CBWCHC), a federally qualified community health center with three clinical sites in New York City. CBWCHC serves predominantly low-income Chinese Americans with limited English proficiency. Mental health services at CBWCHC are provided by the integrated mental health Bridge Program in the internal medicine clinic (13). In 2003 CBWCHC had 32,531 patient encounters at the internal medicine clinic from which the analytic sample was drawn. The project and analysis plan reported here was reviewed and approved by the New York University School of Medicine Institutional Review Board.

Instruments

Depression screening at CBWCHC uses a two-tier screening approach that includes the PHQ-9. The PHQ-9 is the depression screening module of the full Patient Health Questionnaire, a self-administered version of the Primary Care Evaluation of Mental Disorders (PRIME-MD) diagnostic instrument for common mental disorders (14). The PHQ-9 consists of nine questions based on the nine DSM-IV criteria for a major depressive episode. Each of the questions asks patients to select the frequency of the depressive symptoms that they experienced in the two weeks before survey administration. Scores for each item range from 0, not at all, to 3, nearly every day. Scores between 10 and 14 indicate a moderate level of depressive symptoms, scores between 15 and 19 indicate moderately severe major depression, and scores 20 and above indicate severe major depression (10).

A modified version of the two-item PHQ (PHQ-2) was used for initial screening. The PHQ-2 is a depression screener developed and validated by Kroenke and colleagues (15). The PHQ-2 asks about the frequency of depressed mood and anhedonia over the past two weeks. Considering cultural factors, we added an additional item to increase the sensitivity in the CBWCHC community. Because Chinese-American patients are known to somatize psychological distress (16,17), a third question was added to the PHQ-2 that asked whether in the past two weeks the patient had been "feeling tired or having difficulty sleeping." Patients who endorsed at least one positive symptom on this initial tier of screening were then given the full PHQ-9.

The initial three-item screen and the PHQ-9 were translated into Chinese and back translated into English by the mental health and health education teams at CBWCHC. The process of translation and back translation was repeated according to methodology outlined by previous research (18) until the clinicians felt that the Chinese version corresponded closely with the English version and would be meaningful to Chinese patients at CBWCHC.

Procedure

Patients who arrived at the CBWCHC for their routine physical examination were given the three-item initial screen by patient service representatives at the reception desk to fill out by themselves in either English or Chinese. They were informed that the purpose of the form was to help their doctor understand their health condition better. For patients with a positive symptom, the nursing staff would contact a care manager to administer the PHQ-9. Patients who were able to read would fill out the form themselves. A care manager (bachelor's-level social worker) would read the PHQ-9 questions to those who had difficulty with reading. The care manager then provided the results to the primary care provider before the clinic visit. If any of the patients scored 10 or higher on the PHQ-9, the primary care provider would review the PHQ-9 results with the patient, confirm the diagnosis, and decide on an action plan. Depending on depression severity, treatment planning options included a range of clinical activities, such as watchful waiting, initiation of an antidepressant medication, or referral of the patient for on-site mental health specialty treatment. The treatment planning guidelines used at CBWCHC are consistent with those in the MacArthur Depression Management Tool Kit (19), which are shown in Table 1.

Statistical analysis

Each patient's PHQ-9 depression score was entered into a registry for clinical monitoring and follow-up. The registry was converted into a SPSS format for statistical analysis. Basic descriptive statistics (frequencies, percentages, and means) and bivariate statistical analyses (chi square tests and t tests) were performed with SPSS version 12.

General screening results

As shown in Table 2, among the 3,417 individuals screened with the initial three items, 44.8 percent were male and 55.2 percent were female. The age range for this group was between 18 and 87 years; the mean age was 42.8 years. Of the screened individuals, 98.6 percent were Chinese and .5 percent were Vietnamese. The remaining .9 percent were white, African American, or other. The complete demographic characteristics of the individuals included in the analysis are listed in Table 2.

Of the 3,417 patients, 973 (28.5 percent) screened positive for depression on the initial three-item screen and were administered the full PHQ-9 scale at the same visit. Of these 973 individuals, 454 (46.7 percent) were male and 519 (53.3 percent) were female. PHQ-9 scores were not available for seven men and five women. The PHQ-9 score data analyses were based on the valid data of 961 individuals, which is 98.8 percent of all the individuals screened with PHQ-9. Women had significantly higher mean±SD PHQ-9 scores than men (6.7±4.7 compared with 5.7±4.5; p<.005). In terms of age and PHQ-9 scores for this group that initially screened positive, patients aged 55 to 64 years had the highest scores (mean of 7.1±5.6) and patients older than 74 had the lowest scores (mean of 4.6±3.1). The difference between these two age groups was statistically significant (p<.05).

A total of 141 of the 3,417 patients (4.1 percent) had clinically significant depressive symptoms (PHQ-9 score of 10 or higher). Sixty-eight (2.0 percent) had PHQ-9 scores between 10 and 14 (moderate depressive symptoms, or minor depression), and 45 (1.3 percent) had PHQ-9 scores between 15 and 19 (moderately severe major depression). A total of 28 (.8 percent) had PHQ-9 scores of 20 or above (severe major depression). The demographic characteristics of patients with clinically significant depressive symptoms are detailed in Table 3.

Women were more likely than men to have a PHQ-9 score of 10 or higher (89 women, or 17.3 percent, compared with 52 men, or 11.6 percent; χ²=8.35, df=3, p<.05). Figure 1 shows the gender differences in depression. Women were also more likely to be moderately depressed than men (47 women, or 9.1 percent, compared with 21 men, or 4.7 percent; χ²=7.31, df=1, p<.01).

Monitoring treatment progress

Of this group of 141 patients with at least moderate depressive symptoms (PHQ-9 score of 10 or higher), 114 patients (80.9 percent) agreed to and had treatment plans initiated by their primary care physician consistent with the guidelines presented in Table 1. Patients who received treatment had at least one follow-up PHQ-9 assessment by eight weeks. At this follow-up, 60 patients (52.6 percent) had a minimal clinically important difference, defined as 5-point drop of PHQ-9 score (10,20,21). The average decrease in PHQ-9 score by eight-week follow-up was 5.2±2.9. Forty of the 114 individuals treated (35.1 percent) had a 50 percent or greater decrease in their PHQ-9 score by eight weeks. Seven of these patients (6.1 percent) were found to be in remission, defined as a PHQ-9 score less than 5 (20).

Primary care providers mainly used the general term "depression" as a diagnosis. About 65 percent of the 114 patients had a diagnosis written as "depression" by their primary care physician. Twelve patients (10.5 percent) were referred to the on-site mental health professionals for consultation and treatment and received one of the DSM-IV major depressive disorder diagnoses. In terms of treatment received, six (5.3 percent) received antidepressant treatment from their primary care physician. Primary care physicians used watchful waiting as a way of follow-up with 42 patients (36.8 percent). For the rest of the patients (54 patients, or 47.4 percent), it is not clear what type of treatment was provided because the documentation is lacking in the progress notes.

Utility of the PHQ-9 in a primary care setting

The depression screening project in CBWCHC was started as a quality improvement initiative. Our experience with this project shows that the PHQ-9 is an efficient criterion-based screening measurement tool for the detection and management of depressive disorders among Chinese-American primary care patients. The PHQ-9 scoring system helps to establish the likelihood of depressive disorder diagnoses as well as grade depressive symptom severity for treatment planning. Primary care providers can continue to monitor treatment by performing reassessments with the PHQ-9. Having a consistent monitoring system may help primary care providers or mental health clinicians to monitor the effectiveness of their treatment.

The two-tier screening system that was established at CBWCHC and the brevity of the PHQ-9 caused minimal disruption of clinical operations. We were able to screen more than 3,000 Chinese-American patients in ten months in a busy urban community health center. We found that the PHQ-9 could be administered in ten minutes or less while the patient was waiting to be seen by his or her primary care provider. The length of a screening tool is an important consideration for its utility in primary care. On the basis of their experience of surveying Chinese Americans for depression in primary care, Yeung and colleagues (22) found that a shorter scale had a positive impact on patient participation. Our experience also showed that the refusal rates for the two screens were extremely low. The screening activities were originally built into our process for performing a routine physical examination. Our physicians and nursing staff were trained to discuss with patients the fact that they were given the screens as a way to better understand how stress might affect overall health. In addition, a research study is currently under way in CBWCHC in which potential participants are routinely screened with initial questions. More than 5,500 patients were given the initial screen, and the refusal rates for both screens have been less than 1 percent.

The PHQ-9 was also useful for monitoring treatment progress. Patients were readministered the PHQ-9 at least eight weeks after starting treatment. This score gave primary care physicians a clear idea about the nature of patients' depressive symptoms and gave both the patient and the physician an indication of treatment progress. Changes in PHQ-9 scores of 5 or more are considered clinically significant changes in general populations; however, future studies will need to confirm that a 5-point change also corresponds with a clinically significant change among Chinese Americans.

Prevalence of depressive symptoms among Chinese Americans

Our findings showed that 4.1 percent of Chinese-American primary care patients had moderate to severe levels of depression. The rate of depression that we found is generally consistent with that found in an epidemiologic study of major depression in a Chinese-American community (rate of 3.4 percent of 1,747 participants over 12 months) (23). However, the rate that we found is lower than that reported in the one previous study of Chinese Americans in primary care clinics. In their study, Yeung and colleagues (5) reported a current prevalence of major depressive disorder among Chinese Americans in a community health center in the Boston area of 19.6 percent (N=503); the Chinese version of the Beck Depression Inventory was used in this study (5).

There are several possible explanations for this discrepancy. Because the PHQ-9 is based on DSM-IV criteria, it is possible that the PHQ-9 had a more stringent basis of identifying depressive cases than the Beck Inventory. Also, we screened patients who came to the health center for an initial visit or annual physical examination. This group may be a healthier population than those found in Yeung's study. Also, compared with participants in our study, those in Yeung's study were older (mean age of 50±17.0) and were more likely to be female (60.4 percent).

Monitoring outcomes of treatment

The PHQ-9 was also used to monitor the outcomes of patients' treatment in this project. A total of 114 of the 141 patients who screened positive for depression agreed to receive and received treatment. At the eight-week follow-up, 52.6 percent of these patients had a minimal clinically significant improvement, 35.1 percent had a 50 percent reduction in PHQ-9 score, and 6.1 percent were in remission. These response and remission rates are comparable to clinical outcomes described in an article detailing treatment outcomes of depression using a medication-based algorithm, the Texas Medication Algorithm Project, in public-sector outpatient mental health settings (24). Of the 118 patients followed in that study, only 26.3 percent had a clinical response by the one-year follow-up, and 11.0 percent were in remission.

Limitations

There are several limitations of this analysis. Because the project reported in this article started as a quality-improvement initiative rather than as a research project, our results may not be broadly generalizable. But the PHQ-9-based depression screening system described in our article should be replicable in similar primary health care settings, if primary care physicians, nursing, and care management staff are appropriately prepared and ready for the task (1).

The individuals scoring 10 or greater on the PHQ-9 were clinically confirmed to have depression by primary care physicians before treatment was initiated. There was no formal validation of these cases with instruments such as the Structured Clinical Interviews for DSM-IV. In addition, primary care clinicians did not diagnose any additional psychiatric conditions that may have been present among those patients. The rates of clinically important depressive symptoms described in this analysis may not reflect the prevalence rate of depression for Chinese Americans because patients coming in for urgent or follow-up appointments were not screened at CBWCHC. The second limitation for this analysis is that follow-ups were limited to at most eight weeks in our registry. The treatment outcome rates reported here should be considered acute-phase outcomes. At our center a current randomized controlled trial is being conducted of the treatment of Chinese-American patients with depression in primary care; this trial will assess outcomes with the PHQ-9 at later time points.

Because Asian Americans generally underutilize specialty mental health services, screening for depression in primary care is crucial for early detection and treatment. The PHQ-9 is a useful tool for screening depression among Chinese-American primary care patients, because of its brevity and its capability to help establish a DSM-IV-based diagnosis of major depression. The two-tier procedure outlined in our article can be implemented as part of routine depression screening in busy primary care settings. In addition to making criteria-based diagnosis of depression, the PHQ-9 can also be used as a measure of depression severity to monitor treatment progress and guide clinicians' treatment decisions.

The authors thank Sarah Yip, B.A., for her assistance in care management using the PHQ scales. This analysis was supported by the Center for the Study of Asian American Health and sponsored by grant 5-P60-MD0005-38 from the National Center for Minority Health and Health Disparities, National Institutes of Health.