Screening for Mental Illness in a Veterans Affairs Women's Health Clinic

A total of 225 consecutive patients of a VHA women's health clinic were invited to complete a demographic questionnaire and a psychiatric screening instrument during routine health care visits in 1999. Screening was conducted with the diagnostic portion of the Primary Care Evaluation of Mental Disorders Patient Health Questionnaire (PRIME- MD PHQ) (7,8). The study was reviewed and approved by our human studies subcommittee.

The PRIME-MD PHQ is a self-administered instrument that screens for eight psychiatric disorders, divided into two categories: threshold disorders, which conform to specific DSM-IV diagnoses, and subthreshold disorders, which require fewer criteria than DSM-IV diagnoses. In the former category are major depressive disorder, panic disorder, other anxiety disorders, and bulimia nervosa. In the latter category are somatoform disorders, other depressive disorders, binge-eating disorder, and probable alcohol abuse or dependence (7). Women who screened positive for a somatoform disorder were clinically evaluated by their primary care provider to determine whether there was an adequate biomedical explanation for their somatic complaints. The PRIME-MD PHQ includes a self-report scale of social and occupational impairment due to the psychiatric symptoms. Possible responses range from 1, no impairment, to 4, severe impairment. Information about the women's active medical problems was obtained by record review.

A total of 223 of the 225 patients agreed to fill out the questionnaires; 209 (93 percent) completed questionnaires. Only completed questionnaires were analyzed. The mean±SD age of the respondents was 47.7±15.1 years (range, 21 to 93 years). A total of 142 (68 percent) of the patients were non-Hispanic white, 40 (19 percent) were African American, ten (5 percent) were Native American, and eight (4 percent) were Hispanic. Sixty-seven women (32 percent) were married, 79 (38 percent) were separated or divorced, 44 (21 percent) had never been married, and 19 (9 percent) were widowed. The mean±SD number of years of education was 13.6±2.0, and 142 women (68 percent) reported having had some college education. The mean±SD number of dependents living with the women was .61±.91. Seventy-seven (37 percent) of the women were employed full-time, 27 (13 percent) were employed part-time, and 105 (50 percent) were unemployed.

Ninety-four women (45 percent) had a service-connected disability. Current medical problems included musculoskeletal conditions (144 women, or 69 percent), hypertension (77 women, or 37 percent), gastroesophageal reflux disease (44 women, or 21 percent), and obstructive pulmonary disease (40 women, or 19 percent). A total of 186 women (89 percent) were established patients of the clinic.

Ninety-four of the women (45 percent) screened positive for at least one PRIME-MD PHQ disorder, 46 (22 percent) for two or more coexisting disorders, and 40 (19 percent) for subthreshold conditions only. Fifty (24 percent) screened positive for somatoform disorder, 38 (18 percent) for panic disorder, 31 (15 percent) for other anxiety disorders, 19 (9 percent) for other depressive disorders, 17 (8 percent) for binge-eating disorder, 15 (7 percent) for major depressive disorder, 13 (6 percent) for probable alcohol abuse or dependence, and two (1 percent) for bulimia nervosa.

Women who were under the age of 50 years were more likely to have a PRIME-MD PHQ disorder (F=4.03, df=2, p=.019), as were women who had a service-connected disability (χ²=10.75, df=4, p=.029). Greater functional impairment was associated with a higher number of screen-positive disorders (χ²=97.85, df=6, p<.001).

The prevalence of psychiatric disorders was high among the female veterans in this study relative to rates in some community primary care samples, particularly in the case of somatoform and anxiety disorders (4,5,7,8). In the PRIME-MD 1000 study, Linzer and associates (5) showed that the prevalence of mental disorders among female primary care patients was 46 percent, with 26 percent of the patients having two or more disorders. However, compared with women in the PRIME-MD 1000 sample, the patients from the VHA women's health clinic had a lower prevalence of mood disorders (16 percent and 31 percent, respectively) and a higher prevalence of anxiety disorders (33 percent and 22 percent), somatoform disorders (24 percent and 18 percent), and probable alcohol abuse or dependence (6 percent and 2 percent).

In a study by Spitzer and colleagues (8), 27 percent of female primary care patients had a PRIME-MD PHQ disorder. Our study found a prevalence of PRIME-MD PHQ disorders of 34 percent, excluding somatoform disorder, which was not measured in the study by Spitzer and associates. The women in our study had a higher prevalence of anxiety disorders (33 percent compared with 11 percent), which may account for the difference in the prevalence of disorders between the studies (8). The PRIME-MD PHQ validation studies also found that the prevalence of disorders varied considerably among sites and may reflect demographic factors (7,8).

The finding that more than 50 percent of the patients who screened positive for a mental disorder also screened positive for multiple disorders emphasizes the importance of screening for comorbid conditions. Women who screened positive for multiple mental disorders also reported higher levels of impairment in social and occupational functioning. Women under the age of 50 and those with a service-connected disability were significantly more likely to have mental disorders. Although younger women may have more problems, it is also possible that older women are less likely to volunteer information about psychiatric symptoms, in which case other approaches to screening older women for mental disorders may be needed. Nearly half the women in this study had a service-connected disability, and these women were more likely to screen positive for a mental disorder, indicating that they may have been more vulnerable to mental health problems.

Limitations of our data include the relatively small sample and the use of self-report measures that were not confirmed with a structured clinical interview. Because the PRIME- MD PHQ is a screening instrument, positive results need to be evaluated further. Finally, the functional impairment scale provides only a tentative estimate of a patient's level of functioning.

The high prevalence of self-reported mental disorders and associated functional impairment in this clinic sample highlights the need for integrated delivery of mental health and primary care services. Further research is required to examine the unique needs of female veterans in primary care settings and to compare their needs with those of male veterans.

The authors thank Tim Hancock, Ph.D., former psychology intern, for his assistance in data analysis, and Barbara Zicafoose, M.S.N., A.N.P., for her valuable assistance in the women's health clinic.

Dr. Bader is director of consultation-liaison psychiatry at the VA Medical Center in Salem, Virginia, and assistant clinical professor of psychiatric medicine at the University of Virginia School of Medicine, Roanoke-Salem Program. Dr. Ragsdale is a clinical psychologist with the Veterans Affairs Medical Center in Salem, Virginia. Dr. Franchina is a professor in the psychology department at Virginia Polytechnic and State University in Blacksburg. Send correspondence to Dr. Bader at the VA Medical Center (116A7), 1970 Roanoke Boulevard, Salem, Virginia 24153 (e-mail, [email protected]).