Underuse of antidepressants in major depression: prevalence and correlates in a national sample of young adults.

BACKGROUND: Epidemiologic studies have reported disturbingly low rates of treatment for major depression in the United States. To better understand this phenomenon, we studied the prevalence and predictors of antidepressant treatment in a national sample of individuals with major depression. METHOD: Between 1988 and 1994, 7589 individuals, aged 17-39 years and drawn from a national probability sample, were administered the Diagnostic Interview Schedule as part of the National Health and Nutrition Examination Survey. Interviewers asked about prescription drug use and checked medication bottles to record the name and type of medications. RESULTS: A total of 312 individuals, or 4.1% of the sample, met DSM-III criteria for current major depression. Only 7.4% of those with current major depression were being treated with an antidepressant. Among individuals with current major depression, being insured and having a primary care provider each predicted a 4-fold increase in odds of antidepressant treatment; telling the primary provider about depressive symptoms predicted a 10-fold increase in treatment. CONCLUSION: The study's findings support the notion that a serious gap exists between the established efficacy of antidepressant medications and rates of treatment for major depression in the "real world." Underreporting of depressive symptoms to providers and problems with access to general medical care appear to be 2 major contributors to this problem.

Cross-system service use among psychiatric patients: data from the Department of Veterans Affairs.

This study examines the cross-system use of non-Department of Veterans Affairs (VA) services in a sample of psychiatric patients from the VA in 1990. Data were collected over a two-week period on all mental health outpatients and included demographic information, diagnoses, and self-reported use of non-VA services in the previous two weeks and six months. In the entire sample, 10.6% and 23.3% reported cross-system use in the previous two weeks and six months, respectively. Predictors of cross-system use were lower VA utilization, a nonschizophrenic diagnosis, not having a VA service-connected disability, and being female. These data indicate that a substantial proportion of VA mental health patients are using non-VA services. Utilization patterns indicate that they may be substituting non-VA for VA services. These results are unlikely to be unique to VA, and rates of cross-system use will likely increase in all health care systems as financial restrictions increase.

Diversion from jail of detainees with substance abuse: the interaction with dual diagnosis.

The objective of this retrospective study was to determine if a jail diversion program reduced days of incarceration over the year following arrest in a sample of detainees with substance abuse disorders who had been arrested for a minor (misdemeanor) crime (n = 252) compared to a group (n = 95) who were not diverted. Particular attention was paid to comparing those singly diagnosed to those dually diagnosed. Analyses indicated that 1) detainees with substance abuse alone were less likely to be diverted than those with a dual diagnosis; 2) jail diversion reduced incarceration time during the next year; and 3) the effect of diversion differed depending upon the level of criminal charge. Diversion significantly reduced jail time only among those who were arrested for the more serious of the minor offenses that are associated with longer jail sentences.

The effects of a jail diversion program on incarceration: a retrospective cohort study.

UNLABELLED: The objective of this retrospective cohort study was to determine whether a jail diversion program significantly reduced the number of days of incarceration over the year following arrest in a sample of seriously mentally ill (SMI) people who had been arrested for a minor crime. METHODS: A group of 314 SMI detainees were diverted out of jail and into mental health treatment. They were compared with a sample of 124 people who would have been eligible for diversion but were not diverted. For each group, the authors compared the total days incarcerated in the year after index arrest. Analyses indicated that jail diversion significantly reduced incarceration time during the next year (40.51 versus 172.84 days, p = .0001). However, the effect of diversion differed depending upon the level of criminal charge: diversion significantly reduced jail time only among those who were arrested for more serious offenses. Those arrested for Class D felony and Class A misdemeanor charges and diverted into mental health treatment spent significantly less time in jail in the next year than those not diverted (260 and 110 fewer days, respectively; p = .0001 for both). Those arrested for Class B or C misdemeanors had similar days incarcerated regardless of diversion. These results remained after adjusting for age, race, gender, and diagnosis. This study presents the first evidence that jail diversion may produce positive longitudinal criminal justice outcomes for SMI people. It also suggests that diversion may not reduce incarceration in all sub-groups of SMI people who are arrested for minor crimes.

The cost of treating substance abuse patients with and without comorbid psychiatric disorders.

OBJECTIVE: Data from a national sample of patients with a primary diagnosis of a substance use disorder were analyzed to examine whether having a comorbid psychiatric diagnosis-a dual diagnosis-was associated with increased costs of health services over a six-year period and whether dually diagnosed patients used particular types of services more frequently. METHODS: A national sample of substance abuse patients being treated in Veterans Affairs (VA) facilities were classified into two groups-those with a dual diagnosis (N=3, 069) and those with a single diagnosis of a substance use disorder (N=9,538). The sample was identified from two sources during a two-week period in 1990: outpatients in specialty substance abuse clinics and inpatients discharged with a substance-related diagnosis. Administrative data were used to track VA health care utilization and costs between 1991 and 1996. RESULTS: Dual diagnosis was associated with a significantly increased total cost of care over the six years, which was primarily explained by increased utilization of outpatient psychiatric and substance abuse services. Costs for both groups decreased over time, but they decreased faster among dually diagnosed patients. CONCLUSIONS: Having a comorbid psychiatric diagnosis appears to consistently increase the cost and utilization of services among patients with a primary diagnosis of a substance use disorder. These results are consistent with previous findings for dually diagnosed patients with a primary psychiatric diagnosis. The increased cost may simply reflect the greater severity of illness among dually diagnosed patients, but it may also indicate fragmented and inefficient service delivery.

Mental illness as a predictor of satisfaction with inpatient care at Veterans Affairs hospitals.

OBJECTIVE: Patient satisfaction ratings are increasingly being used as an indicator of quality of care. However, satisfaction scores do not account for differences in satisfaction among patients that may be attributable to sociodemographic characteristics or type of illness rather than to the quality of service delivery. This study examines the role of psychiatric diagnosis in satisfaction with inpatient care delivered at Department of Veterans Affairs (VA) hospitals. METHODS: Data were taken from a large national VA customer feedback survey of patients discharged from VA hospitals between June 1 and August 31, 1995 (N=38,789). Analyses examined whether patients discharged with a psychiatric diagnosis were less satisfied with care than those discharged with a nonpsychiatric diagnosis. RESULTS: Patients with a psychiatric diagnosis were less satisfied with their care, regardless of whether they were treated in a psychiatric treatment program or a medical unit, and this relationship remained after the analysis adjusted for other determinants of satisfaction. The association between low satisfaction and psychiatric diagnosis was more pronounced among nonblack patients. CONCLUSIONS: The results suggest the need for caution in using patient satisfaction measures to compare mental health programs and other health care programs.

The quality of VA mental health services.

This study examines whether male and female veterans differ on either subjective or objective measures of the quality of VA mental health care. The study sample were all discharged with a psychiatric diagnosis from a VA inpatient mental health program. Results indicate that women were less satisfied with inpatient care, but were mixed on their satisfaction with overall mental health care. Males and females did not differ on the quality of inpatient care, but women had significantly higher quality of outpatient care. These results not only highlight the need to stratify or adjust quality measures by gender, but also highlight the potential confounding effect of sub-group specific health behaviors on measures of quality.

The use of VA and non-VA mental health services by female veterans.

OBJECTIVES: This study compared Department of Veterans Affairs (VA) and non-VA mental health service use among male and female veterans. Because female veterans are a distinct minority in VA, it was hypothesized that they would be less likely to use VA mental health care than would male veterans. METHODS: Using data from a national sample of Vietnam and Vietnam-era male (n = 1,612) and female (n = 736) veterans, the following were examined: (1) gender differences in use of VA mental health services, (2) gender differences in use of non-VA mental health services, and (3) differences in utilization of mental health services across subgroups defined by psychiatric diagnosis, adjusting for sociodemographic and need variables. RESULTS: Female veterans were equally likely to use VA mental health services as male veterans, but were substantially more likely to use non-VA mental health services. This suggests that the demand for mental health services overall is substantially greater among female veterans than among male veterans and, by implication, that the equal levels of observed VA service use actually represent underutilization of VA services on the part of female veterans. Logistic regression models showed that these utilization patterns were consistent across diagnostic subgroups. CONCLUSIONS: Special efforts, such as the development of women's specialty mental health clinics, may be needed to enhance the acceptability of VA mental health services to female veterans. This study also highlights the importance of considering overall demand for services in addition to more objective diagnostic data in evaluating the adequacy of service delivery and its accessibility.

Using recidivism rates as a quality indicator for substance abuse treatment programs.

Recidivism is a widely used quality indicator for inpatient substance abuse care. However, unadjusted recidivism rates do not account for important confounding variables, which may lessen their usefulness as a quality indicator. Using a study of a statewide network of inpatient substance abuse services in Connecticut, the authors present a method for sampling existing administrative data and adjusting recidivism rates. The method can be used by managers of provider networks to assess whether patient subgroups with different demographic or geographic characteristics have equal access to care; to check for potential weaknesses in services, facilities, or systems; and to identify programs with unusually high or low recidivism rates for improvement or replication.

Female veterans' use of Department of Veterans Affairs health care services.

OBJECTIVES: As access of women to mental health services has become increasingly important, empirical research has begun to examine the determinants of mental health care utilization across gender. This article examines the effect of being an extreme minority on utilization of Department of Veterans Affairs (VA) health services by female veterans. METHODS: Data were collected on a representative national sample of veterans in 1992 as part of the National Survey of Veterans. These data included information on sociodemographic variables, military service variables, physical health and disability, and health services utilization. The authors examined whether women who used health services in 1992, and who were eligible for VA care, differed from men on the likelihood of using any VA health services and on the likelihood of use of VA outpatient and inpatient health services. In addition, we compared VA health care utilization among subgroups of veterans with physical and mental disorders, and compared self-reported reasons for choice of health care provider, across gender. RESULTS: Results indicated that female veterans were less likely than male veterans to use VA health services. This difference was explained by lower utilization by women of VA outpatient services, since inpatient admission rates were the same across gender. The lower outpatient utilization was specific to women with self-reported mental disorders. Women with physical conditions did not differ from men with similar conditions in their VA outpatient utilization. Finally, men and women did not differ on their reasons for choosing VA or non-VA care. CONCLUSIONS: The authors conclude that extreme gender minority status appears to affect outpatient utilization rates at the VA among women with mental disorders, perhaps because of the more personal or sensitive nature of the services involved. Further research is needed to understand why certain women may be underutilizing VA outpatient services and on the consequences of minority gender status for health service utilization, more generally.

Long-term patterns of service use and cost among patients with both psychiatric and substance abuse disorders.

OBJECTIVES: This is a longitudinal study designed to determine: (1) if patients dually diagnosed with psychiatric and substance abuse disorders incur higher health care costs than other psychiatric patients and (2) if higher costs can be attributed to particular subgroups of the dually diagnosed or types of care. METHODS: Two cohorts of veterans treated in Veterans Affairs mental health programs at the start of fiscal year 1991 were followed for 6 years: one cohort of inpatients (n = 9,813) and the other of outpatients (n = 58,001). Data were analyzed on utilization of all types of Veterans Affairs health care. Repeated measures analysis of variance was used to examine cost differentials between dually diagnosed patients and other patients. RESULTS: Dually diagnosed outpatients incurred consistently higher health care costs than other psychiatric outpatients, attributable to higher rates of inpatient psychiatric and substance abuse care; however, this difference decreased with time. Costs were substantially higher in the inpatient cohort overall, but there were no differences in cost between dually diagnosed and other patients. CONCLUSIONS: In an atmosphere of cost cutting and moves toward outpatient care, the dually diagnosed may lose access to needed mental health services. Possibilities of developing more intensive outpatient services for these patients should be explored.

Preventive health care for mentally ill women.

Utilization of preventive medical care was compared for two low-income groups--47 women with serious mental illness in an urban mental health center and 17 women patients at a primary care center. Appropriate preventive care was defined as at least one physical examination, a Pap test, and a breast examination in the past five years and a mammogram if the patient was over age 40. Receipt of preventive care by women in both settings was similar. Histories of physical and sexual abuse were prevalent in both groups, and a history of abuse was associated with less frequent receipt of preventive care. Results indicate that procedures to identify and provide services to women with abuse histories should be further developed.

Trends over a decade for a general hospital psychiatry unit.

This study examines the composition and delivery of services in a general hospital inpatient psychiatry unit during a 10-year period. Multiple regression techniques were used to assess the association of clinical, insurance, and demographic data with length of stay and likelihood of readmission for all admissions from 1985-1993. Two variables became progressively associated with readmission--Medicaid and psychotic diagnosis. The results indicate that: (1) the hospital is increasingly treating a poorer, sicker group of patients with shorter lengths of stay and more readmissions, and (2) the rise in readmissions, particularly within vulnerable populations, could represent an inadequate length of initial treatment. Future research should further investigate the generalizability of these results and implications for quality of inpatient care.

Utilization of mental health services by women in a male-dominated environment: the VA experience.

OBJECTIVE: The study examined whether women in the Veterans Affairs system use mental health services at a lower rate than men because the system is geared to treat a mostly male population. METHODS: Data were obtained on a national cohort of patients utilizing specialty mental health services in the VA during a two-week period in fiscal year 1991 (N = 70,979). Analyses included comparison of the proportion of women among treated veterans with the age-adjusted proportion of women among all veterans, comparison across gender of the likelihood of use of any general psychiatric services or substance abuse care in 1991, comparison of the intensity of services used (inpatient days and outpatient contacts) by service users in 1991, and comparison of the likelihood of receiving care and the intensity of mental health services received two years later. RESULTS: Overall, 3.95 percent of veterans who used VA mental health specialty services were women; 4.02 percent of all veterans were women. No significant differences between genders were found in use of general psychiatric services, either in the likelihood of any use or the intensity of services used. However, women were significantly less likely to receive substance abuse care (16.3 percent of women versus 71.2 percent of men); once receiving care, they used a similar intensity of substance abuse services. CONCLUSIONS: Being a woman does not appear to have a substantial effect on overall access to VA mental health services or use of general psychiatric services; however, women use VA substance abuse treatment services at a lower rate than men.

Mental disorder as a risk factor for human immunodeficiency virus infection in a sample of veterans.

People who suffer from mental disorders are at increased risk for becoming infected with HIV. There have been no studies that show whether particular psychiatric disorders present an increased risk for HIV infection in samples of nonpatients. This article uses data from the 1992 National Survey of Veterans to determine if veterans with posttraumatic stress disorder (PTSD), or with other mental or emotional problems, are at increased risk for HIV infection. The results indicate that the combination of PTSD and substance abuse increased the risk of HIV infection by almost 12 times over those without either. This is evidence of a particular psychiatric disorder increasing risk for HIV. Although cross-sectional, these data allow some conjecture about the timing of the onset of PTSD in relation to HIV infection. These results present powerful evidence that mentally ill persons such as those with PTSD, who may be underserved for health services including AIDS prevention efforts, should be targeted as an at-risk group.

Subjective ratings of emotional health as a risk factor for major depression in a community sample.

BACKGROUND: Although subjective ratings of health have been shown to be accurate predictors of physical health outcomes, there is little research on the association between subjective emotional health (SEH) and psychiatric outcomes. METHOD: This paper utilises data from the Epidemiologic Catchment Area study to explore the relationship between baseline SEH and the risk for major depression in the next year. Both recurrent and incident episodes of depression are outcomes of interest. RESULTS: The age- and gender-adjusted relationship between SEH and depression is quite significant, and remains so after adjusting for other factors associated with major depression. The more positive the SEH rating, the lower the risk of an episode of depression in the next year. CONCLUSIONS: Some possible explanations for this association are explored, including possible confounders that were not accounted for and the possibility that SEH ratings pose an independent risk for major depression.

Depression in former school-age mothers and community comparison subjects.

This paper examines the lifetime prevalence and correlates of major depressive disorder in a cohort of former school-age mothers and compares the prevalence and correlates to those found in women of the same age, ethnicity, and geographical location as the school-age mothers. The sociodemographic characteristics show some striking differences. The former young mothers were less likely to be on public assistance (19% vs 42%), but were more likely to be working (78% vs. 55%), to have completed high school or college and to meet the DSM-III criteria for depression (10.7% vs. 4.9%) than the sample of community women. The only factor related to depression in the former school-age mothers was a diagnosis of drug/alcohol abuse or dependence.

The effects of cognitive impairment on 9-year mortality in a community sample.

Using 9-year mortality data on a community sample of 3,560 adults aged 40 and over, this study assessed the effects of cognitive functioning and one-year declines in cognitive functioning on mortality controlling for comorbid chronic medical illness, physical disability, and psychiatric illness. The study determined the 9-year vital status and, among the decreased, date of death of respondents who were first interviewed in 1980 by the New Haven Epidemiologic Catchment Area Study. Mortality risk by cognitive functioning, as assessed by the Mini-Mental State Examination (MMSE), was estimated using Cox Proportional Hazards Models controlling for baseline assessments of physical and mental health. For both men and women, lower scores on the MMSE decreased the risk of survival, although the effect was stronger for younger respondents than older respondents. Decline in MMSE scores over the course of one year had no additional effect on mortality beyond the resulting MMSE score. Cause-specific mortality was also examined.

Social and physical health risk factors for first-onset major depressive disorder in a community sample.

Using prospective data on 3,170 respondents aged 18 years and over who were at risk for a first-onset major depression from the New Haven site of the Epidemiologic Catchment Area (ECA) study, these analyses assessed the effects of social status, physical health status, and social isolation on first-onset depression in a 1-year period, controlling for demographic characteristics and baseline psychiatric factors. Among the assessed potential risk factors, poverty status [odds ratio (OR = 2.034, P < 0.05)] and confinement to a bed or chair (OR = 4.015, P < 0.05) were independently associated with an increased risk for a first-onset depressive episode when controlling for gender, age, past history of substance abuse, and subclinical depressive symptoms. The effects of poverty, and to a lesser degree homebound status, were substantially reduced when controlling for degree of isolation from friends and family, suggesting that social isolation mediates some of the relationships between social and physical statuses and major depression.

Psychiatric status and 9-year mortality data in the New Haven Epidemiologic Catchment Area Study.

OBJECTIVE: This study examined the effects of nine axis I psychiatric disorders, as assessed by the Diagnostic Interview Schedule, on the risk of mortality over a 9-year period among a community sample of 3,560 men and women aged 40 and older. METHOD: The study identified the vital status as of Oct. 1, 1989, of respondents who were first interviewed in 1980 by the New Haven Epidemiologic Catchment Area study. Mortality risk by psychiatric status was estimated by using Cox proportional hazards models. RESULTS: Nine years after the baseline interview, it was confirmed that 1,194 (33.5%) of the respondents were deceased and 2,344 (65.8%) survived; the vital status of 22 (0.6%) remained unknown. When the relative risk of mortality was adjusted for age, several disorders--major depression, alcohol abuse or dependence, and schizophrenia--increased the likelihood of mortality. CONCLUSIONS: These data are further evidence of the negative outcome of some psychiatric problems even when assessed in community samples. The relatively high prevalence of depression and alcohol disorders indicates the far-reaching impact that these problems have on community health in general.