The association between body mass and health care expenditures.

With several studies estimating the health care costs attributable to obesity-related medical conditions, the economic consequences of being overweight are beginning to come into focus. The present study complements this growing body of literature by directly estimating health care costs across a broad range of body mass index values. Data were obtained from the 1987 National Medical Expenditure Survey (NMES) public use data tapes and retrospective analyses conducted on NMES Household Survey data only. The analyses included a total of 16,217 individuals between the ages of 18 and 65. Four classes of health care utilization and expenditures were derived using the NMES data: (1) use of any health care service and total health care expenditures, (2) use of inpatient services and inpatient expenditures, (3) use of outpatient services and outpatient expenditures, and (4) use of prescription medication and medication expenditures. Estimates based on our findings suggested strong relationships between body mass and the likelihood of using health care services and between body mass and average annual health care expenditures for both men and women. Increased body mass was associated with increased expenditures. However, this association was greater among males than among females and did not hold for individuals in the lowest body mass category. Ideal body mass was associated with 6.3% to 36.1% lower annual health care expenditures among females and 3.6% to 18.2% lower health care expenditures among males. The results of this set of analyses suggest that health care expenditures increase as weight deviates from the ideal-that is, health care expenditures among both underweight and overweight individuals in the United States were increased in relation to ideal weight. Separate analyses including weight-related diseases such as diabetes and hypertension indicated that body mass increased health care expenditures largely by increasing the risk for these costly chronic medical conditions.

Reliability of paper-pencil assessment of drug use severity.

This research examines whether self-reported information about drug use severity can be obtained as reliably using a paper/pencil format as the traditional interviewer format. A sample of 67 patients seeking treatment for substance use disorders was recruited from a Veterans Administration Medical Center. Subjects self-reported information related to drug use severity using both paper/pencil and interview formats. The results of comparisons of the two approaches indicate that method of test administration does not affect the test-retest reliability for most questions tested. Test-retest reliability estimates for these relatively brief indicators of drug use severity generally ranged from good to excellent. Although assessing drug use severity using a paper/pencil format is certainly not appropriate for all individuals with substance use disorders, for many individuals and situations it may prove to be a cost-effective alternative to the interview format. Further research is required to determine if parallel paper/pencil versions of widely used interviews can be developed.

Comparison of DSM-III-R symptoms for alcohol dependence between patient self-report and clinician interview or the Structured Clinical Interview for DSM-III-R.

This study sought to determine which Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R), symptoms of alcohol dependence were most sensitive to under-reporting by 78 inpatients from alcohol treatment programs. We hypothesized that patients would be more reluctant to report social/behavioral symptoms (lost time, hazardous behavior or major role interference, and reduced activities) than psychological or physiological symptoms. Patient endorsement of symptoms on a self-administered diagnostic questionnaire was compared with parallel items on the Structured Clinical Interview for DSM-III-R (SCID) and clinician interview. Bias and Prevalence Adjusted Kappas for individual symptom agreement ranged from -.02 to .87. Subjects endorsed fewer symptoms per category than either the SCID or clinician interview. Using a Student-Newman-Kuels post-hoc analysis at the p < .05 level, we found that mean agreement for the social/behavioral category was significantly lower than mean agreement for the other categories.

Correlates of patterns of substance abuse among patients with schizophrenia.

OBJECTIVE: Most previous research on substance abuse among patients with schizophrenia has not considered the types of substances abused or the confounding influences of polysubstance abuse. The authors' goals were to identify patterns of substance abuse among a sample of subjects with schizophrenia and to determine demographic and clinical correlates of these patterns. METHODS: Subjects with schizophrenia were identified from the Epidemiologic Catchment Area study data base, which also provided information on the types of substances abused and on selected demographic and clinical characteristics. Latent class analysis was used to group subjects based on their patterns of substance abuse. Logistic regression was used to identify demographic and clinical correlates of the patterns. RESULTS: Three latent classes based on patterns of substance abuse were identified: no substance abuse, abuse of alcohol and cannabis, and polysubstance abuse. Subjects with either of the two patterns of substance abuse were more likely to be younger and male and to have depressive symptoms. CONCLUSIONS: The results do not support a link between abuse of specific substances and specific clinical symptoms but do suggest a general link between substance abuse and affective disturbance among schizophrenic patients.