Barbara Dickey

Persons with dual diagnoses of substance abuse and major mental illness: Their excess costs of psychiatric care.

OBJECTIVES This study examined the costs of psychiatric treatment for seriously mentally ill people with comorbid substance abuse as compared with mentally ill people not abusing substances. METHODS Three different sources of data were used to construct client-level files to compare the patterns of care and expenditures of 16,395 psychiatrically disabled Medicaid beneficiaries with and without substance abuse: Massachusetts Medicaid paid claims; Department of Mental Health state hospital inpatient record files; and community support service client tracking files. RESULTS Psychiatrically disabled substance abusers had psychiatric treatment costs that were almost 60% higher than those of nonabusers. Most of the cost difference was the result of more acute psychiatric inpatient treatment. CONCLUSIONS Although the public health and financial costs of high rates of comorbidity are obvious, the solutions to these problems are not. Numerous bureaucratic and social obstacles must be overcome before programs for those with dual diagnoses can be tested for clinical effectiveness.

CONNECT: A Measure of Continuity of Care in Mental Health Services

This paper introduces a measure of continuity of care (CONNECT) developed for mental health services research. CONNECT addresses qualities of interpersonal interaction in service-user/practitioner relationships through 13 scales and one single-item indicator. The scales are grouped into five domains: knowledge, flexibility, availability, coordination, and transitions. Domains were derived from ethnographic data. Service users rate responses to items using 5-point scales. The measure is administered in interview format. CONNECT was developed for use with persons who have serious mental illness. Preliminary testing included cognitive interviews and two pilot studies. The results of a field test in which 400 persons with serious mental illness completed CONNECT indicate that the measure is easily administered and produces well-distributed responses. Five scales meet the .80 criterion for internal-consistency reliability for group-level research. Estimates of 2-week test-retest reliability indicate fair- to-good agreement. A broad initial validation strategy including known groups and convergent validity assessments produced results that will inform and focus future efforts. Next steps in the measure development process are discussed.

Housing costs for adults who are mentally ill and formerly homeless

The goal of this study was to evaluate the costs, under two different housing conditions, to the state mental health agency of caring for adults who are homeless and mentally ill. One hundred and twelve clients of the Massachusetts Department of Mental Health, living in psychiatric shelters, were randomly assigned to one of two housing types: Evolving Consumer Households or Independent Living apartments. For the next 18 months each client was followed so that the cost of treatment, case management, and housing could be collected and compared. The authors found that treatment and case management costs did not vary by housing type, but housing costs were significantly higher for those assigned to Evolving Consumer Households. Regardless of original housing assignment, treatment costs were lower for clients who remained where they were originally placed. The authors conclude that providing support for clients that increases housing stability reduces their need for treatment and that independent living arrangements may be a more costeffective policy choice.

Mental health cost models. Refinements and applications.

This paper describes a 2-year study whose goal was to refine Burton Weisbrods cost model for public programs for the chronically mentally ill. The authors made comprehensive cost assessments of all the resources, including treatment programs, used by a small sample of severely disturbed chronically ill patients. Refinement of the model included a method to assess capital costs of public facilities. The use of disaggregated patient information permitted analysis of cost differences between patients when adjusted for case mix. Patient costs over the study period ranged from

Associations between adherence to guidelines for antipsychotic dose and health status, side effects, and patient care experiences

24,000 to

Schizophrenia, substance use disorders and medical co-morbidity

99,000. Patient characteristics and change in clinical status account for 30% of the variance.

ASSESSING OUTCOMES IN CLINICAL PRACTICE

Background:One approach to improving quality of care is to encourage physicians to follow evidence-based practice guidelines. Examples of evidence-based guidelines are the PORT recommendations for the treatment of schizophrenia. However, few studies have examined the relationship between adherence to guidelines and patient outcomes in clinical settings. Objective:The purpose of this article is to report the relationship between guideline adherence to antipsychotic medication dose and self-reported health status, side effects, and perceptions of care. Research Design:This report is based on a subsample of patients from a larger prospective observational study of disabled Massachusetts Medicaid beneficiaries treated for schizophrenia. Subjects:Participants were 329 acutely ill, vulnerable, high-risk Medicaid adult beneficiaries enrolled after visiting any 1 of 8 psychiatric emergency screening teams for hospital admission evaluation. Measures:Dose levels, symptoms, and functioning from medical records; self-reports as data collected from BASIS-32, SF-12, and CABHS; and paid health benefit claims for psychiatric treatment were measured. Results:Approximately 40% of the patients in this study had daily antipsychotic doses well above the recommended range, but there was no evidence that their health status was better than those on doses below 1000 CPZ units recommended for acute episodes. High-dose levels had no relationship to baseline symptom profile or referral source. Conclusions:There was no evidence that health status was better on higher-than-recommended doses, but we cannot conclude that lower doses for some would have led to poorer outcomes. Physicians who believe that higher doses are more therapeutic for patients need to demand rigorous effectiveness research that tests whether there are benefits of higher doses and determine the ratio of those benefits to the clinical costs, including the risk of side effects.

Assessing consumer perceptions of inpatient psychiatric treatment: the perceptions of care survey.

OBJECTIVES: This study compared medical treatment costs of adults with schizophrenia to adults with both substance use disorders and schizophrenia. METHODS: This cross-sectional observational study used a paid claims data base to identify 6884 adults treated for schizophrenia. Twenty percent of these also had substance use disorder. We report the costs and likelihood of hospitalization for eight common medical diseases, and the categories of injuries and poisoning, and ill defined conditions. Multivariate analyses were used to adjust rates of treatment for age and sex differences in the comparison groups. RESULTS: There were higher rates of treatment for five of the eight medical disorders, higher treatment costs for two of the medical disorders and much higher costs for psychiatric treatment among those with comorbid substance use disorders. Both groups had high rates of treatment in the categories of injury and poisoning and ill defined conditions. CONCLUSIONS: Closer working relationships among mental health and medical professionals are needed to care for those with schizophrenia and substance use disorders: first, greater attention to the treatment of substance use disorders may improve the health status of those with schizophrenia, reduce their costly medical and psychiatric care and stabilize their psychiatric condition, and second, continuity of care among professionals may promote willingness to seek medical attention or alleviate misunderstandings when adults with schizophrenia present with medical problems.

Cost-Shifting in Managed Care

We are witnessing a remarkable explosion in interest and activity in quantifying outcomes and using these measures to enhance the value of clinical care. Outcomes assessment has become an imperative for clinical practice. This paper first will offer criteria for an ideal system of outcomes assessment. The paper will then review the principal domains of assessment for psychiatric practice and provide examples of instruments available in each domain. We will then describe the use of two instruments, one for clinical outcome and one for interpersonal aspects of patient satisfaction, developed and used at McLean Hospital. The relation between outcomes assessment and outcomes management will then be discussed. Finally, we will discuss the fundamental questions a clinical group or facility might consider in choosing outcomes measurement instruments.

Using health status measures with the seriously mentally ill in health services research.

BACKGROUND Consumer perceptions of behavioral health care are widely recognized as important quality indicators. This article reports the development and use of the Perceptions of Care (PoC) survey, a standardized public domain measure of consumer perceptions of the quality of inpatient mental health or substance abuse care. The goals were to develop a low-cost, low-burden survey that would address important quality domains, allow for interprogram comparisons and national benchmarks, be useful for quality improvement purposes, and meet accreditation and payer requirements. METHODS The sample was composed of 6,972 patients treated in 14 inpatient behavioral health or substance abuse treatment programs. The PoC survey was given to patients by program staff in the 24-hour period before discharge. RESULTS Aggregate reports and ratings of care identified areas that are highly evaluated by consumers, as well as areas that provide opportunities for quality improvement. Factor analysis identified four domains of care, and a 100-point score was developed for each domain. Regression analyses identified significant predictors of perceptions of care for use in computing risk-adjusted scores. Unadjusted and adjusted scores were presented to demonstrate the impact of risk adjustment on quality of care scores and relative ranking of programs. Examples were given of how programs used survey results to improve the quality of care. DISCUSSION Results demonstrated that the PoC survey is sensitive to detecting differences among inpatient behavioral health programs and can be useful in guiding quality improvement efforts. However, risk adjustment is important for appropriate interpretation of results.