Page 1 of 1

Compendium of Research Related to Mental Health (continued)


PI: Kenneth B. Wells, M.D., M.P.H., RAND
Grant No.: R01 HS08349
Grant Period: 4/1/95-3/31/01
Title: Improving the Cost-Effectiveness of Care for Depression—Patient Outcomes Research Team (PORT-II)

Description: Using a quasi-experimental design, this project evaluated the cost-effectiveness of alternative approaches to improving care for depression in prepaid group practices and obtained information on the linkages between processes of care and patient outcomes in primary care settings. Patients in general medical clinics were randomly assigned to their usual care or to one of two interventions (improving medication management or improving counseling for depression). Treatment decisions were up to the usual care providers and their patients, but clinical protocols, changes in practice structure, and patient and provider education were used to enable more appropriate care. The interventions were tested in 46 primary care clinics in 6 U.S. managed care organizations, which adapted the interventions to their own policies.

Findings: A quality improvement (QI) program of evidence-based materials was developed and tested and is available as a resource toolkit designed for clinicians, nurse specialists, psychotherapists, and patients in most types of managed care organizations. This QI program was found to improve care for depressed patients without requiring any services beyond those already covered by the patients' health plans. Study findings also indicated that overall employment rates can be positively affected with improved treatment. Full or partial toolkits are available at

When the outcomes of the two study interventions were compared to the outcomes of usual care, patients who were enrolled in the QI-therapy group had better results over the 2-year observation period. The QI-therapy group had fewer role limitations than the QI-medications group, but both groups had fewer role limitations than those in usual care.

PORT researchers examined the relationship between patient ratings of interpersonal patient-provider relationship (PPR) and both satisfaction with health care and technical quality of care among a sample of depressed primary care patients and found that interpersonal PPR is a critical element of health care quality. Patients are more likely to disclose concerns and details about their illness when the doctor asks them about specific problems such as depression, shows empathy, discusses treatment choice, and listens well, which are all elements of quality PPR.

PORT findings have also shown that despite effective psychotherapy and antidepressant medication, about three-fourths of primary care patients with depression do not receive appropriate care. Yet most (83 percent) of these patients, regardless of the severity of their symptoms, wanted to be treated for their depression, and most preferred counseling over medication. Specific treatment preferences vary by race, sex, income, and knowledge about treatments.

The PORT quality improvement approach can help the primary care physician with depression patient assessment, treatment, and followup. This study found that when clinical leaders of local managed care practices were trained in multimodal quality improvement programs, practices achieved above 70 percent adherence rates for most intervention components. For example, they were near 100 percent for hiring depression nurse specialists and reducing copayments for psychotherapy. Adherence rates were lower in some areas, however, with only 55 percent of patients on antidepressants followed by nurses to monitor symptoms, side effects, and medication compliance for the full expected duration (6-12 months).

When compared to patients with chronic medical conditions such as back problems, arthritis, and hypertension, depressed patients had worse mental health-related quality of life (HRQOL) and poorer physical functioning than those with a chronic condition. Because rates of detection and treatment of depression are moderate at best, primary care physicians should place more emphasis on helping patients recover from depression.

PORT researchers also found that patient preferences for mental, physical, and social health may strongly shape their treatment decisions. Patients may prefer to maintain a certain tolerable quality of life rather than risk additional suffering that could result from high-technology treatments. While concerns about physical health explained 35-55 percent of variation in patient treatment preferences, patients also placed a high value on their mental health (20-42 percent of preference variation). Mental health was strongly related to health preferences, even among patients with chronic medical conditions, such as arthritis or diabetes.

High copays or strict limits on visits and other types of health care services may dissuade patients from seeking care, while low copays or unlimited visits may have the opposite effect. Apparently, patients suffering from depression often perceive their mental health coverage to be more generous than it actually is, which could lead to unexpected out-of-pocket costs. In contrast, patients usually have a fairly accurate idea of their coverage for medical visits and prescription copays. Overall, depressed patients reported better mental health benefits than they actually had. Depressed patients who had used health care services in the past 6 months had a more accurate picture of medical benefits but not of mental health benefits. Patients with depressive disorder and patients who were less satisfied with their health care reported mental health copays more accurately than those who had depression symptoms only and more satisfied patients, respectively. Also, white patients and patients who had fewer chronic health problems were more accurate in reporting mental health visit limits than patients from ethnic minorities and sicker patients. More effective dissemination of information about mental health benefits by health maintenance organizations, health care plans, and employers could correct misperceptions, enhance trust, and improve access to and quality of mental health care.

PI: Whitney P. Witt, M.P.H., Johns Hopkins University (Dissertation Grant)
Grant No.: R03 HS11254
Grant Period: 9/1/00-6/30/01
Title: Family Influences on Children's Health and Health Care

Description: This study examined the impact of two major family stressors, childhood disability and maternal depression, on the health status of children and on health care and mental health care services utilization among children. The study also investigated the impact of maternal mental health care, respite care, and child care services on the health status of children. The researchers used data from the 1994 National Health Interview Survey Supplement on Disability (NHISSD), a questionnaire of biological mothers concerning about 20,000 children under age 18, including about 3,200 disabled children.

Findings: No findings are available at this time.

Return to Contents


PI: Deborah L. Ackerman, Ph.D., University of California, Los Angeles (UCLA)
Grant No.: R03 HS09551
Grant Period: 9/30/97-7/31/98
Title: Outcomes and Costs of Inpatient Antidepressant Drugs

Description: This inpatient study retrospectively reviewed billing data and clinical outcomes data from the UCLA Neuropsychiatric Hospital to evaluate the relative costs and benefits of different treatment regimens and to provide descriptive information about antidepressant drug use. Other outcomes were evaluated, including length of stay, discharge to the medical center, and readmission.

Findings: The higher initial cost of the selective serotonin-reuptake inhibitors (SSRIs) for treating depression may be offset by higher expenses associated with monitoring and managing side effects of tricyclic antidepressants (TCADs). SSRIs were the most commonly prescribed antidepressants. The atypicals (trazondone, bupropion, and nefazodone) and SSRIs were associated with the highest charges: $2,000-$3,000 higher than for the TCADs. The atypicals were associated with $500-$1,000 higher charges for services and procedures, specifically for electroconvulsive therapy, which was billed more often and for more procedures per patient. Readmission rates were similar across drug classes.

PI: Frank M. Ahern, Ph.D., Pennsylvania State University
Grant No.: R01 HS09075
Grant Period: 8/1/96-7/31/00
Title: Impact of Prospective Drug Use Review on Health

Description: The purpose of this study was (1) to conduct a descriptive epidemiological analysis of psychotropic drug use, prescribing patterns, and yield of Prospective Drug Utilization Review (ProDUR) screening and (2) to evaluate the independent effects of two different procedures on health outcomes, including both outcomes from changes in prescribing practices and differential mortality, morbidity, and health services utilization. The study examined two 24-month periods, before and after the ProDUR interventions, and two outpatient populations in which ProDUR has been implemented: the Pennsylvania Medicaid program and the Pennsylvania PACE program, a State-level outpatient pharmaceutical assistance program for the elderly.

Findings: No findings are available at this time.

PI: Rajender R. Aparasu, Ph.D., South Dakota State University
Grant No.: R03 HS10813
Grant Period: 9/1/00-2/28/02
Title: Evaluation of Inappropriate Psychotropic Use in Elderly

Description: The objective of this research was to determine the factors leading to and the effects of inappropriate psychotropic use in the elderly. The researchers utilized data from the Medical Expenditure Panel Survey (MEPS) on the community-dwelling elderly in the United States aged 65 years and older who used psychotropic medications. The researchers specifically assessed the characteristics of the psychotropic-using elderly, the quality and degree of inappropriate psychotropic use, and the impact of inappropriate psychotropic use on health care costs and utilization.

Findings: No findings are available at this time.

PI: Viktor E. Bovbjerg, M.P.H., University of Washington (Dissertation Grant)
Grant No.: R03 HS08469
Grant Period: 9/1/94-8/31/95
Title: Lipid-Lowering Medications and Risk of Injury

Description: The purpose of this study was to examine the association between the use of lipid-lowering medications and injury, including motor vehicle accidents, suicide or attempted suicide, and homicide or assaults.

Findings: No association between lipid-lowering medications and elevated risk of injury was suggested. The researchers found no increased injury risk among current users of lipid-lowering medications or past users, after adjustment for behavioral disorders, medical conditions, and health status. Current use of these medications was associated with lower risk of injury among patients who did not have behavioral disorders, while current use was not associated with injury risk among patients with behavioral disorders.

PI: Jeffrey Brown, Heller School, Brandeis University (Dissertation Grant)
Grant No.: R03 HS10016
Grant Period: 9/30/98-9/29/00
Title: Clozapine's Impact on Schizophrenic Medicaid Enrollees

Description: The objective of this study was to compare the use of clozapine as treatment for schizophrenia patients in the Medicaid populations of New York and New Jersey and to compare the prescribing of the drug by various providers. The study compared New York, which excluded clozapine from its Medicaid drug formulary, and New Jersey, which included it, to estimate the effects of clozapine on access, utilization, and expenditures.

Findings: This dissertation is not available to date.

PI: William Campbell, M.D., University of North Carolina, Chapel Hill
Grant No.: U18 HS10397-01
Grant Period: 9/30/99-9/29/02
Title: Rational Therapeutics for the Pediatric Population

Description: The University of North Carolina Center for Education and Research on Therapeutics (CERTs) builds on the recent work of the Program on Health Outcomes and brings together a diverse clinical, research, and education community around the needs of children in achieving optimum outcomes of drug and device therapy. The CERTs is organized into four core teams that are pursuing a coordinated education and research agenda focusing on Rational Therapeutics for the Pediatric Population. For their initial and phased studies, the core teams have selected 16 projects based on the skills of the team and the CERTs' partners. The methods range from a literature search and review to pharmacokinetic studies. The partnerships include representation from the pharmaceutical industry, a nonprofit arm of a managed care organization, several contract research organizations, an integrated health care delivery system, and other academic medical centers.

Two of the projects selected for study are related to mental health. They are: (1) prescribing patterns of psychotropic drugs for adolescents and (2) the attention deficit-hyperactivity disorder (ADHD) project. The first project is a retroactive analysis administrative database that is addressing use of psychotropic drugs to treat depression and ADHD. The second project is designing and testing toolkits and process improvement strategies for practitioners to use in the diagnosis and management of ADHD.

Findings: Researchers reviewed and analyzed 1995-99 claims data of the UnitedHealth Group in Minnesota for children under age 20 in six independent practice association health plans from four different geographic regions. They calculated the prevalence of use of four psychotropic drug classes: central nervous system stimulants (CNSSs) such as Ritalin, selective serotonin-reuptake inhibitors (SSRIs), trycyclic antidepressants (TCAs), and other antidepressants (OADs). Over the 5-year period, use of OADs increased 195 percent and use of TCAs decreased 21 percent. Pediatricians made up half and family doctors 20 percent of the first prescribers of CNSSs in 1995, compared with 13 percent of psychiatrists (which increased to 18 percent by 1999). Psychiatrists were most likely to prescribe SSRIs (56 percent in 1995, declining to 44 percent by 1999). However, the proportion of pediatricians and family doctors prescribing this class of drugs increased from 7 percent in 1995 to 23 percent in 1999 for pediatricians and from 13 percent in 1995 to 28 percent in 1999 for family doctors. These findings underscore the importance of training and expertise among primary care physicians in the use of psychotropic medications in youths.

PI: Alan M. Garber, M.D., Ph.D., Stanford University
Grant No.: R01 HS07818
Grant Period: 3/1/93-9/30/96
Title: Preference Assessment for Pharmaceutical Evaluation

Description: The purpose of this research was to develop a health-state classification system, present descriptions of possible health states to patients for rating, and assess patient preferences toward alternative states of health for use in cost-effectiveness analysis of pharmaceutical therapies. Subjects were randomized to exposure to either multimedia or text descriptions of the health state and tested for recall and recognition of its features with regard to two important pharmaceutical therapies: alglucerase, for Gaucher disease, and clozapine, an antipsychotic medication.

Findings: Multimedia subjects had better recall and better recognition of health state features than did text-only subjects. The research suggests that multimedia presentations result in a more accurate and complete understanding of the described health states than do text-only presentations.

PI: Judith M. Garrard, Ph.D., University of Minnesota
Grant No.: R01 HS07772
Grant Period: 3/1/93-2/28/99
Title: Patient Outcomes Associated with Antidepressant Drugs

Description: The purpose of this study was to examine the association between patient outcomes and antidepressant treatment of people with depressive symptoms. The subjects of the study were elderly people (65 years of age or older) living in the community who were enrolled in a social health maintenance organization (SHMO). The patient outcomes studied were health services utilization; health and functional status, including mortality; and quality of life.

Findings: Primary care physicians miss identifying depressive feelings in half of the elderly community-dwelling patients they see, especially men. Elderly people who have minor depression or are not depressed generally rate their physical health better as their level of illness declines, while elderly patients who are suffering from serious depression often provide assessments of their health that are inconsistent with clinical evidence. The researchers concluded that one way to improve clinical detection of depression, and hence the quality of life of community-dwelling elderly people, might be periodic screening for depression using an instrument such as the Geriatric Depression Scale. Secondly, when a patient's report is inconsistent with the clinical condition, evidence of minor depression should not preclude further investigation of inconsistencies between a patient's report and clinical evidence, while there was no such inverse association for those with serious depression. The report is available for a fee from the National Technical Information Service, 800-553-NTIS. The NTIS accession no. is PB2001-106533.

PI: Joseph T. Hanlon, Ph.D., Pharm.D., Duke University Medical Center
Grant No.: R01 HS07819
Grant Period: 3/1/93-5/31/95
Title: Cognitive Impairment and Medication Appropriateness

Description: The objective of this study was to determine if the use of psychotropics, nonsteroidal anti-inflammatory drugs (NSAIDs), or anticonvulsants is associated with cognitive impairment in the community-dwelling elderly. The study examined patterns of medication use among three groups of community-dwelling elderly: the demented cognitively impaired, the cognitively impaired without dementia, and the cognitively intact.

Findings: Among the community-dwelling elderly, cognitively impaired subjects (including demented individuals) are less likely to use over-the-counter (OTC) medications and analgesics than cognitively intact individuals. The more cognitively impaired a community-dwelling elder is, the less apt that individual is to use OTC, cardiovascular, and analgesic, as well as prescription medications. The higher use of analgesics by those who were not demented may represent a protective effect of certain diseases, such as rheumatoid arthritis, or the medications used to treat them, such as anti-inflammatories. There was no compelling evidence to suggest that NSAID use is associated with either deterioration or improvement in the level of cognitive function among community-dwelling elderly. However, current use of benzodiazepines, either with a short half-life or a long half-life, was associated with memory impairment, and use of higher doses of benzodiazepines was associated with increasingly worsened memory function. The researchers concluded that the risk of cognitive impairment should be balanced against the clinical benefit of benzodiazepine use in the elderly and that elders should be given prescriptions for smaller total daily doses. Previous benzodiazepine use was unrelated to memory problems, and current and previous benzodiazepine use was unrelated to other aspects of cognitive functioning. Researchers observed a relationship between digoxin and decreased cognitive function but suggested that additional research is needed to assess the role of digoxin as opposed to underlying disease.

PI: David A. Katz, M.D., University of Wisconsin
Grant No.: R01 HS10466
Grant Period: 9/1/00-8/31/02
Title: Improving the Evidence for Unstable Angina Guidelines

Description: The purpose of this research is to develop a statistical technique by which to measure costs and patient health outcomes jointly, incorporating the influential factors that were often neglected in previous analyses. To test its performance and sensitivity, the technique will be applied to real data, including an investigation of the determinants of health care costs and cost and use of pharmacologic treatments in 9,000 Medicaid recipients with diagnosed attention deficit hyperactivity disorder.

Findings: No findings are available at this time.

PI: Barbara M. Maloney, University of Massachusetts (Dissertation Grant)
Grant No.: R03 HS07954
Grant Period: 9/1/93-8/31/95
Title: Psychoactive Drug Use Post Omnibus Budget Reconciliation Act (OBRA): Massachusetts

Description: The objective of this study was to evaluate the effectiveness of OBRA's Federal legislation regulating psychoactive drug use in nursing homes. The researchers reviewed data from the Massachusetts Medicaid Claims File and nursing home resident records to determine the effects of the regulations on the use of psychoactive drugs (including antipsychotics, hypnotics/sleepers, long-acting benzodiazepines, and anxiolytic drugs), on prescriber compliance with the guidelines, and on any changes in survival, hospitalization rates, activities of daily living functioning, and behavioral stability of nursing home residents.

Findings: No findings are available at this time.

PI: Vincent Mor, Ph.D., Brown University
Grant No.: T32 HS00011
Grant Period: 9/30/86-6/30/03
Title: Institutional National Research Service Award

Description: The Brown University Center for Gerontology and Health Care Research received an AHRQ-funded postdoctoral training grant that has become an integral part of this highly productive research group. Since its inception in 1986, seven physicians and seven PhDs have been trained in the interdisciplinary environment. Several of these individuals have already begun to influence health services research in the areas of health status assessment, geriatrics, and long-term care, and others launched promising careers in the field. The goals of the training program are: (1) to teach health services research methods as well as content material relevant to gerontology and long-term care; (2) to train both physicians and nonphysician researchers; and (3) to give trainees didactic as well as "hands on" supervised experience in health services research.

Findings: This project has shown that psychoactive drugs are used more frequently with the elderly in nursing homes than with the elderly living in the community, and residents under 65 years of age were more likely to be given psychoactive drugs than were older residents. These drugs include anxiolytics (agents that relieve anxiety), antidepressants, hypnotics, and neuroleptics and can cause complications in the frail elderly and increase the risk of falling when inappropriately used. Nearly 62 percent of the residents under age 65 received psychoactive drugs, compared with 56 percent of the residents aged 75-84 and 50 percent of those 85 and older. Residents admitted from other nursing homes and from psychiatric facilities were more likely to be given psychoactive drugs than were persons transferred from hospitals or private homes. Disruptive and more cognitively impaired residents also had a higher risk of being given psychoactive drugs. Moreover, many residents take psychotropic medication with drugs for diabetes, hypertension, Parkinson's disease, and other conditions. This potentially harmful polypharmacy sparked regulation of medications dispensed in nursing homes in 1987 and probably warrants similar regulation in board-and-care facilities.

The 1987 Nursing Home Reform Act (NHRA) directed nursing homes to reduce their use of physical restraints; it also detailed nursing home residents' rights to be free of physical and chemical restraints used simply for discipline or convenience. As a result, the number of restraint-free homes increased nationwide from 1 percent in 1989 to 8 percent by 1995. Restraint-free homes differ in several ways from homes that use restraints. Restraint-free homes are 30 percent more likely to have a high ratio of full-time equivalent Registered Nurses per resident than homes that use restraints. Restraint-free homes have a lower average occupancy rate—68 percent compared with an overall rate of 87 percent in other facilities. Restraint-free homes also are more likely to have residents less in need of restraint; that is, their residents have fewer problems in performing activities of daily living, and the homes have a lower average number of residents who are bowel or bladder incontinent. Restraint-free homes also differ in organizational characteristics, as they are more likely than other homes to be not-for-profit facilities located in urban and more competitive areas, to be smaller, and not to be members of chains (perhaps suggesting that change may be delayed in bureaucratic organizations such as nursing home chains). Restraint-free homes are less likely to have special care units for Alzheimer's patients but are more likely to have other special care units.

With the passage of the 1987 NHRA, nursing homes are no longer allowed to use psychoactive medications except for justifiable clinical reasons. The Federal law seems to have had some impact, as this study found that although the use of antipsychotic drugs increased a bit from 1990 to 1993 (14 percent to 17 percent), clinical factors were slightly more significant in 1993 in the use of both antipsychotics and antianxiety/hypnotic drugs. This suggests that the use of psychotropic drugs is more clinically motivated than in the past. In 1990, nursing home residents who were more physically impaired, older, or had a prior stroke were significantly less likely to receive antipsychotic drugs. Those who were more cognitively impaired or suffered from dementia, depression, or a history of psychiatric problems were significantly more apt to receive antipsychotic drugs. In 1993, a largely similar pattern of effects was observed, with the addition of anxiety disorders as a factor increasing the likelihood of receiving antipsychotic drugs. In 1993, nursing home residents with a history of psychiatric problems were three times more apt to be given antipsychotics than other residents. Those with an anxiety disorder were four times more apt to receive antianxiety/hypnotic drugs than other residents in 1993. Nursing home facility factors also were less important in influencing psychotropic drug use in 1993 than in 1990. However, the changes in psychotropic drug use may not have been specifically related to the NHRA but may instead have been due to increased knowledge about the appropriate indications for psychotropic drug use.

PI: Wayne A. Ray, Ph.D., M.S., Vanderbilt University
Grant No.: R01 HS03222
Grant Period: 9/1/78-8/31/80
Title: Claims-File Research: Pediatric Health Care Quality

Description: This objective of this study was to review a data set in order to describe psychotropic drug use and prescribing in children. The researchers reviewed the records of 341,422 children under age 18 enrolled in the Tennessee Medicaid Aid to Families with Dependent Children program from 1977 through 1981. The study specifically reviewed incidence and utilization rates of prescribing, frequent use of psychotropic drugs, exposure to multiple agents, doctor shopping, and 5-year trends for children, as well as relationships between physicians' training and practice and their prescribing of psychotropic drugs. The researchers also studied differences in prescribing based on age and gender of the children. Prescribing of psychotropic drugs—including minor tranquilizers, sedatives/hypnotics, antidepressants, and antipsychotics—as well as of controlled-analgesic drugs and of opioid-containing drugs was studied.

Findings: For the 341,422 children, an average of 3.9 office visits per child were made and an average of 4.8 prescriptions per child were received. A total of 18,023 children received 54,123 psychotropic drug prescriptions: 55 percent were for sedatives/hypnotics, 17 percent were for tranquilizers, 16 percent were for antidepressants, and 8 percent were for antipsychotics. Along gender-specific lines, proportions of male and female children receiving psychotropic drug prescriptions were comparable through age 11. From age 11, the proportions of male and female children receiving psychotropic drug prescriptions diverged. By ages 16 and 17, almost twice as many females (6.3 percent) as males (3.3 percent) received prescriptions for psychotropic drugs. From age 11, prescribing of psychotropic drugs was significantly higher for females than males for comparable diagnoses, including mental illness. The pattern of psychotropic drug prescriptions per office visit for mental illness was similar: by age 17, females received 95.8 prescriptions per 1,000 office visits, while males received 47.0 prescriptions per 1,000 office visits. Total number of prescriptions received per child in any drug class was independent of gender. The researchers suggest that the sex-specific difference in frequency of psychotropic drug prescriptions is a result of factors that begin at puberty, but no earlier. However, they also suggest that the phenomenon of greater prescribing for females goes beyond strictly female diagnoses, such as gynecologic and genitourinary disorders.

PI: Betsy L. Sleath, M.S., University of Wisconsin (Dissertation Grant)
Grant No.: R03 HS07499
Grant Period: 9/1/92-9/30/93
Title: Patient Gender and Psychotropic Prescribing

Description: The objective of this research was to test an interactional model of psychotropic drug prescribing believed to be capable of explaining the variability that exists in psychotropic prescribing to male and female patients. The researchers reviewed about 550 existing audiotapes of physician-patient interactions from 11 different ambulatory care settings in the United States and Canada to examine the influence of differences in patient-physician interaction on psychotropic prescribing.

Findings: Forty-two percent of psychotropic prescriptions were initiated by patients rather than their primary care physicians. These patients had one or more chronic conditions, had seen their physicians at least twice before, and in many cases, had received the psychotropic medication before. Seventeen percent of the patients received prescriptions for one or more psychotropic medications; 47 percent of repeat psychotropic prescriptions and 20 percent of new ones were initiated by patients. Sixty-nine percent of prescriptions for low-income patients were physician-initiated, while nearly 90 percent of psychotropic prescriptions for high-income patients were patient-initiated. Fifty percent of patients who had been to see their physician seven or more times initiated psychotropic prescribing, compared with 29 percent of patients who had seen their physician two to six times previously. The patients' sex, race, age, and rating of physical and emotional health were not significantly related to whether physicians or patients initiated psychotropic prescribing. However, 20 percent of white patients received prescriptions for one or more psychotropic medications, while only 13.5 percent of black patients received prescriptions. For white patients, both patient expression of emotional symptoms and physician perceptions of patient emotional health significantly influenced psychotropic prescribing, while for patients who were not white, only patient expression of emotional symptoms significantly influenced psychotropic prescribing. For both white and other patients, neither patient expression of physical and social symptoms nor physician perceptions of patient physical health and social problems influenced psychotropic prescribing.

PI: Stephen B. Soumerai, Sc.D., Harvard University
Grant No.: R01 HS05947
Grant Period: 7/1/88-6/30/90
Title: Unintended Outcomes of Health Care Cost Containment

Description: The objective of this study was to determine the effects of outpatient drug cap policies on physician visit rates, hospital admissions, institutionalization, and deaths. This study compared 42 months of Medicaid claims data and clinical records from two community mental health centers (CMHCs) and the State psychiatric hospital in New Hampshire, a State with outpatient drug cap policies, to 42 months of Medicaid claims data in New Jersey, a State with no such caps. The goal was to determine the effects of a three-prescription monthly limit on the use of psychotropic drugs and acute mental health services by noninstitutionalized schizophrenia patients.

Findings: The cap resulted both in immediate reductions of 15 to 49 percent in the use of psychotropic drugs such as antidepressants, lithium, and antipsychotic drugs and in coincident increases of one to two visits per month to CMHCs (with administration of antipsychotic drugs) and a sharp increase in the use of emergency mental health services and partial hospitalizations. After the cap was discontinued, the use of medications and most mental health services reverted to pre-cap levels. The researchers concluded that limiting Medicaid reimbursement to three drug prescriptions per patient per month markedly increased emergency and acute care services for noninstitutionalized schizophrenic patients who were unable to pay for the additional psychotropic drugs they needed. This cost-cutting initiative cost Medicaid 17 times more than it saved in prescription drug costs.

Return to Contents
Proceed to Next Section


Page last reviewed January 2003
Internet Citation: Compendium of Research Related to Mental Health (continued). January 2003. Agency for Healthcare Research and Quality, Rockville, MD.