Beyond the Water's Edge: Charting the Course of Managed Care for People with Disabilities - Conference Resource Book. Managing the Care of Schizophrenia: Lessons From a 4-Year Massachusetts Medicaid Study


Barbara Dickey, Ph.D.; Sharon-Lise T. Normand, Ph.D.; Edward C. Norton, Ph.D.; Hocine Azeni, M.A.; William Fisher, Ph.D.; and Frederic Altaffer, Ph.D.
Archives of General Psychiatry 53:945-952 (October 1996)

Background: In 1992, Massachusetts launched a statewide managed care plan for all Medicaid beneficiaries.

Methods: This retrospective, multiyear, cross-sectional study used administrative data from the Massachusetts Division of Medical Assistance and Department of Mental Health, consisting of claims for 16400 disabled adult patients insured by Medicaid in Massachusetts between July 1, 1990, and Jun 30, 1994. The main outcome measures include annual rates of hospitalization, emergency department utilization, and follow-up care 30 days after discharge; length of inpatient stay; and per-person inpatient and outpatient expenditures.

Results: Between 1991 and 1994, the likelihood of an inpatient admission decreased from 29% to 24% and was accompanied by a slight reduction in length of stay (median number of bed-days per admission dropped by 3.3 days). There was a slight decrease in the number of patients who sought care in general hospital emergency department utilization. However, there was a small increase in the fraction of patients readmitted within 30 days of discharge. Medicaid and Department of Mental Health expenditures for mental health per treated beneficiary decreased slightly, from $11060 to $10640, during the 4-year study period.

Conclusions: Although per-person expenditures dropped and most patient patterns of care remained the same, longer-term study is recommended to assess whether the trends can be maintained.

The treatment of schizophrenia remains a major clinical challenge to health care providers.1 The behavioral problems and thought disorders that are characteristic of schizophrenia make management complex and expensive. For example, in the United States, even though slightly more than 1% of adults have the disorder, treatment expenditures account for more than 2.5% of all health care expenditures.2, 3 It is not unusual for those with schizophrenia to have disabilities that lead to loss of employment and private health insurance. When this occurs, government becomes the primary health care insurer. Almost two-thirds of all the expenditures for schizophrenia treatment come from federal, state, and local government sources.2

With so much government money at stake, it is not surprising that reforms are rapidly changing the provision of government services to the mentally ill. All but 6 states are pursuing managed care for Medicaid beneficiaries, including those with severe mental illness, such as schizophrenia. In some states, Medicaid managed care plans tap into existing health maintenance organization networks.4 In other states, Medicaid contracts with mental health managed behavioral care companies to provide administrative functions and direct beneficiaries to a local provider network.

There are 2 fundamental issues in the evaluation of managed care for the severely mentally ill. First, can managed care succeed in providing quality care to psychiatrically disabled patients, especially those diagnosed as having schizophrenia?5, 6, 7, 8 These individuals are at high risk for catastrophic psychiatric and medical care but seldom are able to navigate effectively within the health care system and often lack advocates on their behalf. Even though providers surveyed in Massachusetts have reported that quality has not been compromised,9 both critics and advocates would like to have more evidence before accepting this conclusion.

The second issue is whether managed care actually reduces costs of shifts costs onto families, other state agencies, or medical care providers. When there are strong financial incentives to reduce acute hospital admissions, managed care plans will be financially motivated to divert beneficiaries to the long-term care system run by the state mental health agency. Individuals with schizophrenia are likely to be eligible for both Medicaid benefits and a state-funded long-term care system of community and hospital-based services. Shifting costs to the state mental health agency may result in greater profits for managed care plans but may not improve continuity of care; moreover, the societal costs may be higher. To date, there are no studies of how the reduction in mental health expenditures for acute care might shift costs to long-term care.

Evaluations of managed care plans are in an early stage of development, and little descriptive information is available to provide benchmarks against which to compare different approaches to cost containment.10 Earlier reports 9, 11 of the Massachusetts plan studied only the first year after implementation. In addition, these studies were limited to claims for mental health treatment.

We developed a database on adult Medicaid beneficiaries with schizophrenia to examine access to care, use of services, and treatment costs associated with schizophrenia before and after the introduction of managed care. Because care for our disabled population is not limited to services reimbursed by Medicaid alone, the data are drawn from 2 state agencies: the Division of Medical Assistance (Medicaid) and the Department of Mental Health (DMH). Data cover 2 years of the plan after implementation and include nonpsychiatric medical care, pharmacy, transportation, and dental care. These additional data are important to include because they account for roughly 40% of the total expenditures. We also examined incident patients, those not treated for schizophrenia before managed care was introduced.



The Massachusetts Managed Mental Health Program

In 1992, Massachusetts received a 1915b waiver from the Health Care Financing Administration. Under this plan, all beneficiaries were asked either to enroll in a local health maintenance organization or to select a Medicaid-approved primary care clinician. Virtually all psychiatrically disabled beneficiaries chose a primary care clinician. Medicaid contracted with a single proprietary vendor, Mental Health Management of America, a division of First Mental Health, Boston, MA, to management the provision of mental health benefits. The vendor had 4 specific cost-containment strategies: (1) negotiation of reimbursement rates with a network of providers who would be paid on a fee-for-service basis, (2) implementation of an aggressive utilization management plan, (3) development of community-based alternatives to hospitalization, and (4) collaboration with the DMH to fund emergency service teams to screen patients for appropriateness of inpatient admission, with a view toward diverting many of them to alternative treatment sites.

Under the terms of the contract with Medicaid, the vendor was required to make available to recipients all the mental health and substance abuse benefits: acute inpatient treatment, crisis stabilization, outpatient evaluation and treatment, psychiatric day treatment, residential detoxification, and methadone treatment. The vendor was directed to add diversionary services, including acute residential treatment programs, family stabilization teams, and partial hospitalization programs. The contract further specified that the vendor would be responsible for the centralized functions of utilization review, claims processing, systems support, and provider relations, and for decentralized regionally based case management and network management. The contract with the vendor excluded payment for long-term nursing home care, mental health services provided by the DMH, and any medical treatment or outpatient pharmacy. In addition, it did not include members of health maintenance organizations and those who had Medicaid as a second payer. Disabled beneficiaries were covered at a higher rate than other beneficiaries, and providers were reimbursed by the vendor on a fee-for-service basis.

Data Sources

We used administrative data obtained from Medicaid and the DMH. Together, these files provided information regarding patient sociodemographic status, reimbursed inpatient and outpatient care, discharge diagnoses, and timing of services.

Definition of Cross-Sectional Cohorts

We created 4 separate cohorts, 1 for each fiscal year of the study, that together described treatment spanning the period from July 1, 1990 (the start of fiscal year 1991) through June 30, 1994 (the end of fiscal year 1994). The members of each cohort consisted of all adult Massachusetts Medicaid beneficiaries, aged 18 to 64 years, who were disabled and treated, either as inpatients or outpatients, for schizophrenia (International Classification of Diseases, Ninth Revision, Clinical Modification, primary diagnostic code of 295) at least once during the fiscal year. The cross-sectional cohorts were created by assigning patients with a schizophrenia claim to the fiscal year in which the claim was submitted; for this reason, it was possible for patients to appear in more than 1 cohort.

With the use of the patient's unique Medicaid identification number, patient-level files for each fiscal year were constructed by identifying paid claims for all psychiatric and substance abuse care (claims with a primary diagnostic code of 290-315), medical care (claims with any primary diagnostic code excluding V codes and 290-315 or any claim with a mental health Current Procedure Terminology procedure code), and other services, such as pharmacy, transportation, and dental care. Finally, to ensure that we had a complete record of service use for each patient, we merged state hospital admissions from the DMH inpatient files with the administrative Medicaid information by means of unique patient identification numbers.

Sociodemographic and Comorbidity Data

We used Medicaid membership files to identify the date of birth, sex, race, and residence ZIP code for each patient in our study cohort. To measure the degree of substance abuse in our sample, we assumed that if a patient was ever diagnosed as a substance abuser (primary or secondary diagnostic International Classification of Diseases, Ninth Revision, Clinical Modification code of 291, 292, 303.00, 303.90, 304, or 305) in a given year, then the patient had a drug or other alcohol abuse problem in the given year.

Admission Type

During the study, the DMH had contracted with a few general hospitals for inpatient beds to replace some of the beds in state hospitals closed as part of a larger deinstitutionalization plan. Thus, DMH-funded admissions occurred in both state and general hospitals. To differentiate between beds funded by Medicaid and those funded by the DMH, we classified each mental health inpatient admission as either a DMH admission or a Medicaid admission. The admission policy for Medicaid recipients to a DMH bed required that beds be available to forensic patients and to patients with behavioral management requirements that could not be met in general hospital psychiatric units or in freestanding psychiatric facilities otherwise reimbursed by Medicaid.

Evaluation of Access to Care

We defined access to care in each year as the number of Medicaid beneficiaries with a primary diagnosis of schizophrenia who had at least 1 Medicaid-paid claim. We also examined the number of incident patients in each year. We classified a patient as incident if there was no mental health claim with a primary diagnosis of schizophrenia for the patient in the previous year(s). Because we did not have Medicaid data before 1991, we were unable to identify new patients in 1991 and consequently may also have overestimated the number of incident patients in the remaining years.

Mental Health Inpatient Utilization

We defined mental health inpatient utilization as hospital admissions with primary mental health discharge diagnoses corresponding to schizophrenia, or any other psychiatric and substance abuse disorder. Because we hypothesized that hospital admissions would drop as a result of the screening and diversion programs of the managed care plan, we examined the distribution of mental health inpatient admissions in each year. We also estimated the likelihood of having any mental health inpatient admission by the percentage of patients who had at least 1 such admission in a given year. Finally, for those patients who had at least 1 mental health inpatient admission in a given year, we examined the number of bed-days per admission.

Continuity of Care

To describe follow-up care after discharge, we first defined inpatient transfers as admissions to another hospital within 24 hours after a discharge and then linked information from the transfers to form a complete inpatient episode of care for each patient in our cohorts. We then categorized each discharge into 1 of 4 mutually exclusive categories: discharges for which there was no outpatient or inpatient contact within 30 days, discharges for which there was outpatient contact within 30 days, discharges resulting in rehospitalization within 30 days, and discharges for which both an outpatient contact and a rehospitalization within 30 days resulted. Outpatient contact included a visit to any hospital outpatient department or to a clinic; a visit to a physician office; or the provision of any 1 of a set of mental health services, such as psychological testing, case management, or day treatment.

Because we believed that, ideally, continuous patient care should be rendered from 1 provider, we also calculated the number of unique hospitals to which patients with more than 1 hospitalization were admitted. Finally, we estimated the distribution in each year of general hospital emergency department visits. We did not calculate the distribution for state hospitals because they do not have emergency departments.

Assessment of Expenditures

We derived costs for Medicaid services from the paid claims that indicated the amount reimbursed. Although we could not determine whether paid claims overestimated or underestimated the true cost of treatment, these costs represented public expenditures through this entitlement program. Because the DMH operates on a fixed budget and records only use of services, we estimated costs for inpatient care by means of the per diem for state hospitals calculated by the DMH. These estimated per diem costs are based on accounting costs, calculated by dividing total annual inpatient expenditures, including capital costs, by the number of actual patient bed-days in each facility annually. Hospital-specific per diem costs were used to estimate the episode costs for each person admitted to a DMH facility by multiplying the calculated per diem by the number of days in the episode.

Inpatient mental health expenditures (psychiatric or substance abuse care) were dichotomized into Medicaid and DMH admissions to examine the extent of cost shifting between these 2 government agencies. All inpatient expenses are clustered together so that room and board, ancillaries, and physician fees are included.

Outpatient mental health expenditures included hospital outpatient department or clinic services, physician services, and other mental health services provided in free-standing mental health agencies.

Non-mental health expenditures were composed of medical care, pharmacy, transportation, and dental care. Claims for the last 3 categories did not report diagnoses, and consequently we were unable to distinguish mental health--related from non-mental health--related expenditures.

We aggregated the inpatient and outpatient mental health expenditures and then added all categories for total expenditures. Within each category of expenditures we report the number and percentage of persons with any expenditures, the average expenditure per person with any expenditure, and the total expenditure. For inpatient care we also report expenditure per admission.

Statistical Analyses

We computed simple univariate summary statistics by year. For continuous-valued variables, we calculated sample means and SDs; we also constructed box plots12 to display the center and spread of the distributions of the observations. Inpatient utilization was stratified by admission type (DMH or Medicaid). All expenditure figures are reported in 1994 dollars by adjusting expenditures in 1991, 1992, and 1993 for inflation by means of the gross domestic product deflator.13



Description of the Cross-Sectional Cohorts

Between July 1, 1990, and June 30, 1994, we observed 16400 disabled adults who contributed a total of 32135 annual observations. Despite changes in the number of treated beneficiaries during the 4-year period, we found that the sociodemographic characteristics remained virtually unchanged (Table 1): approximately half the beneficiaries were female and the majority were white, with a mean age of 41 years. Eleven percent were ethnic or racial minorities. Comorbid substance abuse increased as a proportion of the total study population. This increase might be a coding artifact resulting from changes in diagnostic practice or an increased awareness of substance abuse. Reimbursement of the treatment of substance abuse received much attention from providers because of the emphasis placed by the managed care vendor on outpatient rather than inpatient detoxification. It is also possible, given increases in the level of alcohol and other drug dependence in the general population, that there is increasing substance abuse among many of these patients, and the increase is being documented by providers.

Access to Care

The number of Medicaid beneficiaries with schizophrenia being treated increased from 6614 in 1991 to 7541 in 1994 (Table 1). However, in 1993, the year managed care was introduced in Massachusetts, the number of treated beneficiaries increased by more than 3000 from the previous year. This 1993 increase occurred despite a decrease in the number of providers. The increase might be an epidemiological phenomenon, but the stable demographic characteristics suggest that is not the source of the increase. More likely, the increase can be attributed to the advocacy work of the mental health provider community and family members who wanted to ensure that those who met the eligibility criteria were actually enrolled in Medicaid. In fact, the incident patients in 1993 were more likely to be older, white, female, and substance abusers than new patients in the remaining years (Table 1). The most striking demographic change for new schizophrenia patients was the higher percentage, in all years, of substance abusers.

Mental Health Inpatient Utilization

The percentage of patients who had at least 1 inpatient admission dropped by 4 percentage points during the study period, from 29.8% in 1991 to 25.4% in 1994 (Table 2). The decrease in the likelihood of a mental health admission during the study period was larger for DMH admissions than for Medicaid admissions (Table 2). Even though the probability of an admission decreased, the total number of DMH admissions remained almost the same in all 4 years. Medicaid admissions dropped 65% in 1993 but returned to the pre-managed care level by 1994. For those admitted, the median number of bed-days per admission decreased by about 3.3 days; there was a drop of 2.5 days for Medicaid admissions and of 3 days for DMH admissions (Figure 1).

Continuity of Care

We found little evidence of change in continuity of care. Rapid readmissions were up slightly, from 22.1% in 1991 to 24.2% in 1994 (Table 3), but there was essentially no change in the absence of outpatient follow-up contact, with the proportion of discharges without any follow-up contact remaining at 29%. (We were unable to identify in these data follow-up that may have occurred through DMH-funded community support services.) For patients with multiple admissions, there was an increase in the percentage who were admitted to more than 1 hospital (Table 3) but no increase in the percentage of patients who used emergency departments.

Mental Health Expenditures


Medicaid inpatient expenditures dropped dramatically in 1993, in the first year of managed care, but rose the following year (Table 4). The savings in Medicaid inpatient expenditures were largely offset, however, by increased expenditures on DMH hospital admissions. The annual Medicaid per-inpatient costs dropped below pre-managed care levels, while DMH per-inpatient costs were higher after implementation of the managed care program.


Both total expenditures and pre-treated beneficiary expenditures rose from their pre-managed care levels. The large increase in the number of beneficiaries in 1993 led to a dip in per-treated outpatient expenditures, but by 1994 this effect had disappeared. Given the large reductions in inpatient treatment, it was expected that outpatient treatment would expand, although these increases suggest only a modest cost shifting from inpatient to out-patient treatment.

Non-Mental Health Expenditures

Total expenditures of inpatient and outpatient medical and surgical care rose with the influx of new beneficiaries in 1993 (and then dropped as the number declined in 1994), but the per-person treated costs were about the same across all 4 study years. These data do not provide evidence that mental health treatment was shifted to the non-mental health sector. Other non-mental health expenditures include transportation and dental costs, which remained essentially the same during the study period. Pharmacy costs doubled both in total expenditures and per person treated.

Total Expenditures

The total Medicaid and DMH expenditures for mental health per treated beneficiary fell slightly from $11090 to $10600 during the 4-year study period. When all other Medicaid reimbursed care is added to mental health care, there is a slight increase in the total per-person expenditures (Table 4). The total dollar expenditures fluctuated with the number of treated beneficiaries across the study years, but it was decreasing, not increasing, at the end of the study period. The total Medicaid and DMH dollar expenditure for mental health care after managed care reversed the upward trend and stabilized at about $80 million in 1993 and 1994 (Figure 2).



Using a unique database of patient-level mental health treatments constructed from 2 sources, Medicaid and the DMH, we found that managed care was associated with some gains in continuity of care but a slight increase in rapid readmissions. Furthermore, there were reductions in mental health expenditures at the per-person level, primarily because fewer inpatient bed-days were reimbursed by Medicaid. The use of DMH inpatient beds for these beneficiaries remained about the same. Total mental health expenditures during the 4-year study period were contained, despite a growth in the number of treated beneficiaries.

There are several general conclusions from this study. First, because our measures of access and continuity of care are based on administrative data, they are limited in their scope and sensitivity. Furthermore, we have no way of knowing whether the pre-managed care levels were appropriate. Finally, for those with chronic illnesses, examination of short-term results is not an adequate indicator of the value of managed care. Information regarding the appropriateness of processes of care, such as the adequacy of discharge planning, or knowledge regarding patient well-being is crucial in judging the adequacy of quality. Because there is no clear evidence about the effectiveness of managed care plans to provide services needed by the most seriously mentally ill, we believe that continued research is essential to document the benefits or risks to clients.

Second, although we established that there were cost savings under managed care, we cannot be certain of the actual magnitude of the savings. Our assessment before and after managed care allows only 2 types of comparisons. The first type simply focuses on levels and directions of trends observed before and after implementation. The second approach compares observed postintervention levels with the levels that would have been expected in the absence of the intervention.14 From the perspective of the first approach, the vendor appears to have achieved a net reduction in expenditures and service use. The reduction in Medicaid inpatient expenditures was a function of 3 factors: the negotiated rates with the network hospitals, the reduced number of admissions, and the reduction in the total number of bed-days.

Third, the introduction of this managed care plan resulted in an unanticipated increase in the number of beneficiaries treated for schizophrenia. In this study we observed an increase in additional patients in 1993 who had a profound effect on the system, at least in the first year. The increase in treated patients in 1993, which shrank in 1994, tells us less about access to care and more about diagnostic variability in mental health. Rather than roughly 3000 members losing their coverage, as it appears, we found that they remained enrolled and were being treated for other mental illnesses. The marked increase in the number of beneficiaries in 1993 is real, regardless of diagnostic category. However, it creates a denominator problem: comparing percentages across years may be misleading, and per-person mean costs may be lower because individuals who need less intensive treatment are added to the membership. For example, the proportion of treated beneficiaries who had 1 or more admissions to a DMH inpatient bed during a year appears to drop from 15% to 10%, but the actual number of admissions did not change. This suggests that many of the new patients were among those less seriously ill. Trends such as these have been exhibited in a range of evaluations in a number of divergent fields.15, 16 Their ubiquitous character suggests the need for caution on the part of administrators and providers who would attempt to learn in the first few months after implementation what the ultimate effects of managed care will be on savings or service use.

Our final conclusion relates to cost shifting. We did find some evidence of cost shifting in this study. For example, one striking finding is the doubling of pharmacy costs. The increases in pharmacy costs observed might raise concern that psychosocial treatments are too often replaced by pharmacological interventions, but what seems more likely is that pharmaceutical costs have risen, especially for patients who are taking newer antipsychotic medications. Additionally, the growth in medical expenditures might signal cost shifting to that sector, and the fact that per-person medical care costs increase slightly might signal such a shift. The growth in medical expenditures are important because they compose about a third of all the health care dollars spent by Medicaid and the DMH on treatment for those with schizophrenia.

This report must be considered carefully in the light of its limitations. The mental health environment in Massachusetts at the time of this study was in transition. Reforms that are a response to fiscal and social problems are rarely unidimensional. In Massachusetts, prepaid managed care was only 1 aspect of a more global effort to privatize the Massachusetts mental health service system in the early 1990s. This effort entailed the closing of 3 state hospitals and the expansion of community-based services provided by vendors under contract to the DMH. The current study design does not rule out secular trends.

Although this study raises many questions, it also provides preliminary findings about the relationship of managed care with service use and with expenditures for seriously mental ill adults with schizophrenia. Future studies of managed care will need to continue to explore the trade-off between quality of care and costs, cost shifting between government agencies, and the difference between short-term and long-term effects.


From the Departments of Psychiatry (Dr. Dickey) and Health Care Policy (Dr. Normand), Harvard Medical School, Boston, MA; Mental Health Services Research, McLean Hospital, Belmont, MA (Dr. Dickey and Mr. Azeni); Department of Biostatistics, Harvard School of Public Health, Boston (Dr. Normand); Center for Economics Research, Research Triangle Institute, Research Triangle Park, NC (Dr. Norton); Center for Psychosocial and Forensic Services Research, University of Massachusetts Medical School, Worcester (Dr. Fisher); and Mentor, Inc., Boston (Dr. Altaffer).



Accepted for publication June 21, 1996.

This research was supported by grant RO1-MH54076 from the National Institute of Mental Health, Rockville, MD.

We thank the Massachusetts Department of Mental Health and the Division of Medical Assistance for providing data and technical support. Thanks also to Richard Lindrooth, Research Triangle Institute, Research Triangle Park, NC, for providing technical assistance.

Reprints: Barbara Dickey, Ph.D., McLean Hospital, Administrative Building, 115 Mill Street, Belmont, MA 02178-9106.



  1. Lehman, A.F., W.T. Carpenter, H.H. Goldman, and D.M. Steinwalchs. "Treatment Outcomes in Schizophrenia: Implications for Practice, Policy, and Research." Schizophr Bull, 1995; 21:669-675.
  2. Rupp A., and S.J. Keith. "The Costs of Schizophrenia." Psychiatr Clin North Am, 1993; 16:413-423.
  3. Regier, D.A., W.E. Narrow, D.S. Rae, R.W. Manderscheid, B.Z. Locke, and F.K. Goodwin. "The De Facto US Mental and Addictive Disorders Service System." Arch Gen Psychiatry, 1993; 50:85-94.
  4. Hurley, R.E., D.A. Freund, and J.E. Paul. Managed Care in Medicaid: Lessons for Policy and Program Design. Ann Arbor, MI: Health Administration Press; 1993.
  5. Sharfstein, S.S. "Utilization Management: Managed or Mangled Psychiatric Care?" Am J Psychiatry, 1990; 147:965-966.
  6. Tischler, G.L. "Utilization Management of Mental Health Services by Private Third Parties." Am J Psychiatry, 1990; 147:967-973.
  7. Schlesinger, M. "On the Limits of Expanding Health Care Reform: Chronic Care in Prepaid Settings." Milbank Q, 1986; 64:189-215.
  8. Schlesinger, M. "Striking a Balance: Capitation, the Mentally Ill,and Public Policy." In: D. Mechanic and L.H. Aiken, eds. Paying for Services: Promises and Pitfalls of Capitation, New Directions for Mental Health Services. San Francisco, CA: Jossey-Bass, Inc.; 1989; 43:91-115.
  9. Callahan, J.J., D.S. Shepard, R.H. Beinecke, M.J. Larson, and C. Cavanaugh. Evaluation of the Massachusetts Medicaid Mental Health/Substance Abuse Program. Waltham, MA: Brandeis University; January 1994.
  10. Wells, K.B., B.M. Astrchan, G.L. Tischler, and J. Unutzer. "Issues and Approaches in Evaluating Managed Mental Health Care." Milbank Q, 1995; 73:57-75.
  11. Dickey, B., E.C. Norton, S.L. Normand, H. Azeni, W. Fisher, and F. Altaffer. "Massachusetts Medicaid Managed Health Care Reform: Treatment for the Psychiatrically Disabled." Adv Health Econ, 1995; 15:99-116.
  12. Statistical Sciences Inc. S-PLUS: User's Manual. Seattle: WA:Statistical Sciences, Inc., 1991; 1:5-51.
  13. U.S. Department of Commerce. Survey of Current Business. Washington, DC: U.S. Department of Commerce, 1995.
  14. Cook, T.D., and C.T. Campbell. Quasi-Experimentation: Design and Analysis for Field Settings. Chicago, IL: Rand McNally, 1979.
  15. Rossi, P.H., and H.E. Freeman. Evaluation: A Systematic Approach. 5th ed. Newbury Park, CA: Sage Publications, 1993.
  16. McCleary, R., and R.A. Hay. Applied Time Series Analysis for the Social Sciences. Beverly Hills, CA: Sage Publications: 1980.
TABLE 1. Demographic Characteristics of Cross-Sectional Cohorts*
  All Patients, No. (%) Incident Patients, No. (%)
FY 1991
FY 1992
FY 1993†
FY 1994†
FY 1992
FY 1993
FY 1994
Age, y
- 18-21
- 22-39
- 40-64
- Mean±SD
172 (2.6)
3234 (48.9)
3208 (48.5)
184 (2.5)
3503 (48.0)
3608 (49.5)
331 (3.1)
4843 (45.3)
5511 (51.6)
179 (2.4)
3538 (46.9)
3824 (50.7)
118 (4.7)
1302 (51.5)
1108 (43.8)
246 (4.5)
2480 (45.2)
2758 (50.3)
104 (5.9)
934 (52.6)
736 (41.5)
Female 3182 (48.1) 3486 (47.8) 5808 (54.4) 3470 (46.0) 1201 (47.5) 3291 (60.0) 764 (43.1)
- African American
- American Indian
- Asian American
- Hispanic
- White
- Unknown
618 (9.3)
3 (0.0)
19 (0.3)
86 (1.3)
5865 (88.7)
23 (0.3)
699 (9.6)
3 (0.0)
25 (0.3)
125 (1.7)
6413 (87.9)
30 (0.4)
891 (8.3)
9 (0.1)
28 (0.3)
188 (1.8)
9537 (89.3)
32 (0.3)
772 (10.2)
4 (0.1)
22 (0.3)
102 (1.4)
6464 (85.7)
177 (2.3)
260 (10.3)
1 (0.0)
14 (0.6)
76 (3.0)
2164 (85.6)
13 (0.5)
430 (7.8)
5 (0.1)
15 (0.3)
126 (2.3)
4891 (89.2)
17 (0.3)
187 (10.5)
0 (0.0)
12 (0.7)
38 (2.1)
1377 (77.6)
160 (9.0)
Substance Abuse 636 (9.6) 775 (10.6) 3108 (29.1) 1023 (13.6) 355 (14.0) 1875 (34.2) 324 (18.3)
* Patients are disabled Massachusetts Medicaid beneficiaries treated for schizophrenia. FY indicates Fiscal Year.
† Managed care plan years.
‡ Number of treated beneficiaries
§ Number of new treated beneficiaries.


TABLE 2. Mental Health Inpatient Utilization* for Disabled Patients With Schizophrenia
  FY 1991 FY 1992 FY 1993† FY 1994†
No. of treated beneficiaries 6614 7295 10685 7541
Total No. of admissions 3937 4624 2486 3870
Distribution of hospital admissions, No. (%)‡
- 0
- 1
- 2
- 3
- 4
- >5
4690 (70.9)
991 (15.0)
459 (6.9)
198 (3.0)
125 (1.9)
151 (2.3)
5120 (70.2)
1120 (15.4)
471 (6.5)
252 (3.5)
148 (2.0)
184 (2.5
9146 (85.6)
1001 (9.4)
318 (3.0)
114 (1.1)
58 (0.5)
48 (0.4)
5623 (74.6)
1085 (14.4)
388 (5.1)
202 (2.7)
108 (1.4)
135 (1.8)
>1 inpatient admission, No. (%)‡
- All admissions
- DMH admissions
- Medicaid admissions
1924 (29.1)
1082 (16.4)
1170 (17.1)
2175 (29.8)
1098 (15.1)
1466 (20.1)
1539 (14.4)
1038 (9.7)
608 (5.7)
1918 (25.4)
915 (12.1)
1232 (16.3)
* Hospital admissions for treatment of mental illnesses or substance abuse.
† Managed care plan years.
‡ Percentage was calculated with number of treated beneficiaries used as the denominator. FY indicates fiscal year; DMH, Department of Mental Health.


TABLE 3. Continuity of Care for Disabled Patients With Schizophrenia*
  No. (%)
FY 1991 FY 1992 FY 1993† FY 1994†
Follow-up care within 30 d of a discharge
- None
- Outpatient contact
- Rehospitalization
- Rehospitalization and outpatient contact
- Total No. of Discharges
1077 (29.2)
1797 (48.7)
142 (3.8)
677 (18.3)
3693 (100.0)
1199 (27.3)
2178 (49.7)
127 (2.9)
880 (20.1)
4384 (100.0)
932 (38.0)
1159 (47.2)
124 (5.1)
238 (9.7)
2453 (100.0)
1117 (29.7)
1729 (46.0)
215 (5.7)
696 (18.5)
3757 (100.0)
Distribution of unique hospitals‡
- 1 hospital
- 2 hospitals
- 3 hospitals
- 4 hospitals
- >5 hospitals
- Total No. of Patients With 2 Hospitalizations
361 (38.7)
364 (39.0)
145 (15.5)
42 (4.5)
21 (2.3)
933 (100.0)
342 (32.4)
431 (40.9)
171 (16.2)
71 (6.7)
40 (3.8)
1055 (100.0)
227 (42.2)
213 (39.6)
64 (11.9)
24 (4.5)
10 (1.9)
538 (100.0)
298 (35.8)
381 (45.7)
113 (13.6)
29 (3.5)
12 (1.4)
833 (100.0)
Emergency department visits
- 0
- 1
- 2
- 3
- 4
- >5
- Total
5626 (85.1)
545 (8.2)
192 (2.9)
92 (1.4)
44 (0.7)
115 (1.7)
6614 (100.0)
6219 (85.3)
577 (7.9)
221 (3.0)
97 (1.3)
69 (0.9)
112 (1.5)
7295 (100.0)
9894 (92.6)
562 (5.3)
140 (1.3)
44 (0.4)
22 (0.2)
23 (0.2)
10585 (100.0)
6653 (88.2)
435 (5.8)
210 (2.8)
82 (1.1)
43 (0.6)
118 (1.6)
7541 (100.0)
* FY indicates fiscal year.
† Managed care plan years.
‡ Distribution of patients with 2 or more hospitalizations categorized by the number of unique hospitals to which they were admitted during the fiscal year. In FY 1991, 38.7% of the 933 patients who had at least 2 hospitalizations went to a single hospital, 39.0% were admitted to 2 distinct hospitals, 15.5% were admitted to 3 distinct hospitals, 4.5% were admitted to 4 distinct hospitals, and the remainder (2.3%) were admitted to 5 or more distinct hospitals.


TABLE 4. Annual Expenditures for Disabled Patients With Schizophrenia*
  FY 1991 FY 1992 FY 1993† FY 1994†
Mental Health Expenditures‡
Medicaid mental health inpatient admissions
- No. of inpatients
- Average annual expenditure per inpatient, $
- Total annual expenditure, x$1000
DMH mental health inpatient admissions
- No. of DMH inpatients
- Average annual expenditure per DMH inpatient, $
- Total annual expenditure, x$1000
Outpatient mental health care
- No. who are outpatients
- Average annual expenditure per recipient, $
- Total annual expenditure, x$1000
Non-Mental Health Expenditures
Inpatient or outpatient non-mental health care
- No. who received non-mental health care
- Average annual expenditure per recipient, $
- Total annual expenditure, x$1000
- No. who use pharmacy
- Average annual expenditure per recipient, $
- Total annual expenditure, x$1000
Transportation or dental care
- No. who use transportation or dental care
- Average annual expenditure per recipient, $
- Total annual expenditure, x$1000
Total Expenditures
No. of patients
Average annual expenditure per patient, $
Total annual expenditure, x$1000
* FY indicates fiscal year; DMH, Department of Mental Health.
† Managed care plan years.
‡ All expenditures have been converted to 1994 dollars by means of the gross domestic product deflator. The cumulative inflation rates from 1991 until 1994 were 7.1%, 4..1%, and 2.1% respectively.




View full report


"96cfpk01.pdf" (pdf, 147.17Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report


"96cfpk02.pdf" (pdf, 45.07Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report


"96cfpk04.pdf" (pdf, 50.53Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report


"96cfpk05.pdf" (pdf, 52.03Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report


"96cfpk06.pdf" (pdf, 1.02Mb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®