CMH/API 2000 Project

Proposal to

Add PACT to Anchorage’s Community Mental Health Services System

Introduction

An important shift has taken place in efforts to support the downsizing of API by enhancing existing and creating new community-based services. The attempt to place 16 adult acute care psychiatric beds in private hospitals by expanding the state’s designated evaluation and treatment (DET) program to the Anchorage area was unsuccessful. Anchorage hospitals are not interested in providing this level of inpatient psychiatric care – at least not for now. As a result, the CMH/API 2000 Project team is no longer looking to replace API adult acute care beds but instead is now concentrating on ways to reduce the need for them.

This demands a leap of faith: By delivering the right services in the right ways at the right times, we can provide the assistance people with chronic and serious mental illness need to help them to substantially reduce how often they experience a mental health crisis and become a danger to themselves or others or become unable to care for themselves. Further, it calls for a strategy for effectively concentrating the availability of program resources on the individuals for whom a marked decrease in their incidence of psychiatric crisis will result in a marked decrease in the demand for adult acute care hospital beds -- sufficient to accomplish the API downsizing.

In its ongoing effort to decrease the demand for API beds, the policy committee of the CMH/API 2000 Project is considering a proposal for increasing the capacity of the Anchorage community-based mental health system with the addition of a new program, the Program for Assertive Community Treatment, PACT. Based on nearly 20 years of experience with PACT programs in successfully reducing inpatient hospitalization throughout the nation, the National Alliance for the Mentally Ill, NAMI, launched a campaign in 1999 to establish the PACT in every state. The policy committee is considering the PACT for Anchorage after reviewing a recently completed analysis by project staff of how API’s acute care units have been utilized over the last five years.

The analysis of API acute care utilization revealed that a small number of people, 96 on average, or about 12% of the total users of API acute care services, account for a majority of the use of the Denali and Susitna units, 61%, from 1995 to 1999. Hence, by increasing the capacity of the existing array of community-based services at the high end of the intensity spectrum and by making these services available to people who would otherwise be the highest users of inpatient care, we are most likely to impact the demand for acute care hospitalization. Attached is a summary of the acute care hospitalization analysis, Attachment 1.

Program for Assertive Community Treatment

The five-year analysis of the impact of the small number of high users on the demand for API acute care beds confirms that we have good reason to look for ways to reduce the size of this group and the number and severity of psychiatric crises they experience in order to decrease the number of beds at API. Project staff has surveyed API files for 25, just under 30 percent, of the 87 persons who spent 30 days or more on the API acute care units in 1999. Based on information in the files alone, of this group of 25, 19 appear likely to be high API users in the future;13 of these have returned to API already this year. Of the 19, there are17 whom staff believe would be well served by a program for assertive community treatment located in Anchorage.

PACT is designed to maintain consumers in the community and to break the cycle of hospital re-admissions. The model calls for a community-based team that operates seven days a week, 24 hours a day, to provide a full range of medical, psychosocial, and rehabilitation services. PACT team members directly deliver the majority of services, with minimal referral to outside providers, in the natural environment of the person served. PACT services are comprehensive and highly individualized, and are modified as needed through an ongoing assessment and person-centered services planning process. Services vary in intensity based on the needs of the individual. For consumers with a dual diagnosis of mental illness and substance abuse, PACT integrates treatment for both with the goal of improving treatment engagement and compliance by focusing on the individual’s practical needs and by delivering services in community settings where learning may generalize more readily to daily living (Clark, 1998).

Most of the 400 plus PACT teams in the United States are provided by the public sector or publicly supported community mental health organizations and funded primarily by Medicaid, public mental health block grants and tax revenues (Salkever, 1999). PACT teams are working now in more than half of the states in the union, with statewide programs in Michigan, Wisconsin, Delaware and Rhode Island and local programs in Connecticut, the District of Columbia, Idaho, Illinois, Maryland, Missouri, North Carolina, New Jersey, New Mexico, South Dakota, Texas and Ontario Canada. The Veterans Administration is also developing PACT-related programs for the inner city and elsewhere (Grahl, 1998).

A recent review of the plethora of research on the effectiveness of PACT teams found that the PACT approach has succeeded in reducing hospitalizations by 13 to 94 percent fewer inpatient days for the clients served (Latimer, 1999). The highest success rates have been attributed to programs that adhere closely to the essential features of the PACT model pertaining to staffing size and qualifications and service delivery. The impact of PACT on hospitalizations appears stronger for older consumers but unaffected by the consumer’s race (Salkever, 1999).

Researchers find that the PACT model is most cost-effective when implemented to treat the most disabled clients in need of intensive services. Money is saved through reducing costly hospitalizations for these clients (Grahl, 1998). To be cost effective, PACT teams should be limited to those clients who are so severely ill as to be heavy users of hospital services (Essock, 1999). One study suggests an average of 50 inpatient days over the past year for consumers served by a PACT in order for the program to breakeven with more traditional and less intense case management services[1] (Latimer, 1999).

A PACT team consists of 10 to 12 clinical and rehabilitation staff, including a team leader, team psychiatrist, psychiatric nurses, vocational specialists, substance abuse specialists, social workers, registered nurses, vocational and rehabilitation therapist and a peer specialist. Each team serves as many as 100 to 120 clients. Because of the intensity of services delivered around the clock to PACT consumers, PACT veterans warn that full staffing is essential to ensure quality service and effective response to consumer needs and to avoid staff burn-out.

Briefly, what PACT could offer individuals of the high inpatient user group that is not available to them today is a team of medical, psychosocial and rehabilitation staff, working seven days a week and available 24-hours a day, to deliver services directly to them where they live. Most importantly, the consumer’s PACT team would be their first responder when the consumer is nearing or in crisis. The goal is for professionals with whom the individual is familiar to assist them to stabilize where they are and to avoid hospitalization. With a daily one-on-one contact relationship and crisis mitigation plan developed and adopted by the consumer and the PACT team, the incidence of crisis will be reduced for each individual served.

References

Allness, D. J., & Knoedler, W.H. (1998). The PACT model of community-based treatment for persons with severe and persistent mental illnesses: A manual for PACT start-up. Arlington: Programs of Assertive Community Treatment, Inc. (NAMI).

Clark, R. E., Teague, G.B. & others. (1998). Cost-effectiveness of assertive community treatment versus standard case management for persons with co-occurring severe mental illness and substance use disorders. Health services Research, 33(5), 1285-1307.

Essock, S. M., Frisman, L.K. & Kontos, N.J. (1998). Cost effectiveness of assertive community treatment teams. American Journal of Orthopsychiatry, 68(2), 179-191.

Grahl, C. (1998). PACT: Is it the answer? Interview with Mary Ann Test, Ph.D., Program for Assertive Community Treatment (1998). Behavioral Health Management, November/December, 26-28.

Latimer, E.A. (1999). Economic impacts of assertive community treatment: A review of the literature. The Canadian Journal of Psychiatry, 44, 443-455.

Salkever, D., Dominnnno, M.E, et al. (1999). Assertive Community Treatment for people with severe mental illness: The effect on hospital use and costs. Health Services Research, 43(2), 577-600

Attachment I

Summary: Analysis of Inpatient Acute Care Demand

A total of 869 adults were admitted one or more times to the Denali or Susitna Units in 1999. The Denali and Susitna Units are where people who are in crisis are cared for at API. They are referred to as the “acute care units.” Eighty-seven, or 10% of these people, were in what is referred to here as the “high users group.” Individuals in the “high users group” spent 30 days or more as patients on the acute care units in 1999 through one or more admissions. This 10% of acute care patients accounts for 50% of all the “patient days” spent on the Denali and Susitna Units in 1999. A “patient day,” a way of measuring hospital use, is one person occupying one bed for one day or less. We wondered if it was an anomaly that in 1999 a small percent of API acute care users accounted for a large percent of the total use of those services. We found it was not. Table 1 displays similar findings for the years 1995, 1996, 1997 and 1998.

Table 1. Numbers of adults accessing API acute care services by lower and high users groups, 1995 through 1999.

	Adults accessing API acute care
	Total	Lower users: persons accumulating < 30 inpatient days # %		High users: persons accumulating > 30 inpatient days # %
1999	869	782	90	87	10
1998	821	718	88	103	12
1997	767	664	87	103	13
1996	740	645	87	95	13
1995	635	545	86	90	14

Perhaps more interesting though, we found that the utilization of inpatient acute care in the year 1999 was different from that of the other years in a way that is important to the goals of the CMH/API 2000 Project. In 1999, the use of API acute care services in total was less than it had been for any one of the previous four years.

Finding out why use of API acute care hospital beds was down for 1999 has given us a much better understanding of how changes in the demand posed by a small number of very high users impacts the overall demand for acute care beds. These findings support concentrating project efforts and resources on the group of especially high users, and potential future very high users, in order to perpetuate the decline in demand for API acute care patient beds experienced in 1999.

A. What the numbers tell us.

The number of adults (unduplicated admissions) accessing API acute care services (Denali and Susitna Units) has steadily increased, a total of 37 percent between 1995 and 1999. The number of admissions for acute care services has also steadily increased, a total 33 percent for the same period.

Surprisingly, the number of total patient days has decreased, not so steadily, but by 8 percent between 1995 and 1999. Hence, while admissions to the acute care units is higher in 1999 than for any of the preceding four years, the total number of patient days spent on Denali and Susitna in 1999 is the lowest experienced in this five year period.

The utilization rate of 32[2] API acute care beds has gone from 97 percent in 1995 to a high of 136 percent in 1996 to the low of 89 percent last year. The number of persons in the lower user group, those who use API acute care services less than 30 days over the course of a year, has grown steadily each year, a total increase of 43 percent. In contrast, the number of persons in the high users group is nearly flat with a decline in 1999.

B. Patient days, not admissions, will measure project progress.

Comparing 1996, the highest year for acute care inpatient days, to 1999, the lowest year, emphasizes the importance of relying on number of patient days, rather than the number of admissions, to evaluate our progress in affecting a decrease in the demand for API acute care services. In 1999 there were15 percent more API admissions than in 1996, but in 1999 the total acute care patient days were 35 percent less than in 1996; that is, while admissions were higher in 1999 than in 1996, bed use was much lower. So how can this be?

Table 2. Average numbers of annual admissions per person, patient days per admission, and total inpatient days per person compared for high and lower users groups, 1995 through 1999.

	Average number admits/person			Average number patient days/admit			Average number total patient days/person
	All	Persons receiving < 30 days	Persons receiving > 30 days	All	Persons receiving < 30 days	Persons receiving > 30 days	All	Persons receiving < 30 days	Pesons receiving > 30 days
1999	1.3	1.2	1.8	9	7	31	12	7	59
1998	1.3	1.2	2.1	13	7	42	17	7	91
1997	1.3	1.2	2.3	12	8	32	17	8	74
1996	1.3	1.2	2.2	16	8	52	22	8	113
1995	1.3	1.2	2.2	13	8	35	18	8	78

Table 2 shows that the answer lies in the amount of variability in the numbers of patient days attributable to high users versus the stability in the numbers of patient days for the majority of users. The average number of admissions per person is fairly stable for both groups. Average number of patient days per admission and per person for the majority is also fairly stable. But this is not so for the high users group: the average number of patient days per admission for the high users varies dramatically from a high of 52 days to a low of 31 days, as does the average number of total days spent at API per high user, from 59 to 113. Charting these fluctuations illustrates the dramatic differences between groups.

C. Conclusion

Compared to the previous four years, the year 1999 was a remarkable one for API. It was the year of the highest number of adult acute care admissions, the year of the highest number of adults accessing acute care services, and the year of the lowest number of adult acute care patient days. This decrease in patient days was not brought about because time spent in API for the majority of acute care users was less; but because the number of patient days spent at API by the minority, the highest users, was less.

Further, in 1999, the number of people falling into the high user group,

those spending 30 days or more in acute care at API through one or more admissions, was the lowest it had been in five years, while the number of adults spending less time in adult acute care at API was the highest recorded for the same time period.

The decreases in demand for API acute care services by the small group of high users offset increases in both the numbers of and admissions by the majority of API acute care users, those spending less than 30 days there in a year. We can conclude that the high users group’s impact on the demand for acute care services is the result of three variables: number of high users, the average number of patient days per admission of high user, and the average number of inpatient days per high user. Our ability to impact these three variables for this small group of people is pivotal to our ability to reduce the number of acute care beds at API.

Table 3. Numbers of patient days of API adult acute care services provided and number of admissions, by high and lower users groups, 1995 – 1999.

	Patient days of acute care provided					Admissions
	Total	Persons receiving < 30 days # %		Persons receiving > 30 days # %		Total	Persons receiving < 30 days # %		Persons receiving > 30 days # %
1999	10,356	5,227	50	5,129	50	1,142	977	86	165	14
1998	14,170	4,813	34	9,357	66	1,075	853	80	222	20
1997	12,762	5,120	40	7,642	60	1,036	798	77	238	23
1996	15,938	5,205	33	10,733	67	989	781	79	208	21
1995	11,311	4,333	38	6,978	62	856	655	77	201	23

---

[1] The group of 87 that accounts for 50 percent of adult cute care patient days at API averaged 60 days total in API in 1999.

[2] Reported API acute care capacity.