Getting chronic mental patients
out of hospital:
Big savings and a better result
The U.S. experience as seen in Springfield, Mass.
Possible Montreal Applications
A report by Jim Farquhar MD
Psychiatrist, Montreal, Quebec, Canada
September, 1995
To Internet readers: James Farquhar MD wrote this fact-finding report in summer 1995, for hospital and mental health people in the Montreal area. This was when very few people in the Montreal area had heard of Assertive Community Treatment. It serves as an introduction to the A.C.T. idea, and also to a type of highly-staffed group home which offers a much better alternative to "back wards" for the "very chronic" psychiatric patients seen in every psychiatric hospital. This article was circulated widely but was not published in a journal.
CONTENTS:
Testing the limits of de-institutionalization
Para-professionals (Community Mental Health Workers)
Highly-staffed group homes in Springfield
A visit to a hightly-staffed group home
P.A.C.T. in action: Two impossible hospital patients living on their own
How P.A.C.T. is Organized in Springfield
Applying the models in Montreal
Recommendations for the Montreal area
Appendix I: Rough cost-benefit analysis of PACT model
Rigorous cost-benefit analysis of one real PACT program
Highly staffed group homes can be a cheaper and more effective location for chronically mentally ill people to live, compared to hospitals.
And, surprisingly, it is even cheaper and more effective to give some of the most difficult patients their own apartments, with intensive visiting by supportive para-professional staff. This is known as the Program for Assertive Community Treatment (P.A.C.T.) model.
These two complementary housing solutions have permitted a high-quality and cost-effective mental health care system in Springfield, Massachusetts and more than 300 other American communities. Hospitals and mental health professionals are still important there. However the need for hospital beds has shrunk dramatically because of the community-based supports.
Both these options permit more widespread use of Clozapine outside hospital, for the many patients who could now be expected to take weekly blood tests reliably, guaranteed by the high level of supervision.
The move to community-based community resources is now a main-stream trend in the U.S. supported by 20 years of outcome research. Much of this is summarized in the July, 1995 issue of Psychiatric Services (formerly known as Hospital and Community Psychiatry). I am appending these articles to this report, for reference.
In Montreal, we are now challenged, indeed obliged, to cut hospital-based costs. Budget cuts, bed cuts, reduced positions in nursing and other disciplines, are all imposed by external authority -- the provincial government and the Régie Régionale. The same goes for the accordion-like expectation to squeeze the patients of the Reddy Memorial and the Queen Elizabeth hospital into our shrinking anglophone hospital system.
These powers are also imposing on us an approach to develop community-based resources to enable us to perform these miracles.
Against this backdrop, I decided to gather information to help our group look realistically at what can be achieved with community-based resources. My interest took me to Springfield, where this approach has been taken to the limit.
I was pleased to see that patient care quality is high, and so is the level of comfort of professionals and para-professionals working in the system.
Can the Springfield model be applied in Montreal? I definitely believe that these simple housing concepts can be introduced comfortably and gradually, to achieve our goals of quality care and lower cost.
This report gives a look at how the Springfield resources function. The reader can picture how the models might work in Montreal.
With hope and enthusiasm, I invite you to read the whole report so that we may discuss these two housing options together.
TESTING THE LIMITS OF DE-INSTITUTIONALIZATION
A journal article with this title caught my interest in June (1).
It described the mental health care system of Western Massachusetts, centred in Springfield. To quote the musical "Oklahoma", "They've gone about as far as they can go."
Since 1978, a court judgment motivated people to reduce psychiatric hospital beds from about 550 in 1978 to 68 today, for a population of more than 850,000. This includes 40 acute-care beds and 28 chronic beds.
After contacting the mental health administrators in Springfield, I was invited to come and see for myself how they keep patients out of hospital. I went, on Aug. 2, 1995, with Steve Wohl, chief of the community resources unit with offices at the Douglas Hospital.
Springfield has gone as far as any area in the U.S. to maintain the mentally ill in the community, according to the administrators.
They have found that a few key concepts can make a community-based approach a success, from the viewpoint of patients as well as professionals.
Many patients are stable in board-and-care arrangements, similar to our foster homes (familles d'accueil). But almost 260 patients need more support than that.
They find it in highly-staffed group homes (60 patients) and a PACT approach (which they call "Outreach" or "supported housing"), serving about 195 more patients.
Both these options cost considerably less than hospital beds.
These two services are provided by community-based agencies, one for each geographical sub-region of Western Massachusetts. They report to the state Department of Mental Health rather than to a hospital. They are grouped under an umbrella agency called the Center for Human Development.
These two approaches evolved and survived, while some other ideas were abandoned.
In particular, staffed group homes with more than four patients were found to be cost-ineffective and also not favored by the patients (called "consumers" in Springfield).
Conflicts with the rules, and between patients, took up too much costly staff time in the larger group homes for chronic patients.
Staffed group homes now serve four people or fewer, with staffing around the clock, sometimes one and sometimes 2 staff per shift.
These homes are for people chronically disturbed, confused, disinhibited, or regressed. In Montreal these represent the most chronic and disabled in-patients, excluding those who are dangerous even in hospital, or who need frequent non-psychiatric medical attention.
PARA-PROFESSIONALS (COMMUNITY MENTAL HEALTH WORKERS)
Besides the two housing concepts, the key to quality and savings in the Springfield system seems to be extensive use of para-professionals, rather than more highly-paid doctors, nurses, occupational therapists.
These mental health workers may or may not have a university degree. They rarely have formal qualifications in nursing, social work, or other professions. The para-professionals form the backbone of the group-home and PACT systems.
They are chosen for intelligence, a caring attitude, people skills, and other good qualities we also see in their Montreal counterparts.
Their pay is equivalent to $21,000 to $24,000 Canadian per year, with good benefits. Although this is in the low end of the middle class for the U.S., managers have had no trouble recruiting in recent years because there are few such interesting jobs elsewhere for similar pay.
(It seems likely we would have to pay higher than this in Montreal because of higher living expenses and higher taxes here.)
All things considered, the knowledge base required of these staffers is much narrower than for nurses, occupational therapists, doctors, and other hospital-based mental health professionals. A nurse or other professional would be expensive as a person to spend the morning helping a client shop in a grocery store.
However, nurses and social workers are team leaders or members in other similar PACT-type programs in the U.S.
The para-professionals get 70 to 100 hours classroom training in their first year, in addition to learning on the job. This includes 16 hours of formal training about medications and their side effects. They pass an exam to be certified in helping administer the medications, just as foster home care-givers give medication in our system.
"Para-professionals often are the ones who develop, write and implement the treatment plan."
Training of the para-professionals is not done centrally in the Western Massachusetts area. Each sub-region organizes its own training. Sub-region manager Audrey Higbee was able to give us an outline of the core formal training given to para-professionals in the first four months of being hired, while they were also learning on the job, at staffed apartments or in Outreach.
There are weekly 90-minute group training sessions over 16 weeks, covering the following topics:
- Mental Illness: Overview
- The Agency and Recovery Models
- Mental Illness: Empathy and Empowerment (parts I and II)
- Treatment approaches (parts I and II)
- Communication with clients and other staff
- Dual Diagnosis (mental illness and substance abuse) (two parts)
- Hallucinations and Delusions as Metaphor (listening to feelings through psychotic phenomena)
- Violence and Aggression
- Crisis intervention with violent or aggressive clients
- Sexual Abuse in childhood - consequences for adult behavior
- Suicide and Despair
- Familes (two parts)
In addition, there is a separate series of 16 hour-long classes on psychiatric medication and side effects. If the staffer passes an exam at the end of this, they become certified to give out medications (not to prescribe, but to administer) to clients in residences and at the clients' homes.
They are taught first aid and also CPR.
As in most organizations, there are regular meetings and educational talks for sharing and learning new information, thus a kind of ongoing training.
A separate agency exists for training, called The Training Consortium. I had the impression that different regions made use of this resource to different degrees. Largely the agency seemed to play a consultation role rather than acting as a school for new employees.
The para-professionals are not unionized. (However, some of them are reported to be unionized in other areas.)
HIGHLY-STAFFED GROUP HOMES IN SPRINGFIELD
How can community group homes house the most chronically disturbed and regressed patients usually seen on "back wards"?
Structure is the key allowing Springfield's highly-staffed group homes to house patients who live in hospitals in Montreal. And structure has its price.
Structure of the physical environment includes a single room for each patient. This offers privacy and dignity. It is also a retreat when the patient feels too stimulated, and a "time-out" place for use in behavioral programs or in case of agitation.
Four is the maximum number of patients in one home. This reduces friction between patients to a minimum while maintaining economies of scale. These economies are increased when there are two "homes" (apartments or duplexes) in a single building or neighborhood.
Structure in human terms is given by the para-professionals who are there 24 hours a day. They ensure a reassuring routine including meals, activities, bed-times, and enforcement of rules.
Patients may have some chores or no chores, depending on their ability or disability. A given home does not rigidly apply all rules to all patients. Each patient has an individualized care plan including expectations, entitlements and, sometimes, a behavior program.
Patients are not pushed to perform above a certain level to "graduate" from the home. Rather, it is recognized that many or most of the patients will need the structure of such a home for the rest of their life.
Different homes have different numbers of staff, depending on the difficulty of the patients' behavior. Because the homes are in a network, staffing can be increased or reduced in any home if there is more or less need over a time. The network is large enough to provide for replacement workers if staff is sick or on vacation.
Staff are supported by a central administrator, and by the other resources in the community. This includes the hospital, psychiatrists who may visit in some cases, and a crisis-management resource. Depending on the network, this may be a more highly-staffed home, an agency such as TRACOM, or the emergency room.
Because the homes are in a small network, one para-professional can do most of the grocery shopping for the whole network. Meal preparation is usually done by a para-professional. However, there is a flexibility in the network. Some meals could possibly be prepared for all homes in the network by one of the kitchens, or by a cook contracted for the work.
Housework is done by patients as part of their programs. Failing this, some housework is done by staff. An earlier effort to use low-cost cleaning persons on a daily basis was economical but caused discomfort and conflicts regarding status and job boundaries.
In one of the sub-regions, a cleaning crew has been set up to help with the heavier housework for all the group homes. They visit each home regularly. This involves one full-time cleaning person working with two half-time people recruited from among the more functional clients. This appears to be a good solution for a system involving about 70 para-professionals (including group home and PACT).
Who takes patients to their appointments? It depends on the network. Staff of the home may do this if their presence is not needed every minute in the home. Elsewhere, other arrangements are made, sometimes involving volunteers or para-professionals designated for errands.
About overhead, highly staffed group homes have an "overhead" cost similar to a foster home in Montreal. There is rent (or mortgage), food, heating, maintenance and so on.
A VISIT TO A HIGHLY STAFFED GROUP HOME
A "group home" we saw gave us an idea of how very regressed and psychotic clients, perhaps the most difficult in the system, could be kept in the community.
We were told that most of the clients here simply stayed at home all day or might go out to a store at times. A few of them spent up to 1.5 hours on weekdays in a sheltered workshop or drop-in center.
The home was a large duplex in a spread-out, residential part of the Springfield area. It houses eight clients who are famous within their system for dramatic behavior difficulties.
The clients we met appeared to be similar to the very schizoid and disorganized patients one sees on some wards in the Douglas. They were able to be friendly and hospitable to us with guidance, but could tolerate the stimulation of others for only short times.
The home was organized as two separate homes housing four clients each, on two floors. Each client had his own private room which, on our visit, was clean and neat. (My own house should be so tidy!) There was a cheerily decorated common area with TV, although the client we met also had his own color TV in his room. There was a separate dining area and also a modern kitchen. There were at least two bathrooms for four clients.
This particular home was remarkably spacious. I gathered it had been built in the last 10 years with just such a purpose in mind, on a fairly large lot with a big lawn and some trees and bushes marking the edges of the territory. It was a lovely house with grey-painted aluminum siding common to the area.
The rooms seemed to be about 10 by 10 feet plus walk-in closets about 3 by 8 feet at least. Inexpensive but attractive carpeting was everywhere. Clients decorated their own rooms. The man we spoke to most was a sports fan and had put up posters of some of his heroes.
I gathered he was a severe chronic schizophrenic with residual thought disorder. It was hard for him to give the names of his own sports heroes even though the names were written on the posters.
Another resident was a woman in her late thirties, rather overweight, who was known for her extreme disinhibition. I understood that this meant she had called negative attention to herself in the past for sexual, aggressive and other behaviors related to her mental illness. I supposed she might have been schizo-affective but she appeared to have a severe regressed residual syndrome.
My impression was that the luxurious space of this home was helpful, but was not the key to its success at keeping these clients stable.
However, all leaders we met are convinced that single rooms are necessary for their clientele. For Montreal, we would probably want to continue shared rooms in existing foster homes, but use single rooms only for a highly-staffed home proposal.
Smoking rules did exist in the home for the regressed. Smoking was forbidden except in a designated area. Most residents preferred to smoke outside. (Their winter is milder, of course.) "We find that exhaust fans really don't work to clear the air in a living room."
Group home staff were paid on the same scale as Outreach (PACT) workers, about $16,000 to $18,000 US ($21,000 to $24,000 Cdn) per year. The house manager got about $22,000 to $24,000 US ($29,000 to $32,000 Cdn).
This particular group home had a total of 10 staff for eight clients. They worked eight-hour shifts, typically one person per floor per shift, but sometimes overlapping for the busier times. Such a two-in-one group home would have the advantage of being able to deal with sudden unwanted behavior or crises in one client by doubling up the staff on the spot, while calling for backup.
(This is a similar ratio to what we see in hospitals, but the workers are paid less than hospital professionals. This is a saving to the system, and perhaps a disadvantage to the professionals who are displaced.)
Each client has an individualized treatment plan, mostly decided and administered by the team running the home. These clients were definitely not allowed simply to vegetate. They are encouraged to take part in activities inside and outside the home, within the limits of their disabilities.
Money management programs are fairly common for clients and these can be integrated into the treatment plans for the rewards.
Management of Social Security (Welfare) cheques by a third party, usually not the case manager, is usually voluntary. However, it can also be imposed by a psychiatrist in conjunction with the Social Security authorities.
As a routine, a person outside the immediate treatment team is asked to control the money, to prevent quarrels between the client and his day-to-day helpers.
The atmosphere at the home seemed quiet and pleasant, much like a Douglas Hospital chronic ward during a stable time, except that the physical surroundings were more homelike and pleasing.
The lovely hardwood furniture we saw in the dining room had been made in a sheltered workshop we saw later.
TWO IMPOSSIBLE HOSPITAL PATIENTS LIVING ON THEIR OWN
Tony and Fred are two burly young men about 30 who share a two-bedroom apartment.
They are remarkable examples of the success of the "supported housing" or PACT concept in helping people with difficult mental illness, personality, and drug abuse problems.
Both have been violent while psychotic in the past. Fred has beaten up people a few times while in a paranoid state, and he once set a house fire that almost claimed a couple of lives. His psychotic states can be dramatic, with vivid visual hallucinations of a ball of bright light from which he hears dangerous voices. The voices give infuriating insults, leading him to lose control. Or they may urge him to do dangerous or aggressive things. These problems occur, of course, when he stops taking his medication, and they are worsened by use of cocaine or other drugs, including frequent drunkenness.
Thanks to Clozapine, he has been well for more than a year. Today he appears mildly shy but fun-loving and in a good mood.
He is considered schizo-affective and has also struggled with cocaine and alcohol abuse. He has been well with medication for more than a year, and abstinent from drugs for seven months.
Today, at his best, he is smiling and friendly, but has just-noticeable pressure of speech and is more intrusive than most people would be comfortable with. All the more since he has a wrestler's build.
"We're the Odd Couple," says Fred with a smile. "I'm Felix and I forget the other guy's name on TV."
Their apartment has plenty of space, and it is tidy and nicely decorated with wall hangings and posters. There are a few framed family pictures in Tony's room.
They are not in a run-down neighborhood. Their place is in a two-storey low-rise apartment building which might also house young families and seniors.
"We don't have any neighborhood resistance to mental patients moving in, because they are just about invisible," says our host, director of supported housing (called Outreach, like PACT) for the sub-region.
"They were just two guys renting an apartment. If they had trouble here they could move somewhere else."
Tony and Fred are on their own, stable and able to function outside hospital largely because of frequent visits and meetings by Springfield's Outreach staff -- usually the same person.
In Montreal's system, they would probably be among the most difficult and feared patients in some hospital's revolving door, admitted two or three times a year for several months at a time.
At first, they required intensive attention at home, up to 20 hours a week, by a mental-health para-professional. Attention to detail was important. The young men needed to be reminded of their appointments or their weekly blood tests for the clozapine routine, and often they were driven there by the worker. They needed guidance and supervision in handling money, in part to discourage drug abuse, which was always tempting.
They needed to learn new habits of thinking and doing things. At first, one or another worker was visiting several times a day -- to supervise medication, food preparation, and other things that can go wrong and lead to re-hospitalization.
It has been a learning process for both, and there have been setbacks. Now the worker visits only once or twice a day, for a total of four to eight hours a week.
The worker's comfort and confidence in helping the pair is helped by knowing of the large system of drop-in centers, activities, respite homes, and other resources that form the backup. Hospitalization remains an option, but is not usually required for the mentally ill in the system.
The resources are mainly people and not buildings.
The whole system's budget is taken up by more than 70 per cent in salaries. The balance reflects some housing that has been bought and mortgaged, mileage costs for the many workers on the road, and other non-employee expenses.
Both Tony and Fred are very satisfied with their new life, especially compared to previous living situations provided by the mental health system.
An acquaintance, George, tells of how he hated living in a residence for "dual-diagnosis" patients -- mentally ill people with substance abuse problems.
"I lived with three other clients, and I had my own room in a two-floor apartment. You'd think that I would be happy with that, but it felt like jail to me.
"In fact, it was a kind of jail. I got in trouble when I was ill and intoxicated, and a judge sent me to that residence.
"There were a lot of rules, and it was hard not to break the rules.
"I didn't really get along well with a couple of the clients. They were noisy and I had spent a few years just living in the woods west of here.
"Now I'm in my own place with a friend and I really feel like it's mine. I make most of my own rules and no one bugs me if I don't get up on time.
"I feel like I don't have to hassle with the system. I just enjoy my life and I get a lot of support from the people who visit."
Our host noted that the independent-but-supported living option costs the system about half as much as living in the residence, if you take account of all system costs. (This one of a few confusing estimates of savings we got that day, but everyone agreed that PACT was cheapest for difficult patients not requiring constant supervision.)
Once the client gets used to living in an apartment with supportive visits, employee time and related costs are much less than in a supervised residence.
"I don't care if it costs less, I like this way of living more," says Fred. "If you ask me, the consumer's choice should count."
"When I see the same person visiting all the time, I get to trust him. That's important to me."
But he admitted: "I had to test them a while." He smirked and a trace of pain crossed the system supervisor's face.
"But they were there all the time for me. Someone would come in the middle of the night if I was feeling real nervous, you know."
Fred too had lived in a large apartment complex designated for mentally ill people, called Maple House.
"It's not good to be around other people who are going off their medication and taking cocaine or smoking a joint. There was sex going on there that didn't make me feel right.
"You get together with someone and smoke up, and then you go and see the counselor one floor below, and you're high. You get the counseling all right, but you still might end up in the hospital.
"I found it disgusting. It was almost like institutionalization. Too many people too close together.
"In a big building like that you don't feel like it's your own home, so you don't care about what it looks like.
"Now I'm in my own apartment, I do care. I fix it up the way I want. I'm really living.
The manager of mental health services for Springfield region agrees. "The congregate living situations cause more problems than they solve," said Jim Goodwin.
(Their experience mirrors that of east-end francophone Montreal, where the larger "pavilions", with more than 10 residents, have largely been dismantled in favor of smaller group homes.)
George had a caution: "People have trouble getting used to the loss of the routine in an institution at first. I had to find a new routine. Now I have an apartment with another man, I go to a day program and I also have some part-time work."
George, now 37, spent a total of 13 years in hospital. With schizo-affective illness, his various psychotic episodes have ravaged his brain like lightning storms.
George has been a celebrity because he once commandeered a bulldozer in an urban airport and trashed three airplanes before he was finished. He was off medications and was acting on delusional beliefs and fears, with manic energy. He has stopped medication all too often, paying for it with a burnt-out concentration.
Yet he still has an amazing vocabulary and wit.
"Little things matter," he says. "When I was first discharged directly to an apartment, it didn't work, because I didn't have the energy to go up four flights of stairs several times a day.
"The second time I got in on the ground floor, you could say, and that has been fine."
One mental health client we met, Louise, said she once definitely preferred a supervised residence. She feels fragile, but believed she was at her maximum of independence in a home structured by someone else.
She is an obese woman, highly myopic, in her early 30s who loses the thread of her thoughts at times and is visibly nervous about attending a meeting for us, the inquiring visitors. In Montreal's system I would guess she would be in the hospital half the time, and unable to live in an ordinary foster home (famille d'accueuil).
In Springfield she now lives with another woman in an apartment supported by visiting. She attends a few outside activities.
A staffer told us: "Louise became dependent, and now is proud she can manage her own money. But it was only because of a relationship that was built. Teaching the skills is easy once you have the relationship, and that can take a long time.
The clients we met now have a high degree of independence. "Yet at first they didn't do the things they knew how to do, like use a bank, because of fear and hopelessness. Our program builds their confidence to do these simple things, and we're right there with them, step by step."
How P.A.C.T. is Organized in Springfield
I had to visit Springfield to understand that PACT is not just a flurry of visiting to pamper patients who should be in group homes.
Rather, PACT removes from hospital the patients who fail in group homes, and it keeps them out. This saves much money, needless months of illness and disability.
Springfield PACT patients are like Montreal patients hospitalized once or more per year, because of non-compliance with medication and poor social supports. These "revolving door" patients include those we consider our worst management problems, for example chronic psychosis with residual disability, drug or alcohol dependency, personality disorders and family problems.
There are sheaves of proof (2) from across the U.S. that PACT patients have much shorter and less frequent hospital stays.
PACT teams of para-professionals tend to be organized in groups of 10 or 11, including a supervisor who is paid somewhat more.
Each team member is responsible for ten ordinary patients, or fewer if the case-load is time-consuming. One patient is typically visited two to five times a week, for a total of 2 to 20 hours.
For clients with intensive needs, a team of five staff and a supervisor take care of about 10 clients, with visits to each client several times a day.
The team keeps a file on each patient, including the individual care plan. A given patient might be supported by visits, phone calls, and involvement with other resources with the aid of the primary para-professional. This person usually takes the role of case manager for the patient within the entire mental health care system.
As in a hospital ward, the PACT team works in shifts, with the most staffing during the weekdays. There is always someone on duty, or at least on call.
The most problematic patients are discussed in a daily morning meeting of all staff and the supervisor. Then most staff spend the rest of the day on the road and on the phone.
They visit patients to make sure medications are taken, to drive them for a Clozapine blood test, to go shopping or banking with the patient, and to help them remain organized in other ways. The goal is always to get the patient to do more for himself. But if a continuing follow-up is needed even for years, the PACT team provides it.
As for any out-patient treatment, PACT may fail, and the patient must be re-hospitalized. However, people in the PACT program averaged only 12 days in hospital per year over a six-year period (1987-1993).
(This may be compared to 60 to 90 days, or more, per year in the hospital for similar patients in Montreal. At $300 to $500 per hospital day, one can see how PACT brings big savings. See Appendix I.)
Springfield apartment rents for patients are comparable to those in Montreal or even more expensive. The geographical region of Western Massachusetts is quite wide, involving some clients who live many miles from the centre. This again is similar to the Montreal area.
Overhead costs are minimal. Springfield PACT teams don't need a separate office for each staffer. Staff are in their cars all day, and just need some desk space, voice mail and regular mail at the end of shift. Two large office rooms are enough for the ten staffers, similar in total area to the 4-east nursing station at Montreal General Hospital.
Staffers use their own cars, and are paid 33 cents a mile on the job. The total car expenses vary according to how spread-out the clients are in a region. Equipment costs are modest. Cellular phones seem to save more in time and money than they cost.
This is different from a hospital in-patient setting, where unionized housekeeping staff clean the floors daily, dietitians supervise meal preparation, pharmacists spend meticulous hours keeping track of medications, and psychiatrists are paid $18.15 per day just to say hello to the patient (visite de controle), or to exchange a few minutes' conversation. Nurses are obliged to document every medication and a summary of the patient's behavior each shift. Much nursing time goes to enforcing rules, particularly smoking rules, which are not necessary in a supported-housing, PACT, setting. There, the client does not inconvenience others by having his own pace and schedule.
These factors, and others, probably explain why the PACT approach has been adopted in so many places in the U.S.
It is "our model of choice", says Jim Goodwin, PACT administrator in Springfield.
"Many clients do better in Outreach (PACT) than in staffed apartments or group homes. In the group settings they get into conflicts, perhaps for their own personality reasons even more than their mental illness. They get into conflict with other people, staff or clients, and with the rules. They may feel that the group setting is another institution which they don't belong in. In their own home, they don't have anyone to have a conflict with, and they have a sense that they are in their own environment."
For example, a person with obsessive-compulsive disorder who spends hours a day cleaning himself in the bathroom, will be the focus of endless conflict in a traditional foster home. Others have to use the bathroom, and there will be arguments. But support such a person in his own apartment, and this useless conflict ends. He may be just as symptomatic, but he is happier and no longer at odds with his environment.
A portion of the system's savings are due to increased use of Clozapine in the last few years, our hosts told us. Their impression was that Clozapine gave a superior result for the most mentally ill patients, and was better tolerated and liked by the patients. In staffed group homes and in PACT, the system could depend on patients getting their weekly blood tests, which were a priority for the staff. In Montreal, the most difficult patients usually don't get Clozapine because no one can ensure their compliance with the tests. All the managers were enthusiastic about Clozapine as a key element in improving the patients' quality of life, and in keeping them out of hospital.
A recent review article in the American Journal of Psychiatry strongly recommends Clozapine for most chronic psychotic patients who have been treatment-resistant.(7)
The Springfield region has a parallel program for mentally handicapped adults, in Westfield, with the same philosophy of visiting and supporting. Of course, there is almost no trouble with medication compliance, so it is largely "education-oriented".
APPLYING THE MODELS IN MONTREAL
Some reflections emerged from the Springfield visit, my reading, and discussion with Steve Wohl and others.
I am enthusiastic about both models. PACT in particular has a solid success record for effective treatment and reduced costs. (3,5) Of various interventions developed to cope with de-institutionalization of the severely mentally ill, PACT is "by far the most carefully defined, well documented, and successful."(2)
Its cost-effectiveness has been excellently documented (4) in many settings, although I cannot find specific reports of how much it may save in dollars and cents. I believe this vagueness reflects the difficulty of measuring the savings.
A recent article even describes in detail how to implement PACT teams even in the absence of new funding. (6)
About money, we should not forget that the government has promised a new recurrent budget of $10 million starting in 1995-96, for community-based mental health services on Montreal Island. Louis-H. Lafontaine gets a separate new budget of $6.5 million to de-institutionalize 130 patients. Of the $10 million, $5 million are earmarked for new residential services and community-based rehabilitation and crisis-response.
For the Douglas, Royal Victoria and Montreal General alone, our needs will probably get 20 to 25 per cent of the $5 million, thus $1 to $1.25 million. This is logical since they serve sectors containing 400,000, or about 20 per cent, of the island's population. The Douglas should get an even bigger share since it houses very difficult patients who have accumulated over the years from all over Quebec. Others may be able to refine this estimate. This is an approximation of the main source of funding for new projects.
I imagine that our group will want to support proposals for both options, a PACT team and one or more highly-staffed group homes. If other hospitals or Montreal regions like these ideas, we can benefit from sharing experiences and consultations between regions. It seems sensible to organize the resources by regions, as in Springfield, perhaps roughly corresponding to the current sectors of one or more hospitals.
PACT keeps revolving-door, barely-functioning patients out of hospital, while the highly-staffed group home serves patients even less functional.
The two ideas are complementary, not competing with one another.
One colleague, Steve Wohl, has presented an organisational model for a four-person highly staffed home. With a single room for each patient, and a staff of six mental health workers to cover all shifts for the four patients, it would cost about $160,000 per year. This includes rent or mortgage, food, other overhead costs, and salaries for six full-time community workers.
This figure of $40,000 per patient per year seems high at first. However, note that L-H Lafontaine is getting $50,000 per patient per year to de-institutionalize some of its hardest clients, 130 of them. That should be plenty of financing, with money left over for day programmes and support services.
A colleague at a large psychiatric hospital says that the government has to pay an average of $75,000 per patient per year to keep them in pavilions on the hospital grounds. Heating and maintenance are big expenses. So the $40,000 per year gives a saving of almost 50%. However, you would have to close the pavilion entirely to get rid of all its costs.
I have not been able to get solid figures for current costs for chronic regressed patients now in general hospitals. It is hard to determine the total of nursing costs, maintenance costs, food, and much invisible overhead. Again, a big saving can be achieved by closing an entire ward, but savings are less for a few beds. I am told that the costs from the hospital budget are sometimes estimated at from $50,000 to $150,000 per year. Those figures are all more than $40,000 for the highly staffed homes.
I imagine that many ideas could be tried: slightly more structured group homes, ordinary group homes reinforced by extra staff, and so on. In these cases, the community agencies that organize the highly staffed homes could place one community mental health worker per day, seven days a week, in an ordinary foster home (famille d'accueil). With this presence, that home might be able to tolerate perhaps two more difficult patients among their residents. The extra structure and direction from the workers might help them stick to their medication and abide by the rules. This would cost about $36,000 per year in salaries, thus $18,000 per year for those two hard patients. (It might also help keep the other, easier, patients out of hospital.)
By comparison, in the official anglophone foster home system, the average government subsidy (on top of the patient's rent contribution from welfare) is about $3,000. But if $18,000 seems more than $3,000 per year, then $18,000 is still much less than keeping these "impossible-to-place" patients year-round in mental hospital beds costing $75,000 or whatever.
I must emphasize that Mr. Steve Wohl did not feel that this "partial" solution would work.
And in Springfield, after 17 years of experience in pushing de-institutionalization to the limit, there are only two models beyond foster homes: highly staffed group homes and PACT.
In order to create a PACT programme, you need a community-based agency, according to the clear terms of the new government financing available. Mr. Wohl says it's just a question of creating a non-profit organization (French: OSBL), such as a government-subsidized community agency. This is done in the framework of a partnership with the Régie Régionale and other interested parties. Hospitals can ensure their satisfaction by having representatives on the boards of directors of these agencies. In some regions, such an agency already exists, which could be adapted to provide this service.
There are many good community mental health managers with experience who would probably be glad to make the highly-staffed group home system work. In our region, the RESSORT agency might easily adapt to the new task.
This is a simple model. We merely have to apply the experience and expertise which already exists.
Montreal General Hospital is already ahead, having submitted a proposal (March 1995) for a Community Link Service based on the PACT philosophy. It would have served 20 patients at a time with 1.5 nurses and 1.5 community mental health workers, and a half-time psychiatrist.
Despite the many hours invested in creating this proposal, I believe it is unlikely to be approved by the Régie Régionale. Mainly, it relies on "high-priced help" such as doctors and nurses. Also, it is not based in the community but appears to be an extended service of one hospital.
If I can claim any clairvoyance on the matter, it iss just that I have been keenly following the health-care budget cutting for two years now. One of my conclusions is that government's unstated agenda, or wish, is to replace highly-paid workers with lower-paid workers, wherever this can be done. Many of us would deplore such an attitude, yet it seems to be a key for getting projects approved.
I and some others now think the Régie Régionale would probably fund a PACT kind of programme entirely staffed by community mental health workers.
Mr. Chris McFadden, a veteran organizer of community mental health programs, has become very interested in PACT. He is starting to draft a proposal for a 10-person team to attend to up to 100 patients. This would serve our three hospitals at least. It is a hard planning job, since (in early September 1995) it is hard to gauge the money available, the political interest from above, and the motivation from hospitals and other groups.
As noted in Appendix I, this program would eliminate the need for possibly 25 acute-care beds, more or less, in our three hospitals, at a cost of $5,000 to $8,000 per patient served per year (cost depending on difficulty of patient). The Douglas has been asked to reduce beds by 30 in the next year. The RVH must reduce by 13, and the MGH by 9, over the next three years (as I understand it).
Highly-staffed group homes, or enlargement of the PACT program, or other measures, could accommodate many other bed closures.
If a PACT project were financed with, for example, $750,000 per year, our region (3 hospitals) would still have $500,000 or possibly $750,000 more to create highly staffed group homes. This would give us 12 to 18 beds for very regressed patients.
In my opinion, if the government pushes the Douglas to go faster in closing beds, it should endow the Douglas with a special budget to do this, as it has done for L-H Lafontaine. These are points for discussions with the leaders concerned. A publicity strategy has a good chance of getting these extra funds.
If PACT were even modestly a success, possibly over the next few years we would see further de-institutionalization with these models. If so, we must insist that the government would use further bed closures to transfer all or most of the savings toward these community programmes.
Otherwise, we will see a "failure of de-institutionalization" which has been described a few years ago by Dr. Gaston Harnois, a main leader of the movement in the last 20 years.
We need a strong advocacy approach, as one sees in the United States. An inter-hospital, inter-agency committee could ensure a long-term presence of these issues in the media. This is how governments are most successfully influenced when it comes to funding. (Many hospitals already to this kind of permanent media campaigning. It is not such a special or new approach.)
It is important to note that a PACT team would not just free up beds. It would also reduce the workload of grinding details and crisis-intervention for social workers and psychiatrists. This will free up hospital resources for integrating the cast-off patients of the Reddy Memorial and Queen Elizabeth hospitals.
In my opinion, it is in the interest of all our three hospitals to promote the creation of BOTH a PACT program AND highly-structured group homes, in the next few months.
RECOMMENDATIONS FOR THE MONTREAL AREA
1. Hospitals and community groups should make a commitment to work together to create new resources. Deciding to work together is more important than accepting any particular model. However, this report may give a useful focus to discussions about what you want.
2. Examine the models in small and large groups. Everyone working in a hospital could see the video "Hospital Without Walls" (9) to get an idea of how difficult patients can stay out of hospital. Invite a speaker from the U.S. to explain in detail how it might work in your region. In Montreal, we have invited Dr. Malcolm Diamond, from the "original" PACT program in Madison, Wisconsin. Dr. Laurie Curtis of Burlington, Vermont, was a stimulating speaker at the Douglas Hospital.
3. Gather consensus and enthusiasm for the solutions you propose. Make up preliminary proposals and discuss them with everyone concerned, including the patients and their families. Hold a large conference. Do the plans fit the needs? Can they bring savings? Ask government representatives for advice and feedback about funding. Who will lead and staff the new resources? At some point, the newspapers and broadcast media will become very interested in a helpful way. This can be easily managed by skilled and experienced people like myself and some others in the system.
The book "Return to Community" by Paul J. Carling (8) is an excellent source of information and inspiration, for people re-thinking community resources for the mentally ill.
I do not ask anyone to endorse these ideas solely on the basis of my report.
I do propose that we embark on a process of getting new community resources, using these two ideas as a focus of discussion.
We seem to have the means, motive, and opportunity -- to get some new valuable resources.
Together, we can achieve that goal.
Rough cost-benefit analysis of PACT model
The PACT model of mental health care delivery aims to reduce hospital utilization of the more difficult "revolving door" patients, by having "community mental health workers" visit them very frequently in their homes.
Though this seems at first glance to be an expensive mode of treatment, let us consider the likely costs and benefits to the psychiatric system.
Costs
A typical PACT team consists of 10 staff with an average caseload of 10 patients, for a total of 100 patients served at a given time.
In Springfield, costs are about 70 per cent for staff and 30 per cent for overhead (car costs, rent of premises and equipment, etc.).
In Montreal, staff now typically cost $30,000 per year, possibly a bit more, including wages and payroll taxes and costs. Thus annual costs would be around $300,000 for staff, let us say $350,000 on the outside. Then overhead would likely be 3/7 of that, or $150,000.
This brings us to a total rough estimate of $500,000 per year to run such an agency. This is an average cost of $5,000 per client. This compares favorably with average government subsidies of about $3,000 per year per patient for an ordinary foster home (famille d'accueil).
In Springfield, a PACT agency serving 98 clients costs about $5,000 to $6,000 U.S. per patient, according to manager Jim Goodwin. This would be like $6,700 to $8,000 in Canadian funds per patient per year. Also, a separate PACT agency for intensive-need patients needing 20 hours or more per week, costs $26,000 (Hartford, Ct.) to $33,000 U.S. per year per patient. However, it was considered that these patients cost up to $180,000 per year in hospital, so the system was satisfied that there was a significant saving.
For this analysis, let us say that the true final cost of an agency for 100 patients was ($8,000 per patient x 100 patients =) $800,000 per year.
Benefits
These revolving-door patients are not living in the hospital year-round, though some of them almost do. In my experience, they tend to spend about three months out of 12,in the hospital. Thus 100 patients under PACT might be expected to save the use of 3/12 or 25% of 100, or 25, acute-care (not chronic) beds per year in the system.
Four such patients would occupy a hospital bed for a total of about a year. Therefore, the PACT costs to eliminate the need for that bed would be (4 x $8,000 =) $32,000 a year.
Sources estimate that it costs $50,000 to $150,000 (hospital budget) per year to run an acute-care psychiatric bed. In the PACT model, we are paying $32,000 per year to eliminate the need for a hospital bed. There may be hidden costs, but clearly costs will be reduced by 35 to 80 per cent of what they were, for this group of 100 patients, for in-patient services alone.
The usual calculations for this kind of analysis only look at the costs to the hospital budget. But in Quebec, there are doctor's fees in hospital, coming from the "invisible" MD's salary budget controlled directly from Quebec City. As an in-patient psychiatrist, I have estimated these doctor's fees as about $15,000 per year per hospital bed. This is based on an average of 30 minutes time spent with the patient per day on an average ward, with the $18.15 "control visit" on top of that.
Thus the real cost per hospital bed must be between $65,000 and $165,000 per year.
With the PACT model, we pay $32,000 to eliminate the need for a bed. This leads to a real saving of between 50 and 81 per cent of the bed's cost.
Consider also that these patients will become much less burdensome for our out-patient services. Their need for social work interventions will drop dramatically. This is because the PACT staffers help patients budget their money, get along with their families, and keep them out of crises that now occupy so much social work time.
Psychiatrist's time spent with this heavy population should also be reduced. In my experience, many hours of my week go to supporting these patients in person and on the phone, and in sorting out the many details of their social crises.
This time saving will also be a cost saving to the government, indeed it would probably be measurable through RAMQ statistics. Also, it would free up psychiatrist's time to do better, more efficient in-patient (and out-patient) care. I believe that this would shorten hospital stays for patients who do come to hospital.
Overall, the PACT model is an attractive cost-saving option, easily saleable to governments in my opinion.
The financial benefits of PACT are numerous. It's an idea which comes along just at the right time.
Rigorous cost-benefit analysis of one real PACT program
In the last few days before giving the final copy of this report, Mr. Jim Goodwin sent me real-life figures about one PACT programme in Hartford, Connecticut. All figures are in U.S. dollars.
This is a medium-sized city only 50 miles from Springfield, where the PACT idea has only been used since 1989, with supervision from Springfield.
Mr. Goodwin chose a "heavy" PACT program which serves patients needing several visits a day, possibly for years. Unlike typical PACT programs costing $5,000 to $10,000, this one costs about $26,000 per patient per year, because of the high staffing costs. The clients include the worst drug-abusing, personality-disordered, mentally ill people in the system, who are not so regressed as to need a highly staffed group home.
An recent study shows outcomes for patients from the start in July 1989 to July 1995.
On average, patients spent 57.6 per cent of their time in hospital before PACT. Now they spend only 9.5 per cent of their time there, thanks to PACT. This is a difference of 48.1 percentage points. Or you could say they spend only one-sixth the time in hospital compared to their former pattern.
A hospital bed costs between $400 and $500 a day there, thus between $146,000 and $182,500 per year. For this analysis, let's assume it is between these two figures, say $165,000 per year.
Thus before PACT, we had 48 patients spending 57.6 per cent of their time in hospital, for a yearly cost of (48 x 57.6% x $165,000 =) $4,562,000 for the mental health system.
With PACT, we have 48 patients costing $26,000 per year each, and also occupying hospital beds for 9.5 per cent of the year. The cost to the system is then: (48 x $26,000) + (48 x 9.5% x $165,000) = $2,000,000.
The net saving for just 48 people is $2,562,000. The PACT system allows for reduction down to 44 per cent of the original cost for care.
Jim Goodwin says it is common experience and knowledge in the PACT network in the U.S. that there are similar cost savings whether the patients are "light", merely too-frequent revolving-door visitors, or "heavy" patients with multiple problems who are unpleasant in hospital and make staff feel hopeless.
At present, this same Hartford PACT program has been downsized for 1995-96. It currently has only 37 clients, and its budget is about $1 million. This continues the tradition of about $27,000 per patient per year, for these more difficult patients with intensive needs.
It is accepted that there is a range of costs from about $5,000 U.S. to about $26,000 or $33,000 per year per patient under PACT, depending simply on the difficulty of the patients served by the program.
For Montreal, it will probably be wiser to start with the easier revolving-door patients while we gain experience in making such a system work. A formal consultant, possibly on loan from the U.S. for many months, might be well worth the investment to help us get a program on its feet. Other Montreal regions could learn from our experience.
It seems likely that adopting these models could lead to a gradual downsizing of our in-patient services, transferring the savings to create expanded community resources of this type.
Clearly, PACT programs generally bring savings to mental health systems, or at least they cost no less than existing programs for the same patients. This finding is borne out by at least ten separate research studies in the literature. PACT is the best-researched community mental health approach since de-institutionalization began.
1. Okin RL, Testing the limits of deinstitutionalization. Psychiatric Services, June 1995, 46:6, 569-574
2. Drake RE and Burns BJ. Special Section on Assertive Community Treatment: an Introduction. Psychiatric Services, July 1995. 46:7, 667-668
3. Deci PA, Santos AB, Hiott DW, Schoenwald S, Dias JK. Dissemination of Assertive Community Treatment Programs. Psychiatric Services, July 1995. 46:7, 676-683
4. Burns, BJ, and Santos, AB. Assertive Community Treatment: an update of randomized trials. Psychiatric Services, July 1995. 46:7, 669-675
5. Santos AB, Henggeler SW, Burns BJ, Arana GW, Meisler N, Research on field-based services: models for reform in the delivery of mental health care to populations with complex clinical problems. American Journal of Psychiatry, (Special Article), Aug. 1995, 152:8, 1111-1123
6. Essock SM, Kontos N, Implementing Assertive Community Treatment Teams. Psychiatric Services, July 1995, 46:7, 679-683
7. Carpenter WT, Conley RR, Buchanan RW, Beier A, Tamminga CA, Patient response and resource management -- another view of clozapine treatment of schizophrenia. American Journal of Psychiatry, June 1995, 152:6, 827-832
Also recommended:
(8) Book: Paul J. Carling, Return to Community: Building support systems for people with psychiatric disabilities. The author is a PhD psychologist who has guided at least two communities into a community-based mental health care system, emphasizing PACT. The book is clearly and elegantly written. It provides helpful hints, even a kind of road map, for stimulating the process towards effective community resources. Available through Copp-Clark Canada (Toronto). Published by the Guilford Press, New York and London, 1995. In Canada, copies can be obtained for $28.95 by phoning 1-800-263-4374.
This book assumes the role of the Bible if one compares the spread of PACT to an evangelical movement.
(9) VIDEO: Hospital Without Walls, a 48-minute documentary illustrating daily life in a PACT team and the people they help, in a South Carolina city. Produced by Duke University. There are a few copies available in the McGill network.