Many of you have asked me about my departure from CHEO. Many of my patients and their families have asked as well. There is a long and complicated story behind this, which I will not share at this time.

However, I would like to discuss an important factor influencing my departure from CHEO: The economic arguments that have contributed to the change in the mandate of the clinical service to which I belonged.

I was the psychiatrist on the Dual Diagnosis service. 'Dual Diagnosis' is a term referring to the subspecialty of psychiatry focussed on the mental health needs of individuals with developmental disabilities. Individuals with intellectual disabilities and/or autistic spectrum disorders are at a very high risk for mental illness and severe behavioural disorders. In child and adolescent psychiatry, a proportion of dually diagnosed individuals constitute what could be argued to be the most challenging clinical presentations in the field with chronic and complex needs that require highly specific expertise. The service I belonged to was able to provide interdisciplinary care in a manner that was effective, but required involvement in the treatment of each patient for many years. A few months ago, the mandate of the service was changed so that ongoing treatment was no longer available beyond 3 or 4 months, and the bulk of what the service offered would change to single-visit consultations. Consequently, the patient, along with the assessment and recommendations of the consultant would be sent back to the referring physician.

So the real question that should be asked is: Why does the system believe that consultations (with recommendations to primary care physicians) or, at the most, very short-term interventions are the best way to go?. The classic argument usually made is: "This is the only approach we can afford".

I beg to differ.

I will try to share with you my understanding of how these decisions are made and why these conclusions are reached. This understanding, I must caution you, is both biased by my specific experiences with the system and limited by the fact that I was never in a bureaucratic position that would have allowed me to develop a more comprehensive understanding of all the factors involved. However, I have knowledge of at least some of the more important factors involved and will try to convey them to you from the perspective I had over the past few years.

There is basically a short list of statistics that influence the decision-making involved in funding mental health services here in Ontario, and I suspect these would be similar to other provinces in Canada.

Wait-lists
Among the top items on that list are wait-lists. The longer someone has to wait to access a service, the lower the likelihood that the elected public officials involved would be re-elected. It does not really matter what shortening the wait-list would lead to, as long as it is shortened. While wait-lists are generally horrible things to have, they unfortunately are more potent in forcing a more efficient use of services than anything else out there. If you have a condition that truly requires a specialist, you would be more likely to remain on the wait-list than someone who might benefit from less-specialized services, and if you were referring a patient, you would only refer to someone with a long wait-list if you felt that was the only option.

To be clear, we should not have long wait-lists, ever. But I just wanted to point out that they result in a selection process that increases efficiency of use of services. Absent another mechanism that would enable more appropriate and efficient use of services, such as very strict criteria for referral and increasing the transparency of the processes involved in determining suitability for referrals and how cases are prioritized, what happens is that specialists receive referrals that generalists could manage, and the great number of such referrals makes it impossible for the specialists to dedicate enough of their clinical time to doing what they have been specifically trained to do.

Let me give you an example: One day I referred a patient who had chronic constipation to a gastroenterologist. I had my reasons for doing so, and didn't think the family physician was the appropriate person to manage this fairly common condition because of specific clinical features this patient had. Regardless, I got a quick, nearly automated response from the gastroenterology department that said something like: "We do not accept such referrals, but you may want to consider referring to the one of the following physicians..." and listed a number of community-based (i.e. private practice) pediatricians with experience in treating chronic constipation. That was efficient! That would reduce the likelihood of clogging the wait-list for individuals that can only be treated by a gastroenterologist, such as patients with Crohn's disease, liver cirrhosis or bleeding gastric ulcers. Would it have been better for my patient to see a gastroenterologist? In an ideal world, yes. But when there are limited resources, the answer is no. When services are run with such clarity and provide high quality specialized care that is efficient in the manner that really matters (to be explained below), the only real method of minimizing wait-lists is to increase the number of service providers. So if the gastroenterology wait-list was long, we need to hire more gastroenterologists.

In order for the gastroenterologist to see every case of chronic constipation, the respirologist to see every case of chest infection, or the neurologist to see every case of headache, it would mean they couldn't spend the time necessary to treat the more complex cases they were trained to treat. It would also deprive the generalist from the practice needed in treating common conditions they are best suited to treat. If this were to be the norm, the gastroenterologist would not develop the clinical experience needed for treating the most complex cases, simply because s/he would be inundated with single-visit consultations around common conditions a generalist ought to be able to manage. Two things happen here: The patient with the chronic and complex condition (let's say treatment-resistant ulcerative colitis, intractable epilepsy, or a progressive form of pulmonary fibrosis) does not get the ongoing specialized care they desperately need, and the specialist does not develop or maintain the expertise needed to treat such a patient. Everyone looses as a result.

This is precisely what would be achieved when you change a service, which was providing specialized care for individuals with severe and chronic conditions, to a consultative model, which will perform assessments for almost all referrals, and provide little or no specialized treatment. It's an effective solution for wait-lists, but not a good solution for patients with needs for specialized care and not a good use for the expertise of a specialist. And that was precisely the change that took place for the service I used to belong to at CHEO.

Patient Flow
Another item on the list is patient flow, a term variably used to describe how many patients are served. One of the ways this is most often calculated is to count the number of "new referrals" seen. "New" here does not necessarily mean what you might think it does. It means anyone who was not a patient in that service just before the referral was received. That patient may have been 'discharged' from the service the day before the referral was received, it still counts as a new referral! So when we provide a consultation on the same patient we were consulted about a week ago, we have had 2 "new" consultations!

When many of the physician's patients have chronic and complex conditions that require ongoing specialized care, it is difficult for the physician to see many new cases, because his/her time is consumed with the care of existing patients. The reality of such specialists is that there is a certain limit to the number of new referrals they can see because of the fact that each new patient will occupy some of the doctor's time for a number of years. In other words, providing ongoing care, even when it is absolutely essential, will reduce patient flow.

The result of short-term interventions and consultative approaches to complex and chronic conditions is that many patients receive inadequate treatment and are subsequently referred for treatment again. This often results in a continuous cycle we sometimes refer to a 'revolving door syndrome'. There is no evidence of lasting improvement in the quality of life of the patient in many, if not most such situations, but there is greater patient flow!

As a matter of fact, I believe patient flow progressively increases with such a system, because the needs of the community are not actually met in a manner that makes a lasting difference, which results in its own vicious cycles that raise the burden of mental illness in the whole community. Inadequate or partial treatment for acute conditions, or short interventions for chronic conditions means that the intensity of the problem is reduced somewhat and often only temporarily, but the extent of the dysfunction the problem produces for the patient, the patient's family, and the patient's community is not significantly or permanently reduced, and the ripple-effects may include many more individuals than just the patient and the patient's immediate circle of caregivers. So the apparent efficiency of the service provides the impression that more people are being served. But since no one is actually examining the realities I just described, the progressive increase in the demand for services by the community appears to be happening in spite of 'improved efficiency'.

Another method of calculating patient flow is to count the number of visits received per full-time-equivalent (FTE). FTE refers to the proportion of 'full-time' that an employee works. If I work full-time, I am at 1.0 FTE. If I work 3 days a week, that means that I'm at 0.6 FTE. The sum of FTEs of all the employees in a service is an indication of the manpower needed to run that service. You take that number and divide the number of visits by it to get a superficial idea about numerical efficiency of that service. In other words, services that treat a higher number of patients per FTE would appear to be more efficient numerically. Obviously, this has nothing to do with the actual ability of the service to meet the needs of the population it serves. If a patient is better served by having two members of the team see him/her simultaneously during a visit, numerical efficiency is reduced because there is less patient flow per employee. If the care of a patient requires longer visits, and the average length of visits for a service is longer as a result, that would lead to a reduction in the total number of visits the service had and, consequently, the numerical efficiency of the service, despite improving the quality of care the patient gets.

So for services such the one I used to be part of at CHEO, where long visits involving multiple members of the interdisciplinary team were needed, and few new patients are seen because we had no 'revolving door syndrome' and were providing care over a long period of time for patients in need for ongoing, highly specialized care, good clinical outcomes don't really matter. What matters is that it appearsthat the service is inefficient.

As a result of the requirement of better patient flow, these parameters dictate how services are provided. It is not the quality of care that patients receive or the clinical efficiencyof the service in managing the needs of patients that determines how services are provided.

Another result is that service providers have data that they can use to support demanding more funding, more professionals, and more services. After all, they are 'efficient' according to the standards required of them, yet are inundated by the progressively increasing needs of the community. But no amount of additional funding, staffing or services would have a lasting impact because the cycle continues. It is a setup that ensures there will alwaysbe a shortage of services. With a perpetual shortage, the administrators of these services are in a strong negotiating position with the government. Yet, the government is happy with their apparentfiscal and operational efficiency. Politicians are also happy because the wait-lists are short and the numbers of people served (see definition of 'new patients' above) are greater, so they can boast that they have made things better. In other words, it is an ideal setup for service providers and for the government, and terrible for the people that these services are meant to help.

Cost per visit
Another item on the list is cost per visit, which essentially refers to how expensive each visit by the patient is. A number of factors affect this, some of which we have already discussed above, such as the length of the visit and the number of professionals involved in each visit. In some cases, although this is not consistently done, the cost of the visit may also be calculated based on the compensation rate for the professionals involved. A physician is paid more per unit of time than a nurse, a psychologist more than an occupational therapist, and so on. Probably more important in the calculations of the cost per visit, as performed by the bureaucrats that look at these statistics, are factors related to overhead. When certain types of equipment are used or specific space requirements are met, the cost per visit is affected.

What is absolutely astonishing to me is that the 'cost per visit' is a relatively meaningless concept clinically, yet it influences decisions related to funding. So let us suppose that I spent 90 minutes meeting a patient and her/his family, along with an occupational therapist (or 'OT'). If I spent 30 minutes with the family, instead of 90 minutes, and asked them to come 2 additional times, and the OT did the same, the family would have to visit the service 6 times to receive similar service. Doing things this way adds incredible inconvenience to the family, which might include lost work and school hours as well as travel & parking costs, and ignores the fact that spreading of the intervention over a protracted period of time could also limit its effectiveness. However, the bureaucrats will view these 6 visits as having much lower cost per visit, and therefore more desirable. It also appears, on paper, to indicate higher patient flow, because there are more visits over all.

On the service I used to belong to, we needed to have more than one professional at each visit for numerous reasons, including safety concerns (some patients are violent), case-management requirements, as well as the need for utilization of the expertise of the specific disciplines these professionals represent. We also focussed on reduced 'burden of visit' on the patient on family, and tried to address the complexities of their clinical presentations adequately, which necessitated longer visits. We often needed specific space and equipment to do our work. So, the cost per visit was very high. It did not matter that we had statistical evidence of the very high effectiveness of our methods in reducing the burden of illness on the patient/family in a manner that could not have been achieved absent these specialized interventions.


How are these statistics used?

Apparently, these statistics explained above, as well as related others (such as average length of hospital stays, average emergency room waiting time, etc), are collected by the ministries involved and used to compare services across regions or even the entire province. Hospitals are required to meet certain 'benchmarks', and services must try to compete with others in the region in order to maintain the funding they receive. So there is a race to reduce wait-lists and cost per visit and to increase patient flow.

Apparently, the government does not try to evaluate the impact of these policies on the communities affected. There is no attempt, to my knowledge, to see if such policies reduce the demand for services in the community or improve the quality of the care provided.

What everyone knows by now is that the demand for services has been increasing progressively. There is, as far as I know, no attempt to examine whether these policies contributed to this situation.


Side effects for the dually diagnosed

We often talk about the side effects of medications, but I'm referring to the side effects of such myopic economic perspectives on health care.

There is a phrase from an old movie that I will quote: "If you build it, they will come". The population served by the Dual Diagnosis service prior to the change in the mandate was among the most disadvantaged in the community. While many of them were "super-users" of healthcare, social and educational services, many had long disappeared in residential placements or in under-privileged conditions and have given up hope for clinical interventions that could improve their quality of life. When our service was proving to be consistently effective with such challenging clinical presentations, referrals for even more complex cases began to progressively flood us. These were cases that people, including pediatricians and family doctors, did not think would benefit from a service such as ours. These patients "came out of the shadows" when a service that could help was made available. Now that the service has changed, one wonders what will happen to these patients. If all that will be offered is a consultation, why should they bother coming? As a matter of fact, these highly challenging patients have, on average, received 4.2 consultations by the time I got involved in their care. 

When we look at the needs of the dually diagnosed, we could very easily be overwhelmed. The numbers are staggering. Around 40% of intellectually disabled individuals suffer mental illness or major behavioural issues. However, looking more closely, the majority can be managed by social and educational services, and by their primary care physicians. A significant proportion requires the care of specialists, such as developmental pediatricians, neurologists and general child psychiatrists, and that is sufficient in most cases. However, a few are so complex and difficult to treat that a specialist in dual diagnosis is needed. Dual Diagnosis is an area of expertise that attracts fewer than one in seventy psychiatrists, and requires the acquisition of expertise through actual treatment over long periods of time of severe and complex cases. So it would stand to reason to only involve a dual diagnosis specialist when the patient cannot be managed by other services.

Asking dual diagnosis psychiatrists to perform single-visit consultations would result in a return to the shadows of many patients, and a net loss of existing service. The reason why it is a loss is that we are offering them what they already get: consultations, and taking away what they cannot find elsewhere: treatment. It is true that more patients with dual diagnosis will see a psychiatrist, but it is unclear whether this would translate into significant clinical change for the most needy among them, given the chronic and complex nature of their illness. As for the ones that do not have such intense needs, it is unclear whether the new service model provides anything different than a consultation with a general child psychiatrist who is not a dual diagnosis specialist.

"Falling through the cracks of the system" is a phrase I hear caregivers saying all the time. That is the story of these patients. To deliberately create another crack in the system for the population that is the most vulnerable in our society is beyond unethical.

So how else could we improve the efficiency of our services?

As far as I am concerned, I am convinced that there may be meaningful ways to improve the efficiency of services without minimizing or eliminating their usefulness. However, most of such improvements would have to be in the actual content of the interventions provided by the clinical service. While some administrative and time-management approaches can be helpful, the biggest factor in the efficiency of a service is whether what the service provides to the community is actually effective and helpful.

If that approach is used, and we still have long wait-lists, then the solution is in hiring more people to perform the service. When there is no will to force the government to pay more people to do the job that few cannot do, there will always be long wait-lists. The idea that fixing fundamental funding problems by trying to save pennies here and dimes there only leads to gradually and progressively bigger and more costly problems.

Now there are situations where there is a dearth of professionals with the necessary expertise. Dual Diagnosis is such an area of expertise. So what should we do in these situations? Well, one thing we should not do, in my opinion, is to throw away the expertise of the rare specialist because we need to reduce the length of wait-lists. That will result in fewer physicians being attracted to that field, and perpetuate the problem, while not substantially meeting the needs of the community.

We tend to forget that we are talking about real people, with real problems that are having real effects on their lives. Humans, not accounting books. And there is another fundamental truth that is always overlooked even when thinking about the economics of healthcare: It is not how costly an intervention is that matters. What matters is the overall cost to society of the problem we are trying to solve. Even if better parameters are used to examine the fiscal efficiency of clinical services, we cannot simply look at cost of service provision in isolation from clinical effectiveness. We cannot measure clinical effectiveness, or burden of illness, or overall cost to society, simply from the perspective of one government agency or ministry. What affects the mental health of a citizen, affects education, work productivity, utility of social services, involvement of charitable organizations, and forensic systems, in addition to healthcare. The effects are not limited to the individual involved, and the ripple effects in the community and society at large have to be taken into account.

It is very painful to see how the illness of a child could break up a family, traumatizing the siblings and impairing the mental health of the parents, increasing the lost days of work by both parents and reducing days spent at school by the patient and the siblings, increasing the debt of the family, utilizing emergency rooms, inpatient beds on psychiatric and non-psychiatric wards, involving the police, and resulting in physical disabilities for some peers of the patient and some hospital workers with all the ramifications of such disabilities. Does this seem too dramatic? I have seen many patients where the above description applies and in some cases only scratches the surface. A service that can reduce many of the different facets of dysfunction and burdens to the society of such devastating conditions must never be viewed as 'expensive'.

@yasseraddabbagh
14/06/2011
Back to Top