As noted in Anatomy of an Epidemic, the rationale for promoting long-term treatment with neuroleptics for people who have been diagnosed with schizophrenia was based on the results of the relapse studies.  In these studies, people who had been treated with neuroleptics during the acute phase of their illness were randomized to continue on either medications or placebo. In most studies, they were then followed for 1-2 years.  The relapse rate in the group who was not on medication was much higher.

However, there is more to this story.  To summarize the findings of a large meta-analysis (Gilbert, P, et al. Arch Gen Psych. 1995;52:173-188):

16% of individuals relapsed despite being on medications.

47% of people who are not taking medications do not relapse.

Therefore, maintenance medications reduced relapse rate in less than 40% of  people.

In an accompanying commentary, Ross Baldessarini suggested that in some people relapse was due to the abrupt discontinuation of medications and more people would have done well if the dose reduction had been gradual (Arch Gen Psych. 1995; 52:189 – 191). If this is true, the percentage of people who are deriving benefit from remaining on medications is probably even lower.  It is difficult, however, to determine who these individuals are.

As I re-consider this problem, I have a clearer recognition of how expectations influence  clinical decision making.  The bias for those who have confidence in the overall efficacy of medications is to believe that when someone does well it is the result of the medication and when someone does not do well it is the result of the underlying illness.  A patient of mine is extremely distressed by her voices which she hears continuously.    Periodically, she gets so annoyed by them that she shouts out loud in response to their taunting.  I reduced the dose of medications to try to minimize side effects.  The next time she had a bad day of shouting at her voices both she and the people who live with her were convinced that it was the dose reduction that had caused this episode.  It was the belief that medications are helpful that led this patient and her support system to conclude that it was the dose reduction that had caused the shouting episode even though she had these episodes at regular intervals before the change was made.

Many years ago, I conducted a small study of people who were admitted to a hospital because they had experienced an increase in psychotic symptoms. The people in the study had all been taking neuroleptics prior to the hospital admission. The common practice was to increase the dose of these medications. Most people improved and would be discharged on a higher dose of medication. The assumption was that they got better because of the medication change.  In this double-blind study, some people received an increased dose of medications and others remained on the same dose they had been taking prior to admission. We found that at 10 days, most people  felt better regardless of whether or not their dose was increased (Steingard S, et al. Journal of Clinical Psychiatry  1994; 55 :90  470-472.). This small study suggested that the common practice of increasing the dose of medication was not necessary. It also demonstrates how one’s bias can lead one to form an incorrect conclusion.

At the same time, however, I would posit that those who do not believe that medications are of benefit have their biases as well.  When they encounter someone who is taking medications and is not doing well, this is assumed to be the result of the medications.  When they meet someone who is on medications and doing well, I suspect they tend to believe they are doing well despite their use of medications. My point is that in any given individual it is hard to be sure what the impact of the medication is and we tend to provide explanations that fit our frame of reference.

Although for many years I have practiced with the knowledge that it is important to minimize exposure to neuroleptic medications,  I still do not know who among my patients will be most successful in tapering medications.  My current plan is to systematically track and record my experiences.  I am hoping that by doing this, I will be less likely to be influenced by the dramatic events that might tend to have a disproportionate impact on my thinking (the person who comes off medications completely and does well or the person who gets much more distressed after having done well for many years).  I hope that I can collect enough data to be able to make a contribution to our understanding of neuroleptic taper.