Tag Archives: Reliability

Robert Spitzer

Robert Spitzer (1932 to 2015) died last Friday (25th December).

Plausibly, Robert Spitzer could be described as the most influential figure in psychiatry of the last 50 years. He made contributions in many different areas. Most media reports focus significantly on his contribution in removing homosexuality as a diagnosis, something which certainly deserves praise. His biggest contribution, which I will discuss, is his influence on the DSM.

Robert Spitzer was the chairman of the task force for the DSM III and DSM III R. This process has been discussed by historians, some of whom go into specific details the about exact steps Spitzer took (Kirk and Kutchin’s The Selling of the DSM is a good overview). Rather than discuss that, I wish voice my thoughts about whether DSM III was a positive step or not (it is important to note that Spitzer was partially reacting to and redirecting existing desires for change, he was not fully responsible for all the changes).

DSM III (compared to earlier editions) brought much more standardisation to classification of mental illness. The diagnostic criteria in DSM III were much more detailed than those of earlier DSMs, the description of each mental illness being assigned many more pages. This was partly intended to improve reliability, to increase the degree psychiatrists agree about which mental illness the patient exhibits.

DSM III also went someway to reduce the impact of psychoanalytical approaches. DSM III intended to be a-theoretical (though it failed in this), but it certainly placed the emphasis on describing symptoms and how they cluster rather than basing classification on causes (be them psychoanalytical, cognitive psychological or biological). This was partly intended to then allow studies which might improve validity, finding the actual underlying causal basis (rather than speculative psychoanalytical causes).

These two DSM approaches, broadly present also in later editions, resulted in major problems. Firstly, it contributed to removing the subjectivity of the patient from psychiatry, standardised criteria not capturing the aspects of patient which were specific to himself or herself (including symptoms not mentioned by the diagnostic criteria). Secondly, by making the diagnostic criteria more specific, it then meant alternative formulations of psychiatric classifications, potentially superior ones, stopped being employed for diagnosis or for studies. Thirdly, very few DSM psychiatric diagnosis have had their causal basis established. Fourthly, it resulted in high levels of co-morbidity with associated difficulties of establishing which symptoms should be part of which diagnosis.

Despite these problems, I believe Spitzer’s influence was a step in the right direction, but his influence went too far. Standardisation is generally a good thing for creating scientific reliable generalisations, which can then be modeled to specific contexts in a more reliable way. So in this sense I think the DSM III had to happen. I think it went wrong by blocking out alternative psychiatric classifications. We need modify the psychiatric classifications we employ and try new approaches, modeling symptoms and how they cluster in many different ways. This would be ideally done using multiple classifications system at the same time, though each variation needs standardisation. So Spitizer’s contribution of standardisation was a positive one, but the psychiatric classifications should not have been set in stone. Instead, we needed to develop further sophistications in our approach to classification, rather than stopping with the important step of standardisation. If Spitzer had to happen then we now need to move beyond him, take his contribution and use it in new ways, further innovating how we approach classification.