{"id":60222,"date":"2015-09-30T19:56:45","date_gmt":"2015-09-30T23:56:45","guid":{"rendered":"http:\/\/1boringoldman.com\/?p=60222"},"modified":"2015-10-01T01:45:39","modified_gmt":"2015-10-01T05:45:39","slug":"to-obey-the-scout-law","status":"publish","type":"post","link":"https:\/\/1boringoldman.com\/index.php\/2015\/09\/30\/to-obey-the-scout-law\/","title":{"rendered":"to obey the scout law…"},"content":{"rendered":"
A 2002 article<\/a> by Stanford Biostatistician Helena Chmura Kraemer et al summarizes the "characteristics of a well-performed RCT<\/font><\/em>" [referencing the 1986 classic, Clinical Trials: Design, Conduct, and Analysis<\/a>, by Curtis Meinert]:<\/div>\n
\n
Considerable progress has been made in the development and evaluation of treatments, both pharmacologic and psychological, for a variety of different psychiatric disorders. This research has emphasized the use of the randomized clinical trial [RCT], which is widely regarded as the gold standard of evaluation of efficacy and effectiveness in medicine. The characteristics of a well-performed RCT are well established. They include the following features:<\/div>\n
    \n
  1. \n
    A well-defined and justified population, with a representative sample of sufficient size, to yield power to detect clinically significant differences between treatments and to provide accurate estimates of the effect sizes in that population on which to base considerations of clinical or policy significance.<\/div>\n<\/li>\n
  2. \n
    One or more control or comparison groups, with protocols for treatment in each group specified well enough to permit replication in the clinic or another research project.<\/div>\n<\/li>\n
  3. \n
    Randomization to treatment and control or comparison groups to avoid confusing selection effects with treatment effects.<\/div>\n<\/li>\n
  4. \n
    A few a priori<\/font><\/strong><\/em>, well-chosen, and justified outcome measures, selected in advance of the trial, obtained either blinded to treatment group or otherwise with measurement bias controlled to avoid confusing the opinions or expectations of patients or researchers with treatment effects.<\/div>\n<\/li>\n
  5. \n
    Analysis performed by intention to treat [ie, all randomized subjects are included in the analysis of outcome]. Only those subgroups specified and justified in the a priori<\/font><\/strong><\/em> hypotheses [eg, baseline severity] or in the design [eg, sites in a multisite study] are addressed in the primary analysis<\/font><\/strong><\/em>.<\/div>\n<\/li>\n
  6. \n
    A valid test for statistical significance and estimates of effect sizes<\/font><\/strong><\/em> informative enough to guide consideration of clinical and policy significance.<\/div>\n<\/li>\n<\/ol>\n
    The knowledge derived from such RCTs is of direct relevance to health care system reform and the growing demands for accountability…<\/div>\n<\/blockquote>\n
    There’s nothing particularly new about doing Randomized Clinical Trials [RTCs]. Each point is clear – the rationale intuitively obvious. One has the sense that the analyses of trial efficacy data could be programmed from an a priori<\/font><\/strong><\/em> protocol before commencing the trial. The problems haven’t come from the algorithms used to crunch the numbers. They’ve come from the two legged critters that use them. <\/div>\n

    In psychiatry, we’re primarily focused on the problems from two classes of drugs – the "SSRI Antidepressants<\/font>" and the "Atypical Antipsychotics<\/font>" – both variations on older themes. We thought of their predecessors as dangerous and self limiting, used in dire situations. Their toxicity was on the front page – too many everyday adverse effects to expect people to take them long term even when recommended. One way to frame the modern problems doesn’t have to do with the Clinical Trials that got them on the market but rather with the dramatically broadened indications<\/font><\/em>. With the "SSRI Antidepressants<\/font>," the DSM-III had handed indications<\/font><\/em> to the pharmaceutical companies on a silver platter. Some was the focus of the neoKraepelinians on the biological aspects of mental illness. But the main gift was the creation of the diagnosis – Major Depressive Disorder [MDD] – <\/font><\/em>that could be expanded to fit the majority of patients showing up for treatment. The combination of the SSRI’s greater tolerability and the catch·<\/strong>all MDD<\/font><\/em> diagnosis was a recipe for success. These are not potent drugs nor is MDD a solid diagnosis, so the challenge was getting through the Clinical Trials for approval.<\/p>\n

    With the "Atypical Antipsychotics<\/font>" there were different bridges to cross. At least in terms of felt adverse effects, the incidence of neurological side effects was more in the background than with the older drugs. The dangers were metabolic and insidious – often taking longer to develop than the duration of the usual Phase 3 Clinical Trials. Antipsychotics are potent drugs and Risperdal was no exception. So unlike the Antidepressants, getting a treatment effect in an RCT wasn’t really a problem. But in the words of J&J Sales Manager, Tone Jones, "You can’t make a blockbuster<\/font><\/em> [> $1 B\/YR] out of a 1% disease <\/font><\/em>[incidence  of Schizophrenia]." They weren’t handed much of a market, unlike the Antidepressant makers. They had to create their own. They aimed for a particular set of situations where the drugs were sought by someone other than the patients themselves<\/font><\/em>: mentally impaired children<\/font><\/em>; autistic children<\/font><\/em>; the kids with "Super angry\/grouchy\/cranky irritability" that Joseph Biederman called Bipolar Kids<\/font><\/em>; elderly people in long term care<\/font><\/em>; psychotic and others in government facilities including prisons<\/font><\/em> – something like captive audiences<\/font><\/em>.<\/p>\n

    They put Alex Gorsky in charge of Risperdal sales. Gorsky was a lifelong Boy Scout, still active in the national organization:<\/div>\n