Establishing efficacy and effectiveness in the treatment of personality disorders

Abstract

Establishing efficacy and effectiveness in the treatment of personality disorders 'The consequence was, that the most sudden and visible good effects were perceived from the use of oranges and lemons', wrote James Lind in his description of the first published randomised controlled trial, one showing the prevention of scurvy (Lind, 1753) carried out in 1747 with sailors in warships patrolling the English Channel to prevent Spanish marauders. Yet it took 100 years for this evidence to be absorbed into practice. But when one contemplates the size of the difference between scurvy prevention and treatment efficacy in personality disorder, a new period of 100 years for personality disorder treatment to be similarly summarised does not seem excessive. Scurvy, caused by a simple deficiency of vitamin C, is now perceived as a simple medical disorder; personality disorder is one of the most complex.But the James Lind legacy has persisted in the universal embrace of the randomised controlled trial across the whole of therapeutics. While there are many preliminary enquiries needed before a definitive answer to a therapeutic question can be answered, the gold standard, sometimes looking a little tattered, of the randomised trial still holds sway, even for the most complex disorders. But Austin Bradford Hill, one of the pioneers of controlled trials, frequently emphasised the need for researchers to have an array of experimental data in advance to allow a precisely framed question to be tested in the context of the definitive trial. The tendency in many complex interventions is to try and test several hypotheses simultaneously, but 'to start off without thought and with all and sundry included, with the hope that the results can somehow be sorted out statistically in the end, is to court disaster' (Hill, 1966).This important observation needs to be remembered by all researchers in the psychological sciences. Because almost all the interventions we test are complex interventions, we need to have robust models, grounded in theory (Muthukrishna & Henrich, 2019) and derived from qualitative and other methodologies (Campbell et al., 2000) that enable one very precise hypothesis to be tested.