Major depressive disorder (MDD) is a substantial healthcare concern. Conventionally conducted meta-analyses support the efficacy of both psychological and pharmacological interventions for MDD, but methodological limitations of meta-analyses may obfuscate rather than clarify the clinical efficacy of available interventions. The thesis begins with a systematic review of meta-analyses of high quality psychological treatment studies for MDD. The results of the systematic review indicated that 48% of patients achieved remission after a course of psychological treatment. However, approximately 70% of remitted patients relapsed within 3 years after the discontinuation of psychological therapy. Consistent methodological limitations were identified in the primary outcome studies contributing to the meta-analyses. The primary studies typically published insufficient evidence on treatment fidelity. There was considerable variability in the overall treatment duration, the mean severity of samples and the definition of clinical significance. These factors pose a risk to the validity of meta-analytic results of psychological interventions for MDD. The next component of the thesis investigated the impact of idiosyncratic clinical significance definitions on the published conclusions of studies that used the Beck Depression Inventory (BDI) or Hamilton Rating Scale for Depression (HRSD) to assess outcome. The availability of individual patient data (IPD) for 7 published studies enabled the empirically-based Jacobson Method of clinical significance to be used as a standard definition of recovery across IPD studies. Comparisons of published and Jacobson method clinical significance rates for each IPD study showed that idiosyncratic outcome definitions typically overestimated treatment efficacy. Moreover, treatment efficacy was confounded with the definition of clinical significance employed. This indicates that to reduce the risk of bias in meta-analysis, a standard and empirically-based definition of clinical significance should be used across primary MDD treatment studies. Subsequently, the moderating role of pre-treatment severity on clinical significance rates was investigated via individual patient data meta-analysis. It was found that being male and having higher pre-treatment severity both predicted a significantly reduced likelihood of achieving clinical recovery. It is evident that between-study methodological differences means that even high quality conventional meta-analyses of psychological treatments for MDD remain at risk of bias. The novel finding that gender significantly moderated treatment outcome indicates that IPD meta-analyses are both more powerful and flexible than conventional meta-analyses based on summary data. Ideally, future meta-analyses of primary MDD treatment studies should be based on individual patient data.
Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:617399 |
Date | January 2013 |
Creators | Connor, Martin |
Contributors | Fisher, Peter; Dickson, Rumona; Kinderman, Peter |
Publisher | University of Liverpool |
Source Sets | Ethos UK |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Source | http://livrepository.liverpool.ac.uk/12733/ |
Page generated in 0.0014 seconds