The five year outcome of major depression: Effects of baseline variables and type of treatment.

BACKGROUND: Depression is commonly a relapsing or chronic disorder. Long-term outcome is therefore important. We report on the outcome of major depression five years after receiving treatment with medication or psychotherapy. METHODS: 472 patients were treated in three consecutive randomised controlled trials in one clinical research centre. 298 were followed up at five years. Of these, 106 patients were treated with medications, while the remaining 192 were given psychotherapy. The a priori outcome measure was mood symptoms in the two years prior to the assessment. RESULTS: The majority (56%) of patients had no depressive symptoms in the prior two years. One third (32%) had fluctuating depression, while 12% were chronic depressed. Predictors of outcome were few; baseline severity, suicidality, personality pathology, and type of treatment. Those receiving medication did somewhat worse, even when adjusted for this group's higher depression severity, suicidality and personality pathology at baseline. CONCLUSIONS: Long-term depressive symptoms are common after evidence-based treatment, although over half the patients appear to recover. Psychotherapy may be superior to medication in reducing the level of symptoms in the longer term. Personality remains one of the few baseline predictors of long-term outcome.

A behaviourally anchored rating scale for evaluating the use of the WHO surgical safety checklist: development and initial evaluation of the WHOBARS.

BACKGROUND: Realising the full potential of the WHO Surgical Safety Checklist (SSC) to reduce perioperative harm requires the constructive engagement of all operating room (OR) team members during its administration. To facilitate research on SSC implementation, a valid and reliable instrument is needed for measuring OR team behaviours during its administration. We developed a behaviourally anchored rating scale (BARS) for this purpose. METHODS: We used a modified Delphi process, involving 16 subject matter experts, to compile a BARS with behavioural domains applicable to all three phases of the SSC. We evaluated the instrument in 80 adult OR cases and 30 simulated cases using two medical student raters and seven expert raters, respectively. Intraclass correlation coefficients were calculated to assess inter-rater reliability. Internal consistency and instrument discrimination were explored. Sample size estimates for potential study designs using the instrument were calculated. RESULTS: The Delphi process resulted in a BARS instrument (the WHOBARS) with five behavioural domains. Intraclass correlation coefficients calculated from the OR cases exceeded 0.80 for 80% of the instrument's domains across the SSC phases. The WHOBARS showed high internal consistency across the three phases of the SSC and ability to discriminate among surgical cases in both clinical and simulated settings. Fewer than 20 cases per group would be required to show a difference of 1 point between groups in studies of the SSC, where α=0.05 and ß=0.8. CONCLUSION: We have developed a generic instrument for comprehensively rating the administration of the SSC and informing initiatives to realise its full potential. We have provided data supporting its capacity for discrimination, internal consistency and inter-rater reliability. Further psychometric evaluation is warranted.