CollAborative care and active surveillance for Screen-Positive EldeRs with subthreshold depression (CASPER): a multicentred randomised controlled trial of clinical effectiveness and cost-effectiveness

Efforts to reduce the burden of illness and personal suffering associated with depression in older adults have focused on those with more severe depressive syndromes. Less attention has been paid to those with mild disorders/subthreshold depression, but these patients also suffer significant impairments in their quality of life and level of functioning. There is currently no clear evidence-based guidance regarding treatment for this patient group. To establish the clinical effectiveness and cost-effectiveness of a low-intensity intervention of collaborative care for primary care older adults who screened positive for subthreshold depression. A pragmatic, multicentred, two-arm, parallel, individually randomised controlled trial with a qualitative study embedded within the pilot. Randomisation occurred after informed consent and baseline measures were collected. Thirty-two general practitioner (GP) practices in the north of England. A total of 705 participants aged ≥ 75 years during the pilot phase and ≥ 65 years during the main trial with subthreshold depression. Participants in the intervention group received a low-intensity intervention of collaborative care, which included behavioural activation delivered by a case manager for an average of six sessions over 7-8 weeks, alongside usual GP care. Control-arm participants received only usual GP care. The primary outcome measure was a self-reported measure of depression severity, the Patient Health Questionnaire-9 items PHQ-9 score at 4 months post randomisation. Secondary outcome measures included the European Quality of Life-5 Dimensions, Short Form questionnaire-12 items, Patient Health Questionnaire-15 items, Generalised Anxiety Disorder seven-item scale, Connor-Davidson Resilience Scale two-item version, a medication questionnaire and objective data. Participants were followed up for 12 months. In total, 705 participants were randomised (collaborative care = 344, usual care = 361), with 586 participants (83%; collaborative care 76%, usual care 90%) followed up at 4 months and 519 participants (74%; collaborative care 68%, usual care 79%) followed up at 12 months. Attrition was markedly greater in the collaborative care arm. Model estimates at the primary end point of 4 months revealed a statistically significant effect in favour of collaborative care compared with usual care [mean difference 1.31 score points, 95% confidence interval (CI) 0.67 to 1.95 score points;