Modelling diabetes and depression in Pakistan: using economic modelling to inform intervention design and a clinical trial of a behavioural activation intervention

Abstract

Objectives: the ‘Developing and evaluating an adapted behavioural activation intervention for depression and diabetes in South Asia (DiaDeM)’ trial investigates a psychological intervention, behavioural activation (BA), on people with both diabetes and depression in Bangladesh and Pakistan. This study aimed to aid the intervention and trial design.

Design: this was a modelling study using microsimulation to assess the intervention's cost-effectiveness. Diabetes was modelled using the UK Prospective Diabetes Study (UKPDS) model based on Pakistani patients and depression was modelled using Patient Health Questionnaire-9 trajectories allowing for multiple depressive episodes. It was assumed that diabetes-related adverse events increased depression recurrence, while depression impacted HbA1c, increasing diabetes-related events.
The model estimated (i) maximum cost of BA which would be cost-effectiveness (headroom analysis) to inform intervention design, and (ii) value of reducing uncertainty around different measures (value of information analysis) to prioritise data collection in the DiaDeM study

Setting: analysis was conducted from a Pakistani healthcare perspective over a lifetime with costs and outcomes discounted at 3%.

Interventions: BA plus usual care was compared against usual care. BA involved six sessions by a trained (non-mental health) facilitator. The usual care comparator was the prevailing mix of pharmacological and non-pharmacological treatments used in Pakistan.

Primary and secondary outcome measures: the primary outcome was disability-adjusted life-years (DALYs). Secondary outcomes included life years, healthcare costs and the rate of depression and diabetes-related events.

Results: over their lifetime, individuals receiving BA plus usual care avoid 3.2 (95% credible interval: 2.7, 3.8) years of mild depression and experience fewer diabetes-related events. BA plus usual care resulted in an additional 0.27 (0.03, 0.52) life years, 0.98 (0.45, 1.86) DALYs averted, and had incremental healthcare costs of -$97 (-$517, $142), excluding BA costs. The maximum cost per BA course at which was cost effective is $83 ($9, $214). Value of information analysis found the most important measures to include in the trial are impact of depression on diabetes and PHQ-9 over time.

Conclusions: this is the first model to jointly model depression and diabetes for South Asia and uses novel methods to reflect the diseases and inform intervention and trial design. This evidence has helped to inform the design of the DiaDeM intervention and the trial to evaluate it.

More information

Identifiers

ORCID for Richard Ian Gregory Holt: orcid.org/0000-0001-8911-6744

Catalogue record

Export record

Altmetrics