Predicting prognosis for adults with depression using individual symptom data: a comparison of modelling approaches.

J E J Buckman,E Watkins,D Kessler,G Ambler,N Wiles,C O'Driscoll,T.C Eley,R Saunders,Z D Cohen,G Lewis,A J Peel,C Rayner,S D Hollon,S Pilling,E I Fried,R J Derubeis,S Gilbody,T Kendrick

doi:10.1017/s0033291721001616

Abstract

This study aimed to develop, validate and compare the performance of models predicting post-treatment outcomes for depressed adults based on pre-treatment data. Individual patient data from all six eligible randomised controlled trials were used to develop (k = 3, n = 1722) and test (k = 3, n = 918) nine models. Predictors included depressive and anxiety symptoms, social support, life events and alcohol use. Weighted sum scores were developed using coefficient weights derived from network centrality statistics (models 1-3) and factor loadings from a confirmatory factor analysis (model 4). Unweighted sum score models were tested using elastic net regularised (ENR) and ordinary least squares (OLS) regression (models 5 and 6). Individual items were then included in ENR and OLS (models 7 and 8). All models were compared to one another and to a null model (mean post-baseline Beck Depression Inventory Second Edition (BDI-II) score in the training data: model 9). Primary outcome: BDI-II scores at 3-4 months. Models 1-7 all outperformed the null model and model 8. Model performance was very similar across models 1-6, meaning that differential weights applied to the baseline sum scores had little impact. Any of the modelling techniques (models 1-7) could be used to inform prognostic predictions for depressed adults with differences in the proportions of patients reaching remission based on the predicted severity of depressive symptoms post-treatment. However, the majority of variance in prognosis remained unexplained. It may be necessary to include a broader range of biopsychosocial variables to better adjudicate between competing models, and to derive models with greater clinical utility for treatment-seeking adults with depression.

Highlights

Depression affects ∼320 million people worldwide every year (Thornicroft et al, 2017; Vos et al, 2016)
Prognoses generated by the models developed here could be informative for depressed patients seeking treatment in primary care
There were few differences between the models, with no clear advantage in using individual items over sum scores, or in using network models or factor analytic models to weight individual items, in order to derive prognostic predictions. This may represent a limitation of the available data, or of the modelling approaches

Summary

Introduction

Depression affects ∼320 million people worldwide every year (Thornicroft et al, 2017; Vos et al, 2016). There are a lack of accurate, validated prognostic models for adults in treatment for depression (Cohen & DeRubeis, 2018). This study aimed to develop, validate and compare the performance of models predicting post-treatment outcomes for depressed adults based on pre-treatment data. All models were compared to one another and to a null model (mean post-baseline Beck Depression Inventory Second Edition (BDI-II) score in the training data: model 9). Any of the modelling techniques (models 1–7) could be used to inform prognostic predictions for depressed adults with differences in the proportions of patients reaching remission based on the predicted severity of depressive symptoms post-treatment. It may be necessary to include a broader range of biopsychosocial variables to better adjudicate between competing models, and to derive models with greater clinical utility for treatment-seeking adults with depression

Objectives

Results

Discussion

Conclusion