Prognostic models for predicting relapse of depression: A Cochrane Prognosis Review and implications for primary care

Conference:

SAPC ASM 2020 - cancelled due to pandemic

Talk Code:

P.8

Presenter:

Andrew Moriarty

Twitter:

@A_S_Moriarty

Co-authors:

Nick Meader, Simon Gilbody, Rachel Churchill, Bob Phillips, Shehzad Ali, Richard D Riley, Dean McMillan, Carolyn A Chew-Graham

Author institutions:

Hull York Medical School, University of York, Keele University, Western University

Problem

Most people with depression are managed in primary care. After a first episode, approximately half of patients will relapse, and this risk increases for every subsequent episode. Depression severity and treatment-resistance increase with each successive episode, highlighting the potential benefits of intervening early to prevent relapse. Factors associated with increased risk of relapse include adverse childhood events, previous episodes of depression and residual symptoms. Combining several prognostic factors within a multivariable prognostic model can result in improved individualised risk predictions. Our goal is to develop a prognostic model, to be implemented in primary care, to identify patients at increased risk of relapse and allow more effective allocation of relapse prevention interventions to those individuals.This is the first systematic review to set out to identify prognostic models for relapse of depression, a recommended first step in prognostic model development. The two main aims of this review are:To summarise the predictive performance of prognostic models developed to predict the risk of relapse and related outcomes in patients who meet criteria for remission. To summarise the value of updating an existing prognostic model or identify whether the development of a novel prognostic model is required.  

Approach

The methodology is informed by most recent guidance in prognosis research. We searched a wide range of electronic medical databases and used the following eligibility criteria:Population — Adult patients (18 years and over) diagnosed with depression and meeting criteria for remission.  Index model —Prognostic models predicting relapse and related outcomes in patients with depression.  Comparator — NoneOutcomes — Relapse and related outcomes (recurrence, sustained remission or recovery) in depression.  Timing — Start-point is the point at which a patient has responded to treatment and is identified as meeting criteria for remission. Setting — Primary, secondary or community care.Data were extracted using the Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS) and risk of bias assessed using the Prediction model risk of bias assessment tool (PROBAST).

Findings

The review is ongoing (full text screening stage) and SAPC ASM 2020 will be an early opportunity to share the results.

Consequences

The results will inform future work to improve risk stratification in primary care. If an existing model performs satisfactorily, we will update and refine this, with input from key stakeholders, including patients and the public, for implementation in a UK primary care setting. If no existing model has sufficient predictive performance or clinical acceptability, we will use information from this review to develop a novel prognostic model. The longer-term goal of this study is to improve clinical outcomes and quality of life for patients, as well as facilitating more targeted use of NHS resources in primary care.

Submitted by:

Andrew Moriarty

Funding acknowledgement:

This report is independent research supported by the National Institute for Health Research (NIHR Doctoral Research Fellowship, Dr Andrew Moriarty, DRF-2018-11-ST2-044). The views expressed in this publication are those of the authors and not necessarily those of the NHS, the National Institute for Health Research or the Department of Health and Social Care.