Empirical Assessment of Minimum Clinically Important Differences in Network Meta-Analysis

Accepting Students

The minimal clinically important difference (MCID) is the smallest change in a treatment outcome above which clinicians, patients, and researchers would identify as important and would consider a change in the patient’s management.

To date, approaches to determining MCIDs are based on previous studies or expert knowledge. MCIDs are part of the assessment of certainty of evidence obtained from a meta-analysis comparing two interventions or network meta-analysis (NMA) comparing at least three interventions (CINeMA [Confidence In Network Meta-Analysis] or GRADE [Grading of Recommendations Assessment, Development and Evaluation]).

NMA combines the available evidence from a network of trials in a single model, and hence simultaneously evaluating the efficacy or safety of multiple interventions. Certainty of evidence in NMA is assessed using CINeMA or GRADE, which evaluate within-study bias, indirectness, imprecision , heterogeneity inconsistency, and reporting bias.

In this project we aim to evaluate the interpretation of NMA results according to different MCID thresholds using both CINeMA and GRADE assessments.

We will develop a protocol, and overall will use our database of 658 published NMAs comparing pharmacologic interventions published up to July 2018. We will prioritize NMAs according to the most frequently encountered disease categories, i.e., categories with at least 40 NMAs per category. We will establish the MCID thresholds by conduct a scoping review, contacting targeted clinicians and experts in the underlying fields, and conducting an online modified Delphi. The MCID values will be assessed through CINeMA and GRADE to assess credibility in NMA results.

Accepting Students

Student(s) assigned to this project will have the opportunity to work with a multidisciplinary team. The student will assist the research team in coordinating and conducting the scoping review, contacting clinicians and experts, designing of the Delphi exercise, including support with writing the manuscript.

Lead Faculty

Affiliated Faculty