Incorporating single-arm evidence into a network meta-analysis using aggregate level matching: Assessing the impact

Area of Science:

• Health Technology Assessment
• Pharmacoeconomics
• Biostatistics

Background:

• Single-arm studies are increasingly submitted for health technology assessment (HTA) for drug reimbursement.
• Randomized controlled trials (RCTs), the gold standard, are often unavailable for newly licensed drugs.
• Alternative strategies are needed to formally assess single-arm evidence within HTA.

Purpose of the Study:

• To examine and propose formal approaches for incorporating single-arm evidence into the HTA evaluation process.
• To investigate methods for matching single-arm data with comparator arms or trials for network meta-analysis.
• To address the challenge of evaluating new drugs with limited evidence.

Main Methods:

• Matching aggregate-level covariates from comparator arms/trials to single-arm studies.
• Incorporating matched evidence into a network meta-analysis framework.
• Two matching approaches: direct inclusion of matched arm and using matched trial's baseline odds as a plug-in estimator.

Main Results:

• The synthesis of evidence is sensitive to between-study variability and prior formulation.
• The weight assigned to single-arm evidence and the extent of bias significantly impact results.
• A flowchart for synthesis and recommendations for sensitivity analyses are provided.

Conclusions:

• Formal methods can incorporate single-arm evidence into HTA, but results are sensitive to various factors.
• Careful consideration of bias, variability, and weighting is essential for reliable assessments.
• The proposed methods were illustrated using a hepatitis C dataset.