A global medical device company with an innovative targeted interventional approach was seeking market access and provider reimbursement in western European countries. However, the open-label, single-arm pilot study currently available was insufficient to support reimbursement in some jurisdictions.
To evaluate requests for device reimbursement, payers in Europe increasingly expect high-level evidence of efficacy, that is, a completed confirmatory randomised controlled trial (RCT) proving equal or incremental benefit compared to standard of care, measured in patient-relevant outcomes, for a specific indication. Demonstrated clinical efficacy and safety will allow the device to be included in the service catalogue of the public health insurance (market access) and achieve sufficient cost-covering remuneration for the facilities utilising the device (provider reimbursement). However, in some cases, RCTs may not be feasible or ethical. To overcome this issue, naïve indirect treatment comparisons are commonly used, relying exclusively on aggregate data. However, indirect treatment comparisons suffer from critical limitations, namely, cross-trial differences in patients’ baseline characteristics and differences in outcome definitions, which can bias comparisons. Moreover, leading European HTA agencies (National Institute for Health and Care Excellence [NICE] in England; Haute Autorité de Santé [HAS] in France, Institute for Quality and Effectiveness in Health Care [IQWiG] in Germany) have independently classified naïve indirect treatment comparisons as invalid and inappropriate methodologies. In certain situations, in which direct comparative studies are not feasible or ethical, European HTA agencies will consider indirect comparisons if the appropriate methods are applied.
In the absence of a head-to-head randomised trial, HTANALYSTS proposed to conduct a MAIC of treatments across separate trials for the evaluation of the device for reimbursement in jurisdictions that require MAICs to demonstrate comparative effectiveness.
MAIC is grounded on the idea that the method can adjust for between trial differences in baseline characteristics and thus provide a valid estimate of relative treatment effects when standard indirect comparisons are inappropriate. MAIC combines individual patient data (IPD) from clinical trials for one treatment with published summary data (aggregated data) for comparator treatments, adjusting for cross-trial differences in baseline characteristics, reducing sensitivity to effect measures and resolving differences in outcome definitions. The treatment outcomes can then be compared across the balanced trial populations. The general analytic steps of an unanchored MAIC of survival data are 1) balancing the IPD with the aggregate data to obtain weights; 2) digitisation to reconstruct comparator IPD; and 3) pooling the IPD and the reconstructed comparator IPD to conduct weighted survival analysis.
The analysis leveraged a previous systematic literature review conducted for the company’s regulatory submissions. The previous search was updated and re-screened to identify appropriate trials using a priori inclusion criteria. The IPD trial was assessed to ensure the inclusion/exclusion criteria were equally or more inclusive than those in trials with aggregate data. Trial characteristics, patient inclusion criteria and characteristics, and efficacy outcomes were then extracted from the comparator studies. Weights were obtained to balance the baseline characteristics of the IPD with the aggregate data from the comparator studies. For the survival outcomes, comparator study Kaplan-Meier curves were then digitised to reconstruct comparator IPD. Finally, for the survival outcomes the weighted IPD were pooled with the RIKM from the comparator studies to conduct weighted survival analyses; and for the binary outcomes the weighted IPD were pooled with the aggregate data from the comparator studies to conduct weighted generalised linear regression analyses.
As with most novel treatments, there is generally a paucity of evidence about the relative effectiveness compared with its competitors. However, demonstrating comparative effectiveness is a critical consideration in reimbursement decisions.
In the absence of a head-to-head trial, comparative evidence can be generated through indirect comparisons. However, such comparisons are subject to limitations such as cross-trial differences in patients’ baseline characteristics which affects the reliability of the results. By combining IPD with published aggregate data, the MAIC analysis reduced observed cross-trial differences. The findings of the MAIC study allowed a robust assessment of the comparative effectiveness of the client’s novel treatment approach to standard of care and provided decision makers with timely comparative evidence.
Overall, our MAIC study demonstrated, in the absence of a head-to-head clinical trial, our client’s innovative targeted interventional approach could improve upon current standard of care.