OBJECTIVES: Indirect treatment comparisons (ITCs) have been developed to generate relative treatment effects vs all relevant comparators beyond the ones selected in the pivotal trial. However, in some cases the data available is non-comparative and standard methods used for ITCs become not appropriate. We suggest alternatives to conduct ITCs on non-comparative data based on previous challenges observed.

METHODS: Recommendations from the NICE Technical Support Documents (TSDs) were reviewed, a targeted literature review was conducted and challenges faced during previous ITCs were considered to present cases where non-comparative data limited the use of standard ITCs.

RESULTS: Different situations were encountered when assessing the feasibility of ITCs in the presence of non-comparative data. Non-comparative data included differences in terms of study designs, of populations between trials and presence of single arm trials. Alternatives considered and presented corresponded to the use of population-adjusted ITCs as matching-adjusted indirect comparisons (MAICs), simulated treatment comparisons (STCs), propensity score matching and use of external comparator arm to connect a single-arm trial. In addition to the alternatives, different considerations in terms of identification of factors, heterogeneity assessment, sensitivity analyses and assessment of validity of results can help building a strong comparison.

CONCLUSIONS: This study identifies acceptable and preferred methods to be considered when conducting ITC of non-comparative data.