OBJECTIVES : This analysis sought to examine trends in Canadian oncology health technology assessment (HTA) between 2016 and 2021, over which time the Canadian Agency for Drug and Health Technology (CADTH) updated their guidelines.

METHODS : The pan-Canadian Oncology Drug Review reimbursement review database was systematically searched, including all oncologic pharmaceutical submissions received by CADTH. The tumor type, drug class, and modelling and comparative effectiveness strategies were analyzed. The results were presented as proportions or trends. The Cochran-Armitage test was used for trend analysis of categorical data to assess an association between a binomial and an ordinal (year) variable.

RESULTS : During the extraction period, 109 oncologic pharmaceutical submissions were completed. Nineteen tumor types were represented, of which the four most common submissions were lung (18.4%), leukemia (15.6%), lymphoma (12.8%), and breast (9.2%), accounting for 56% of the overall total. This analysis found that 43.1% of the submissions were for antibody treatments; 59.6% were modelled using partitioned survival models (PSMs), while 40.4% were (semi-) Markov. Moreover, 54.1% used within-trial comparisons, and 52.3% included indirect comparisons, of which 18.4% were network meta-analyses (NMAs), 7.3% matched indirect treatment comparisons (MAICs), and 2.8% simulated treatment comparisons (STCs); 13.8% included naïve comparisons. Trends over time show significant increases in indirect treatment comparisons (p=0.01) driven mostly by NMA among MAIC and STC (p=0.05), while decreases were seen in within-trial comparisons (p=0.02). Modelling strategies show non-significant trends (p=0.06) towards increasing use of Markov models.

CONCLUSIONS : This analysis shows trending topics and strategies between 2016 and 2020 in Canadian HTA. Of note, only a few tumor types accounted for a disproportionate number of the submissions, and PSM was the more prevalent modelling strategy. Lastly, broader implementation of global guidelines is suspected to have influenced the trends toward increased comparative effectiveness matching in comparators.