Association of Research and Development Investments With Treatment Costs for New Drugs Approved From 2009 to 2018

Abstract

Drug companies frequently claim that high prices are needed to recoup spending on research and development. If high research and development costs justified high drug prices, then an association between these 2 measures would be expected. To examine the association between treatment costs and research and development investments for new therapeutic agents approved by the US Food and Drug Administration (FDA) from 2009 to 2018. This cross-sectional study analyzed 60 drugs approved by the FDA between January 1, 2009, and December 31, 2018, for which data on research and development investments and list or net prices were available. Data sources included the FDA and SSR Health databases. The primary independent variable was estimated research and development investment. The outcome was standardized treatment costs (ie, annual treatment costs for both chronic and cycle drugs, and treatment costs for the maximum length of treatment recommended for acute drugs). Standardized treatment costs were estimated separately using list and net prices obtained from SSR Health at the time of launch and in 2021. To test the association between research and development investments and treatment costs, correlation coefficients were estimated and linear regression models were fitted that controlled for other factors that were associated with treatment costs, such as orphan status. Two models were used: a fully adjusted model that was adjusted for all variables in the data set associated with treatment costs and a parsimonious model in which highly correlated variables were excluded. No correlation was observed between estimated research and development investments and log-adjusted treatment costs based on list prices at launch (R = -0.02 and R2 = 0.0005; P = .87) or net prices 1 year after launch (R = 0.08 and R2 = 0.007; P = .73). This result held when 2021 prices were used to estimate treatment costs. The linear regression models showed no association between estimated research and development investments and log-adjusted treatment costs at launch (β = 0.002 [95% CI, -0.02 to 0.02; P = .84] in the fully adjusted model; β = 0.01 [95% CI, -0.01 to 0.03; P = .46] in the parsimonious model) or from 2021 (β = -0.01 [95% CI, -0.03 to 0.01; P = .30] in the fully adjusted model; β = -0.004 [95% CI, -0.02 to 0.02; P = .66] in the parsimonious model). Results of this study indicated that research and development investments did not explain the variation in list prices for the 60 drugs in this sample. Drug companies should make further data available to support their claims that high drug prices are needed to recover research and development investments, if they are to continue to use this argument to justify high prices.