Forecasting drug utilization and expenditure: ten years of experience in Stockholm

Abstract

Background: Operating under constrained budgets, payers and providers globally face challenges in enabling appropriate and sustainable access to new medicines. Among payer initiatives aiming to improve preparedness of healthcare systems for the introduction of new medicines, drug utilization and expenditure forecasting has played an increasingly important role. This study aims to describe the forecasting model used in Region Stockholm and to evaluate the accuracy of the forecasts produced over the past decade.

Methods: In this repeated cross-sectional study, we compared the predicted pharmaceutical expenditure with actual expenditure during the entire available follow-up period (2007-2018) both for overall drug utilization and for individual therapeutic groups. All analyses were based on pharmaceutical expenditure data that include medicines used in hospitals and dispensed prescription medicines for all residents of the region.

Results: According to the forecasts, the total pharmaceutical expenditure was estimated to increase between 2 and 8% annually. Our analyses showed that the accuracy of these forecasts varied over the years with a mean absolute error of 1.9 percentage points. Forecasts for the same year were more accurate than forecasts for the next year. The accuracy of forecasts also differed across the therapeutic areas. Factors influencing the accuracy of forecasting included the timing of the introduction of both new medicines and generics, the rate of uptake of new medicines, and sudden changes in reimbursement policies.

Conclusions: Based on the analyses of all forecasting reports produced since the model was established in Stockholm in the late 2000s, we demonstrated that it is feasible to forecast pharmaceutical expenditure with a reasonable accuracy. A number of factors influencing the accuracy of forecasting were also identified. If forecasting is used to provide data for decisions on budget allocation and agreements between payers and providers, we advise to update the forecast as close as possible prior to the decision date.

Keywords: Drug utilization; Forecasting; Pharmaceutical expenditure.

Fig. 1

Illustration of the forecasting model. a Actual expenditure during 4 years with predicted growth for 2 years based on linear regression. b Adjusted growth in expenditure to account for a patent expiry and introduction of generics. c Adjusted growth in expenditure to account for a new medicine to be launched

Fig. 2

Selection of individual therapeutic groups

Fig. 3

Predicted and actual change in pharmaceutical expenditure (%). Dashed line indicates the forecasts published within the same forecasting report (i.e. same and next year forecast)

Fig. 4

Predicted and actual change in pharmaceutical expenditure (M SEK) for individual therapeutic groups. Dashed lines represent linear regression for predicted and actual change. Black lines represent the threshold for further analyses (SEK 25 M error in forecast). Letters in quadrants indicate A. predicted decrease but actual increase, B. predicted and actual increase, C. predicted and actual decrease, and D. predicted increase but actual decrease in expenditure

Fig. 5

Individual therapeutic groups with major errors in the same year forecast. Predicted and actual change in expenditure (M SEK). Dashed line indicates the forecasts published within the same forecasting report (i.e. same and next year forecast)

Fig. 6

Forecast (individual therapeutic groups) for pharmaceutical expenditure (M SEK) in Region Stockholm (data as of May 2019). ATC 1st level groups. ATC groups J, L, and N are divided in subgroups