Health economic evaluations of shoulder pain, colorectal cancer and scoliosis

Key conclusions\nIn Paper I, a cost-of-illness (COI) study on shoulder pain in Sweden, showed that the mean health care cost per patient was €326 during 6 months, and physiotherapy treatments accounted for 60% of this cost. The mean annual total cost was €4 139 per patient. Of this, sick leave accounted for 84% of the cost, but different methods for estimating sick leave cost can provide very different results.\nIn Paper II a semi-Markov model with 70 health states was presented and validated. We tracked age and time since specific health states using tunnels and a three-dimensional data matrix. The structure and complexity of the model and the variety of data sources implied that we faced parameter and methodological uncertainty, as well as modelling uncertainty. Therefore, the model was validated using face, internal, cross and external validation. The main result from Paper II was the validation, and this revealed a satisfactory match with other models and empirical estimates of both the cost of colorectal cancer treatment and survival time, which are the two main outcomes of the model. We performed no preceding calibration of the model.\nIn Paper III, we found that altered decisions about palliative treatment can increase the average CRC cost substantially. Reducing the recurrence rate by better surgery and implementing preventive efforts like screening of asymptomatic persons could have a considerable cost-effectiveness potential. Further, we saw that expectations about the future are important for cost and survival estimates. Because many evaluations have time horizons of 20-40 years, PSA that is based on parameter probability distributions estimated from “yesterday’s data” can be misleading.\nIn Paper IV, we compare costs in screening and non-screening scenarios using a cost-minimization analysis. Many relevant factors can be assumed to differ from country to country. We found that the cost-effectiveness of screening is heavily dependent on (i) the percent of the non-screened that receive some kind of treatment (surgery or bracing) for their scoliosis, and (ii) the share of surgery versus bracing, in both screened and non-screened children. We also found that it is more cost-effective to screen girls only rather than screening all children.