• C Ten Thoren, C Balg, J Gibbert, S Mostardt, M Ripoll, D Schierbaum, S Schiller, and A Schwalm.
• Ressort Versorgung und Gesundheitsökonomie/Dept. of Health Care and Health Economics, Institut für Qualität und Wirtschaftlichkeit im Gesundheitswesen (IQWiG), Im Mediapark 8, 50670, Köln, Germany. corinna.ten-thoren@iqwig.de.
• Eur J Health Econ. 2020 Aug 1; 21 (6): 881-893.

ObjectivesDossiers submitted for early benefit assessments in Germany also provide information on the precise determination of the target population (patients eligible for a drug). The situation is complex for non-small-cell lung cancer (NSCLC) due to highly specific therapeutic indications. Our aim was to compare the different methodological steps applied to determine the target population in dossiers on drugs for NSCLC.MethodsWe analysed NSCLC dossiers assessed by the German Institute for Quality and Efficiency in Health Care (IQWiG) between 01.01.2011 and 31.12.2017. Methodological details regarding the determination of the target population were extracted and compared.ResultsWe analysed 23 NSCLC dossiers. In all dossiers, the target population was determined using the number of all patients with lung cancer as the basis for calculations. This patient population was further reduced in several successive steps by assuming proportions of patients with a specific characteristic (e.g. disease stage). The most important calculation steps were patients with NSCLC (n = 23 dossiers), with a specific disease stage (n = 23), with a specific tumour mutation (n = 14), with a specific tumour histology (n = 7), without prior treatment (n  = 15), with pretreatment in second or further treatment lines (n  = 17), and/or with specific pretreatments (n  = 9). The proportions of patients determined within the same calculation step varied considerably between dossiers.DiscussionThe calculation methods applied and the target population sizes reported in NSCLC dossiers vary considerably. A consensus with regard to the databases and calculation methods used to determine the target population in NSCLC would be helpful to reduce variations.

23 keywords...

hide…