Cost-description and multiple imputation of missing values: the (SAT) study

Abstract

Aim: This article reports on a retrospective quarterly cost description (CD) performed on 401 patients with stable chronic obstructive pulmonary disease (COPD) at enrolment in the national, multicenter, noninterventional SATisfaction and adherence to COPD treatment (SAT) study. Methods: Examined from the perspective of the Italian National Health Service, health care resources were costed via published sources and expressed in euro (€) 2017 on a per patient basis. Incomplete data were addressed via multiple imputation (MI) with 20 completed datasets. Results: 384 patients were evaluable (moderate COPD: 76.00%). Most patients (61.47%) reported no COPD exacerbations during the year before the enrolment. Under Available Case Analysis the mean quarterly cost was €328.59 (standard deviation – SD: €537.25); 187 patients (48.69%) reported missing completely at random (MCAR) or missing at random (MAR) data. Their mean quarterly cost (€191.30; SD: €356.86) was significantly lower (bootstrap p-value < 0.0001) than that of patients with complete data (€458.90; SD: €638.85). Under MI the mean quarterly cost reached €662.48 (SD: €1476.03). Oxygen therapy (265.73; SD: €807.19) and drugs (€179.27; SD: €103.49) added up to 67.17% of the total quarterly cost. The mean quarterly cost for ICS/LABA/LAMA drugs (€295.03; SD: €64.48) was 64.57% higher than the overall mean quarterly cost for drugs. The difference between mean quarterly cost for severe and moderate BPCO patients was +179.90%. Conclusions: Preliminary results of the SAT study highlight the need to further investigate appropriate and sustainable drug treatment for moderate and severe COPD. MI increases the reliability of CD estimates when MCAR and MAR data are taken into account.