Healthy survivor bias in patients with idiopathic pulmonary fibrosis in clinical registries

Abstract

<b>Background:</b> Antifibrotic treatment prolongs survival of patients with idiopathic pulmonary fibrosis (IPF). However, imperfect follow-up in real-world registries may obscure the benefit due to differences in patient groups. <b>Aims:</b> Aim of the study was to compare selected baseline characteristics of patients treated with pirfenidone (PIR) and patients without antifibrotic treatment (OT), who remained under follow-up at different time points from baseline. <b>Methods:</b> The study included patients from the EMPIRE registry enrolled between 2015 and 2018. Index date (baseline) was defined as either PIR therapy initiation (PIR group), or enrolment for the registry (OT group). Baseline FVC (%) and DLCO (%) were analysed in patients, who remained under follow-up at 12th and 24th month. <b>Results:</b> There were 769 patients in the PIR group and 633 in the OT group at baseline; baseline FVC (%) 73.7 vs 83.2, DLCO (%) 46.3 vs 51.0. In the PIR group, 345 and 126 pts remained under follow-up at 12th and 24th month; in the OT group, 208 and 62 pts. Patients from the OT group remaining under follow-up after 1 and 2 years had the most favourable baseline characteristics – baseline FVC (%) 90.2 (12 m) and 92.7 (24 m) and DLCO (%) 56.3 (12 m) and 59.0 (24 m). The shift was less apparent in the PIR group – baseline FVC (%) 73.8 (12 m) and 72.6 (24 m) and DLCO (%) 48.0 (12 m) and 51.2 (24 m). <b>Conclusions:</b> There was a different drop-out of patients in the PIR and OT group, with a higher drop-out rate of patients with less favourable baseline FVC (%) and DLCO (%) in the OT group. Interpretation of real-world data from registries must be always cautious, taking into consideration the real number of patients remaining under follow-up.