Article Text


S16 Outcomes in idiopathic pulmonary fibrosis: a meta-analysis from placebo controlled trials
  1. CP Atkins1,
  2. YK Loke2,
  3. AM Wilson1
  1. 1Norfolk and Norwich University Hospital, Norwich, Norfolk
  2. 2Norwich Medical School, University of East Anglia, Norwich, Norfolk


Background Much of the data regarding the natural history of Idiopathic pulmonary fibrosis (IPF) pre-dates current clinical criteria. Despite high mortality debate has been sparked recently regarding the appropriateness of mortality as an endpoint for clinical trials. Additionally, respiratory infections are seen to have a role in the natural history of IPF. We aimed to evaluate the frequency of mortality and respiratory infections in the placebo arms of IPF trials.

Methods We identified randomised, placebo-controlled trials of IPF though a systematic review of MEDLINE, EMBASE, AMED and Cochrane central. Outcomes were mortality and lower respiratory tract infection (LRTI). We standardised event rates and compared differences using incidence rate ratios between subgroups of trials, according to disease severity inclusion or use of low-dose corticosteroids).

Results Thirteentrials were included, involving 1631 patients (2067.7 patient-years of follow-up). Standardised mortality was 91.9 deaths per 1,000 patient-years (range 18.8 to 224.5 per 1,000 years), though rates were higher in trials permitting low-dose immunosuppression and lower in trials excluding severe disease (table 1). Rates of respiratory tract infections were higher in trials including severe disease and in trials including patients receiving immunosuppression (table 1)

Abstract S16 Table 1.

Conclusion Mortality rate was heterogenous and dependent on entry criteria, with rates in studies excluding severe disease half that of unselected studies. Rates of infection in IPF are high, even without concurrent use of immunosuppression. Given low mortality rates, consideration should be given to alternative primary outcomes to mortality in future IPF trials of patients with mild-moderate disease.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.