Methods We performed a systematic literature review based on the PRISMA guidelines following a pre-specified protocol. Pubmed and Embase databases were searched for studies reporting the use of Surgical Lung Biopsies (SLB) in the diagnosis of adults with Interstitial Lung Disease (ILD). Randomised controlled trials, case control studies and case series with more than 20 subjects were included, restricted to papers published from 2000 till September 2015 taking into consideration changes in surgical techniques and diagnostic criteria. All relevant abstracts were assessed by two independent reviewers utilising EPPI reviewer 4, an online software tool for research synthesis. Full papers were obtained for those deemed potentially eligible, and two reviewers agreed the final set of review papers. Primary outcomes were 90 day mortality and complications while secondary were diagnostic yield, mean length of stay and change of treatment following biopsy.
Results (see Figure 1). 24 studies were included reporting on the use of SLB in 2600 patients. The overall quality of the reports was moderate to poor with mainly retrospective case series available. Mean mortality was 4.9% (CI 90% −0.04 −0.14) with a wide range of 0 – 22.4%. Complication rates were reported in 19 of the studies. Mean overall complication rate was 19.4% (CI −0.05 – 0.48) with a range from 7.1% to 65.7%. Mean length of stay adjusted for patient numbers was 5.4 days and diagnostic yield for definite pathological diagnosis was 89%. Eight studies recorded treatment change following SLB in a total of 588 patients out of 869. Mean percentage of patients in whom treatments was changed on the basis of the SLB result was 60% (CI 90% 0.35–0.87).
Conclusions High-quality data on the outcomes of SLB in ILD diagnosis are sparse. Comparison between different studies is difficult due to heterogeneous patient populations (e.g. acute vs elective cases) and differences in outcome reporting. Nonetheless, the overall mortality and morbidity rates are similar to a recent analysis of a US national database. SLB in ILD remains a useful diagnostic tool but carries significant mortality and morbidity. More prospective data and evaluation of surgical risk stratification is required.