Article Text

S45 How Successful Are Medical Thoracoscopists At Predicting Malignancy?
  1. RJ Hallifax,
  2. JP Corcoran,
  3. I Psallidas,
  4. NM Rahman
  1. Oxford Centre for Respiratory Medicine, Oxford, UK


Introduction Use of medical thoracoscopy by physicians to diagnose malignant pleural disease is increasing. Thoracoscopy is also used therapeutically to pleurodese (talc poudrage), at the time of biopsy, to minimise pleural effusion recurrence. However, this relies on the physician being confident of their diagnosis macroscopically.

A survey was conducted of regular thoracoscopists to establish the current practise related to medical thoracoscopy and a video of previous thoracoscopies (with known histology) was used to assess the accuracy of macroscopic evaluation of pleural disease.

Methods 20 video clips recorded during thoracoscopy in Oxford were combined into a short video (30 s per clip): including a selection of cases with malignant (13) and benign disease (7), intentionally including some cases which were deemed "unclear". A survey of thoracoscopists was conducted via email. Respondents were asked to state whether each clip showed malignant or benign disease, scoring their confidence in their diagnosis (out of 10), whether they predict trapped lung and if they would perform pleurodesis. Gold standard of diagnosis was the histology result.

Results Procedural survey: 16 physicians responded from 12 tertiary referral centres: 13 consultants and 3 specialist registrars. 15 (94%) had performed >30 thoracoscopies each. Four institutions (25%) perform between >10 thoracoscopies per month; 12 (75%) between 1–10 per month. Only 6 (38%) perform thoracoscopies as day cases. All perform rigid rather than flexible thoracoscopies.

Video survey: Of the 20 video clips, the mean number of correct answers was 12.4 (62%). Respondents were more confident of their answers (out of 10) when correct (7.1/10) than incorrect (6.1/10). In cases deemed malignant, 69% would have performed talc pleurodesis; however, 17% would have pleurodesed patients who had benign disease (See Table). Respondents only correctly predicted trapped lung in 2.6/20 cases (13%).

Conclusion Despite being experienced thoracoscopists, only 62% correctly diagnosed malignant or benign on video clips. The majority would appropriately perform pleurodesis, but 17% may have inappropriately pleurodesed benign disease. There are limitations to this small survey using short thoracoscopy clips, but this data suggests caution is required when considering making diagnosis on macroscopic appearance and deciding whether the lung is trapped.

Abstract S45 Table 1

Survey results: mean number of correct answers with mean level of confidence scores for the answers (those correct vs those incorrect), and whether the respondent would perform pleurodesis or not

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.