PT  - JOURNAL ARTICLE
AU  - A. Kahlon
AU  - M.E. Phitidis
AU  - A. Ponnuswamy
AU  - S. Scott
AU  - J.P. Finnerty
TI  - Bronchoscopic inter-operator diagnostic yield for lung malignancy: benchmarking for trainees
DP  - 2012 Sep 01
TA  - European Respiratory Journal
PG  - P1272
VI  - 40
IP  - Suppl 56
4099  - http://erj.ersjournals.com/content/40/Suppl_56/P1272.short
4100  - http://erj.ersjournals.com/content/40/Suppl_56/P1272.full
SO  - Eur Respir J2012 Sep 01; 40
AB  - IntroductionBritish Thoracic Society guidelines suggest that in cases of endoscopically visible malignancy, bronchoscopists should aim for a minimum diagnostic yield of 80%. There is no recommendation however for the diagnostic yield that trainees should achieve before becoming independent bronchoscopists. We conducted a preliminary study in our bronchoscopy unit, comparing inter-operator diagnostic yield for cases of definite endoscopically visible malignancy, aiming to introduce a benchmark for trainees.MethodWe retrospectively looked at 775 bronchoscopies performed by three experienced bronchoscopists. Data was collected on the type of abnormality seen, the likelihood of malignancy, sampling modalities used and positivity rate of samples (histology/cytology) i.e. the diagnostic yield.ResultsA total of 114 bronchoscopies were reported as definite tumours, of which 103 were diagnosed as lung malignancies. The diagnostic yield when a definite tumour was seen on bronchoscopy was 90.9% (60/66) for operator 1, 72.8% (8/11) for operator 2, and 94.6% (35/37) for operator 3, with an overall average of 86.1%. The standard deviation for bronchoscopist success ratio was calculated at 9.55, making the lower limit for success (percentage of positive diagnostic yield) 75%.ConclusionLocally our experienced bronchoscopists are exceeding the 80% diagnostic yield in cases of a definite tumour. Using the lower limit for success, we suggest that 95% of trainee bronchoscopists should be achieving a minimum diagnostic yield of 75%. More data is needed in order to obtain more robust statistical data which can be developed into realistic benchmarks and assessment tools to measure trainee competency.