Interobserver-Agreement zwischen Pneumologen und dem Zytopathologen, die identische TBNA-Ausstriche bewertet haben

Pulmonolgists ability to review specimens obtained by TBNA - A real life study

by Tobias Anslinger
Doctoral thesis
Date of Examination:2019-08-08
Date of issue:2020-07-02
Advisor:Prof. Dr. Heyo K. Kroemer
Referee:Prof. Dr. Stefan Andreas
Referee:Prof. Dr. Hans-Ulrich Schildhaus
Persistent Address: http://dx.doi.org/10.53846/goediss-8072

 

 

Files in this item

Name:Promotion Tobias Anslinger.pdf
Size:1.13Mb
Format:PDF

The following license files are associated with this item:

Abstract

English

Introduction: Cytology has become increasingly important in lung cancer as mediastinal staging via EBUS guided TBNA has virtually replaced mediastinoscopy. The aim of this study was to compare two independent cytological laboratories with different degrees of experience and training. Material and methods: Results of two cytology labs for all patients in two consecutive years (2012 and 2013) who underwent TBNA were retrospectively assessed. Lab A consisted of two pulmonologists with 6 and 10 years experience in cytology respectively, each reviewing 1000 specimen a year. Lab B consisted of a single very experienced, board certified, cytologist reviewing 36.000 specimens a year. The results were compared in three steps: 1. malignant vs. benign 2. Of all lung cancers NSCLC vs. SCLC 3. Of all NSCLC Adenocarcinoma (ADC) vs. squamous cell carcinoma (SCC) and vs. NSCLC not otherwise specified (NOS). If available histological data was compared to verify cytology results. Results: 1052 TBNA-results were compared. Agreement for malignant vs. benign was substantial (92%, κ=0.79) with 232 diagnosed as malignant by both labs. Of these 211 were lung cancers. Agreement was almost perfect (96%, κ=0.89) for SCLC vs. NSCLC with 158 diagnosed as NSCLC by both labs. Diagnosed as SCC/ADC/NOS respectively were: Lab A 18/85/55; Lab B 38/84/37. 68 resected mediastinal lymph nodes confirmed by histology were compared to preoperative cytological results. Lab A had 7 false malignant results but none false benign. In Lab B one result was false malignant and one false benign. Conclusion: In this "real life" study interobserver agreement was high. Trained pulmonologists can reliably review cytological specimen obtained by TBNA.
Keywords: TBNA; cytology; lung cancer; NSCLC; SCLC; diagnostic; pulmonologist; interrater; interobserver; histology
Schlagwörter: TBNA; Zytologie; Bronchialkarzinom; NSCLC; SCLC; Diagnostik; Bronchoskopie; Interobserver; Histologie
 
Statistik