Hospital variation in bronchoscopy and esophagoscopy rates during head and neck cancer diagnostic evaluation.

Abstract

6044 Background: Hospital variation in bronchoscopy and esophagoscopy rates while diagnosing head and neck cancer may reflect professional uncertainty about the effectiveness of these procedures and a lack of clinical guidelines on best practices. Furthermore, high-volume hospitals have demonstrated better outcomes for select oncologic procedures. We examined the association between hospital case volume and diagnostic bronchoscopy and esophagoscopy rates. Methods: This retrospective cohort study used the 2006-2010 Michigan State Ambulatory Surgery Databases, capturing all outpatient surgical cases in Michigan. Eligible cases included head and neck cancer patients who underwent laryngoscopy, bronchoscopy, and/or esophagoscopy. The primary outcome measure was the likelihood that a patient who underwent laryngoscopy during head and neck cancer diagnostic workup also underwent either bronchoscopy or esophagoscopy. We used hierarchical, mixed-effect logistic regression to measure the association between the primary outcome and hospital case volume (<100, 100-999, or ≥1,000 cases/hospital) while adjusting for patient-level variables such as age, sex, race, insurance status, and household income. Results: Of 17,828 head and neck cancer patients, 9,218 underwent diagnostic laryngoscopy. The 50 low-volume and 40 medium-volume hospitals performed significantly more concurrent bronchoscopies and esophagoscopies compared to the 2 high-volume hospitals (both p<0.001). After adjusting for patient characteristics, medium-volume and low-volume hospitals respectively had 9.3-fold and 7.8-fold higher odds of performing esophagoscopy relative to high-volume hospitals (p=0.003), although the association with bronchoscopy was no longer statistically significant. Conclusions: The proportion of head and neck cancer patients undergoing diagnostic laryngoscopy with concurrent esophagoscopy, but not bronchoscopy, varies significantly by hospital volume. A robust discussion of the comparative effectiveness of comprehensive and selective endoscopy will require further research into whether endoscopic volume correlates with tumor staging, survival, and other outcomes data.