Comparative Five-year Revision Risk Following Balloon Dilation versus Endoscopic Sinus Surgery: Analysis of 670,000 Procedures

Introduction: Balloon sinus dilation is increasingly used for chronic rhinosinusitis (CRS), yet comparative revision risk relative to traditional functional endoscopic sinus surgery (FESS) remains uncertain.

Methods: The PearlDiver Mariner claims database was queried to identify adults with CRS undergoing initial maxillary, sphenoid, or frontal sinus procedures with at least five-year follow-up. Procedures were categorized as maxillary antrostomy versus balloon dilation, sphenoidotomy versus balloon dilation, and frontal sinusotomy versus balloon dilation. Revision was defined as a subsequent procedure on the same sinus after the index operation. Unadjusted revision rates were compared using chi-square testing. Multivariable logistic regression modeled revision for each sinus, adjusting for age, sex, average percent insured by zip code, surgeon volume, asthma, nasal polyps, eosinophilic granulomatosis with polyangiitis, migraine, cystic fibrosis, tobacco use, atypical facial pain, trigeminal neuralgia, perioperative steroid use, and time from CRS diagnosis to index procedure.

Results: A total of 327,509 maxillary, 145,222 sphenoid, and 197,493 frontal sinus procedures were identified. Five-year revision rate was 9.35% after maxillary balloon dilation versus 8.06% after maxillary antrostomy (χ²=90.8, p< 0.001), 5.48% after sphenoid balloon dilation versus 6.71% after sphenoidotomy (χ²=62.4, p< 0.001), and 6.73% after frontal balloon dilation versus 8.07% after frontal sinusotomy (χ²=100.2, p< 0.001). After adjustment, balloon dilation was associated with higher odds of revision for maxillary (OR 1.54, 95% CI 1.47–1.61), sphenoid (OR 1.14, 95% CI 1.06–1.23), and frontal procedures (OR 1.09, 95% CI 1.04–1.15).

Conclusions: In this cohort, balloon sinus dilation was associated with higher adjusted odds of revision compared with FESS across maxillary, sphenoid, and frontal procedures, despite lower unadjusted revision rates in sphenoid and frontal procedures. Further investigation is required to identify which covariates moderate this relationship.