Variation in Postoperative Debridement Patterns in Endoscopic Sinus Surgery: A Retrospective Cohort Study

Abstract

Endoscopic sinus surgery (ESS) is recommended for cases of chronic rhinosinusitis refractory to medical management. Postoperative care following ESS is important for achieving successful surgical outcomes and reducing infection rates, inflammation, and synechiae [1, 2]. Indeed, a 2012 survey of 265 otolaryngologists found that 87.9% of them routinely perform postoperative debridements (PDs) [3]. Evidence on the optimal timing, frequency, duration, and intensity of PDs, however, is limited. Nonetheless, between 2000 and 2016, average annual growth rate in the total PD number and cost billed to Medicare Part B was 6.0 and 8.4%, respectively [4].

The American Academy of Otolaryngology states the frequency of PDs is a clinical decision best determined by the surgeon and on a case-by-case basis [5]. No study has explored factors which affect PD utilization. Therefore, we used a commercial claims database to characterize variation in PD patterns, highlighted by a random-effects analysis to test the hypothesis that variation in the number of PDs was dependent on the surgeon who performed the ESS.

We conducted a retrospective cohort study in the Merative™ MarketScan Commercial Database.

We queried every claim between 2003 and 2022 with an ESS procedural code (CPT: 31241, 31253-31257, 31259, 31267, 31276, 31287, 31288, 31295–31298). We selected the first claim for each patient and excluded patients younger than 18 years. We used a clinician identifier variable which became available in 2015 to group patients to surgeons—thus, our cohort was restricted to initial ESS between 2015 and 2022. We counted number of days on which patients had an outpatient PD (31237) within 90 days of their ESS. Patients with fewer than 90 continuous days of coverage following ESS were excluded.

Descriptive statistics of patient characteristics and number of PDs were calculated. Ordinary linear regression analyzed the association between number of PDs and patient characteristics, including bilateral versus unilateral ESS, septoplasty versus no septoplasty, age, sex, geography, year, insurance plan, and state. A random-intercept linear regression determined whether the number of PDs was clustered by the surgeon who performed the ESS [6]. Upon identifying clustering, the intraclass correlation was calculated, which quantified how much of the variation in the number of PDs not explained by the covariates was instead explained at the surgeon level [6].

We identified 69,170 patients undergoing initial ESS between 2015 and 2022 (Table 1). The average age was 44.0 years (standard deviation: 12.7), 52.6% were female, and 82.9% resided in urban geographies. In total, 70.1% of patients underwent bilateral ESS, 49.2% underwent concurrent septoplasty, and 35.6% underwent bilateral ESS and concurrent septoplasty. The 69,170 patients were seen by 11,147 unique otolaryngologists, with each surgeon operating on an average and median of 6.2 (standard deviation: 14.8) and 2 (interquartile range: 1–6) patients, respectively. The number of PDs ranged from 0 to 15 within 90 days of ESS, with 72.2% of surgeons performing a PD on at least one of their patients. Overall, 38.8% of patients had zero PDs, 25.9% had one PD, 21.3% had two PDs, 10.2% had three PDs, and 3.8% had four or more PDs.

Bilateral ESS and concurrent septoplasty were associated with increased number of PDs (p < 0.001; Table 2). Older age was also associated with increased number of PDs (p < 0.001), while rural compared with urban geography was associated with fewer PDs (p < 0.001). Random-intercept regression identified significant clustering in the number of PDs by surgeon. The intraclass correlation was 0.387, such that 38.7% of the unexplained variation in the number of PDs was at the surgeon level.

This retrospective cohort study identified significant variation in the number of PDs performed following ESS. Some of this variation was explained by the laterality of the ESS, whether concurrent septoplasty was performed, and the geography of the patient; however, approximately 39% of the unexplained variation was explained at the surgeon level. These results underscore a lack of consensus regarding how many PDs are appropriate to perform after ESS.

This study is the first to demonstrate that much of the variation in PD patterns depends on surgeon practice preferences. Follow-up research is necessary to determine which characteristics of a surgeon explain this variation; however, we considered the following: first, there is variation in surgical beliefs and attitudes regarding the indications and frequency for PD, as is often described in literature explaining geographic variation in clinical care [7]. These beliefs and attitudes may be secondary to the institutional preferences surgeons develop during training. Second, there is geographic variation in patient demand for ESS and ensuing PDs, to which surgeons respond accordingly [4]. Indeed, we found that patients in rural areas received fewer PDs. Third, there is variation in compensation structures among surgeons, as ESS does not have a global period [5]. Surgeons whose compensation structure is productivity-based may have stronger financial incentives to perform PDs.

This study has limitations. First, we did not identify PDs beyond 90 days of ESS. Second, advance practice providers are performing more PDs [8], and those PDs were only captured if they were indirectly billed [9]. Third, the MarketScan database comprises privately-insured patients, so results may not generalize to publicly-insured patients. The database also does not include patient race/ethnicity or surgeon characteristics such as years in practice, training background, and compensation structure; furthermore, the database may be subject to coding inaccuracies (e.g., tumor surgery) and lack granular clinical information (e.g., polypoid disease). Finally, results may not generalize to revision ESS [10]. Whether debridement patterns increase risk of subsequent revision ESS remains an open question; future studies should investigate this issue.

Conception and design: M. H. and C. R. Data acquisition: M. H. and C. R. Analysis and interpretation: M. H., N. K., J. P. and C. R. Drafting the manuscript: M. H., N. K., J. P. and C. R. Critical revision: M. H., N. K., J. P. and C. R. Guarantor: C. R.

The data were previously collected, statistically de-identified, and are compliant with the conditions set forth in Sections 164.514(a)-(b)(1)ii of the Health Insurance Portability and Accountability Act of 1996 Privacy Rule; therefore, approval from an institutional review board was not sought.

The authors declare no conflict of interest.

The authors have nothing to report.