Revision Rates of Septoplasty in the United States

Abstract

Background:

Large-scale studies characterizing septoplasty revision rates are lacking.

Objectives:

To identify rates of septoplasty revision in the United States.

Methods:

Patients undergoing initial septoplasty between January 1, 2007 and December 31, 2013 were identified using the IBM^® MarketScan^® Commercial Database. Patients were excluded if they had nasal vestibular stenosis, rhinoplasty, or costal cartilage grafts for the initial surgery, or did not have either septoplasty, nasal vestibular stenosis, rhinoplasty, and/or costal cartilage grafts for the second surgery.

Results:

295,236 patients received an initial septoplasty, and 3213 (1.1%) patients underwent a revision. Among the revision group, 178 (5.4%) patients received a septorhinoplasty, among which 13 (7.3%) required a costal cartilage graft. Older patients were less likely to need revision surgery (RS). Patients in the Northeast and West were significantly more likely than patients in the Midwest to undergo RS. Insurance plans such as comprehensive and point-of-service were associated with greater odds of RS, whereas others such as high-deductible health plans were associated with lower odds.

Conclusion:

Septoplasty revision rates are relatively low at 1.1% but influenced by age, region, and insurance plan.

KEY POINTS

Question: What is the septoplasty revision rate in the United States?

Findings: Patients undergoing a septoplasty in the United States rarely have revision surgery.

Meaning: In this study, patients who were younger, from the Northwest or Eastern United States, or with certain insurance plans were more likely to have a revision septoplasty, but the rate remained very low at 1.1%.

Introduction

Septal deviations are extremely common among the general population, with studies reporting incidence rates from 22% to 37%.^1,2 When symptomatic, a deviated septum most often presents with a nasal obstruction that decreases quality of life, but has also been associated with obstructive sleep apnea, nasal crusting, sinusitis, and epistaxis.^3–5 Although medical management for a deviated septum with a nasal obstruction often includes nasal sprays and local corticosteroids, up to a quarter of these patients end up opting for surgical treatment, making septoplasties the most frequently performed procedure in the field of otolaryngology.^3,6–9

Recently, a randomized control trial by van Egmond et al. demonstrated that surgical management is superior to medical management and leads to better outcomes.^6,10 However, success rates, defined through various metrics including the Nasal Health Survey and Nasal Obstruction Septoplasty Effectiveness Scale, range from 43% to 85%, which lead some patients to opt for a revision procedure to improve quality of life and overall health.^5,11,12 From an anatomical standpoint, persistent deviation is often seen in the middle septum and caudal septum, whereas from a surgical standpoint, persistent obstruction may be due to iatrogenic surgical deformities as well as pathological processes that were not corrected during the primary procedure.^13,14

Regardless of the cause, septoplasty revisions are a significant issue in the field of otolaryngology, with revision rates ranging from 1.7% to 9.6%.^15–17 Although these studies provide evidence of the relatively low revision rate for septoplasties, they only involve single institutions that do not adequately represent the diverse field of otolaryngology.^18–20 In addition, in both septoplasty and numerous other medical services, revisions and patient outcomes are influenced by a myriad of factors, including technique, age, gender, region, and insurance status.^13,20–29 Thus, we performed a retrospective population-based cohort study among patients undergoing septoplasty to identify septoplasty revision rates across the United States.

Methods

This retrospective population-based cohort study was conducted using the IBM^® MarketScan^® Research Databases. Data for this project were accessed using the Stanford Center for Population Health Sciences Data Core. This study was approved by the Institutional Review Board and informed consent was waived as it is a population database study.

The IBM MarketScan Commercial Database contains inpatient and outpatient data from >20 million individuals and their dependents who are covered by employer-sponsored private health insurance in the United States.³⁰ Individuals between the ages of 18 and 65 years were included if they received a septoplasty without nasal vestibular stenosis (NVS), rhinoplasty, or a costal cartilage graft as determined by the Current Procedural Terminology (CPT) codes (CPT codes Supplementary Table S1). Among this cohort, revision septoplasty was defined as having had a second surgery, including septoplasty, NVS, or rhinoplasty with or without cartilaginous grafts (Supplementary Table S1).

Although the full database available to our institution consists of claims from January 1, 2007 to December 31, 2016, we examined a subset of claims where the initial surgery took place from January 1, 2007 to December 31, 2013 to allow for a minimum follow-up period of 3 years from the initial procedure. Patients were excluded if they did not meet the criteria for the first surgery (septoplasty without NVS, rhinoplasty, or a costal cartilage graft) or were missing demographic or essential variables (age, gender, insurance plan type, and region). The first 12 CPT codes for each unique patient encounter were utilized for this analysis.

The primary outcome of the study was to determine the revision rate of septoplasty. Potential associations between demographic variables, including age, gender, insurance plan type, and region, and the rate of revision septoplasty, were evaluated. The secondary outcomes of the study were to determine the rate of conversion of septoplasty to septorhinoplasty and the percentage of septorhinoplasties requiring a costal cartilage graft.

Univariate descriptive analyses were performed for demographic variables for each surgical cohort (RS vs. no RS). Continuous and categorical variables were tested for differences using t-tests and chi-square tests, respectively. Univariate and multivariate logistic regression analyses were performed to examine predictors of revision septoplasty. Predictor variables included age, gender, insurance plan type, and geographic region. Crude and adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were calculated. Statistical analysis was performed using RStudio Version 1.4.1103 with p < 0.05 set as the threshold for statistical significance.

Results

Patient characteristics

A total of 320,082 patients met the inclusion criteria for the study. After excluding patients with missing data points, 295,236 patients with a mean age of 40.95 (standard deviation [SD] 12.5) were included in the analyses. Within this cohort, 3213 (1.1%) received a revision septoplasty (Table 1). Patients who underwent RS were significantly younger (mean age 38.17, SD 12.8) than patients who did not undergo revision surgery (NS) (mean age 40.97, SD 12.5).

Table 1.

Characteristics of study sample by revision septoplasty status, IBM MarketScan commercial database 2007–2013

	Surgical cohort		p
	Revision septoplasty	No revision surgery
Patients (n)	3213	292,023
Age, mean (SD), years	38.71 (12.8)	40.97 (12.5)	<0.001
Age group, n (%), years
18–34	1264 (39.3)	96,200 (32.9)	<0.001
35–44	759 (23.6)	74,460 (25.5)
45–54	765 (23.8)	71,577 (24.5)
55–64	425 (13.2)	49,786 (17.0)
Gender, n (%)
Female	1425 (44.4)	125,572 (43.0)	0.129
Insurance plan type, n (%)
PPO	2065 (64.3)	199,037 (68.2)
Comprehensive	82 (2.6)	4739 (1.6)	<0.001
EPO	49 (1.5)	5930 (2.0)
HMO	453 (14.1)	38,270 (13.1)
POS	343 (10.7)	23,577 (8.1)
POS with capitation	29 (0.9)	1940 (0.7)
CDHP	136 (4.2)	11,432 (3.9)
HDHP	56 (1.7)	7098 (2.4)
Region, n (%)
South	1242 (38.7)	119,472 (40.9)
Northeast	559 (17.4)	43,154 (14.8)	<0.001
Midwest	670 (20.9)	72,605 (24.9)
West	742 (23.1)	56,792 (19.4)

CDHP, consumer-directed health plan; EPO, Exclusive Provider Organization; HDHP, high deductible health plan; HMO, Health Maintenance Organization; POS, point-of-service; PPO, Preferred Provider Organization; SD, standard deviation.

There was no significant difference in the gender composition among RS cohort (55.6% male) and NS cohort (57.0% male). Between the RS and NS cohorts, there were significant differences in the distribution of insurance plans and regions where the patients were from Table 1. Insurance plan type variation within regions was also found (p < 0.001).

Primary and secondary outcomes

A total of 3213 (1.1%) patients underwent RS. Among those who underwent RS, 178 (5.4%) patients received a septorhinoplasty, with 13 (7.3%) of these patients requiring a cartilaginous graft.

Univariate and multivariable analyses

Univariate and multivariate logistic regression analyses were performed to evaluate predictors of septoplasty revision (Table 2). In univariate analyses, patients in older age groups had a significantly lower risk of RS compared with patients who were in the youngest age group (ages 18–35): 35–44 age group (odds ratio [OR] 0.776 [95% CI 0.708–0849]), 45–54 age group (OR 0.813 [95% CI 0.743–0.890], and 55–64 age group (OR 0.650 [95% CI 0.581–0.725]).

Table 2.

Characteristics associated with septoplasty revision, IBM MarketScan commercial database 2007–2013

	Unadjusted odds ratio (95% CI)	Adjusted odds ratio (95% CI)
Age group
18–34	1.000 (reference)	1.000 (reference)
35–44	0.776 (0.708–0.849)	0.777 (0.709–0.850)
45–54	0.813 (0.743–0.890)	0.811 (0.740–0.887)
55–64	0.650 (0.581–0.725)	0.638 (0.570–0.712)
Gender
Male	1.000 (reference)	1.000 (reference)
Female	1.055 (0.985–1.133)	1.055 (0.983–1.131)
Insurance plan type
PPO	1.000 (reference)	1.000 (reference)
Comprehensive	1.668 (1.325–2.069)	1.790 (1.422–2.226)
EPO	0.796 (0.591–1.046)	0.732 (0.543–0.962)
HMO	1.141 (1.029–1.262)	1.099 (0.991–1.217)
POS	1.402 (1.248–1.571)	1.391 (1.236–1.559)
POS with capitation	1.441 (0.973–2.043)	1.353 (0.913–1.922)
CDHP	1.147 (0.959–1.360)	1.164 (0.973–1.380)
HDHP	0.760 (0.576–0.982)	0.751 (0.568–0.970)
Region
South	1.000 (reference)	1.000 (reference)
Northeast	1.246 (1.126–1.377)	1.214 (1.096–1.343)
Midwest	0.887 (0.807–0.975)	0.895 (0.814–0.984)
West	1.257 (1.146–1.377)	1.282 (1.168–1.405)

Bolded text indicates statistically significant results. p < 0.05 was set as the threshold for statistical significance.

CI, confidence interval.

Similarly, in multivariate analyses after adjusting for potential confounders, the risk for RS remained lower in older populations compared with the reference 18–34 age group: 35–44 age group (aOR 0.777 [95% CI 0.709–0850]), 45–54 age group (aOR 0.811 [95% CI 0.740–0.887], and 55–64 age group (aOR 0.638 [0.570–0.712]). Gender was not associated with an increased risk for RS in either univariate or multivariate analyses.

Regions associated with increased risk of revision included the Northeast (OR 1.246 [95% CI 1.126–1.377]) and West (OR 1.257 [95% CI 1.146–1.377]) relative to the South. In contrast, the Midwest region was associated with a reduced risk for revision septoplasty (OR 0.877 [95% CI 0.807–0.975]) compared with the South. The multivariate analyses yielded similar results with the Northeast (aOR 1.214 [95% CI 1.096–1.343]) and West (aOR 1.282 [95% CI 1.168–1.405]) regions being associated with a higher risk of revision, whereas the Midwest region (aOR 0.895 [95% CI 0.814–0.984]) was associated with a reduced risk of revision.

Insurance plan types associated with increased odds for revision septoplasty in univariate analyses included comprehensive (OR 1.668 [95% CI 1.325–2.069]), point-of-service (POS) (OR 1.402 [95% CI 1.248–1.571]), and Health Maintenance Organization (HMO) (OR 1.141 [95% CI 1.029–1.262]) relative to Preferred Provider Organization (PPO). In this model, having a high deductible health plan (HDHP) (OR 0.760 [95% CI 0.576–0.982]) reduced the risk for RS. In the adjusted multivariate model, only the comprehensive (aOR 1.790 [95% CI 1.422–2.226]) and POS (aOR 1.391 [95% CI 1.236–1.559]) health plans were associated with greater odds of RS. In contrast, HDHP (aOR 0.751 [95% CI 0.568–0.970]) and Exclusive Provider Organization (EPO) (aOR 0.732 [95% CI 0.543–0.962]) were associated with a reduced risk of revision in the adjusted model. The POS with capitation and consumer-directed health plan (CDHP) insurance health plans were not associated with RS in either the crude or adjusted model.

Discussion

This is the first large-scale study to analyze septoplasty revision rates nationwide. In our cohort of 295,236 patients, 3213 (1.1%) opted for an RS. One hundred seventy-eight (5.4%) patients who had a revision also underwent concomitant rhinoplasty, and within that cohort, 13 (7.3%) patients required a costal cartilaginous graft. Subgroup analysis of age found that older patients were less likely to need an RS. Patients in the Midwest were less likely, whereas patients in the Northeast and West were more likely to undergo an RS relative to the South. Insurance plan type was also correlated with revision rate. Although studies looking at revision septoplasty rates are not common in the literature, the few that exist found revision rates ranging from 1.7% to 9.6%.^15–17

Although similar, our revision rate of 1.1% is lower than what is reported in much of the literature, which may be explained by the fact that the IBM MarketScan Commercial Database we utilized is composed of nationwide data from numerous surgeons who utilize a diverse array of techniques.²¹ For example, Gubisch's single institution study from Germany in 2005 looked at revision rates among extracorporeal septoplasty (a more invasive procedure than standard septoplasty) on 2119 patients and found that 4% of operations performed by attending physicians and 7% of operations performed by trainees resulted in a revision procedure.¹⁷ Meanwhile, Karlsson et al. found a 3.2% revision rate for standard endonasal septoplasties performed with or without inferior turbinate reduction.¹⁵

Variation within techniques must also be considered as Rezaeian et al. reported revision rate differences of up to 7.9% within extracorporeal septoplasties based on the suturing technique used.²² Another possible explanation for our lower revision rate is that the IBM MarketScan Commercial Database is based on insurance claims and excludes out-of-pocket rhinoplasties. This may lead to an underestimation of the septoplasty revision rate as well as the septoplasty to rhinoplasty conversion rate when compared with other studies. This limitation is discussed in more depth in the Limitations section. In addition, within the cohort undergoing a rhinoplasty as their secondary procedure, we found that 7.3% of patients required the use of a costal cartilage graft, which is similar to the findings of a large multistate study conducted by Lee et al.³¹

An important finding from this study was that the odds of undergoing RS were lower for the 35–44, 45–54, and 54–64 age groups compared with the 18–35 age group. A study by Spataro et al. showed similar results for revision septorhinoplasty with patients in the 13–18 age group being associated with higher risks of RS and patients' ages 41 and older being associated with reduced risks of revision compared with patients' ages 19–40.²¹ This finding may be explained by the interaction of multiple factors, including different expectations for outcomes, different healing processes for older patients, and anatomical differences compared with younger populations.^23,32,33 For example, a study by Pedersen et al. on patients undergoing primary septoplasty found that older patients had better odds for improved health outcomes after surgery.²³ Nevertheless, further research is needed to understand the reduced risk of revision with older populations. Although not statistically significant, we found that the majority of patients in both the NS and RS groups were male, which supports the existing literature.^{13,15,16,22,24}

The geographic variation in surgical revision rates observed in this study has been well documented throughout the literature.³⁴ One factor contributing to this variation includes the supply of otolaryngologists in the United States, with one study showing that the average number of otolaryngologists per county is higher in the Northeast, Mountain, and Pacific regions compared with the South Central and Midwest regions.¹⁸ Another study by Lango et al. showed that the number of otolaryngologists in a region correlated with areas of higher mean incomes, education levels, and supply of other specialist physicians.¹⁹ As states in the Northeast and West have higher mean incomes and educational attainment levels, this is consistent with where more specialists and otolaryngologists practice.³⁵

In addition, it is possible that insurance plan variations may also influence the geographical variation in revision rates, as one study found that patients in the Northeast are less likely, whereas patients in the Midwest are more likely to have high deductible health plans that may discourage revision procedures.²⁵ Although our chi-square analysis demonstrated differences between plan type and region, our final multivariate model demonstrated statistical significance while accounting for this possible confounding effect (Table 2). Although these factors can explain part of the variation, the true nature of the geographic variation of septoplasty revision risks are likely to be multifactorial as other otolaryngology procedures such as endoscopic sinus surgery and outpatient pediatric tonsillectomy show the opposite trends with the Midwest and South having higher surgical rates.^20,26

In a systematic review by Reames et al., one factor shown to help reduce geographic variation is the establishment of a clinical guideline for a procedure.³⁶ However, the American Academy of Otolaryngology-Head and Neck Surgery could not reach a formal clinical guideline for septoplasty with or without inferior turbinate reduction due to a lack of evidence.³⁷ As a result, factors such as physicians' beliefs and attitudes about the indication for surgery may play a greater role in the revision risks, and these can vary greatly from practice to practice.³⁴

The differences in risk for revision septoplasty for different insurance plans may be attributable to a variety of factors, including the cost-sharing burden inherent to each insurance plan. We observed comprehensive and POS health plans to have greater odds of revision but HDHP and EPO to have lower odds of revision. This could be explained as comprehensive and POS health plans have lower out-of-pocket cost sharing compared with a plan such as high deductible health plan, which has a higher out-of-pocket cost sharing. These costs can ultimately influence the utilization of health care.

Our results are supported by other studies including one study by Chhabra et al. showing that when looking at bariatric surgeries, comprehensive, EPO, HMO, and PPO health plans had the highest rates of health care utilization and lowest out-of-pocket sharing.²⁵ In contrast, HDHP and CDHPs had the lowest utilization rates and highest out-of-pocket cost sharing.²⁵ Other studies including the RAND Health Insurance Experiment have observed these trends as well.^27–29,38 Although the lower-cost sharing associated with EPO does not align completely with the reduced risk of RS we observed, this may possibly be explained by the need for a referral by a primary care provider in EPO health plans as well as other factors not completely understood.

Limitations

First, this study utilizes an insurance claims database, which requires health care providers to accurately code the CPT codes that were used in this study. In our study, we excluded patients who were missing key variables that may have impacted the composition of the final study population. Second, this is a retrospective study that uses convenience sampling and only examines the rates of revision septoplasty for those reimbursed through employer-based insurance. Thus, individuals with public insurance such as Medicaid or Medicare or those who are uninsured are not included in these analyses, which impacts the generalizability of these findings.

Moreover, since procedures with out-of-pocket payments, such as nonfunctional rhinoplasties, are not reimbursed by insurance and thus are not included in the database, we may not be capturing all septoplasty revisions in our study population.³⁹ Third, patients may no longer be enrolled in the database at any point due to various reasons, including a change in employer, switch to Medicare coverage, or death.⁴⁰ Patients would appear as distinct identifiers every time they had a change in employer or insurance, with the identifiers being unlinkable. Furthermore, initial surgery data were examined from January 1, 2007 to December 31, 2013 to allow for a minimum of 3 years follow-up and a maximum of 10 years follow-up time.

However, because of this, we may not capture all the revisions and thus likely underestimate the lifetime revision rate, although the magnitude of this is impossible to determine. Previous studies have looked at septoplasty or septorhinoplasty revisions with a follow-up period ranging from 7 to 24 years.^15,17,21 Fourth, demographic characteristics such as race and socioeconomic status are not included in the database that was utilized and, therefore, not studied. These could perhaps be confounders or help some of the findings we observed. Fifth, due to the lack of granularity in administrative datasets we cannot reveal the causes or indications for RS.

Conclusion

To our knowledge, our study is the first large-scale study examining the revision rates for septoplasty in the United States. The study findings suggest a septoplasty revision rate of 1.1% and a septoplasty to septorhinoplasty conversion rate of 5.4%, among which 7.3% required a cartilaginous rib graft. Patient characteristics correlating with increased risk of revision include younger age groups, patients residing in the Northeast and West, and patients with comprehensive and POS health plans. Although more research is needed to understand the etiologies and causes for septoplasty RSs, our study nevertheless provides valuable information about the rates of RS and risk factors that may help inform clinical decision-making.

Authors' Contributions

E.X.W., C.K., and S.P.M. conceived the project idea and design. G.Y., J.P.S., E.X.W., C.K., and S.P.M. performed data analysis and interpretation. G.Y. and J.P.S. wrote the article with support from E.X.W., C.K., and S.P.M. All authors critically revised the article for important intellectual content. All authors approved the final version to be published. All authors agree to be accountable for all aspects of the study.

Author Disclosure Statement

No competing financial interests exist.

Disclaimer

The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Funding Information

The PHS Data Core is supported by a National Institutes of Health National Center for Advancing Translational Science Clinical and Translational Science Award (UL1 TR001085) and from Internal Stanford funding.

Supplementary Material

Supplementary Table S1