eTable 1. Codes Used for Selecting Cohort
eTable 2. Surgical Characteristics of Cohort
eTable 3. Rate of VTE Based on the Type of Cancer in Entire Cohort
Customize your JAMA Network experience by selecting one or more topics from the list below.
Cramer JD, Dilger AE, Schneider A, Smith SS, Samant S, Patel UA. Risk of Venous Thromboembolism Among Otolaryngology Patients vs General Surgery and Plastic Surgery Patients. JAMA Otolaryngol Head Neck Surg. 2018;144(1):9–17. doi:10.1001/jamaoto.2017.1768
What is the rate of venous thromboembolism in otolaryngology compared with established-risk surgical fields?
This cohort study of 31 896 adult otolaryngology patients found that most patients undergoing otolaryngology operations were at low risk for venous thromboembolism, with rates that are the same or lower than low-risk surgical fields (plastic surgery); however, a subgroup of patients experienced significantly greater rates of venous thromboembolism.
Guidelines for a population at low risk of venous thromboembolism could be adapted to otolaryngology; however, a subgroup of patients with significantly higher risk of venous thromboembolism should be considered for more aggressive prophylaxis.
Venous thromboembolism (VTE), which includes deep venous thrombosis or pulmonary embolism, is the number 1 cause of preventable death in surgical patients. Current guidelines from the American College of Chest Physicians provide VTE prevention recommendations that are specific to individual surgical subspecialties; however, no guidelines exist for otolaryngology.
To examine the rate of VTE for various otolaryngology procedures compared with an established average-risk field (general surgery) and low-risk field (plastic surgery).
Design, Setting, and Participants
This cohort study compared the rate of VTE after different otolaryngology procedures with those of general and plastic surgery in the American College of Surgeons National Surgical Quality Improvement Program from January 1, 2005, through December 31, 2013. We used univariate and multivariable logistic regression analysis of clinical characteristics, cancer status, and Caprini score to compare different risk stratification of patients. Data analysis was performed from May 1, 2016, to April 1, 2017.
Main Outcomes and Measures
Thirty-day rate of VTE.
A total of 1 295 291 patients, including 31 896 otolaryngology patients (mean [SD] age, 53.9 [16.7] years; 14 260 [44.7%] male; 21 603 [67.7%] white), 27 280 plastic surgery patients (mean [SD] age, 50.5 [13.9] years; 4835 [17.7%] male; 17 983 [65.9%] white), and 1 236 115 general surgery patients (mean [SD] age, 54.9 [17.2] years; 484 985 [39.2%] male; 867 913 [70.2%] white) were compared. The overall 30-day rate of VTE was 0.5% for otolaryngology compared with 0.7% for plastic surgery and 1.2% for general surgery. We identified a high-risk group for VTE in otolaryngology (n = 3625) that included free or regional tissue transfer, laryngectomy, composite resection, skull base surgery, and incision and drainage. High-risk otolaryngology patients experienced similar rates of VTE as general surgery patients across all Caprini risk levels. Low-risk otolaryngology patients (n = 28 271) experienced lower rates of VTE than plastic surgery patients across all Caprini risk levels. Malignant tumors were associated with VTE; however, the rates varied by cancer type and were 11-fold greater for cancers of the upper aerodigestive tract compared with thyroid cancers (odds ratio, 10.97; 95% CI, 7.38-16.31). Venous thromboembolism was associated with a 14-fold higher 30-day mortality among otolaryngology patients (5.1% mortality with VTE vs 0.4% mortality without VTE; difference, 4.7%; 95% CI of the difference, 2.2%-9.3%).
Conclusions and Relevance
Most patients undergoing otolaryngology procedures are at low risk of VTE, indicating that guidelines for a low-risk population could be adapted to otolaryngology. Patients undergoing high-risk otolaryngology procedures should be considered as candidates for more aggressive VTE prophylaxis.
Quiz Ref IDVenous thromboembolism (VTE) includes deep venous thrombosis (DVT) and pulmonary embolism (PE) and is estimated to occur at a 150-fold higher rate in hospitalized patients compared with community residents1; thus, it is the most common cause of preventable death in the hospital.2 There have been multiple initiatives in the United States to reduce the rate of VTE from the Centers for Medicare & Medicaid Services, The Joint Commission, and the Office of the Surgeon General of the United States,3 to the extent that the Centers for Medicare & Medicaid Services now consider appropriate VTE prophylaxis to be a pay-for-performance quality measure.4Quiz Ref ID In line with these efforts has been the development of evidence-based guidelines by the American College of Chest Physicians (ACCP) for the prevention of VTE in surgical patients.5 These guidelines provide sets of recommendations tailored to nearly all surgical specialties; however, no guidelines currently exist for otolaryngology–head and neck surgery.5
Previous single-institution studies have found that the rate of VTE in otolaryngology is significantly lower than that in the general surgical population.6-9 This relatively reduced risk of VTE after head and neck surgery may be attributable to patients being better able to ambulate after surgery of the head and neck and having shorter hospital stays compared with patients undergoing other types of surgery. The overall low rate of VTE observed in otolaryngology suggests that different recommendations may be necessary. Current ACCP guidelines for VTE prophylaxis rely on accurate risk stratification of patients to determine which patients require mechanical and/or pharmacologic prophylaxis. Risk stratification is recommended based on meta-analyses of randomized clinical studies5,10 of prophylactic low-dose subcutaneous heparin in general surgery, urology, and orthopedics, which show that its use is associated with an 18% reduction in overall mortality but a 57% increase in nonfatal major bleeding. Whereas pharmacologic prophylaxis is generally recommended for high-risk patients because of the greater mortality associated with PE, the associated increased risk of bleeding is of particular concern in postoperative head and neck surgery patients because hematoma formation in the neck and bleeding into the airway after aerodigestive tract surgery can cause airway compromise.
There is a need for a large, multicenter data set to better understand the risk stratification for VTE in otolaryngology before guidelines can be applied to this population. However, one challenge in adapting VTE guidelines to otolaryngology is the diversity of major and minor procedures performed in the field and a lack of normative data on the rates of VTE after different operations. To address this need, we used the ACCP guidelines for risk stratification (Table 1) to compare the rate of VTE in otolaryngology populations vs populations that are established to be at low risk (plastic surgery) or average risk (general surgery) using a large, multicenter clinical registry. The objective of this study was to develop normative data for the rate of VTE after specific otolaryngology procedures. Comparing these data to established-risk surgical fields could facilitate the development of more accurate VTE prevention recommendations.
We conducted a cohort study that compared the rate of VTE among otolaryngology patients with rates in an established average-risk group (general surgery) and low-risk group (plastic surgery) using prospectively collected data from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) from January 1, 2005, through December 31, 2013. The NSQIP is a multi-institutional surgical registry that collects short-term perioperative outcomes within 30 days of surgery. The detailed methods of the NSQIP have been described previously.11 Relevant to this study, the incidence of DVT or PE is an outcome variable that is prespecified and defined for data collection and that was prospectively ascertained in this cohort. Venous thromboembolism was considered to have occurred if a DVT or PE was diagnosed within 30 days of surgery. Deep venous thrombosis was diagnosed if a new thrombus was present on a duplex ultrasonogram, venogram, or lower-extremity computed tomographic (CT) scan. Pulmonary embolism was diagnosed if present on a chest CT scan, CT angiogram, or pulmonary arteriogram or if a V-Q scan was interpreted as showing a high probability of PE. The NSQIP does not record the type of perioperative VTE prophylaxis given, and protocols for VTE prophylaxis at the participating institutions may vary. The NSQIP data are deidentified in compliance with the Health Insurance Portability and Accountability Act. Data analysis was performed from May 1, 2016, to April 1, 2017. The Northwestern University Institutional Review Board determined that this study using deidentified data was exempt from institutional review board approval and informed consent.
We compared the rate of VTE based on the surgical specialty of the surgeon of the primary index procedure, including general surgery, plastic surgery, and otolaryngology. We included only patients admitted to the hospital after surgery because these are the primary candidates for mechanical and pharmacologic VTE prophylaxis postoperatively. To identify subgroups of procedures within otolaryngology that were associated with a high or low risk for VTE, we grouped otolaryngology patients based on type of procedure using the primary index Current Procedural Terminology code as listed in eTable 1 in the Supplement. Patients who underwent multiple operations were grouped based on the primary index procedure in the NSQIP and could only be counted once.
Quiz Ref IDThe Caprini score is a VTE risk factor assessment tool that uses 40 preoperative risk factors to estimate a patient’s risk of VTE and is recommended as a means of stratifying patients by risk for VTE in the ACCP guidelines.5 We calculated a Caprini score because a previous study12 validated adaptations of the Caprini score to the NSQIP with good reliability. The NSQIP does not collect information on some rare risk factors, including genetic or autoimmune disorders associated with coagulopathy, personal history of VTE, oral contraceptive or hormone replacement therapy or spontaneous abortion, immobilizing cast or lower-extremity fracture, central venous catheter, and swollen legs or varicose veins. Our adaptation of the Caprini score to the NSQIP would capture 96% of patient risk factors for VTE.12 Patients were assigned 5 points if they had a preoperative stroke within 1 month of surgery; 3 points if they were 75 years or older; 2 points if they were 61 to 74 years of age, their surgery lasted longer than 45 minutes, they had a malignant tumor, or they were immobilized with preoperative hospital admission for more than 3 days; and 1 point if they were 41 to 60 years of age, their surgery lasted longer than 45 minutes, their body mass index (calculated as weight in kilograms divided by height in meters squared) was greater than 25, they were pregnant, they had preoperative sepsis within 1 month, they had lung disease (including chronic obstructive pulmonary disease or preoperative pneumonia), they had acute myocardial infarction or congestive heart failure, or they were immobilized with preoperative hospital admission for 1 to 3 days before surgery.13 Secondary outcomes that were assessed included postoperative bleeding that required transfusion and 30-day mortality. Postoperative bleeding was assessed if a patient received a blood transfusion of packed red blood cells or whole blood within 72 hours after leaving the operating room.
The rates of VTE, the proportion of VTE occurring after hospital discharge, and the 30-day mortality rate for VTE were assessed separately. The 95% CIs were calculated for all comparisons. Caprini scores were used to adjust for differences in risk factors for VTE between groups, and patients were compared across individual strata of Caprini scores. We performed a multivariable logistic regression analysis using clinically relevant variables to determine the influence of individual variables on the rate of VTE. We included the Caprini score, length of hospital stay, and type of surgery as variables in our final model. Statistics were computed using SPSS, version 24 (IBM Inc).
We identified a total of 1 295 291 patients, including 31 896 otolaryngology (mean [SD] age, 53.9 [16.7] years; 14 260 [44.7%] male; 21 603 [67.7%] white), 27 280 plastic surgery (mean [SD] age, 50.5 [13.9] years; 4835 [17.7%] male; 17 983 [65.9%] white), and 1 236 115 general surgery patients (mean [SD] age, 54.9 [17.2] years; 484 985 [39.2%] male; 867 913 [70.2%] white). The demographic and clinical characteristics of the study population are given in Table 2, and the characteristics of the index surgical procedure are given in eTable 2 in the Supplement. As expected based on the different populations treated by the surgical disciplines included in the study, some notable differences were found, including a greater proportion of women undergoing plastic surgery and a greater proportion of patients with a diagnosis of cancer undergoing otolaryngology surgery. Most general surgical procedures were of the abdominal or pelvic cavity, most plastic surgery operations were of the breast or skin, and most otolaryngology procedures were of the upper aerodigestive tract or endocrine system.
Quiz Ref IDOtolaryngology patients had the lowest rate of VTE in the study population, with an overall 30-day rate of VTE of 0.5% for otolaryngology patients compared with 0.7% for plastic surgery patients and 1.2% for general surgery patients. Subgroups of surgical procedures within otolaryngology were determined to be high or low risk if the 95% CI for the rate of VTE was not significantly different from that in the respective comparison group. When otolaryngology patients were analyzed based on the category of surgical procedure, we identified a subset of high-risk procedures within otolaryngology, including free or regional tissue transfer, laryngectomy, composite resection, skull base surgery, and incision and drainage (3625 [11.4%] of all otolaryngology patients). High-risk otolaryngology patients had significantly greater mean lengths of hospital stay compared with low-risk patients (8.8 vs 2.7 days; mean difference, 6.1 days; 95% CI for the difference, 5.8-6.4 days). The rates of VTE for individual otolaryngology procedures compared with the overall rates of VTE in general and plastic surgery are given in Table 3. High-risk otolaryngology patients experienced similar rates of VTE as general surgery patients across all Caprini scores. Low-risk otolaryngology patients experienced significantly lower rates of VTE compared with plastic surgery patients across all Caprini scores.
The rates of VTE based on the type of cancer are given in eTable 3 in the Supplement. The rate of VTE among patients with thyroid cancer and salivary gland cancer was less than that among patients without cancer, whereas the rate of VTE among patients with upper aerodigestive tract, skin, or gastrointestinal tract cancer was greater than that among patients without cancer. Otolaryngology patients were more likely to have active cancer at the time of surgery (38.0% for otolaryngology vs 21.2% for general surgery and 20.0% for plastic surgery). There was an 11-fold difference in the rates of VTE in patients with cancers of the upper aerodigestive tract compared with thyroid cancer (odds ratio [OR], 10.97; 95% CI, 7.38-16.31).
On multivariable logistic regression analysis, we identified that the type of surgery, Caprini score, and length of hospital stay were associated with the rate of VTE (Table 4). After adjusting for Caprini score and length of stay, plastic surgery (OR, 1.30; 95% CI, 1.02-1.66) and general surgery (OR, 1.39; 95% CI, 1.14-1.70) but not high-risk otolaryngology (OR, 1.07; 95% CI, 0.77-1.50) operations were associated with an increased rate of VTE compared with low-risk otolaryngology operations. Quiz Ref IDBecause of concerns that VTE could be both a cause and an effect of prolonged length of hospital stay, we reran the multivariable logistic regression analysis without the length of stay term. In this analysis, type of surgery and Caprini score were also associated with the rate of VTE. However, high-risk otolaryngology (OR, 3.15; 95% CI, 2.26-4.38), general surgery (OR, 3.19; 95% CI, 2.63-3.88), and plastic surgery (OR, 2.35; 95% CI, 1.85-3.00) were associated with an increased rate of VTE compared with low-risk otolaryngology.
The rate of postoperative bleeding that required transfusion was higher for high-risk otolaryngology (15.2%; 95% CI, 14.1%-16.4%) compared with other surgical groups (general surgery: 4.3% [95% CI, 4.2%-4.3%]; plastic surgery: 4.6% [95% CI, 4.3%-4.8%]; and low-risk otolaryngology: 1.6% [95% CI, 1.5%-1.8%]) (Table 5). Specifically, major head and neck surgery, which included free or regional tissue transfer, laryngectomy, composite resection, and skull base surgery, had significantly greater rates of bleeding that required transfusion. Low-risk otolaryngology procedures had a significantly reduced rate of bleeding that required transfusion compared with general or plastic surgery procedures.
Venous thromboembolism was associated with a significantly higher 30-day mortality rate for all surgical specialties (5.1% mortality with VTE vs 0.4% mortality without VTE in otolaryngology [difference, 4.7%; 95% CI of the difference, 2.2%-9.3%]; 2.0% mortality with VTE vs 0.4% without VTE in plastic surgery [difference, 1.7%; 95% CI of the difference, 1.2%-3.1%]; and 7.5% mortality with VTE vs 1.9% without VTE in general surgery [difference, 5.6%; 95% CI of the difference, 5.2%-6.0%]). Most VTE events were diagnosed while the patient was an inpatient for all specialties (152 [96.4%] as inpatient for otolaryngology, 186 [94.3%] as inpatient for plastic surgery, and 14 691 [96.6%] as inpatient for general surgery). For all specialties, most VTE events were DVT (10 254 [67.4%] for general surgery, 127 [64.4%] for plastic surgery, and 91 [57.4%] for otolaryngology).
Despite advances in mechanical and chemical prophylaxis, VTE remains a significant cause of postoperative morbidity and mortality. We provide indirect evidence in a large, multicenter cohort that otolaryngology patients overall represent a low-risk group for VTE; however, the consequences of VTE were also severe in this group, with a 14-fold increased risk of 30-day mortality among those with VTE. Within otolaryngology, we also identified a high-risk subgroup that experienced high rates of VTE and bleeding who require a thoughtful approach to balance the potential benefits of chemoprophylaxis with the risk of bleeding complications. Given the morbidity caused by VTE and the challenge in identifying the patients who would benefit from chemoprophylaxis, there is a need to develop specialized guidelines for the prevention of VTE in otolaryngology.
Varying rates of VTE have been reported for otolaryngology. A retrospective study14 of 1018 patients with head and neck cancer identified an overall 0% rate of VTE and an 8.7-fold higher rate of hematoma with pharmacologic prophylaxis. Other studies15,16 have reported moderate overall rates of VTE of 1.2% to 1.3%. A single-center study16 from Michigan of 2016 otolaryngology inpatients who did not receive pharmacologic prophylaxis for VTE found that patients with a Caprini score of 6 or lower had a 0.5% rate of VTE vs 2.4% among those with a Caprini score of 7 to 8 and 18.3% among those with a Caprini score greater than 8. In a follow-up study,15 the same institution found that among patients who received pharmacologic prophylaxis, only those with Caprini scores of 7 or higher had a significant reduction in the rate of VTE. Similarly, a retrospective review17 of 704 patients admitted to an academic otolaryngology service in which all patients received pharmacologic prophylaxis found that 3% of patients with a Caprini score of 7 to 8 and 13% with a score of 9 or higher developed VTE. In the only prospective study of VTE in otolaryngology that we are aware of, Clayburgh et al18 reported a high 13% overall incidence of VTE. In this study,18 when stratified by Caprini score, patients with a score higher than 8 had a 33% rate of VTE, and those with a score of 5 to 8 had a 9% rate of VTE. These rates differ significantly from other studies in otolaryngology according to a meta-analysis by Moubayed et al,19 which may relate to the population of patients undergoing only major head and neck surgery (80% underwent free tissue transfer with a mean length of stay of 9 days), universal lower-extremity Doppler screening on postoperative days 2 to 3, and rare use of chemoprophylaxis (only 14% of patients with chemoprophylaxis). The variety in observed rates of VTE in these studies14-18 reflects heterogeneity of the otolaryngology patient population and varying rates of screening for VTE because many DVTs are asymptomatic.20 Overall, the reported rates of VTE for otolaryngology were lower than have typically been observed for other surgical populations.5
In accordance with these prior findings, we identified that the rate of VTE in otolaryngology was low compared with that in other surgical disciplines. Furthermore, we identified a subgroup of otolaryngology procedures that were associated with a higher risk for VTE, thus facilitating improvements in risk stratification. We hypothesize that greater length of hospital stay contributed to the higher rates of VTE observed in the high-risk otolaryngology group. In support of this, on multivariable logistic regression, after adjusting for Caprini score and length of hospital stay, there was no difference between the rates of VTE among patients undergoing high- vs low-risk otolaryngology operations. However, VTE may have contributed to the longer length of hospital stay observed.
Maintaining a balance between VTE prevention and hemostasis is challenging. Although Caprini scores were designed to identify patients at high risk of VTE who may benefit from chemoprophylaxis, we observed that high Caprini scores were also associated with increased risk of postoperative bleeding. Patients with high Caprini scores may have also been more likely to receive chemoprophylaxis, which may have contributed to the elevated risk of bleeding observed. Higher levels of evidence from more prospective studies in otolaryngology would be useful to more accurately determine the risks and benefits of chemoprophylaxis in this high-risk group.
In addition to the ACCP guidelines, there are overlapping recommendations from the American Society of Clinical Oncology for the prevention of VTE in patients with cancer that we considered applying to otolaryngology.21 These guidelines recommend that “most hospitalized patients with active cancer require thromboprophylaxis (pharmacologic anticoagulation) throughout hospitalization” and that “a patient undergoing major cancer surgery should receive prophylaxis starting before surgery and continuing for at least 7-10 days,”21(pp 654-655) including after discharge. The American Society of Clinical Oncology VTE prevention guidelines have not been specifically investigated in otolaryngology patients. We found that otolaryngology patients, although at low risk for VTE, were the most likely to have active cancer at the time of surgery. When assessed based on cancer status, however, we found that otolaryngology patients with active cancer had significantly lower rates of VTE (0.8%) than did general surgery patients (1.8%). This finding raises the question of whether these recommendations apply to otolaryngology patients. Furthermore, we found that the risk of VTE in several populations with cancer, including those with thyroid and salivary gland cancer, was less than that among patients without cancer. These findings further challenge the applicability of the American Society of Clinical Oncology guidelines to all forms of cancer and limit their application in otolaryngology.
When interpreting the rates of VTE observed in our study, investigators should consider that many patients received mechanical and/or pharmacologic prophylaxis, which likely contributed to the observed low rates of VTE in all surgical groups. As such, one limitation of our study is that we were unable to adjust for the use of mechanical and pharmacologic prophylaxis. However, the rates of VTE prophylaxis may have been similar, allowing for comparison of these groups. In addition, several instrumental studies5,22 on VTE have used data sets with missing information on mechanical and/or chemoprophylaxis. Previous studies23,24 in which all patients received mechanical prophylaxis have observed similar rates of VTE as our own study. The Venous Thromboembolism Prevention Study, which is the current basis for the guidelines for plastic surgery, was a retrospective multi-institutional study of 1126 plastic surgery patients from 2006 to 2009 in which all patients received mechanical prophylaxis and rates of VTE were found that were similar to those in our study stratified by Caprini score.23 Furthermore, the prospective data collection and rigorous methods used in the NSQIP provided a tool to assess VTE and identifed a 47% greater incidence of VTE compared with other administrative data sources24; thus, we believe that it is unlikely that the rates of VTE are low secondary to errors in data collection. Together our findings provide evidence that most otolaryngology patients are at very low risk of VTE, whereas there is a small subgroup of patients undergoing otolaryngology operations that confer greater risk. A foundational principle of the current VTE guidelines is stratification of patients by risk for VTE, and thus distinguishing low-risk from high-risk otolaryngology procedures allows patients undergoing these diverse operations to be stratified appropriately.
Despite these limitations, the NSQIP offers a powerful tool to assess the risk of VTE in otolaryngology and allows comparison with surgical specialties with established guidelines for VTE prevention. We believe that this facilitates a rational basis to adapt existing ACCP guidelines to otolaryngology that are specific to patient and surgical characteristics. We propose that VTE prevention recommendations for the low-risk population of plastic surgery patients could be extended to low-risk otolaryngology operations, whereas recommendations for an average-risk population from general surgery could be extended to high-risk otolaryngology operations. We acknowledge that given the lack of information on VTE prophylaxis in the present study, these conclusions have significant limitations that must be verified in future studies.
In addition to the aforementioned limitations, several others should be acknowledged when considering our findings. First, smilar to all large clinical registries, the NSQIP may contain inaccuracies; however, rigorous methods were used to minimize this possibility. Second, the NSQIP does not record some of the rarer risk factors that are taken into account when calculating a Caprini score. Given the known implications of these factors on the risk for VTE, it is plausible that some patients had an underestimated Caprini score. An additional limitation is that although the NSQIP records the influence of VTE within 30 days of surgery, it is well established that the influence of VTE can last longer.25 One additional challenge with VTE is that it is an uncommon complication and requires large populations of patients to study its effects. Given the large sample size of the groups in our study, even small differences among the groups were likely to be statistically significant, and one must interpret the clinical significance of these findings with some caution.
We provide normative data on the rate of VTE for various procedures in otolaryngology and found that the rates of VTE in otolaryngology were low for most otolaryngology operations. We further identified a subgroup of operations within otolaryngology that were associated with a higher risk of VTE, including free or regional tissue transfer, laryngectomy, composite resection or skull base surgery, and incision and drainage. We propose a method to adapt the existing ACCP guidelines to otolaryngology that provides a framework for future research.
Corresponding Author: John D. Cramer, MD, Department of Otolaryngology–Head and Neck Surgery, Northwestern University Feinberg School of Medicine, 675 N St Clair St, Ste 1325, Chicago, IL 60611 (firstname.lastname@example.org).
Accepted for Publication: July 17, 2017.
Published Online: October 19, 2017. doi:10.1001/jamaoto.2017.1768
Author Contributions: Dr Cramer had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Cramer, Schneider, Smith, Samant, Patel.
Acquisition, analysis, or interpretation of data: Cramer, Dilger, Schneider, Smith, Samant.
Drafting of the manuscript: Cramer, Dilger, Schneider, Samant.
Critical revision of the manuscript for important intellectual content: Cramer, Dilger, Smith, Samant, Patel.
Statistical analysis: Cramer, Dilger, Samant.
Obtained funding: Smith.
Administrative, technical, or material support: Cramer, Schneider.
Study supervision: Cramer, Smith, Samant, Patel.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: Dr Smith is supported by grant K12HS023011 from the Northwestern University Patient-Centered Intervention and Engagement Training K12 Faculty Scholars Training Program and grant U19 AI106683 from the Chronic Rhinosinusitis Integrative Studies Program.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Disclaimer: The American College of Surgeons National Surgical Quality Improvement Program (NSQIP) and the hospitals participating in the NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.