Customize your JAMA Network experience by selecting one or more topics from the list below.
Chen MM, Roman SA, Sosa JA, Judson BL. Safety of Adult Tonsillectomy: A Population-Level Analysis of 5968 Patients. JAMA Otolaryngol Head Neck Surg. 2014;140(3):197–202. doi:10.1001/jamaoto.2013.6215
Tonsillectomy is one of the most commonly performed otolaryngology procedures. The safety of this procedure in adults is based on small case series. To our knowledge, we report the first population-level analysis of the safety of adult tonsillectomies in the United States.
To characterize the mortality, complication, and reoperation rate in adult tonsillectomy.
Design, Setting, and Participants
Retrospective cohort study of 5968 adult patients who underwent tonsillectomy with records in the database of the American College of Surgeons National Surgical Quality Improvement Program (2005 to 2011).
Main Outcomes and Measures
Outcomes of interest included mortality, complications, and reoperation in the 30-day postoperative period. Statistical analysis included χ2 test, t test, and multivariate logistic regression.
The 30-day mortality rate was 0.03%, the complication rate was 1.2%, and the reoperation rate was 3.2%. Most patients had a primary diagnosis of chronic tonsillitis and/or adenoiditis (82.9%), and the most common complications were pneumonia (27% of all complications), urinary tract infection (27%), and superficial site infections (16%). Patients who underwent reoperation were more likely to be male (54.0% vs 32.4%; P < .001), white (84.8% vs 75.3%; P = .02), or inpatients (24.3% vs 14.3%; P < .001) and to have postoperative complications (5.3% vs 1.1%; P < .001) than those who did not return to the operating room. On multivariate analysis, male sex (odds ratio [OR], 2.30 [95% CI, 1.67-3.15]), inpatient status (OR, 1.52 [95% CI, 1.04-2.22]), and the presence of a postoperative complication (OR, 4.58 [95% CI, 2.11-9.93]) were independent risk factors for reoperation.
Conclusions and Relevance
In the United States, adult tonsillectomy is a safe procedure with low rates of mortality and morbidity. The most common posttonsillectomy complications were infectious in etiology, and complications were independently associated with the need for reoperation.
Tonsillectomy is one of the most commonly performed surgical procedures in the United States. In 2006, the Centers for Disease Control and Prevention reported 297 000 tonsillectomies performed on patients 15 years or older out of 737 000 total tonsillectomies.1 There are limited data examining the safety of tonsillectomy in adults in the United States, with most studies examining the pediatric population or a population made up of both children and adults. Tonsillectomy has been estimated to have a mortality rate of 1 death in every 20 000 procedures.2 In the pediatric literature, tonsillectomy is associated with a 0.5% to 2.1% rate of reoperation.3,4 In single-institution studies, adult patients have been shown to be at higher risk for hemorrhage when compared with pediatric patients, with 1 study reporting a 2.2% rate of reoperation.3,4
The current policy and economic climate has resulted in substantial focus on the economic burden of hospital readmission and reoperation. Under the Affordable Care Act’s Hospital Readmissions Reduction Program, hospitals with excess readmissions will incur financial penalties.5 Given the number of tonsillectomies that are performed annually, it is important to evaluate the rate of complications and reoperation in this population.
There is a paucity of data investigating the safety of tonsillectomy in adult patients. To our knowledge, this is the first population-level analysis of the reoperation rate after adult tonsillectomy in the United States. The aims of our study were to measure mortality, complication, and reoperation rates after tonsillectomy in adult patients and to assess risk factors associated with reoperation.
Quiz Ref IDThe American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) was interrogated to identify adult patients (aged ≥18 years) who underwent tonsillectomy during the years 2005 through 2011. The ACS-NSQIP is a national data set with more than 400 participating hospitals. At each site, a trained surgical clinical reviewer abstracts data on preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity outcomes. Detailed methodology of the ACS-NSQIP database has been described elsewhere and is available from the Participant Use Data File user guide.6
Patients who underwent either tonsillectomy or tonsillectomy and adenoidectomy (T&A) were identified using Current Procedural Terminology codes: 42826 (tonsillectomy) and 42821 (T&A). Patients who had concurrent septoplasty, uvulectomy, pharyngoplasty, rhinoplasty, or lymph node excision and/or biopsy were excluded from the analysis.
Demographic variables included patient sex, age, and race. Race was classified as white, black, and other (Hispanic, Asian, Native Hawaiian or Pacific Islander, American Indian or Alaskan Native).
Patient comorbidities included increased body mass index (BMI, calculated as weight in kilograms divided by height in meters squared), diabetes mellitus, tobacco use, alcohol use, dyspnea, hypertension, chronic steroid use, current wound infection, and bleeding disorder. Tobacco use was defined as any smoking of cigarettes in the year before surgery. Alcohol use was defined as at least 2 drinks per day in the 2 weeks prior to surgery. Dyspnea was categorized as at rest, with moderate exertion, and none.
Clinical variables included inpatient vs outpatient status, presence of a resident during surgery, emergent or nonemergent surgery, prolonged operative time, American Society of Anesthesiologists classification, hospital length of stay (LOS), any postoperative complication, and reoperation. Inpatients were defined as any patient who stayed in the hospital for at least 1 day. Prolonged operative time was defined as any operative time greater than the 75th percentile in this cohort. For inpatients, hospital LOS was categorized as 1 day and more than 1 day (prolonged hospitalization). Postoperative complications included surgical site complications, infections (pneumonia, urinary tract infection [UTI], sepsis, or shock), prolonged ventilator use (>48 h), unplanned reintubation, venous thromboembolism, renal insufficiency or failure, and substantial bleeding requiring more than 4 units of packed red blood cells (PRBCs). Surgical site complications were defined as infections that seem related to the surgical procedure and include superficial site infections (associated with fever, localized swelling, pain, or tenderness), deep site infections (associated with purulent drainage, abscess, or evidence of involvement of deep tissues on radiologic examination or reoperation), and wound disruption. Our primary outcomes of interest were rates of mortality, complications, and reoperations within 30 days of surgery.
Bivariate analysis was used to compare the baseline demographic and clinical characteristics of patients who underwent reoperation in the 30-day postoperative period following tonsillectomy and those who did not. The χ2, Fisher exact, and t tests were used to analyze categorical and continuous variables, respectively. Multivariate logistic regression analysis was used to identify risk factors associated with reoperation. Odds ratios (ORs) and 95% confidence intervals (CIs) were determined for the strength of association between each risk factor and reoperation. All tests were 2 sided, and P < .05 was considered statistically significant.
Data analysis was performed using Statistical Package for the Social Sciences (SPSS) software, version 21.0 (SPSS). The ACS-NSQIP data are publicly available, and patient information is deidentified; thus, this study was granted an exemption from the Yale University institutional review board.
A total of 5968 adult patients who underwent tonsillectomy were identified between 2005 and 2011. The most common primary diagnosis was chronic tonsillitis and/or adenoiditis (82.9%). There were approximately twice as many women as men (Table 1).
Quiz Ref IDAmong all patients, the 30-day mortality rate was 0.03% and the 30-day postoperative complication rate was 1.2%. Among inpatients, there was a 1.5% predischarge complication rate, a 1.1% postdischarge complication rate, and a 2.5% overall postoperative complication rate. The overall rate of reoperation in the 30-day postoperative period was 3.2%.
Quiz Ref IDThe most common postoperative complications experienced in the first 30 days after surgery were due to infections (58%) and surgical site complications (27%). The most prevalent infections were pneumonias (27% of all complications) and UTIs (27%), and the most common category of surgical site complication was superficial site infections (16%) (Table 2). There was a low rate of substantial bleeding necessitating transfusion of more than 4 units of PRBCs, representing 7% of all complications and 0.08% of all patients. Among inpatients, most infections and venous thromboembolisms occurred after discharge, whereas respiratory complications such as unplanned reintubation and prolonged ventilator use primarily occurred prior to discharge. Approximately 42% of all postdischarge complications occurred within the first week after discharge.
Quiz Ref IDIn comparison with those who did not undergo reoperation, patients who had a reoperation were more likely to be male (54.0% vs 32.4%; P < .001), to be white (84.8% vs 75.3%; P = .02), and to have had inpatient surgery (24.3% vs 14.3%; P < .001). Patients who experienced a reoperation were twice as likely to have peritonsillar abscesses than those who did not return to the operating room (4.2% vs 2.0%; P = .06). Among the inpatients, those who had a hospital LOS of more than 1 day were 3-fold more likely to have a reoperation than those with an LOS of 1 day (11.2% vs 3.7%; P < .001). Patients who underwent reoperation were more than 5-fold more likely to have had postoperative complication than those who did not undergo reoperation (5.3% vs 1.1%; P < .001). We did not detect a significant difference in the rates of reoperation between those procedures performed by an attending physician and a resident and those performed by an attending physician alone (3.9% vs 3.1%; P = .45).
Multivariate analysis demonstrated that male sex (OR, 2.30 [95% CI, 1.67-3.15]), inpatient status (OR, 1.52 [95% CI, 1.04-2.22]), and the presence of a postoperative complication (OR, 4.58 [95% CI, 2.11-9.93]) continued to be independent risk factors for reoperation (Table 3).
To our knowledge, our study represents the first population-level analysis of mortality, complications, and reoperations following adult tonsillectomy in the United States. The 30-day mortality rate was low, indicating the safety of adult tonsillectomy. The most common postoperative complications were infections, including pneumonias, UTIs, and surgical site infections. Reoperation was independently associated with male sex, occurrence of a postoperative complication, and inpatient tonsillectomies.
We determined a mortality rate of 0.03%. Given the rarity of posttonsillectomy mortality, there are limited data on a population level regarding mortality after adult tonsillectomy in the United States. A recent survey of 552 otolaryngologists determined an extrapolated mortality rate of 1 in 27 000, or 0.004%.7 Prior European studies have reported mortality rates after adult and pediatric tonsillectomy ranging from 1 in 10 0008 to 1 in 33 921.9 Our study confirms, on a population level, that adult tonsillectomy in the United States is a safe procedure with a low mortality rate.
We observed a 3.2% 30-day reoperation rate, which validates previously published reports in the literature. A recent single-institution study investigating indications for adult tonsillectomy by Hoddeson and Gourin10 demonstrated a 2.2% reoperation rate. The majority of the current literature on reoperation or readmission after tonsillectomy has been limited to pediatric patients3,4,11 or mixed cohorts of both adult and pediatric patients.12-14 The reoperation rate in the pediatric literature has been shown to be 0.5% to 2.1%.3,4 In a national study of 33 921 patients in the United Kingdom, Lowe et al13 demonstrated a 0.9% reoperation rate. Sarny et al12 conducted a multicenter study of 9405 adult and pediatric patients in Austria and determined that the rate of reoperation for tonsillectomy or T&A was 4.6%. Windfuhr et al14 examined 15 218 adult patients in Germany and reported a reoperation rate of 2.9% after tonsillectomy.
Our finding that rates of reoperation were not significantly affected by whether a resident performed the procedure validates the results of a recent retrospective study of 364 adult patients by Muratori et al,15 who demonstrated that there was no difference in complications and the risk for postoperative bleeding between patients operated on by a senior otolaryngologist and those who were operated on by a supervised resident. These findings are also supported in the pediatric otolaryngology literature.16,17 In a review of 4776 pediatric patients, Gallagher et al16 reported that resident surgeon status was not associated with postoperative hemorrhage or reoperation after tonsillectomy. Kim et al18 demonstrated that the risk of postoperative hemorrhage was 1.9-fold greater for tonsillectomies performed by residents, but this difference was not significant after adjustment in their multivariate analysis.
The most common indication for surgery in our population was chronic tonsillitis. This corroborates prior institution-level data, which demonstrated that chronic infection (57%) was the most common indication for tonsillectomy in their institution, followed by upper airway obstruction (27%) and suspected neoplasm (16%).10 We did not observe a statistically significant association between primary diagnosis and risk of reoperation. A previous study by Dünne et al19 reviewed 142 patients after tonsillectomy for peritonsillar abscess and found that 11% experienced postoperative hemorrhage that had to be treated surgically. We did not see a significant association between peritonsillar abscess and substantial bleeding. A study by Suzuki et al20 comparing 103 patients who underwent immediate tonsillectomy for peritonsillar abscess and 99 patients who underwent routine tonsillectomy for chronic tonsillitis did not see any difference in postoperative bleeding between the 2 groups.
Quiz Ref IDIn our cohort, twice as many women as men underwent tonsillectomy. The 2006 National Survey of Ambulatory Surgery reported that females had significantly more ambulatory surgery visits than males and were 1.3 times more likely to undergo tonsillectomy, indicating that our results were representative of national trends.1 We also observed that male sex was independently associated with reoperation. Some single-institution studies in the literature reported no association between sex and postoperative hemorrhage,17,18 whereas others demonstrated that male sex was associated with increased risk of postoperative bleeding.14,21 The multicenter Austrian Tonsil Study found that male sex was independently associated with postoperative hemorrhage (OR, 1.32 [95% CI, 1.12-1.56]).12 In the present study, we did not find an association between sex and substantial postoperative bleeding.
We determined that the overall rate of substantial bleeding in our cohort was 0.08%. A recent analysis of 361 adult patients demonstrated a 5% rate of postoperative hemorrhage.10 In studies looking at both adult and pediatric patients, rates of postoperative bleeding range from 1.5% to 15%.2,12,14,18 Our rate of substantial bleeding only captured postoperative bleeding requiring transfusion of more than 4 units of PRBCs, so the total postoperative bleeding rate in our cohort was likely closer to the reoperation rate of 3.2%.
The rate of pneumonia in the 30-day postoperative period was 0.3%, and pneumonias and UTIs were the most common postoperative complications in our cohort. There is a paucity of data investigating postoperative complications other than hemorrhage after tonsillectomy. Because we focused on adult tonsillectomies, our cohort is more likely to have baseline comorbid dyspnea (2.5%) and respiratory disease than combined adult and pediatric cohorts. As a result of these comorbidities, our cohort may also be more prone to require postoperative mechanical ventilation that may be associated with pneumonia. A recent article by Sun et al22 demonstrated that the rate of mechanical ventilation after pediatric and adult tonsillectomy in 80 hospitals in the United States was 3% and that the use of mechanical ventilation increased costs by more than $30 000 per index hospitalization.
Limitations of our study include those inherent to administrative databases, including coding errors; however, the ACS-NSQIP database is well validated and undergoes regular audits.23 The timing of reoperation was not classified in the database, so we were unable to specify the timing of reoperation within the 30-day postoperative period. A tonsillectomy leaves an open wound that may be subject to interpretation as infection according to the ACS-NSQIP criteria even when this may not be the case. The exact nature of surgical site infections, as recorded in the ACS-NSQIP database, is unclear with respect to tonsillectomy, and additional information about the accuracy of the diagnosis of infection is unavailable. Investigators who pursue further research in posttonsillectomy complications should collect additional information that better clarifies the presence of a posttonsillectomy infection. Information related to postoperative hemorrhage that did not require transfusion of more than 4 units of PRBCs, disease severity, costs, surgeon volumes, hospital-level factors, outcomes at free-standing outpatient centers, and long-term postoperative outcomes were not recorded in ACS-NSQIP and could not be analyzed.
To our knowledge, this is the first population-level study of the safety of adult tonsillectomy in the United States. Adult patients who undergo tonsillectomy in the United States have a low risk of reoperation and mortality similar to that in the pediatric population. Postoperative complications were independently associated with increased risk of reoperation, and the majority of postoperative complications were infectious in etiology (pneumonia, UTI, and surgical site infections). Given the database used, it is possible that this analysis overstates the number of site infections and understates the number of postoperative hemorrhages. Prior meta-analyses and systematic review have demonstrated that there is no evidence in favor of routine administration of perioperative antibiotics in the prevention of posttonsillectomy morbidity.24 Our data indicate that adult tonsillectomy in the United States is a safe procedure; however, additional research into the prevention of postoperative infection in high-risk patients may be useful in continuing to reduce reoperation rates.
Submitted for Publication: August 12, 2013; final revision received October 25, 2013; accepted November 7, 2013.
Corresponding Author: Benjamin L. Judson, MD, Department of Surgery, Yale University School of Medicine, 800 Howard Ave, YPB 425, New Haven, CT 06519 (email@example.com).
Published Online: January 30, 2014. doi:10.1001/jamaoto.2013.6215.
Author Contributions: Ms Chen and Dr Judson had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition of data: Chen.
Analysis and interpretation of data: All authors.
Drafting of the manuscript: Chen.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Chen.
Obtained funding: Chen.
Administrative, technical, and material support: Judson.
Study supervision: Roman, Sosa, Judson.
Conflict of Interest Disclosures: None reported.
Funding/Support: This project was supported by the James G. Hirsch, MD, Endowed Medical Student Research Fellowship at Yale University School of Medicine.
Role of the Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS-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.