Postdischarge Complications Predict Reoperation and Mortality after Otolaryngologic Surgery

Transcription

1 Original Research General Otolaryngology Postdischarge Complications Predict Reoperation and Mortality after Otolaryngologic Surgery Otolaryngology Head and Neck Surgery 149(6) Ó American Academy of Otolaryngology Head and Neck Surgery Foundation 2013 Reprints and permission: sagepub.com/journalspermissions.nav DOI: / Michelle M. Chen 1, Sanziana A. Roman, MD 2, Julie A. Sosa, MD 2, and Benjamin L. Judson, MD 1 Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. Abstract Objectives. (1) Determine procedure-specific rates of postdischarge complications (PDCs) and their risk factors in the first 30 days following inpatient otolaryngologic surgery. (2) Evaluate association between PDCs and risk of reoperation and mortality. Study Design. Retrospective cohort study. Setting. American College of Surgeons National Surgical Quality Improvement Program ( ). Subjects and Methods. We identified 48,028 adult patients who underwent inpatient otolaryngologic surgery. Outcomes of interest included complications, reoperation, and mortality in the first 30 days following surgery. Statistical analysis included chi-square, t tests, and multivariate regression. Results. Laryngectomy, lip, and tongue/floor of mouth surgery had the highest PDC rates (8.0%, 7.4%, and 4.1%, respectively). Within the first 48 hours, week, and 2 weeks post discharge, 10%, 44%, and 73% of PDCs occurred, respectively. Common PDCs included surgical site infections (53.6%), other infections (37.4%), and venous thromboembolic events (7.4%). Multivariate analysis demonstrated that increasing age (odds ratio [OR] = 1.01; 95% confidence interval [CI], ), prolonged operative time (OR = 1.68; 95% CI, ), hospital stay.1 day (OR = 1.49; 95% CI, ), and American Society of Anesthesiologists (ASA) class 3 (OR = 1.45; 95% CI, ) were independently associated with PDCs. Patients with PDCs were more likely to die (0.9% vs 0.1%, P \.001) or have a reoperation (10.4% vs 1.2%, P \.001). Conclusion. This is the first study of overall postdischarge events after otolaryngologic surgery. PDC rates in otolaryngology occur soon after discharge, are procedure specific, and are associated with reoperation and mortality. Targeted procedure-specific triage and follow-up plans for high-risk patients may improve outcomes. Keywords quality, outcomes, complications, NSQIP, post-discharge Received June 3, 2013; revised August 6, 2013; accepted August 23, Introduction Hospital readmissions have been shown to be costly and associated with lower quality of inpatient care. 1,2 Since October 2012, the Hospital Readmissions Reduction Program section of the Affordable Care Act has required the Centers for Medicare and Medicaid Services to levy a financial penalty on hospitals with excess readmissions. 3 Surgical patients have been shown to be at particularly high risk for readmission. 2,4 Otolaryngologic readmissions are associated with higher costs and increased mortality. 5 One factor that may contribute to hospital readmissions and reoperations are postdischarge complications (PDC). Much of the recent literature on postoperative outcomes in otolaryngology have been limited to studies of inpatient morbidity and mortality. 6,7 Recent studies in the general surgery literature have shown that 33% to 42% of postoperative complications occur after discharge. 8,9 Most quality improvement programs only analyze data on in-hospital adverse events, and inclusion of postdischarge data has been shown to change overall hospital quality rankings. 9 There is a paucity of literature on PDCs in otolaryngologic surgery. In order to have effective and targeted quality improvement interventions, it is necessary to assess 1 Department of Surgery, Yale University School of Medicine, New Haven, Connecticut, USA 2 Department of Surgery, Duke University School of Medicine, Durham, North Carolina, USA This article was presented at the 2013 AAO-HNSF Annual Meeting & OTO EXPO; September 29 October 3, 2013; Vancouver, BC, Canada. Corresponding Author: Benjamin L. Judson, MD, 800 Howard Avenue, YPB 425, Department of Surgery, Yale University School of Medicine, New Haven, CT 06519, USA. benjamin.judson@yale.edu

2 866 Otolaryngology Head and Neck Surgery 149(6) procedure-specific rates of complications, unplanned reoperations, and mortality. To our knowledge, this is the first study of overall procedure-specific PDCs after inpatient otolaryngologic surgery. The purposes of our study were: (1) to determine procedure-specific rates of PDCs and risk factors associated with PDCs and (2) to evaluate the association between experiencing a PDC and the risk of reoperation or mortality. Methods Data Source and Study Subjects This study used data from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database for the years 2005 through Demographic, surgical, preoperative, intraoperative, and 30- day postoperative data are collected by centrally trained clinical reviewers at more than 250 sites across the country. A detailed description of the methods of the NSQIP database has been described elsewhere 10 and is available from the Participant Use Data File user guide. 11 Adult patients (aged 18 years) who underwent 1 of 8 groups of inpatient otolaryngologic procedures were identified in the database. An inpatient procedure was defined as any procedure for which the patient stayed in the hospital for at least 1 day. Primary otolaryngology procedures were identified using Current Procedural Terminology codes. We grouped common otolaryngologic procedures into 8 main categories: (1) laryngectomy (total, supraglottic, partial, or pharyngolaryngectomy), (2) neck dissection (suprahyoid, cervical, or modified radical), (3) lip (excision or resection), (4) tongue/ floor of mouth (tongue excision, glossectomy, or composite procedure), (5) palate (uvulectomy or palatopharyngoplasty), (6) salivary gland (parotid, submandibular, or sublingual gland), (7) tonsillectomy/adenoidectomy, and (8) thyroid (lobectomy, total thyroidectomy) (Table 1). Patient demographics included age, gender, race, prehospitalization residence, functional status, and discharge destination. Race was defined as white, black, and other (Hispanic, Asian, Native Hawaiian or Pacific Islander, American Indian or Alaskan Native). Prehospitalization residence was classified into home (including patients arriving from another hospital s emergency department), acute care hospital (inpatient), chronic care facility, and other. Functional status characterized the patient s ability to perform activities of daily living in the 30 days prior to surgery and was grouped into independent, partially dependent, and totally dependent. Discharge destination was categorized into home, facility, separate acute care, and rehab. Patient comorbidities included body mass index (BMI), diabetes, smoker, alcohol use, dyspnea, history of severe chronic obstructive pulmonary disease (COPD), current pneumonia, heart disease, cerebrovascular disease (CVA), sepsis, disseminated cancer, hypertension, and chronic steroid use. Smoker was defined as any patient who had smoked cigarettes in the year prior to admission for surgery. Table 1. Procedure codes for included inpatient otolaryngologic procedures. Procedure Group Current Procedural Terminology Codes Laryngectomy Total 31360, Subtotal supraglottic 31367, Partial 31370, 31375, 31380, Pharyngolaryngectomy 31390, Neck dissection Suprahyoid Cervical Modified radical Lip Excision 40510, 40520, 40525, Resection Tongue/floor of mouth Excision of tongue 41110, 41112, 41113, 41114, Glossectomy 41120, 41130, 41135, 41140, Composite procedure 41150, 41153, Palate Uvulectomy Palatopharyngoplasty Salivary gland Superficial parotidectomy Total parotidectomy 42415, 42420, 42425, Submandibular gland Sublingual gland Tonsillectomy/adenoidectomy Tonsillectomy 1 adenoidectomy Tonsillectomy Adenoidectomy Thyroid Lobectomy 60220, Total thyroidectomy 60240, 60252, 60254, 60260, 60270, Alcohol use was defined as 2 or more drinks per day in the 2 weeks prior to admission. Dyspnea was grouped into at rest, with moderate exertion, and none. Heart disease included congestive heart failure (newly diagnosed or an acute exacerbation within 30 days of surgery), history of myocardial infarction in the 6 months prior to surgery, previous percutaneous coronary intervention or cardiac surgery, or any history of angina in the month prior to surgery. CVA included any patient with history of a transient ischemic attack or cerebrovascular accident. Sepsis included any patient that meets criteria for systemic inflammatory response syndrome, sepsis, or septic shock. Operative variables included surgeon specialty, presence of a resident, emergent or nonemergent surgery, operative time, American Society of Anesthesiologists (ASA)

3 Chen et al 867 Table day unplanned reoperation, postoperative complication, and mortality rates by otolaryngologic procedure type. Unplanned Reoperation a Postoperative Complications b Procedure Type No. of Patients Rate (%) No. of Patients Overall (%) Pre Discharge (%) PDC (%) Ratio of PDC/Pre-Discharge Mortality b Rate (%) Laryngectomy Lip surgery Tongue/floor of mouth Salivary gland Neck dissection Palate Tonsillectomy/adenoidectomy Thyroid , Total 10, , Abbreviation: PDC, postdischarge complication. a Unplanned reoperation available for patients diagnosed in 2011 only. b Postoperative complications and mortality data available for 48,028 patients diagnosed from 2005 to classification, operative wound classification, and hospital length of stay (LOS). Surgeon specialty was classified into otolaryngology and other. Operative time was classified as prolonged if it was in the highest quartile for each procedure group. Hospital LOS was grouped into 1 day and.1 day. Wound classification was categorized into clean/contaminated (representing clean and clean/contaminated classifications) and contaminated/dirty/infected (representing contaminated and dirty/infected classifications). Outcomes of interest included occurrence of postoperative complications (including both pre- and postdischarge complications), unplanned reoperation, and mortality within 30 days of surgery. Postoperative complications were grouped into surgical site complications (superficial, deep, and organ space infections and wound dehiscence), infections (pneumonia, urinary tract infection, sepsis, or shock), prolonged ventilator use (greater than 48 hours), unplanned reintubation, venous thromboembolism (VTE) (pulmonary embolism, deep venous thrombosis, or thrombophlebitis), renal (progressive renal insufficiency or acute renal failure), cardiovascular (stroke, cardiac arrest, myocardial infarction, or bleeding requiring.4 units of packed red blood cells), graft failure (graft, prosthesis, or flap failure), and other complications (peripheral nerve injury or coma.24 hours). A postoperative complication was defined as a PDC if the number of days from operation to complication was greater than the number of days from operation to discharge. Reoperation was characterized as an unplanned return to the operating room at any hospital within the 30- day postoperative period, excluding patients who have follow-up surgeries based on pathology findings from the primary procedure. Statistical Analysis Bivariate analysis was used to compare patients with PDCs to those who did not have a PDC. x 2 and Fisher s exact tests were used for categorical variables, and t tests were used for continuous variables. Multivariate logistic regressions were used to identify risk factors associated with PDCs and unplanned reoperation. Odds ratios (OR) and 95% confidence intervals (CIs) were calculated for the strength of association between each risk factor and the outcomes of interest. 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 Inc, Chicago, Illinois). Permission was granted from ACS- NSQIP to use and disclose the data. ACS-NSQIP data are publically available, and all patient information is de-identified; consequently, this study was granted an exemption from our institutional review board. Results A total of 48,028 adult patients who underwent 1 of 8 inpatient otolaryngology procedures from 2005 through 2011 were identified. The mean age was 52.0 years, and most patients were female (74.3%) and white (77.5%). In our study sample, the most common procedures were thyroidectomy (83.8%), neck dissection (5.1%), and salivary gland procedures (4.5%). Postdischarge Complications The overall 30-day postoperative complication rate was 2.8%, with a 1.7% rate of predischarge complications and 1.2% rate of PDC (Table 2). Laryngectomy, lip surgery, and tongue/floor of mouth surgery had the highest rate of PDCs (8.0%, 7.4%, and 4.1%, respectively). The PDC rate was more than twice the predischarge complication rate for salivary gland (ratio = 2.5) and palate procedures (ratio = 2.1). The majority of PDCs were surgical siterelated or other infections (Table 3). Within the first 48 hours, first week, and first 2 weeks post discharge, 10%, 44%, and 73% of PDCs occurred,

4 868 Otolaryngology Head and Neck Surgery 149(6) Table 3. Percentage of complications that occur post discharge by type of complication. Percentage of PDC a Procedure Type PDC Rate (%) SS Infection Unplanned Intubation Ventilator.48 h VTE Renal CV Flap Other Laryngectomy Lip surgery Tongue/floor of mouth Salivary gland Neck dissection Palate Tonsillectomy/adenoidectomy Thyroid Total Abbreviations: CV, cardiovascular; PDC, postdischarge complication; SS, surgical site infections and wound dehiscence; VTE, venous thromboembolism. Infection indicates pneumonia, urinary tract infection, sepsis, or shock. Flap indicates graft/prosthesis/flap failure. a Row percentages. Rows may not total 100% as some patients had multiple postdischarge complications. Figure 1. Frequency of postdischarge complications among patients who underwent inpatient otolaryngologic procedures in the American College of Surgeons National Surgical Quality Improvement Program database (2005 to 2011). Cumulative percentage is the cumulative portion of postdischarge complications that have occurred by that postdischarge day. respectively (Figure 1). In the first 48 hours post discharge, the most common types of PDCs were pneumonia (24.1%), superficial site infections (16.1%), urinary tract infections (13.8%), and unplanned intubation (11.5%). By 2 weeks post discharge, superficial site infections represented 27.8% of all PDCs. The proportion of complications that occurred post discharge relative to pre discharge was greatest for surgical site complications (74.3%), VTE (52.4%), and other infections (47.3%) (Figure 2). Compared to those without a PDC, patients with a PDC were more likely to be older (mean age 52.0 years vs 57.2 years, P \.001), male (25.5% vs 41.0%, P \.001), admitted from a residence other than home (2.1% vs 0.5%, P \.001), and discharged to a facility (1.0% vs 3.2%, P \.001) (Table 4). Patients with a PDC were more likely to have Figure 2. Proportion of postoperative complications that occurred post discharge among patients who underwent inpatient otolaryngologic procedures, American College of Surgeons National Surgical Quality Improvement Program (2005 to 2011). comorbid diabetes, dyspnea, COPD, pneumonia, hypertension, heart disease, CVA, sepsis, and disseminated cancer. They also were more likely than those without a PDC to have prolonged operative times (38.6% vs 24.6%, P \.001), hospital LOS.1 day (41.9% vs 18.4%, P \.001), and ASA class 3 (53.1% vs 29.4%, P \.001). The majority of patients with predischarge complications did not also have a PDC, but patients with a predischarge complication were 5 times more likely to have a PDC than those without an inpatient complication (5.5% vs 1.1%, P \.001). On multivariate analysis, after adjusting for procedure group, patient age (OR = 1.01; 95% CI, ), prolonged operative time (OR = 1.68; 95% CI, ), hospital LOS.1 day (OR = 1.49; 95% CI, ), ASA class 3 (OR = 1.45; 95% CI, ), wounds classified as contaminated/dirty/infected (OR = 2.16; 95% CI, ), dyspnea (moderate exertion: OR = 1.32, 95% CI, ; at rest: OR = 1.89, 95% CI, ), chronic steroid use (OR = 1.66; 95% CI, ), CVA disease (OR = 1.66; 95% CI, ), and sepsis (OR = 1.93; 95% CI, ) were independently associated with the

5 Chen et al 869 Table 4. Demographic and clinical characteristics of patients undergoing otolaryngologic procedures. No. of Patients (%) a Postdischarge Complications Unplanned Reoperation Characteristics No (N = 47,450) b Yes (N = 578) b P Value No (N = 10,397) b Yes (N = 270) b P Value Mean age in years (SEM) 52.0 (0.1) 57.2 (0.7) \ (0.1) 56.3 (1.0).001 Female 35,282 (74.5) 340 (59.0) 7485 (72.1) 156 (57.8) \.001 Race White 33,673 (77.5) 408 (76.5) 7177 (76.0) 169 (73.8) Black 5319 (12.2) 66 (12.4) 1143 (12.1) 28 (12.2) Other 4448 (10.2) 59 (11.1) 1125 (11.9) 32 (14.0) Functional status \ Independent 46,941 (99.0) 554 (95.8) (99.1) 262 (97.0) Partially dependent 386 (0.8) 19 (3.3) 81 (0.8) 6 (2.2) Totally dependent 73 (0.2) 5 (0.9) 15 (0.1) 2 (0.7) Prehospitalization residence \.001 \.001 Home 47,183 (99.4) 566 (97.9) 10,323 (99.3) 261 (96.7) Acute care 105 (0.2) 4 (0.7) 30 (0.3) 3 (1.1) Chronic care 92 (0.2) 5 (0.9) 18 (0.2) 4 (1.5) Other 66 (0.1) 3 (0.5) 23 (0.2) 2 (0.7) Discharge destination \.001 \.001 Home 10,805 (98.7) 149 (94.9) 10,233 (98.9) 235 (87.4) Facility 113 (1.0) 5 (3.2) 91 (0.9) 24 (8.9) Separate acute care 14 (0.1) 0 (0.0) 10 (0.1) 2 (0.7) Rehab 17 (0.2) 3 (1.9) 10 (0.1) 8 (3.0) Attending/resident Attending and resident 29,825 (69.3) 360 (69.2) 4332 (68.2) 110 (64.7) Attending alone 13,047 (30.3) 157 (30.2) 2016 (31.7) 60 (35.3) Other 170 (0.4) 3 (0.6) 3 (0.0) 0 (0.0) Surgeon specialty \.001 \.001 Otolaryngology 10,314 (21.7) 231 (40.0) 3719 (35.8) 158 (58.5) Other 37,136 (78.3) 347 (60.0) 6678 (64.2) 112 (41.5) Emergent surgery 228 (0.5) 4 (0.7) (0.4) 6 (2.2).001 Prolonged operative time 11,678 (24.6) 223 (38.6) \ (24.0) 96 (35.6) \.001 ASA Class 3 13,928 (29.4) 306 (53.1) \ (34.0) 147 (54.4) \.001 Wound classification \ Clean/contaminated 47,026 (99.1) 560 (96.9) 10,297 (99.0) 261 (96.7) Contaminated/dirty/infected 424 (0.9) 18 (3.1) 100 (1.0) 9 (3.3) LOS.1 day 8710 (18.4) 242 (41.9) \ (21.2) 186 (68.9) \.001 Predischarge complication 753 (1.6) 44 (7.6) \ (1.9) 88 (32.6) \.001 Postdischarge complication 122 (1.2) 28 (10.4) \.001 Abbreviations: ASA, American Society of Anesthesiologists; LOS, length of stay; SEM, standard error of mean. a Percentages are column percentages and may not add up to 100 due to rounding. b Total number of patients for some variables may be less than N due to missing values. occurrence of a PDC (Table 5). Relative to those who had a thyroidectomy, patients who had lip surgery (OR = 6.09; 95% CI, ), laryngectomy (OR = 3.73; 95% CI, ), and tongue/floor of mouth surgery (OR = 2.46; 95% CI, ) had the greatest odds of experiencing a PDC. Unplanned Reoperation In 2011, of the 10,667 patients who underwent an inpatient otolaryngologic procedure, there were 270 patients who experienced an unplanned reoperation. Patients who had an unplanned reoperation were older (mean age 56.3 years vs 52.5 years, P \.001), less likely to be admitted from home (96.7% vs 99.3%, P \.001), and more likely to be discharged to a facility (8.9% vs 0.9%, P \.001). After adjusting for procedure group, predischarge complications (OR = 17.68; 95% CI, ), PDC (OR = 6.55; 95% CI, ), history of COPD (OR = 2.64; 95% CI, ), and disseminated cancer (OR = 1.92; 95% CI, ) were independently associated with unplanned reoperation.

6 870 Otolaryngology Head and Neck Surgery 149(6) Table 5. Multivariate analysis of factors associated with postdischarge complications. Variable Coefficient Standard Error P Value Wald Chi-square Odds Ratio 95% CI Age Procedure Laryngectomy Neck dissection Lip surgery Tongue/floor of mouth Palate Salivary gland Tonsillectomy/adenoidectomy Surgeon specialty Prolonged operative time \ LOS.1 day \ ASA class \ Contaminated/dirty/infected wound Dyspnea Moderate exertion At rest Cerebrovascular disease Sepsis/SIRS Steroid use Abbreviations: ASA, American Society of Anesthesiologists; LOS, length of stay; SIRS, Systemic Inflammatory Response Syndrome. Referents: Procedure type thyroidectomy, ASA class \3, clean wound classification, surgeon specialty other, and for all other references, no or not present. Age was a continuous variable. Mortality Overall, the 30-day mortality rate for inpatient otolaryngologic procedures was 0.1%. The mortality rate was highest for laryngectomy (2.1%). Patients with a predischarge complication were more likely to die in the 30-day postoperative period than those who did not have a predischarge complication (4.9% vs 0.1%, P \.001). Similarly, patients who experienced a PDC were more likely to die in that time period than those who did not have a PDC (0.9% vs 0.1%, P =.001). Discussion This is the first national study of procedure-specific rates and risk factors for PDCs after inpatient otolaryngologic surgery. Most current analyses of postoperative complications focus primarily on in-hospital complications. However, we found that over 40% of postoperative complications following otolaryngologic surgery occurred post discharge. PDC rates varied significantly by procedure; laryngectomy had a ninefold higher rate of PDC than thyroidectomy. A tenth of all PDCs occurred within the first 48 hours after discharge. In the 30-day postoperative period, patients who experienced a PDC were 9 times more likely than those without a PDC to have an unplanned reoperation or die. The overall 30-day PDC rate in our study was 1.2%, which is markedly lower than the PDC rates in general surgery. In an analysis of 12,956 colectomies in NSQIP, Oyetunji et al determined a PDC rate of 8.7%. 12 Similarly, a study of 551,510 patients undergoing general surgery procedures by Kazaure et al reported an overall PDC rate of 6.9%. 8 We found that approximately 40% of complications occurred post discharge, which validates recent findings in general surgery that demonstrated that 33% to 42% of postoperative complications 8,9 and 24% of deaths 9 occurred post discharge. We found that the most common types of PDCs were surgical site infections, pneumonias, and urinary tract infections. PDCs outnumbered predischarge complications, with over 74% of surgical site complications and 47% of other infections occurring in the postdischarge period. In a population-based survey of 622,683 surgical patients, Daneman et al determined that head and neck surgery was an independent risk factor for postdischarge versus inhospital diagnosis of a surgical site infection when compared to abdominal surgery (OR = 1.66; 95% CI, ). 13 They reported a 5.0% rate of postdischarge surgical site infections, with 70% of postoperative site infections occurring post discharge. 13 Reid et al conducted a single-institution study of the infection rate of 1964 clean wounds in the 30-day postoperative period and reported a 7.1% infection rate after head and neck surgery; approximately two-thirds of the wound infections arose after discharge. 14 Surgical site infections carry a significant economic burden. A recent nationwide study of surgical site infections found that these infections, on average, extended hospital LOS by 9.7 days and increased cost by $20,842 per admission. 15 An analysis of postdischarge surgical site infections

7 Chen et al 871 demonstrated that patients with postdischarge infections incurred $3,382 more in total costs than their matched controls. 16 Perioperative antibiotics have been shown to be neither cost-effective nor beneficial in preventing infections in clean uncontaminated otolaryngologic procedures 17,18 but have been shown to be beneficial in clean/contaminated procedures. 19 In a study of postdischarge head and neck wound infections, Reid et al determined that there was no difference in infection rates with the use of antibiotics (7.8% vs 7.9%, P =.99); antibiotics only delayed the occurrence of postdischarge infections (9.8 days vs 11.9 days, P \.001). 14 The current literature regarding prophylactic courses of postoperative antibiotics has been inconclusive since antibiotics may be more likely to be used in cases where physicians believe that they are needed. We observed that the presence of a PDC was associated with unplanned reoperation and increased mortality. In a review of 3044 patients undergoing index general surgical procedures, Birkmeyer et al demonstrated that reoperation was associated with increased mortality rates (15.9% vs 2.3%, P \.001) and total hospital charges ($82,300 vs $17,700, P \.001). 20 The majority of these reoperations were due to surgical site complications (85%). 20 Given the morbidity and economic costs associated with PDC, we identified important risk factors associated with PDC in order to better triage high-risk patients. PDC rates are procedure-specific and patients undergoing laryngectomy, lip, and tongue/floor of mouth surgery are at the highest risk of experiencing a PDC. Additionally, patients who underwent salivary gland or palate procedures have rates of PDCs that are more than twice the rate of predischarge complications and may also benefit from close follow-up. We identified specific patient comorbidities that may aid in the implementation of targeted discharge and follow-up plans, including dyspnea, chronic steroid use, CVA disease, and sepsis. After adjusting for procedure group, prolonged operative time and ASA class continued to be independent predictors of PDCs. Prolonged operative time has been shown to be an important predictor of wound complications in laryngectomy 21 and in general surgery. 8,12,14 Since surgical site infections were a common PDC, prolonged operative time may have contributed to PDCs through prolonged exposure of the surgical site and decreased effectiveness of preoperative antibiotics with increased operative time. These results highlight the importance of careful preoperative patient selection and optimization of operative efficiency without compromising thoroughness. Some of the otolaryngologic procedures we identified, especially thyroidectomies, are performed by both otolaryngologists and general surgeons, so we controlled for this factor by including surgeon specialty in our multivariate analysis. On bivariate analysis, otolaryngologists have a slightly higher rate of PDCs than other specialties; however, this difference is no longer significant on multivariate analysis after adjusting for procedure group and patient characteristics. This suggests that the difference in the rates of PDCs between otolaryngologists and other surgical specialties is likely secondary to differences in the types of otolaryngologic procedures they perform. The immediate postdischarge period was the most susceptible time for patients. Approximately half of all PDCs occurred in the first 10 days after discharge. Close follow-up among high-risk patients may help decrease the rate of PDCs in otolaryngologic follow-up. This follow-up can be in the form of an automated home monitoring system, physician follow-up, or structured follow-up with care coordinators. Graham et al found that an automated home monitoring system reduced the rate of 30-day readmissions by 44%. 22 A review of 1,855,702 Medicare patients by Jencks et al revealed that 50% of patients who were rehospitalized within 30 days of discharge had not seen a physician in that time period. 4 Limitations of our study include those inherent to large databases, such as coding errors and misclassification. The ACS-NSQIP is a well-validated national database that conducts regular audits to ensure reliability of data. 23 Unplanned readmission was coded for fewer than 2% of the patients in our sample, so we were unable to analyze the association between PDC and unplanned readmissions. The timing of unplanned reoperation was not classified in the database, so we were unable to delineate whether reoperation occurred before or after the PDC. ACS-NSQIP is administered by the American College of Surgeons, so a large proportion of the procedures analyzed were thyroidectomies, which are performed by both general surgeons and otolaryngologists. However, on multivariate analysis, we did not find that surgical specialty was associated with an increased risk of postdischarge complications. Information related to antibiotic use, costs, surgeon volumes, hospital type, geographic region, long-term postoperative outcomes, and disease severity was not recorded in ACS-NSQIP and could not be analyzed. In otolaryngologic surgery, PDCs are procedure-specific, occur soon after discharge, and are associated with unplanned reoperation and mortality. Our results demonstrate that targeted triage and follow-up plans are needed in order to best address PDCs. Further research is needed in order to evaluate systems for managing otolaryngology patients after hospital discharge. Continued multiinstitutional data collection on PDCs is needed in order to continue quality improvements in otolaryngologic surgery. Authors Note 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. Author Contributions Michelle M. Chen, contributions to conception and design, acquisition and analysis of data, drafting the manuscript, final approval; Sanziana A. Roman, contributions to conception and design, critical

8 872 Otolaryngology Head and Neck Surgery 149(6) revision for important intellectual content, final approval; Julie A. Sosa, contributions to conception and design, critical revision for important intellectual content, final approval. Benjamin L. Judson, contributions to conception and design, analysis of data, revision for important intellectual content, final approval. Disclosures Competing interests: None. Sponsorships: None. Funding source: James G. Hirsch, MD, Endowed Medical Student Research Fellowship at Yale University School of Medicine. References 1. Ashton CM, Del Junco DJ, Souchek J, Wray NP, Mansyur CL. The association between the quality of inpatient care and early readmission: a meta-analysis of the evidence. Med Care. 1997; 35: Wick EC, Shore AD, Hirose K, et al. Readmission rates and cost following colorectal surgery. Dis Colon Rectum. 2011;54: Readmissions Reduction Program. care/medicare-fee-for-service-payment/acuteinpatientpps/read missions-reduction-program.html. Accessed March 3, Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360: Munoz E, Goldstein J, Lory MH, et al. Hospital readmissions, otolaryngology, and the diagnosis related group hospital payment system. Arch Otolaryngol Head Neck Surg. 1990;116: Gourin CG, Frick KD. National trends in laryngeal cancer surgery and the effect of surgeon and hospital volume on shortterm outcomes and cost of care. Laryngoscope. 2012;122: Chan JYK, Semenov YR, Gourin CG. Postoperative urinary tract infection and short-term outcomes and costs in head and neck cancer surgery. Otolaryngol Head Neck Surg. 2013;148: Kazaure HS, Roman SA, Sosa JA. Association of postdischarge complications with reoperation and mortality in general surgery. Arch Surg. 2012;147: Bilimoria KY, Cohen ME, Ingraham AM, et al. Effect of postdischarge morbidity and mortality on comparisons of hospital surgical quality. Ann Surg. 2010;252: Stachler RJ, Yaremchuk K, Ritz J. Preliminary NSQIP results: a tool for quality improvement. Otolaryngol Head Neck Surg. 2010;143:26 30.e American College of Surgeons National Surgical Quality Improvement Program. ACS-NSQIP user guide for the 2011 Participant Data Use File. uploads/2012/03/2011-user-guide_final.pdf. Accessed February 26, Oyetunji TA, Turner PL, Onguti SK, et al. Predictors of postdischarge complications: role of in-hospital length of stay. Am J Surg. 2013;205: Daneman N, Lu H, Redelmeier DA. Discharge after discharge: predicting surgical site infections after patients leave hospital. J Hosp Infect. 2010;75: Reid R, Simcock JW, Chisholm L, Dobbs B, Frizelle FA. Postdischarge clean wound infections: Incidence underestimated and risk factors overemphasized. ANZ J Surg. 2002;72: De Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D, Vaughn BB. Surgical site infection: incidence and impact on hospital utilization and treatment costs. Am J Infect Control. 2009;37: Perencevich EN, Sands KE, Cosgrove SE, Guadagnoli E, Meara E, Platt R. Health and economic impact of surgical site infections diagnosed after hospital discharge. Emerg Infect Dis. 2003;9: Johnson JT, Wagner RL. Infection following uncontaminated head and neck surgery. Arch Otolaryngol Head Neck Surg. 1987;113: Man L-X, Beswick DM, Johnson JT. Antibiotic prophylaxis in uncontaminated neck dissection. Laryngoscope. 2011;121: Weber RS. Wound infection in head and neck surgery: implications for perioperative antibiotic treatment. Ear Nose Throat J. 1997;76: Birkmeyer JD, Hamby LS, Birkmeyer CM, Decker MV, Karon NM, Dow RW. Is unplanned return to the operating room a useful quality indicator in general surgery? Arch Surg. 2001;136: Schwartz SR, Yueh B, Maynard C, Daley J, Henderson W, Khuri SF. Predictors of wound complications after laryngectomy: a study of over 2000 patients. Otolaryngol Head Neck Surg. 2004;131: Graham J, Tomcavage J, Salek D, Sciandra J, Davis DE, Stewart WF. Postdischarge monitoring using interactive voice response system reduces 30-day readmission rates in a casemanaged medicare population. Med Care. 2012;50: Shiloach M, Frencher SK, Steeger JE, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg. 2010;210:6-16.