Does Resident Participation Influence Otolaryngology Head and Neck Surgery Morbidity and Mortality?

Transcription

1 The Laryngoscope VC 2016 The American Laryngological, Rhinological and Otological Society, Inc. Does Resident Participation Influence Otolaryngology Head and Neck Surgery Morbidity and Mortality? Nicholas B. Abt, BS; Douglas D. Reh, MD; David W. Eisele, MD; Howard W. Francis, MD; Christine G. Gourin, MD, MPH Objectives/Hypothesis: Patients may perceive resident procedural participation as detrimental to their outcome. Our objective is to investigate whether otolaryngology head and neck surgery (OHNS) housestaff participation is associated with surgical morbidity and mortality. Study Design: Case-control study. Methods: OHNS patients were analyzed from the American College of Surgeons National Surgical Quality Improvement Program 2006 to 2013 databases. We compared the incidence of 30-day postoperative morbidity, mortality, readmissions, and reoperations in patients operated on by resident surgeons with attending supervision (AR) with patients operated on by an attending surgeon alone (AO) using cross-tabulations and multivariable regression. Results: There were 27,018 cases with primary surgeon data available, with 9,511 AR cases and 17,507 AO cases. Overall, 3.62% of patients experienced at least one postoperative complication. The AR cohort had a higher complication rate of 5.73% than the AO cohort at 2.48% (P <.001). After controlling for all other variables, there was no significant difference in morbidity (odds ratio [OR] [0.89 to 1.24]), mortality (OR [0.49 to 1.70]), readmission (OR [0.92 to 1.81]), or reoperation (OR [0.91 to 1.80]) for AR compared to AO cases. There was no difference between postgraduate year levels for adjusted 30-day morbidity or mortality. Conclusions: There is an increased incidence of morbidity, mortality, readmission, and reoperation in OHNS surgical cases with resident participation, which appears related to increased comorbidity with AR patients. After controlling for other variables, resident participation was not associated with an increase in 30-day morbidity, mortality, readmission, or reoperation odds. These data suggest that OHNS resident participation in surgical cases is not associated with poorer short-term outcomes. Key Words: Resident, housestaff, attending, morbidity, mortality, surgeon, outcomes, otolaryngology, National Surgical Quality Improvement Program. Level of Evidence: 3b Laryngoscope, 126: , 2016 INTRODUCTION Attending surgeons are faced with a dilemma when patients refuse surgical housestaff participation in an operation, leaving a fundamental choice between patient autonomy and surgical education. 1 Attending surgeons must engage the patient in discussion regarding this issue. Although 86% of general surgery patients indicated an educational willingness in a survey, up to one-third answered they did not want housestaff procedural participation. 2 The crux of this subject is patient safety and how to maximize surgical safety. A primary responsibility of an From the Department of Otolaryngology-Head and Neck Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A. Editor s Note: This Manuscript was accepted for publication February 19, Presented at The Triological Society Combined Sections Meeting, Miami, Florida, U.S.A., January 23, Winner of the first prize for poster presentations. The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Nicholas B. Abt, The Johns Hopkins Hospital, Department of Otolaryngology Head and Neck Surgery, 601 North Caroline Street, Baltimore, MD nabt1@jhmi.edu DOI: /lary attending surgeon is to continually improve residentassisted surgery. The mission is to identify all factors that might affect safety and make resident-assisted surgery as safe as all other surgery within otolaryngology. An average general surgeon receives 13 housestaff operating room exclusion requests each year. 1 A majority of general surgeons responded in a hypothetical survey of a patient s request for housestaff operating room exclusion that they would first protect the resident s educational goals, but make a genuine effort to address all patient concerns. 1 Although many surgical fields have studied whether patient s concerns over housestaff influence on morbidity and mortality are well founded, 3 10 otolaryngology head and neck surgery (OHNS) has yet to examine its own operative volume. Generally, staff surgeons operate faster without residents, 11 but housestaff s impact on complications vary across surgical disciplines, so much as to be called clinically insignificant. 3,4 We sought to determine if OHNS housestaff participation was associated overall 30-day morbidity and mortality following both elective and emergent OHNS operations using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database. 2263

2 MATERIALS AND METHODS Data Source and Patients All adult patients undergoing both elective and emergent OHNS procedures were acquired from the 2006 to 2012 ACS- NSQIP databases. NSQIP is a nationally validated, riskadjusted, prospectively collected database with over 300 participating international academic and nonacademic hospitals. 12 NSQIP perioperative clinical data are collected through a specifically assigned and trained surgical clinical reviewer at each individual hospital on randomly assigned patients. A surgeon champion is also appointed at each hospital to supervise program implementation and ensure data reliability. Patients <18 years old were excluded, as well as those cases that were missing data on resident participation. Specialty-specific data were grouped into otology, head and neck, facial plastic and reconstruction, rhinology, and general OHNS cases (Table I). Postgraduate year (PGY) groups were categorized by interns (PGY- 1), junior residents (PGY-2 and PGY-3), senior residents (PGY-4 and PGY-5), and fellows (PGY-61). Head and neck cases were additionally subdivided into major (total glossectomy, hemiglossectomy, total laryngectomy, hemilaryngectomy, mandibulectomy, and maxillectomy, with or without neck dissection) or minor (total and hemithyroidectomy, total and partial parotidectomy, modified radical or selective neck dissections, parathyroidectomy, partial glossectomy, and sentinel node biopsy) procedures. Variables Cases were classified by the presence of resident participation as an operating surgeon(s). The two cohorts were titled attending only (AO) or attending with resident (AR). The study endpoints were 30-day postoperative morbidity and mortality. Short-term morbidity was defined as an aggregation of all available recorded postoperative complications, including superficial and deep wound infection, systemic infection, cardiac, respiratory, renal, neurologic, or thromboembolic events, and graft/ prosthesis failure (Table II). Independent variables included for adjustment were age, sex, body mass index, smoking status, American Society of Anesthesiologists classification, work relative value units, wound classification, current wound infection, transfusion of >4 units of packed red blood cells <72 hours prior to surgery, Charlson Comorbidity Index score, diabetic status, steroid use for chronic condition, length of operation, history of previous operation within 30 days of surgery, operation year, inpatient/outpatient status, and alcohol use of >2 drinks/ day within the past 2 weeks (Tables III and IV). Additional dependent outcomes analyzed were unplanned readmission, reoperation, and length of stay (Table IV). All readmissions and reoperations occurred within 30 days and were initially recorded in 2011 and 2012, respectively. Statistical Analyses The study population was described with general summary statistics. Group comparisons were made using the t test and Wilcoxon rank sum test for continuous variables and the v 2 test for categorical variables, as appropriate (Tables I and II). Significance was attributed to a P value of <.05. Multivariable logistic regression was used to estimate the odds ratio (OR) for 30-day morbidity and mortality between the AO versus AR cohorts. Generalized linear regression with a log link was used to evaluate length of stay between groups, which was not normally distributed. b coefficients were the mean values representing the change in the value of the intercept mean. Statistical analyses were completed with Stata 12 (StataCorp, 2264 College Station, TX). This study was approved by the Johns Hopkins Institutional Review Board. RESULTS There were 60,823 OHNS surgical patients in the 2006 to 2012 ACS-NSQIP national databases. Of these patients, 27,018 had primary surgeon identification data, with a similar distribution of case type to all OHNS patients within the NSQIP database. The top 10 most commonly performed OHNS operations captured in this dataset were tonsillectomy (17.0%), total thyroidectomy (7.4%), thyroid lobectomy (5.9%), tonsillectomy and adenoidectomy (5.2%), parotidectomy with facial nerve dissection (4.5%), palatopharyngoplasty without tonsillectomy (4.5%), modified radical cervical lymphadenectomy (4.1%), tympanoplasty without mastoidectomy (3.5%), parathyroidectomy/parathyroid exploration (3.2%), and partial thyroid lobectomy (1.7%). A breakdown of resident participation in the most common procedures by specialty is shown in Table I. The mean age of all patients was years, and 55.5% of patients were female (Table II). PGY-level data were available after 2010 for 9,242 cases. There were 318 intern cases, 2,842 junior resident (JR) cases, 4,194 senior resident (SR) cases, and 1,888 fellow cases. Comparing the AO and AR cohorts, the AR cohort had slightly lower mean body mass index, a lower proportion of Caucasian and Latino patients, and fewer smokers. Class I, II, and III obesity rates were all lower in AR patients. The AO group had fewer patients with diabetes, recent alcohol consumption, chronic steroid use, advanced comorbidity, operations within the previous 30 days, and had shorter operation times. The AR patients were less healthy, with 3.1% of patients having Charlson Comorbidity Index scores 3, versus 1.7% of AO patients (P <.001) (Table II). Overall, 3.62% of patients had at least one postoperative complication. The AR cohort had a statistically significantly higher complication rate (5.73%) compared to the AO cohort (2.48%; P <.001) (Table III). Wound infection complications, including superficial surgical site infection (SSI), deep incisional SSI, organ space SSI, and wound dehiscence were all significantly higher in the AR cohort. Systemic infection and respiratory, renal, and cardiovascular complications were also increased in the AR group. There were 64 deaths (0.24%) within 30 days of surgery. A statistically significantly higher rate of mortality in the AR (0.33%) versus AO (0.19%) cohort (P 5.026) was observed. After controlling for all other variables, there was no significant difference in the odds of morbidity (OR ; 95% confidence interval [CI]: 0.89 to 1.24, P 5.554) and mortality (OR ; 95% CI: 0.49 to 1.70, P 5.768) for the AR group, compared to the AO group (Table IV). Elective cases had similar odds of 30-day morbidity (OR ; 95% CI: 0.88 to 1.23) and mortality (OR ; 95% CI: 0.39 to 1.46) compared to urgent cases. There was no significant difference in the odds of morbidity or mortality when stratified by subspecialty or by major versus minor surgical classification (Table IV).

3 TABLE I. Resident Participation in the 35 Most Common Procedures by Subspecialty. Resident and Attending Patients, n (%), N 5 6,510 Attending Only Patients, n (%), N 5 13,651 All Patients, n (%), N 5 20,161 Otology/neurotology 774 1,080 1,854 Tympanoplasty without mastoidectomy or ossiculoplasty 338 (43.7) 595 (55.1) 933 (50.3) Tympanoplasty with ossiculoplasty, homograft 70 (9.0) 125 (11.6) 195 (10.5) Tympanoplasty with ossiculoplasty, prosthesis 110 (14.2) 96 (8.9) 206 (11.1) Tympanoplasty with mastoidectomy, without ossiculoplasty 50 (6.5) 45 (4.2) 95 (5.1) Tympanoplasty with mastoidectomy, with ossicular chain reconstruction 87 (11.2) 86 (8.0) 173 (9.3) Tympanoplasty with mastoidectomy, intact canal wall, 46 (5.9) 54 (5.0) 100 (5.4) with ossicular chain reconstruction Tympanoplasty with mastoidectomy, canal wall down, 73 (9.4) 79 (7.3) 152 (8.2) with ossicular chain reconstruction Head and neck surgery 3,983 5,462 9,445 Total thyroidectomy 651 (16.3) 1,349 (24.7) 2,000 (21.2) Total thyroidectomy, with limited neck dissection 206 (5.2) 183 (3.4) 389 (4.1) Total thyroidectomy, with radical neck dissection 43 (1.1) 72 (1.3) 115 (1.2) Total thyroidectomy, following incomplete previous resection 139 (3.5) 99 (1.8) 238 (2.5) Partial unilateral thyroid lobectomy 183 (4.6) 275 (5.0) 458 (4.8) Total unilateral thyroid lobectomy 521 (13.1) 1,062 (19.4) 1,583 (16.8) Total laryngectomy, without neck dissection 108 (55.7) 45 (43.7) 153 (51.5) Excision of tongue lesion, without closure 86 (44.3) 58 (56.3) 144 (48.5) Excision of tongue lesion, anterior two-thirds, with closure 114 (2.9) 191 (3.5) 305 (3.2) Glossectomy, less than one-half tongue 537 (13.5) 677 (12.4) 1,214 (12.9) Glossectomy, less than one-half tongue, 189 (4.7) 289 (5.3) 478 (5.1) with unilateral radical neck dissection Parotidectomy, total, with unilateral radical neck dissection 72 (1.8) 39 (0.7) 111 (1.2) Submandibular gland excision 170 (4.3) 330 (6.0) 500 (5.3) Modified radical cervical lymphadenectomy 722 (18.1) 373 (6.8) 1,095 (11.6) Parathyroidectomy 18 (0.5) 143 (2.6) 161 (1.7) Complete cervical lymphadenectomy 68 (1.7) 32 (0.6) 100 (1.1) Excision of tongue lesion, without closure 18 (0.5) 84 (1.5) 102 (1.1) Excision of tongue lesion, anterior two-thirds, with closure 45 (1.1) 115 (2.1) 160 (1.7) Glossectomy, less than one-half tongue 179 (4.5) 105 (1.9) 284 (3.0) Glossectomy, less than one-half tongue, 108 (2.7) 44 (0.8) 152 (1.6) with unilateral radical neck dissection Facial plastic and reconstructive surgery Free full thickness flap, nose, ears, eyelids, and/or lips; <20 cm 2 36 (22.0) 69 (52.7) 105 (35.6) Muscle, myocutaneous, or fasciocutaneous flap; head and neck origin 90 (54.9) 26 (19. 8) 116 (39.3) Depressed zygomatic fracture reduction 38 (23.2) 36 (27.5) 74 (25.1) Rhinology and sinus surgery Ear cartilage graft to nose 47 (100) 39 (100) 86 (100) General otolaryngology 1,080 6,432 7,512 Tonsillectomy 723 (66.9) 3,879 (60.3) 4,602 (61.3) Tonsillectomy and adenoidectomy 129 (11.9) 1,288 (20.0) 1,417 (18.7) Adenoidectomy 23 (2.1) 200 (3.1) 223 (3.0) Uvulectomy 11 (1.0) 103 (1.6) 114 (1.5) Palatopharyngoplasty 194 (18.0) 962 (15.0) 1,156 (15.4) Postoperative complications varied by PGY level. Short-term morbidity rates for patients in the AR group increased by PGY year from interns at 3.46%, JR 3.87%, SR 5.94%, and fellows 8.63% (P <.001). All PGY cohorts (JR, SR, and fellows) were compared to the intern cohort as the referent group. Following multivariable regression analysis, all PGY levels had similar odds of 30-day morbidity for JR (OR ; 95% CI: 0.51 to 2.11), SR (OR ; 95% CI: 0.64 to 2.58), and fellows (OR ; 95% CI: 0.68 to 2.80) compared to interns. Similarly, mortality ORs for JR (OR ; 95% CI: 0.04 to 3.72), SR (OR ; 95% CI: 0.08 to 5.37), and fellows 2265

4 TABLE II. General Characteristics of OHNS Patients by Operating Surgeon Status. General Variables All Patients, N 5 27,018 Resident and Attending, N 5 9,511 Attending Only, N 5 17,507 P Value Age, yr, mean 6 SD <.001 BMI, mean 6 SD <.001* Underweight, no. (%) 570 (2.1) 235 (2.5) 335 (1.9) Normal, no. (%) 8,020 (29.7) 3,022 (31.8) 4,998 (28.6) Overweight, no. (%) 8,279 (30.6) 3,090 (32.5) 5,189 (29.6) Class I obesity, no. (%) 5,076 (18.8) 1,662 (17.5) 3,414 (19.5) Class II obesity, no. (%) 2,511 (9.3) 777 (8.2) 1,734 (9.9) Class III obesity, no. (%) 2,562 (9.5) 725 (7.6) 1,837 (10.5) Sex, female, no. (%) 14,982 (55.5) 4,835 (50.8) 10,147 (58.0) <.001 Race, no. (%) <.001* Caucasian 16,906 (62.6) 5,850 (61.5) 11,056 (63.2) African American 1,811 (6.7) 767 (8.1) 1,044 (6.0) Latino 5,741 (21.3) 1,902 (20.0) 3,839 (21.9) Asian/American Indian/Native Hawaiian Pacific Islander 761 (2.8) 336 (3.5) 425 (2.4) Unknown 1,799 (6.7) 656 (6.9) 1,143 (6.5) Current smoker, no. (%) 5,666 (21.0) 1,916 (20.2) 3,750 (21.4).014 Diabetes, no. (%) 2,554 (9.4) 988 (10.4) 1,566 (9.0).001 Noninsulin 1,787 (6.6) 707 (7.4) 1,080 (6.2) Insulin 767 (2.8) 281 (3.0) 486 (2.8) Operation time, mean minutes 6 SD <.001 Alcohol intake in previous 2 weeks, no. (%), 866 (3.2) n 5 26, (5.0) n 5 9, (2.2) n 5 17,404 Steroid use for chronic condition, no. (%) 512 (1.9) 257 (2.7) 255 (1.5) <.001 Operation within previous 30 days, no. (%) 610 (2.3), n 5 26, (3.0), n 5 9, (1.9), n 5 17,426 Charlson Comorbidity Index score, no. (%) <.001* 0 22,654 (84.3) 7,728 (81.7) 14,926 (85.8) 1 3,112 (11.6) 1,246 (13.2) 1,866 (10.7) (1.8) 189 (2.0) 303 (1.7) (2.2) 296 (3.1) 301 (1.7) Data are presented as no. (% of column total). *P values apply to overall category and do not apply to the subcategory. >2 drinks per day. Number of patients with available data. BMI 5 body mass index; no. 5 total number of patients; SD 5 standard deviation. <.001 <.001 (OR: 0.51; 95% CI: 0.06 to 4.49) were not significantly different from those for intern cases. The average length of stay (LOS) for all patients was 1.80 days, with a standard deviation of 7.51 days. The AR cohort had longer LOS at 2.97 days versus the AO group at 1.17 days (P <.001). Generalized linear regression demonstrated the AR cohort was associated an 0.36-day increase in mean LOS (95% CI: 0.07 to 0.66, P 5.017) (Table IV). When divided into specialty-specific data, only general OHNS was associated with increased mean LOS in the AR group. Other specialties had similar mean LOS. When head and neck operations were divided into major and minor cohorts, the effect was no longer observed. Major (coefficient ; 95% CI: to 0.03) and minor (coefficient ; 95% CI: to 0.44) head and neck cases had similar mean LOS between AR and AO groups. Unplanned 30-day readmissions occurred for 3.06% of the total group, with increased rates in the AR versus AO cohort, at 3.72% versus 2.45% respectively (P <.001). Overall, unplanned 30-day reoperation rates were 3.34%, with increased AR rates (3.95%) versus AO rates (2.76%, P <.001). After adjustment, multivariable analysis revealed no differences in odds of readmission (OR ; 95% CI: 0.92 to 1.81) or reoperation (OR ; 95% CI: 0.91 to 1.80). Reoperation odds were similar amongst all specialties. The AR cohort demonstrated increased readmission odds for general OHNS, whereas the other specialties saw no differences. DISCUSSION We found that after adjustment for patient characteristic variance and comorbidities, resident participation in the operating room was not associated with significant differences in the odds of 30-day morbidity or mortality of either elective or emergent OHNS 2266

5 TABLE III. Thirty-Day Complications, Mortality, Length of Stay, Readmissions, and Reoperations in Otolaryngology Head and Neck Surgery Patients by Primary Operating Surgeon Status. All Patients, N 5 27,018 Resident and Attending, N 5 9,511 Attending Only, N 5 17,507 P Value Morbidity Superficial surgical site infection 243 (0.90) 148 (1.56) 95 (0.54) <.001 Deep incisional SSI 86 (0.32) 50 (0.53) 36 (0.21) <.001 Organ space SSI 44 (0.16) 25 (0.26) 19 (0.11).003 Wound dehiscence 105 (0.39) 74 (0.78) 31 (0.18) <.001 Pneumonia 174 (0.64) 103 (1.08) 71 (0.41) <.001 Unplanned intubation 127 (0.47) 63 (0.66) 64 (0.37).001 Pulmonary embolism 31 (0.11) 20 (0.21) 11 (0.06).001 On ventilator >48 hours 77 (0.29) 40 (0.42) 37 (0.21).002 Progressive renal insufficiency 13 (0.05) 9 (0.09) 4 (0.02).010 Acute renal failure 13 (0.05) 7 (0.07) 6 (0.03).159 Urinary tract infection 106 (0.39) 36 (0.38) 70 (0.40).789 CVA/stroke with neurological deficit 23 (0.09) 13 (0.14) 10 (0.06).032 Coma >24 hours 4 (0.01) 1 (0.01) 3 (0.02).669 Cardiac arrest requiring CPR 30 (0.11) 18 (0.19) 12 (0.07).004 Myocardial infarction 25 (0.09) 19 (0.20) 6 (0.03) <.001 Graft/prosthesis failure 60 (0.22) 47 (0.49) 13 (0.07) <.001 DVT/thrombophlebitis requiring treatment 48 (0.18) 26 (0.27) 22 (0.13).006 Sepsis 109 (0.40) 59 (0.62) 50 (0.29) <.001 Septic shock 22 (0.08) 11 (0.12) 11 (0.06).146 Total morbidity events 1, Total no. of patients with 1 morbidity event 979 (3.62) 545 (5.73) 434 (2.48) <.001 Mortality Deceased 64 (0.24) 31 (0.33) 33 (0.19).026 LOS LOS, mean days 6 SD < day unplanned readmission Readmissions, n 5 6,203* 190 (3.06) 111 (3.72) 79 (2.45) < day unplanned reoperation Reoperations, n 5 12, (3.34) 230 (3.95) 171 (2.76) <.001 Data are presented as no. (%). *Unplanned readmissions only recorded in Unplanned reoperations only recorded in 2011 and CPR 5 cardiopulmonary; CVA 5 cerebrovascular accident; DVT 5 deep vein thrombosis; LOS 5 length of stay; SSI 5 surgical site infection. procedures. Although the AR cohort did have increased rates of most postoperative complications and mortality, patients in the AR cohort were significantly more likely to have advanced comorbidity, predisposing unfavorable conditions, and were more likely to undergo more extensive surgical procedures. After controlling for these variables, morbidity and mortality rates were not significantly different for AR cases. These data suggest resident surgeons do not increase complication or death odds. OHNS training programs, like all other surgical specialties, has the goal of training residents to be capable and proficient at diagnostic workups, technical procedures, and postoperative care. 13 OHNS housestaff must become proficient in more than five subspecialties (head and neck, facial plastic and reconstruction, rhinology/laryngology, otology, and pediatrics) before graduation. Our data demonstrate resident participation in surgical cases are not associated with increased postoperative complications or mortality in any of these subspecialties, despite the observation that head and neck surgical cases are more complex, with a greater potential for postoperative complications readmission and reoperation rates. One possibility is that attending surgeon supervision diminishes intraoperative errors that could lead to serious postoperative complications. Additionally, recent key metrics for hospital-wide quality care are readmission and reoperation rates, neither of which was increased due to resident participation. Finally, our data demonstrated there was no morbidity or mortality odds difference amongst JR, SR, or fellows compared to interns. These data suggest the specific postgraduate level of an OHNS trainee is not an independent predictor of complications. One possibility is attending surgeon supervision diminishes intraoperative errors that could 2267

6 TABLE IV. Thirty-Day Morbidity, Mortality, Unplanned Readmission, Unplanned Reoperation Logistic Regression Models, and Length of Stay Generalized Linear Regression Model for All Otolaryngology Head and Neck Surgery Patients. Univariable Regression Multivariable Regression* 30-Day Morbidity Odds Ratio 95% CI P Value Odds Ratio 95% CI P Value All patients to 2.72 < to Otology/neurotology to to Head and neck to to Facial plastics to to Rhinology/sinus 1.00 General otolaryngology to to Day Mortality Univariable Regression Multivariable Regression*, all patients to to Day Unplanned Readmissions Univariable Regression Multivariable Regression* All patients to 3.48 < to Otology/neurotology 1.00 Head and neck to 5.77 < to Facial plastics Rhinology/sinus 1.00 General otolaryngology to 5.18 < to Day Unplanned Reoperations Univariable Regression Multivariable Regression* All patients to 1.77 < to Otology/neurotology 1.00 Head and neck to to Facial plastics to Rhinology/sinus 1.00 General otolaryngology to to Univariable GLM Multivariable GLM* Length of Stay b Coefficient 95% CI P Value b Coefficient 95% CI P Value All patients to 1.04 < to Otology/neurotology to to Head and neck to 0.73 < to Facial plastics to 2.23 < to Rhinology/sinus to General otolaryngology to to *Multivariable analysis adjustment variables found in the Materials and Methods section. Statistical significance. Insufficient number of deaths to calculate odds ratios by specialty. CI 5 confidence interval; GLM 5 general linear regression. 2268

7 lead to serious postoperative complications. Additionally, senior and chief residents supervise all postoperative patients, including PGY-1 cases. The unadjusted increase in morbidity for increasing PGY levels most likely reflects case complexity, and thus a multifaceted interaction between operative length, case complexity, resident surgeon experience, and supervision extent all contribute to postoperative outcomes. Resident participation mostly occurs within teaching hospitals, where higher complexity operations occur, and could explain why there are unadjusted increased complication rates. The increase in observed morbidity and mortality appears to be associated with increased preoperative comorbidity and not associated with resident participation. General surgery, 7,8 laparoscopic surgery, 3 5 orthopedic surgery, 9 and neurosurgery 6,14 have examined their operative volumes and come to similar conclusions. All studies indicated resident participation was not hazardous to postoperative outcomes. Advantages of the ACS-NSQIP database include robust prospective collection of perioperative, surgically tailored variables, which provide a broad base for adjusted analyses. There is also an extensive listing of postsurgical complications and readmission/reoperation data. As with all large national database analyses, however, our study has inherent limitations. Only short-term conclusions can be made, and no extrapolations can be drawn beyond the 30-day time point. Additionally, hospital teaching status was undeterminable, which could shed insight onto the increased raw AR complications rates. Charlson Comorbidity Index scores were close in percentage, limiting detection power of differences between cohorts, especially in index scores >3. Patients with >3 index scores consisted of only 2.2% of the study population, and thus, these data should be interpreted with caution. Although PGY level was provided, exact resident surgeon involvement was not noted. However, multivariable analyses were used to adjust for most possible confounders found within the NSQIP. Adjusted analyses, combined with a high power of over 27,000 patients, were used to limit type II error. Future prospective studies are warranted to authenticate these conclusions beyond the 30-day window. CONCLUSION Otolaryngology resident participation was not associated with increased odds of 30-day elective or emergent morbidity, mortality, length of stay, readmissions, and reoperations. These data suggest that OHNS resident participation in surgical cases is not associated with poorer short-term outcomes, and resident participation should be regarded as safe. BIBLIOGRAPHY 1. Dutta S, Dunnington G, Blanchard MC, Spielman B, DaRosa D, Joehl RJ. And doctor, no residents please! J Am Coll Surg 2003;197: Cowles RA, Moyer CA, Sonnad SS, et al. Doctor-patient communication in surgery: attitudes and expectations of general surgery patients about the involvement and education of surgical residents. J Am Coll Surg 2001;193: Davis SS Jr, Husain FA, Lin E, Nandipati KC, Perez S, Sweeney JF. Resident participation in index laparoscopic general surgical cases: impact of the learning environment on surgical outcomes. J Am Coll Surg 2013; 216: Hernandez-Irizarry R, Zendejas B, Ali SM, Lohse CM, Farley DR. Impact of resident participation on laparoscopic inguinal hernia repairs: are residents slowing us down? J Surg Educ 2012;69: Krell RW, Birkmeyer NJ, Reames BN, et al. Effects of resident involvement on complication rates after laparoscopic gastric bypass. J Am Coll Surg 2014;218: Morgan MK, Assaad NN, Davidson AS. How does the participation of a resident surgeon in procedures for small intracranial aneurysms impact patient outcome? J Neurosurg 2007;106: Raval MV, Wang X, Cohen ME, et al. The influence of resident involvement on surgical outcomes. J Am Coll Surg 2011;212: Relles DM, Burkhart RA, Pucci MJ, et al. Does resident experience affect outcomes in complex abdominal surgery? Pancreaticoduodenectomy as an example. J Gastrointest Surg 2014;18: Schoenfeld AJ, Serrano JA, Waterman BR, Bader JO, Belmont PJ Jr. The impact of resident involvement on post-operative morbidity and mortality following orthopaedic procedures: a study of 43,343 cases. Arch Orthop Trauma Surg 2013;133: Pollei TR, Barrs DM, Hinni ML, Bansberg SF, Walter LC. Operative time and cost of resident surgical experience: effect of instituting an otolaryngology residency program. Otolaryngol Head Neck Surg 2013;148: Puram SV, Kozin ED, Sethi R, et al. Impact of resident surgeons on procedure length based on common pediatric otolaryngology cases. Laryngoscope 2015;125: Fink AS, Campbell DA Jr, Mentzer RM Jr, et al. The National Surgical Quality Improvement Program in non-veterans administration hospitals: initial demonstration of feasibility. Ann Surg 2002;236: ; discussion McCormick ME, Stadler ME, Shah RK. Embedding quality and safety in otolaryngology-head and neck surgery education. Otolaryngol Head Neck Surg 2015;152: Bydon M, Abt NB, De la Garza-Ramos R, et al. Impact of resident participation on morbidity and mortality in neurosurgical procedures: an analysis of 16,098 patients. J Neurosurg 2015;122: