Customize your JAMA Network experience by selecting one or more topics from the list below.
Smith A, Braden L, Wan J, Sebelik M. Association of Otolaryngology Resident Duty Hour Restrictions With Procedure-Specific Outcomes in Head and Neck Endocrine Surgery. JAMA Otolaryngol Head Neck Surg. 2017;143(6):549–554. doi:10.1001/jamaoto.2016.4182
What was the impact on surgery-specific outcomes following duty hour implementation for thyroid and parathyroid procedures in teaching hospitals with otolaryngology residents (THs-OTO)?
In this retrospective analysis of the National Inpatient Sample (2000-2002 vs 2006-2008), THs-OTO contributed a greater share of procedures (18% to 25%). With the earlier period serving as the reference, rates of recurrent laryngeal nerve (RLN) injury, postoperative hematoma, and hypoparathyroidism did not change; however, unintentional vessel lacerations increased and length of stay and mortality decreased.
Although unintentional vessel lacerations increased, RLN injury, hematoma, and hypoparathyroidism did not change. Length of stay and mortality improved following duty hour restrictions.
Graduate medical education has undergone a transformation from traditional long work hours to a restricted plan to allow adequate rest for residents. The initial goal of this restriction is to improve patient outcomes.
To determine whether duty hour restrictions had any impact on surgery-specific outcomes by analyzing complications following thyroid and parathyroid procedures performed before and after duty hour reform.
Design, Setting, and Participants
Retrospective cross-sectional analysis of the National Inpatient Sample (NIS).The NIS was queried for procedure codes associated with thyroid and parathyroid procedures for the years 2000 to 2002 and 2006 to 2008. Hospitals were divided based on teaching status into 3 groups: nonteaching hospitals (NTHs), teaching hospitals without otolaryngology programs (THs), and teaching hospitals with otolaryngology programs (THs-OTO).
Main Outcomes and Measures
Procedure-specific complication rates, length of stay, and mortality rates were collected. SAS statistical software (version 9.4) was used for analysis with adjustment using Charlson comorbidity index.
Total numbers of head and neck endocrine procedures were 34 685 and 39 770 (a 14.7% increase), for 2000 to 2002 and 2006 to 2008, respectively. THs-OTO contributed a greater share of procedures in 2006 to 2008 (from 18% to 25%). With the earlier period serving as the reference, length of stay remained constant (2.1 days); however, total hospital charges increased (from $12 978 to $23 708; P < .001). Rates of postoperative hematoma (odds ratio [OR], 1.21; 95% CI, 1.06-1.38), hypoparathyroidism (OR, 1.27; 95% CI, 1.06-1.52), and unintentional vessel lacerations (OR, 1.36; 95% CI, 1.02-1.83) increased overall with NTHs (OR, 1.26; 95% CI, 1.04-1.52), THs (OR, 1.65; 95% CI, 1.15-2.37), and THs-OTO (OR, 1.98; 95% CI, 1.09-3.61) accounting for these differences, respectively. Overall mortality decreased (OR, 0.66; 95% CI, 0.47-0.94) following a decrease in the TH-OTO mortality rate (OR, 0.34; 95% CI, 0.12-0.93).
Conclusions and Relevance
While recurrent laryngeal nerve injury, hematoma formation, and hypoparathyroidism did not change, length of stay and mortality improved within THs-OTO following head and neck endocrine procedures after implementation of duty hour regulations. This finding refutes the concern that duty hour restrictions result in poorer overall outcomes. Less time available to develop technical competence may play a factor in some outcomes in lieu of recurrent laryngeal nerve injury increasing within THs and accidental injury to vessels, organs, or nerves and hypocalcemia increasing within THs-OTO. Furthermore, head and neck endocrine cases increased at THs with otolaryngology programs.
Although surgery for benign thyroid disease can be performed in an outpatient surgery setting, thyroid disease still frequently requires an inpatient setting.1 Primary hyperparathyroidism is the most common cause of hypercalcemia and the third most common endocrine disorder.2 Owing to the importance of this disease burden, the training of individuals administering care to these patients should be periodically examined. Duty hour reform gives us a unique opportunity for such an examination.
The beginning of the 2005 academic calendar marked implementation of the initial Accreditation Council for Graduate Medical Education (ACGME) duty hour restrictions on all resident physicians. The overall goal was to achieve better patient outcomes by reducing resident fatigue and sleep deprivation. Cognitive and clinical performance declines associated with sleep deprivation could be avoided. Furthermore, personal safety issues, such as needle stick injuries and motor vehicle crashes, could be averted owing to these restrictions with more emphasis placed on resident quality of life.3-7
Otolaryngologists, general surgeons, and endocrine surgeons perform most head and neck endocrine surgical procedures. Which specialty performs more head and neck endocrine procedures as residents is up for debate,8,9 but it seems that otolaryngologists are now managing more surgical thyroid disorders.10 For 2015, general surgery graduates were required to log 8 of any endocrine procedures, while otolaryngology graduates were required to log 22 thyroid and parathyroid cases, suggesting an experiential difference between the specialties. While otolaryngology and general surgery residents are subject to duty hour restrictions, endocrine surgery is a 1-year fellowship following general surgery residency and is not subject to the same duty hour restrictions mandated by the ACGME.
Duty hour restrictions limit a resident to a 24-hour continuous workload with a 6-hour period of transition, every third night call, a 10-hour rest between duty periods, an 80-hour work week, and 1 in 7 days free of patient care responsibilities.11 Studies in neurosurgery, orthopedic surgery, and cardiac surgery programs have looked at outcomes using the National Inpatient Sample (NIS) database with mixed results.12-15 We aimed to characterize any differences in head and neck endocrine procedure-specific complications before and after duty hour reform within the NIS.
The NIS was developed by the Agency for Healthcare Research and Quality (AHRQ) specifically to analyze health care outcomes. This series of databases contains information regarding patient demographics, geographic region, diagnoses, procedures, length of stay (LOS), and hospital characteristics.16 The International Classification of Diseases, Ninth Revision, Clinical Modification(ICD-9-CM), procedure codes correlating to head and neck endocrine procedures were extracted, including unilateral thyroid lobectomy (code 06.2), other partial thyroidectomy (06.3), complete thyroidectomy (06.4), substernal thyroidectomy (06.5), excision of lingual thyroid (06.6), excision of thyroglossal duct or tract (06.7), and parathyroidectomy (06.8). These ICD-9-CM codes were compared before and after the duty hour restrictions took effect. The first period included the years 2000, 2001, and 2002, while the second included 2006, 2007, and 2008 (years 2003-2005 represent the time during which graduated implementation of duty hour restrictions took place).
A teaching status variable exists within the data set; however, this variable is divided only into nonteaching (NTHs) vs teaching hospitals without otolaryngology programs (THs). Previous studies failed to divide teaching hospitals into those with and without a specialty’s respective training program. In our investigation, we found most THs did not have otolaryngology residents. This discovery led to dividing the THs into 2 new groupings: THs and THs with otolaryngology residents (THs-OTO). To create these new groupings, we searched the websites of all known otolaryngology training programs to find the hospitals associated with these programs and sorted institutions based on state, city, and address to correlate with known otolaryngology training programs. Therefore, we then had 3 groupings: NTHs, THs, and THs-OTO. States that did not report hospital-specific information or did not report teaching status were omitted. A summary of those states and number of hospitals included can be found in Table 1.
Discharge weights are available within the database to calculate national estimates; however, these weights were not applied because multiple states were omitted in the hospital stratification process. SAS statistical software (version 9.4; SAS Institute Inc) was used to compare surgical complications individually (Table 2) using multivariate logistic regression for dichotomous variables and multivariate linear regression for continuous variables with significance at P < .05. Cost, LOS, and mortality were also studied. The Romano adaptation of the Charlson comorbidity index (CCI) was built into the analysis as a continuous variable to account for comorbidities (the CCI range was 0-7, while 80% were CCIs of 0-1). Those columns within the data set with fewer than 10 entries were omitted owing to the AHRQ limitations set forth in the Data Use Agreement. Exemption was obtained from the University of Tennessee institutional review board owing to this study’s analysis of preexisting public data.
Nonteaching facilities served as the control with hypoparathyroidism occurring less frequently in THs (odds ratio [OR], 0.64; 95% CI, 0.46-0.91) and hypocalcemia occurring less at THs-OTO (OR, 0.72; 95% CI, 0.62-0.83); however, recurrent laryngeal nerve (RLN) injury occurred more often at THs-OTO (OR, 2.14; 95% CI, 1.58-2.90) (Table 3). Hypocalcemia also occurred more often at THs (OR 1.12; 95% CI, 1.01-1.25). Median LOS and total charges were the same between NTHs, THs, and THs-OTO (2.12 vs 2.08 vs 2.13 days, respectively, and $12 852.98, $12 837.59, and $13 533.77, respectively). Overall, each point increase in CCI was associated with an increased risk for RLN injury (ORs 1.97; 95% CI, 1.75-2.22]), hypocalcemia (ORs 1.62; 95% CI, 1.41-1.86), hypoparathyroidism (ORs 1.52; 95% CI, 1.45-1.60), and mortality (OR, 3.37; 95% CI, 2.81-4.05).
Nonteaching facilities served as the control with no differences noted between NTH and TH. Hypoparathyroidism (OR, 0.68; 95% CI, 0.51-0.92) hypocalcemia (OR, 0.88; 95% CI, 0.79-0.98]), and mortality (OR, 0.33; 95% CI, 0.15-0.76) occurred less frequently at THs-OTO. Compared with NTH within this second period, THs-OTO had more RLN injury (OR, 1.38; 95% CI, 1.07-1.78) and accidental lacerations (OR, 2.03; 95% CI, 1.33-3.08) (Table 4). Median LOS was slightly less at THs-OTO than NTHs and THs (2.02 vs 2.12 vs 2.15 days, respectively). Median total charges were the same among NTHs, THs, and THs-OTO ($23 970.85, $23 389.84, and $23 574.45, respectively). Again, each point increase in CCI was associated with an increased risk for RLN injury (OR, 1.77; 95% CI, 1.62-1.94), hypocalcemia (OR, 1.44; 95% CI, 1.29-1.60), hypoparathyroidism (OR, 1.42; 95% CI, 1.36-1.44), and mortality (OR, 2.78; 95% CI, 2.35-3.30).
The total number of head and neck endocrine procedures was 34 685 for 2000 to 2002 and 39 770 for 2006 to 2008 (14.66% increase overall) with THs-OTO contributing a greater share of procedures in 2006 to 2008 (from 18% to 25%) (Table 5). Overall, there was an increase in the rate of hematomas (OR, 1.21; 95% CI, 1.06-1.38), hypoparathyroidism (OR, 1.27; 95% CI, 1.06-1.52), and unintentional vessel lacerations (OR, 1.36; 95% CI, 1.02-1.83) from 2000 to 2002, to 2006 to 2008. These overall increased frequencies corresponded with an increase in hematoma formation at NTHs (OR, 1.26; 95% CI, 1.04-1.52), an increase in hypoparathyroidism at THs (OR, 0.84; 95% CI, 0.75-0.94), and an increase in unintentional vessel lacerations at THs-OTO (OR, 1.98; 95% CI, 1.09-3.60). Furthermore, THs had an increase in rates of RLN injury (OR, 1.53; 95% CI, 1.11-2.10), but a decrease in rate of hypocalcemia (OR, 0.84 ; 95% CI, 0.75-0.94). Within THs-OTO, hypocalcemia (OR, 1.24; 95% CI, 1.06-1.45) increased. TH-OTO RLN injury did not change between the time periods (OR, 0.79; 95% CI, 0.59-1.07); however, when compared with NTHs in each time period, the OR decreased (from OR, 1.98; 95% CI, 1.46-2.68 in 2000 to 2002 to OR, 1.33; 95% CI, 1.03-1.72 in 2006 to 2008).
Overall mean charges increased from $12 978.29 to $23 708.11 (difference, $10 729.82; 95% CI, $10 312.66-$11 146.98) with consistent increases between each hospital type 9 (NTHs: from $12 852.98 to $23 970.85; difference, $11 117.87; 95% CI, $10 522.14-$11 713.60; THs: from $12 837.59 to $23 389.84; difference, $10 552.25; 95% CI, $9788.50-$11 316.00; THs-OTO: from $13 533.27 to $23 574.45; difference, $10 041.18; 95% CI, $9122.43-$10 959.93). Median LOS did not increase overall (2.1 days) but did show a decrease in THs-OTO (from 2.13 to 2.01 days; P < .002). Mortality rate decreased (OR, 0.66; 95% CI, 0.47-0.94), which corresponded to a decrease in mortality rate of procedures performed at THs-OTO (OR, 0.34; 95% CI, 0.12-0.93). Mortality rates for NTHs (0.02%) and THs (0.02%) remained the same.
The ultimate goal of graduate medical education duty hour restrictions is to improve patient outcomes by reducing sleep deprivation and improving performance of residents in training.
Certain complications and LOS have been shown to increase after the implementation of duty hour restrictions,17 although this finding remains controversial.18 Some teaching programs believe the new standards are barriers to effective training19-21 with advocacy for loosening the restrictions, increasing the length of residency, or restructuring programs. Opponents of the duty hour restrictions fear that patient outcomes will deteriorate rather than improve as a result of residents taking less ownership of patients, less time caring for medically complex patients, and less time improving surgical skills.22 Our findings suggest that duty hour restrictions have not been associated with an overall deterioration in patient outcomes for head and neck endocrine procedures.
Although RLN injury at THs and hypocalcemia and unintentional lacerations to vessels, nerves, or organs at THs-OTO increased slightly and could indicate a decrease in surgical proficiency, these complications could be explained by better coding when a complication occurs. In addition, while medical comorbidity could be adjusted for using the CCI, adjustment for surgical complexity could not be made. Teaching hospitals with otolaryngology programs may represent tertiary referral otolaryngology centers with a higher degree of complexity comparatively and may account for slightly worse outcomes, including unintentional laceration. Furthermore, and most important for THs-OTO, hematoma formation and RLN injury did not increase while mortality and LOS actually improved. Other factors, such as possible increased attending oversight or concentration of head and neck endocrine procedures into THs, could account for some differences noted. Nonteaching hospitals were to serve as a control in this investigation were complications to remain constant; however, an increase in hematoma formation was noted. This finding is worrisome, but it is difficult to fully elucidate the factors leading to it.
The best advantage to using such a robust data set like the NIS is that the power of the study is increased compared with most retrospective reviews. Owing to this major advantage, many investigators have characterized the effect of duty hours on patient outcomes in trauma surgery,23 neurosurgery,24 medicine,25 and orthopedic surgery.13 These studies failed to divide THs into those with and without that specialty’s respective training program. This discovery led to dividing the THs into 2 new groupings: THs and THs-OTO. Even with this distinction, THs-OTO may also have general surgery training programs, thus combining outcomes from both specialties. Moreover, the THs may aggregate outcomes from general surgery trainees and private surgeons. However, such a large data set should still reflect differences between the training specialties.
Large data sets come with several inherent limitations. Only ICD-9-CM procedure codes are used for the coding of procedures; Current Procedural Terminology codes are not found within the data set. Miscoding can occur within any data set that uses ICD-9-CM coding schema; however, miscoding should be present throughout the data set with relative consistency. The number of ancillary staff members, disposition issues following surgery- or patient-specific circumstances could not be assessed, nor could complications following discharge. Since the increase in some complications occurred after the duty hour restrictions took effect, sleep deprivation should not be associated to the same degree as previously; however, adherence to the duty hour policy by residents could not be assessed. There is no way to prove causation based on these findings but rather an association between the duty hour change and complications before and afterward.
Intraoperative nerve monitoring was not available until 2007 in the data sets owing to ICD-9-CM coding limitations (code 00.94 was not available prior to 2007); therefore, this variable could not be assessed in this study.26 Another confounder affecting rates of detection RLN injury and vocal cord paresis is the disparity in performing postoperative laryngoscopy between otolaryngologists and general surgeons. In a Scandinavian database study,27 routine laryngoscopy (which may be performed more often by otolaryngologists) was associated with a higher rate of vocal cord paresis suggesting an observation bias.
Two recent large-scale studies showed no better resident well-being following duty hour restrictions for general surgery residents: the FIRST trial28 and a systematic review29 of 135 articles pertaining to surgical resident duty hours. In addition, the FIRST trial demonstrated that a schedule with more flexible work hours was not inferior to the current mandated duty hour restrictions. These results may not be generalizable to otolaryngology residents.
Some opponents to the duty hour mandate question the ability of residents to perform the minimum number of key indicator procedures (22 thyroid and/or parathyroid procedures) and advocate lengthening otolaryngology residency to a 6-year program if these restrictions continue. An increase in intern time spent on otolaryngology service (from 3 months to 6 months) has already been realized. With steady otolaryngology applicant spots and match rate over the study period, (244 in 2000 vs 273 in 2008 [1.2% of all residency positions]),30-35 this study emphasizes that this key indicator procedure is concentrating into centers that have otolaryngology residents.
While RLN injury, hematoma formation, and hypoparathyroidism did not change, LOS, and mortality improved within THs-OTO following head and neck endocrine procedures after implementation of duty hour regulations. Although confounders within the NIS cannot be assessed, these findings refute the concern that duty hour restrictions result in poorer overall outcomes. Less time available to develop technical competence may play a factor in some outcomes in lieu of RLN injury increasing within TH and accidental injury to vessels, organs, or nerve and hypocalcemia increasing within THs-OTO. Furthermore, head and neck endocrine cases increased at THs-OTOs.
Corresponding Author: Merry Sebelik, MD, Department of Otolaryngology–Head & Neck Surgery, University of Tennessee Health Science Center, 910 Madison Ave, Ste 420, Memphis, TN 38163 (email@example.com).
Accepted for Publication: November 9, 2016.
Published Online: February 9, 2017. doi:10.1001/jamaoto.2016.4182
Author Contributions: Drs Wan and Sebelik 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: Smith, Sebelik.Concept and design: Smith, Sebelik.
Acquisition, analysis, or interpretation of data: Smith, Braden, Wan.
Drafting of the manuscript: Smith, Braden, Wan.
Critical revision of the manuscript for important intellectual content: Smith, Braden, Sebelik.
Statistical analysis: Smith, Wan.
Obtained funding: Sebelik.
Administrative, technical, or material support: Smith.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest, and none were reported.
Meeting Presentation: This article was presented as an oral presentation at the American Head and Neck Society Ninth International Conference on Head and Neck Cancer; July 16-20, 2016; Seattle, Washington.