Machine learning analysis of online patient questions regarding breast reconstruction.

BACKGROUND: Social media has become a dominant educational resource for breast reconstruction patients. Rather than passively consuming information, patients interact directly with other users and healthcare professionals. While online information for breast reconstruction has been analyzed previously, a robust analysis of patient questions on online forums has not been conducted. In this study, the authors used a machine learning approach to analyze and categorize online patient questions regarding breast reconstruction. METHODS: Realself.com was accessed and questions pertaining to breast reconstruction were extracted. Data collected included the date of question, poster's location, question header, question text, and available tags. Questions were analyzed and categorized by two independent reviewers. RESULTS: 522 preoperative questions were analyzed. Geographic analysis is displayed in Figure 1. Questions were often asked in the pre-mastectomy period (38.3%); however, patients with tissue expanders currently in place made up 28.5%. Questions were often related to reconstructive methods (23.2%), implant selection (19.5%), and tissue expander concerns (16.7%). Questions asked in the post-lumpectomy period were significantly more likely to be related to insurance/cost and reconstructive candidacy (p < 0.01). The "Top 6″ patient questions were determined by machine learning analysis, and the most common of which was "Can I get good results going direct to implant after mastectomy?" CONCLUSIONS: Analysis of online questions provides valuable insights and may help inform our educational approach toward our breast reconstruction patients. Our findings suggest that questions are common throughout the reconstructive process and do not end after the initial consultation. Patients most often want more information on the reconstructive options, implant selection, and the tissue expansion process.

Elective Neck Dissection for cT1-4 N0M0 Head and Neck Verrucous Carcinoma.

OBJECTIVE: To investigate the survival benefit of elective neck dissection (END) over neck observation in cT1-4 N0M0 head and neck verrucous carcinoma (HNVC). STUDY DESIGN: Retrospective cohort study. SETTING: The 2006 to 2017 National Cancer Database. METHODS: Patients with surgically resected cT1-4 N0M0 HNVC were selected. Linear, binary logistic, Kaplan-Meier, and Cox proportional hazards regression models were utilized. RESULTS: Of 1015 patients satisfying inclusion criteria, 223 (22.0%) underwent END. The majority of patients were male (55.4%) and white (91.0%) with disease of the oral cavity (67.6%) classified as low grade (90.0%) and cT1-2 (81.8%). The minority of ENDs (4.0%) detected occult nodal metastases. The rate of END increased from 2006 to 2017 for both cT1-2 (16.3% vs 22.0%, p = .126, R2 = 0.405) and cT3-4 (41.7% vs 70.0%, p = .424, R2 = 0.232) disease but these trends were not statistically significant. Independent predictors of undergoing END included treatment at an academic facility (adjusted odds ratio [aOR]: 1.75, 95% confidence interval [CI]: 1.19-2.55), cT3-4 disease (aOR: 3.31, 95% CI: 2.16-5.07), and tumor diameter (aOR: 1.09, 95% CI: 1.01-1.19) (p < 0.05). The 5-year overall survival (OS) of patients treated with and without END was 71.3% and 70.6%, respectively (p = .661). END did not significantly reduce the 5-year hazard of death (adjusted hazard ratio: 1.25, 95% CI: 0.91-1.71, p = .172). END did not significantly improve 5-year OS in univariate and multivariate analyses stratified by several patient, facility, tumor, and treatment characteristics. CONCLUSION: END does not confer an appreciable survival benefit in HNVC, even after stratifying univariate and multivariate analyses by several patient, facility, tumor, and treatment characteristics. LEVEL OF EVIDENCE: Level 4.

Analysis of Care and Outcomes for Epistaxis Weekend Admissions.

OBJECTIVES: To study differences in care of patients admitted for epistaxis during the weekend compared to the weekday. STUDY DESIGN: Retrospective database review. SETTING: 2003 to 2014 National Inpatient Sample. METHODS: Patients admitted for a primary diagnosis of epistaxis were extracted from the National Inpatient Sample from 2003 to 2014. Univariate and multivariate analyses were applied to assess differences in patient demographics, clinical characteristics, treatment, and outcomes between weekend and weekday admissions. RESULTS: A total of 39,329 cases were included in our study cohort, with 28,458 weekday admissions and 10,892 weekend admissions. There was no significant difference in patient race, gender, insurance status, hospital ownership status, or location between weekend and weekday admissions (p > .05). Most weekend admissions were emergent (82.2%) and were treated with packing (51.8%). Upon performing logistic regression, the likelihood of emergent admission (odds ratio [OR] 1.41, 95% confidence interval [CI] 1.32-1.51, p < .001) and prolonged length of stay (OR 1.11, 95% CI 1.05-1.17, p < .001) was higher for weekend admissions versus weekday admissions. Moreover, odds of packing for epistaxis were significantly higher (OR 1.14, 95% CI 1.09-1.19, p < .001) on the weekend, while odds of ligation (OR 0.88, 95% CI 0.80-0.97, p = .013) and endovascular arterial embolization (OR 0.74, 95% CI 0.65-0.84, p < .001) were lower. There were no significant differences in in-hospital mortality, patient discharge disposition, and total hospital charges (p > .05). CONCLUSION: Patients primarily admitted for epistaxis over the weekend were more likely to be emergent, experienced prolonged length of stay, and be treated nonoperatively with packing, than weekday admissions. No significant differences in patient insurance or hospital ownership were identified.

Use of artificial intelligence for the diagnosis of cholesteatoma.

Objectives: Accurate diagnosis of cholesteatomas is crucial. However, cholesteatomas can easily be missed in routine otoscopic exams. Convolutional neural networks (CNNs) have performed well in medical image classification, so we evaluated their use for detecting cholesteatomas in otoscopic images. Study Design: Design and evaluation of artificial intelligence driven workflow for cholesteatoma diagnosis. Methods: Otoscopic images collected from the faculty practice of the senior author were deidentified and labeled by the senior author as cholesteatoma, abnormal non-cholesteatoma, or normal. An image classification workflow was developed to automatically differentiate cholesteatomas from other possible tympanic membrane appearances. Eight pretrained CNNs were trained on our otoscopic images, then tested on a withheld subset of images to evaluate their final performance. CNN intermediate activations were also extracted to visualize important image features. Results: A total of 834 otoscopic images were collected, further categorized into 197 cholesteatoma, 457 abnormal non-cholesteatoma, and 180 normal. Final trained CNNs demonstrated strong performance, achieving accuracies of 83.8%-98.5% for differentiating cholesteatoma from normal, 75.6%-90.1% for differentiating cholesteatoma from abnormal non-cholesteatoma, and 87.0%-90.4% for differentiating cholesteatoma from non-cholesteatoma (abnormal non-cholesteatoma + normal). DenseNet201 (100% sensitivity, 97.1% specificity), NASNetLarge (100% sensitivity, 88.2% specificity), and MobileNetV2 (94.1% sensitivity, 100% specificity) were among the best performing CNNs in distinguishing cholesteatoma versus normal. Visualization of intermediate activations showed robust detection of relevant image features by the CNNs. Conclusion: While further refinement and more training images are needed to improve performance, artificial intelligence-driven analysis of otoscopic images shows great promise as a diagnostic tool for detecting cholesteatomas. Level of Evidence: 3.

Facility volume and survival: Human papilloma virus positive oropharyngeal squamous cell carcinoma.

BACKGROUND: To analyze the impact of facility volume on survival for human papilloma virus positive oropharyngeal squamous cell carcinoma (HPV+ OPSCC) patients. METHODS: Patients treated for HPV+ OPSCC from 2010 to 2017 were queried from the National Cancer Database. Facilities of average annual case volume <50th percentile were categorized as low-volume (LV) and >95th percentile as high-volume (HV). RESULTS: 11,546 were included, with 10,305 patients (89.3 %) treated at LV and 1241 (10.7 %) at HV facilities. A greater proportion of cases involving resection of base of tongue and lingual tonsil were treated at HV (30.3 %) compared to LV (22.3 %) facilities (p < 0.001). Patients treated at a HV facility had greater percentage of clinical T4 (11.2 % vs. 8.6 %, p = 0.001) and N+ disease (90.5 % vs. 85.7 %, p < 0.001) patients. Survival analysis showed no statistically significant difference between five-year overall survival rates by facility volume (p = 0.388) for all patients. On multivariable analysis, facility volume was not associated with survival (HR: 0.968 [0.758-1.235], p = 0.791). These trends were found for both patients undergoing primary surgery or chemoradiotherapy. CONCLUSION: Our data indicates that patients with HPV+ OPSCC do not experience a survival benefit with treatment at HV facility, suggesting these patients may be adequately treated at LV centers.

A machine learning analysis of patient concerns regarding mastopexy.

BACKGROUND: Social media plays an important role in connecting patients and plastic surgeons. We utilized patient inquiries regarding mastopexy from an online social media site to determine the most prevalent patient concerns, while employing a machine-learning algorithm to generate the questions representative of the dataset. OBJECTIVE: This data allow plastic surgeons to better tailor their preoperative consultations to address common concerns, set realistic expectations, and improve overall satisfaction. METHODS: A total of 2,011 inquiries from the mastopexy section of Realself.com were obtained using an open-source web crawler. Each inquiry was manually categorized as preoperative or postoperative and classified into subcategories based upon the free text entry. Lastly, questions were analyzed using machine learning to determine ten questions most representative of the inquiry pool. RESULTS: Of the 2,011 inquiries analyzed, 52.91% were preoperative and 47.09% were postoperative. Most preoperative questions asked about procedure eligibility (309, 29.04%), surgical techniques and logistics (260, 24.44%), and the best type of breast lift for the user (259, 24.34%). Among postoperative questions, questions regarding appearance were the most common (491, 51.85%), followed by symptoms after surgery (197, 19.75%) and behavior allowed/disallowed (145, 15.31%). Appearance was further subcategorized with the most common categories being appearance of the nipple (98, 19.86%), skin discoloration (88, 17.92%), and scarring (74, 15.07%). CONCLUSION: By utilizing the data that social media websites, like Realself.com, provide, plastic surgeons can better understand common patient concerns. This data aid in optimizing the preoperative consultation process to address the common concerns, recalibrate unrealistic expectations, and improve overall satisfaction.

Effect of Hospital Safety Net Burden on Survival for Patients With Sinonasal Squamous Cell Carcinoma.

OBJECTIVE: To examine factors associated with hospital safety net burden and its impact on survival for patients with sinonasal squamous cell carcinoma (SNSCC). STUDY DESIGN: Retrospective database study. SETTING: National Cancer Database from 2004 to 2016. METHODS: SNSCC cases were identified in the National Cancer Database. Hospital safety net burden was defined by percentage of uninsured/Medicaid patients treated, namely ≤25% for low-burden hospitals, 26% to 75% for medium-burden hospitals, and >75% for high-burden hospitals (HBHs). Univariate and multivariate analyses were used to investigate patient demographics, clinical characteristics, and overall survival. RESULTS: An overall 6556 SNSCC cases were identified, with 1807 (27.6%) patients treated at low-burden hospitals, 3314 (50.5%) at medium-burden hospitals, and 1435 (21.9%) at HBHs. On multivariate analysis, Black race (odds ratio [OR], 1.39; 95% CI, 1.028-1.868), maxillary sinus primary site (OR, 1.31; 95% CI, 1.036-1.643), treatment at an academic/research program (OR, 20.63; 95% CI, 8.868-47.980), and treatment at a higher-volume facility (P < .001) resulted in increased odds of being treated at HBHs. Patients with grade III/IV tumor (OR, 0.70; 95% CI, 0.513-0.949), higher income (P < .05), or treatment modalities other than surgery alone (P < .05) had lower odds. Survival analysis showed that hospital safety net burden status was not significantly associated with overall survival (log-rank P = .727). CONCLUSION: In patients with SNSCC, certain clinicopathologic factors, including Black race, lower income, treatment at an academic/research program, and treatment at facilities in the West region, were associated with treatment at HBHs. Hospital safety net burden status was not associated with differences in overall survival.

Impact of Facility Volume on Patient Safety Indicator Events After Transsphenoidal Pituitary Surgery.

OBJECTIVES: To investigate the impact of facility volume on Patient Safety Indicator (PSI) events following transsphenoidal pituitary surgery (TSPS). STUDY DESIGN: Retrospective database review. SETTING: National Inpatient Sample database (2003-2011). METHODS: The National Inpatient Sample was queried for TSPS cases from 2003 to 2011. Facility volume was defined by tertile of average annual number of TSPS procedures performed. PSIs, based on in-hospital complications identified by the Agency of Healthcare Research and Quality, and poor outcomes, such as mortality and tracheostomy, were analyzed. RESULTS: An overall 16,039 cases were included: 804 had ≥1 PSI and 15,235 had none. A greater proportion of male to female (5.8% vs 4.3%) and Black to White (7.0% vs 4.5%) patients experienced PSIs. There was an increased likelihood of poor outcome (odds ratio [OR], 3.1 [95% CI, 2.5-3.7]; P < .001) and mortality (OR, 30.1 [95% CI, 18.5-48.8]; P < .001) with a PSI. The incidence rates of PSIs at low-, intermediate-, and high-volume facilities were 5.7%, 5.1%, and 4.2%, respectively. Odds of poor outcome with PSIs were greater at low-volume facilities (OR, 3.3 [95% CI, 2.4-4.4]; P < .001) vs intermediate (OR, 3.1 [95% CI, 2.1-4.2]; P < .001) and high (OR, 2.5 [95% CI, 1.7-3.8]; P < .001). Odds of mortality with PSIs were greater at high-volume facilities (OR, 43.0 [95% CI, 14.3-129.4]; P < .001) vs intermediate (OR, 40.0 [95% CI, 18.5-86.4]; P < .001) and low (OR, 17.3 [95% CI, 8.0-37.7]; P < .001). CONCLUSION: PSIs were associated with a higher likelihood of poor outcome and mortality following TSPS. Patients who experienced PSIs had a lower risk of poor outcome but increased mortality at higher-volume facilities.

Analysis of the Treatment and Survival of Sinonasal Extramedullary Plasmacytoma.

BACKGROUND: While extramedullary plasmacytomas are infrequently encountered plasma cell malignancies, most cases occur in the head and neck, with a predilection for the sinonasal cavity. Due to the rarity of this disease, the majority of studies on sinonasal extramedullary plasmacytoma (SN-EMP) are case reports or small retrospective case series. OBJECTIVE: To investigate the impact of patient, disease, and treatment factors on the survival of patients with SN-EMP. METHODS: The National Cancer Database was queried for all patients with SN-EMP between 2004-2016 (N = 381 cases). Univariate and multivariate analyses were used to examine patient demographics, tumor characteristics, and survival. RESULTS: The majority of SN-EMP patients were over 60 years old (57.0%), male (69.8%), and white (86.2%). The most common treatment modality was radiotherapy alone (38.6%), followed by surgery plus radiotherapy (37.8%). Five-year overall survival was 74.0% and median survival was 9.1 years. Accounting for patient demographics and tumor characteristics in a multivariate model, the following groups had worse prognosis: 60 and older (HR 1.99, p = 0.031) and frontal sinus primary site (HR 11.56, p = 0.001). Patients who received no treatment (HR 3.89, p = 0.013), chemotherapy alone (HR 5.57, p = 0.008) or radiotherapy plus chemotherapy (HR 2.82, p = 0.005) had significantly lower survival than patients who received radiotherapy alone. Patients who received surgery with radiotherapy (HR 0.57, p = 0.039) had significantly higher survival than patients who received radiotherapy alone. CONCLUSION: In patients with SN-EMP five-year overall survival was found to be 74.0% with decreased survival associated with a frontal sinus primary site and being aged 60 or older. Patients receiving no treatment, chemotherapy alone, or radiotherapy with chemotherapy was associated with lower survival. Receiving surgery plus radiotherapy was associated with the highest five-year overall survival.

Applying Machine Learning to Determine Popular Patient Questions About Mentoplasty on Social Media.

PURPOSE: Patient satisfaction in esthetic surgery often necessitates synergy between patient and physician goals. The authors aim to characterize patient questions before and after mentoplasty to reflect the patient perspective and enhance the physician-patient relationship. METHODS: Mentoplasty reviews were gathered from Realself.com using an automated web crawler. Questions were defined as preoperative or postoperative. Each question was reviewed and characterized by the authors into general categories to best reflect the overall theme of the question. A machine learning approach was utilized to create a list of the most common patient questions, asked both preoperatively and postoperatively. RESULTS: A total of 2,012 questions were collected. Of these, 1,708 (84.9%) and 304 (15.1%) preoperative and postoperative questions, respectively. The primary category for patients preoperatively was "eligibility for surgery" (86.3%), followed by "surgical techniques and logistics" (5.4%) and "cost" (5.4%). Of the postoperative questions, the most common questions were about "options to revise surgery" (44.1%), "symptoms after surgery" (27.0%), and "appearance" (26.3%). Our machine learning approach generated the 10 most common pre- and postoperative questions about mentoplasty. The majority of preoperative questions dealt with potential surgical indications, while most postoperative questions principally addressed appearance. CONCLUSIONS: The majority of mentoplasty patient questions were preoperative and asked about eligibility of surgery. Our study also found a significant proportion of postoperative questions inquired about revision, suggesting a small but nontrivial subset of patients highly dissatisfied with their results. Our 10 most common preoperative and postoperative question handout can help better inform physicians about the patient perspective on mentoplasty throughout their surgical course. Level of Evidence V This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

Impact of Social Determinants of Health on Stereotactic Radiotherapy for Vestibular Schwannoma.

OBJECTIVES/HYPOTHESIS: Linear accelerator (LINAC) and Gamma Knife (GK) are common stereotactic radiation therapies for treating vestibular schwannoma (VS). There is currently limited literature examining specific demographic and socioeconomic factors, which influence the type of stereotactic radiation therapy a patient with VS receives. STUDY DESIGN: Retrospective database review. METHODS: The National Cancer Database was queried for cases of VS between 2004 and 2016. Patient demographic characteristics were compared using chi-squared and t-tests between GK and LINAC treated groups. Multivariate regression analysis was performed to assess predictors of stereotactic radiation therapy received. RESULTS: Of the 6,208 included patients, 5,306 (85.5%) received GK and 902 (14.5%) received LINAC. The mean age of GK patients was significantly lower than that of LINAC patients (58.0 vs. 59.7, P < .001). Individuals treated with GK had greater proportions of private insurance (P < .001) and incomes greater than $63,332 (P = .003). A greater proportion of GK patients were treated in academic centers (P < .001), in high-volume facilities (P < .001), in metropolitan areas (P < .001), and in the Northeastern United States (P < .001). On multivariate logistic regression analysis, region, metropolitan area, facility type, tumor size, and distance traveled by patients independently predict receipt of GK versus LINAC. CONCLUSION: Differences in patient demographics and other social determinants of health influence choice of GK versus LINAC therapy for VS patients. Future studies focused on addressing barriers to care, which may influence postprocedural quality of life and clinical outcomes associated with these two treatments are necessary to better understand the impact of these social differences. LEVEL OF EVIDENCE: 4 Laryngoscope, 132:2232-2240, 2022.

Inpatient Sinus Surgery Patient Morbidity and Outcomes: A National Analysis.

OBJECTIVES/HYPOTHESIS: To investigate the morbidity and mortality of patients undergoing endoscopic sinus surgery (ESS) in the inpatient setting. STUDY DESIGN: Retrospective database review. METHODS: The Nationwide Inpatient Sample was queried for all ESS between 2008 and 2014. Using All Patients Refined Diagnosis Related Groups (APR-DRG) codes, cases with APR-DRG codes under Major Diagnostic Category 3 (Diseases and Disorders of the Ear, Nose, Mouth, and Throat) were designated as patients with primary otolaryngology diagnoses undergoing ESS (ORL), and all other codes were designated as patients with non-otolaryngology pathologies as their primary reason for admission but undergoing ESS (non-ORL). A univariate analysis and a logistic regression were used to compare patient demographics, comorbidities, disease severity, and mortality. RESULTS: There were 8,305 ORL patients and 6,342 non-ORL patients. ORL patients were more likely to be elective admissions (61.3% vs. 48.5%, P < .001), have a deviated nasal septum (17.9% vs. 12.3%, P < .001), nasal polyps (15.8% vs. 5.0%, P < .001), obstructive sleep apnea (10.7% vs. 5.2%, P < .001), and pulmonary disease (15.9% vs. 10.5%, P < .001). Non-ORL patients had a higher likelihood of in-hospital mortality (odds ratio [OR] 6.22, 95% confidence interval [CI] 3.29-11.78, P < .001), length of stay in the highest quartile (OR 2.43, 95% CI 2.16-2.74, P < .001), and a higher proportion had APR-DRG subclasses indicating extreme severity of illness (19.3% vs. 4.3%, P < .001) or extreme risk of mortality (12.5% vs. 2.0%, P < .001). CONCLUSION: Patients undergoing ESS in the inpatient setting have a higher than expected mortality rate which can be associated with a non-otolaryngology pathology as the primary reason for their admission. LEVEL OF EVIDENCE: 4 Laryngoscope, 132:1523-1529, 2022.

Assessing Patient Satisfaction Following Blepharoplasty Using Social Media Reviews.

BACKGROUND: Because patient satisfaction is a significant qualitative consideration in measuring aesthetic surgery outcome, it is important to characterize the individual factors that shape the patient perspective about blepharoplasty. OBJECTIVES: This study analyzed reviews by blepharoplasty patients on the aesthetic surgery social media platform RealSelf.com to determine which aspects of the surgical process have the most significant impact on patient satisfaction. METHODS: Blepharoplasty reviews were gathered from RealSelf.com with an automated web crawler. These reviews were characterized as positive or negative, then given a specific category that more specifically defined the theme of the review. Additional variables, including the specialty of the reviewed physician and any patient self-reported ratings, were documented. RESULTS: A total of 1991 reviews pertaining to blepharoplasty were collected. Among reviews with self-reported "worth it" ratings, 93.5% were positive. Following categorization of all reviews, 1865 (93.7%) were positive and 126 (6.3%) were negative. Of the positive reviews, the most common overall themes were bedside manner (n = 899, 48.2%), aesthetic result (n = 859, 46.1%), and overall comfort (n = 58, 3.1%). Among negative reviews, most pertained to aesthetic result (n = 100, 79.4%), and bedside manner (n = 14, 11.1%). The most frequently encountered physician specialties performing blepharoplasty were plastic surgery (n = 1101, 55.3%), ophthalmology (n = 634, 31.8%), and otolaryngology (n = 69, 3.5%). CONCLUSIONS: The majority of reviews were positive. The most prominent factor driving positive reviews was bedside manner, followed by aesthetic results. Negative reviews were most frequently attributed to suboptimal aesthetic results. Most blepharoplasties in our study cohort were performed by plastic and oculoplastic surgeons.

Trends in HPV Testing for Patients With Sinonasal Squamous Cell Carcinoma: A National Analysis.

OBJECTIVE: The objective of this study was to analyze national trends in human papillomavirus (HPV) testing for patients diagnosed with sinonasal squamous cell carcinoma (SNSCC). STUDY DESIGN: Retrospective database study. SETTING: National Cancer Database (2010-2016). METHODS: Cases from 2010 to 2016 with a primary SNSCC diagnosis and known HPV testing status were extracted from the National Cancer Database. Univariate and multivariate analyses were then performed to assess differences in socioeconomic, hospital, and tumor characteristics between tested and nontested patients. RESULTS: A total of 2308 SNSCC cases were collected, with 1210 (52.4%) HPV tested and 1098 (47.6%) not tested. On univariate analyses, patient age, insurance, income quartile, population density, treatment facility location, and tumor grade were significantly associated with HPV testing status. After multivariate logistic regression modeling, living in a suburban area had lower odds of HPV testing as compared with living in urban areas (odds ratio, 0.74 [95% CI, 0.55-0.99]; P = .041), while tumor grade III/IV had higher odds than grade I (odds ratio, 1.73 [95% CI, 1.29-2.33]; P < .001). HPV-tested patients had a similar 5-year overall survival to nontested patients (48.3% vs 45.3%, log-rank P = .405). A sharp increase in HPV testing rates was observed after 2010 (P < .001). CONCLUSION: Among patients with SNSCC, those with high-grade tumors were more likely to be tested for HPV, while patients with a suburban area of residence were less likely. Additionally, there was no significant survival benefit to HPV testing, with tested and nontested groups having similar overall survival. LEVEL OF EVIDENCE: 4.

Exploratory analysis on the association of mental health disorders with in-hospital postoperative complications and mortality in head and neck cancer surgery.

BACKGROUND: The objective was to assess the association of mental health disorders with in-hospital complication and mortality rates in patients undergoing head and neck cancer surgery. METHODS: In this exploratory retrospective study, the Nationwide Inpatient Sample was queried from 2003 to 2014 for all patients with a diagnosis of head and neck cancer who underwent surgery. Univariate cross-tabulation, logistic regression, and propensity score matching (PSM) were used to compare demographics, procedure-related variables, and in-hospital postoperative complications and mortality between patients with and without selected comorbid mental health disorders. RESULTS: Of 39 600 included patients, 3390 (8.6%) had a selected comorbid mental health disorder diagnosis. After PSM, patients with selected mental health disorders had increased risk of overall medical complications on multivariable analysis (OR 1.28 [CI 1.12-1.46], P < 0.001) but not overall surgical complications or mortality. CONCLUSIONS: Patients with a mental health disorder diagnosis have increased risk of in-hospital medical, certain surgical, and total complications.