Predicting reoperation and readmission for head and neck free flap patients using machine learning

BACKGROUND

To develop machine learning (ML) models predicting unplanned readmission and reoperation among patients undergoing free flap reconstruction for head and neck (HN) surgery.

METHODS

Data were extracted from the 2012-2019 NSQIP database. eXtreme Gradient Boosting (XGBoost) was used to develop ML models predicting 30-day readmission and reoperation based on demographic and perioperative factors. Models were validated using 2019 data and evaluated.

RESULTS

Four-hundred and sixty-six (10.7%) of 4333 included patients were readmitted within 30 days of initial surgery. The ML model demonstrated 82% accuracy, 63% sensitivity, 85% specificity, and AUC of 0.78. Nine-hundred and four (18.3%) of 4931 patients underwent reoperation within 30 days of index surgery. The ML model demonstrated 62% accuracy, 51% sensitivity, 64% specificity, and AUC of 0.58.

CONCLUSION

XGBoost was used to predict 30-day readmission and reoperation for HN free flap patients. Findings may be used to assist clinicians and patients in shared decision-making and improve data collection in future database iterations.

Operative Time Predicts Long-Term Abdominal Morbidity and Complication Requiring Treatment after DIEP Flap Breast Reconstruction

BACKGROUND

 The relationship between skin-to-skin operative time and long-term complications, as well as complications requiring treatment, after deep inferior epigastric perforator (DIEP) flap breast reconstructions has not been thoroughly investigated. The study objective was to evaluate if operative time would independently prognosticate the type and number of treated complications, as well as long-term abdominal morbidity.

METHODS

 Patients who underwent bilateral DIEP flaps from 2010 to 2021 by two senior surgeons with standardized surgical and postoperative protocols were retrospectively reviewed. Inclusion required at least 1 year of postoperative follow-up. Extensive multivariable regression analyses were utilized to adjust for potential confounders, including measures of intraoperative complexity.

RESULTS

 Three hundred thirty-five patients were entered into multivariable analyses. After risk adjustment, there was a significant increase in the risk of all treated complications per hour. Each hour of operation increased risk of complications requiring extended hospital stay for treatment by 25%, flap-site complications requiring readmission treatment by 24%, and flap-site complications requiring extended hospital stay for treatment by 26% (all p < 0.0001). Long-term abdominal morbidity (abdominal bulge) increased by 25% per hour of operative time (p < 0.0001). The number of complications requiring treatment, abdominal donor-site complications, and complications requiring extended hospital stay for treatment had statistically significant linear relationships with the duration of surgery (all p < 0.05).

CONCLUSION

 Operative time predicts long-term abdominal morbidity and complications requiring treatment, especially impacting the odds of developing abdominal bulges and complications requiring extended hospital stay for treatment or readmission treatment. This study emphasizes the importance of reducing operative time to improve DIEP flap breast reconstruction outcomes.

Unplanned reoperation after radical surgery for oral cancer: an analysis of risk factors and outcomes

Background

Unplanned reoperation (UR) after radical surgery for oral cancer (OC) is a health threat for the patients. The aim of the study was to identify the incidence of and risk factors for unplanned reoperation following oral cancer radical surgery, and to explore a potential role for long-term survival.

Methods

The present study followed a retrospective study design. Univariate and multivariate analyses were used to identify risk factors for demographic and clinical characteristics of patients. Survival analysis was performed by the Kaplan–Meier method. The data was analyzed statistically between November and December 2021.

Results

The incidence of UR was 15.7%. The primary cause of UR was reconstructed flap complications. Multivariate logistic regression analyses revealed that diabetes, tumor size, type of reconstruction, and nodal metastasis were independent risk factors for UR. Patients undergoing UR had a longer hospitalization, more post-operative complications, and a higher mortality compared with the non-UR group. UR is negatively correlated with the cancer-specific survival rate of patients (Log-rank test, P = 0.024).

Conclusion

Diabetes, tumor size, pedicled flap reconstruction and cervical nodal metastasis (N2) as independent risk factors for UR was discovered. UR was positively correlated with perioperative complications prolong hospital stay, and increased early mortality, but negatively correlated with the cancer-specific survival rate survival rate.

Postoperative depth of sedation and associated outcomes in free flap transfers to the head and neck

BACKGROUND

To evaluate the impact of postoperative depth of sedation in free flap transfers to the head and neck.

METHODS

A single center, retrospective cohort of 92 patients were stratified by depth of sedation, light sedation (RASS -1 or greater) or deep sedation (RASS less than -1), and analyzed for postoperative flap and medical complications.

RESULTS

Of the 92 patients 45 were included in the light sedation and 47 in the deep sedation group. Flap complication requiring return to the operating room occurred in 8 (22.2%) patients in light sedation compared to 12 (27.7%) (p = 0.450) patients in deep sedation. A composite outcome of flap and medical complications occurred less frequently in the light sedation group 14 (31.8%) compared to deep sedation 32 (69.6%) (p < 0.001).

CONCLUSION

There was no difference in return to the operating room between the two groups. Light sedation had reduced incidence of medical complications compared to deep.

The Utility of Comorbidity Indices in Assessing Head and Neck Surgery Outcomes: A Systematic Review

OBJECTIVE

To evaluate the utility of comorbidity index (CI) scores in predicting outcomes in head and neck surgery (HNS). The CIs evaluated were the Charlson Comorbidity Index (CCI), Elixhauser Comorbidity Index (ECI), Kaplan-Feinstein Index (KFI), American Society of Anesthesiologists Physical Status (ASA-PS), Adult Comorbidity Evaluation-27 (ACE-27), National Cancer Institute Comorbidity Index (NCI-CI), and the Washington University Head and Neck Comorbidity Index (WUHNCI).

METHODS

We report a systematic review according to the 2009 Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Electronic databases (PubMed, Cochrane, and Embase) and manual search of bibliographies identified manuscripts addressing how CI scores related to HNS outcomes.

RESULTS

A total of 116 studies associated CI scores with HNS outcomes. CIs were represented in the literature as follows: ASA-PS (70/116), CCI (39/116), ACE-27 (24/116), KFI (7/116), NCI-CI (3/116), ECI (2/116), and WUHNCI (1/116). The most frequently cited justification for calculating each CI (if provided) was: CCI for its validation in other studies, ACE-27 for its utility in cancer patients, and ECI for its comprehensive design. In general, the CCI and ACE-27 were predictive of mortality in HNS. The ECI was most consistent in predicting >1-year mortality. The ACE-27 and KFI were most consistent in predicting medical complications.

CONCLUSION

Despite inconsistencies in the literature, CIs provide insights into the impact of comorbidities on outcomes in HNS. These scores should be employed as an adjunct in the preoperative assessment of HNS patients. Comparative studies are needed to identify indices that are most reliable in predicting HNS outcomes.

LEVEL OF EVIDENCE

NA Laryngoscope, 2021.

Unplanned reoperations after microsurgical breast reconstruction: Findings from the American College of Surgeons National Surgical Quality Improvement Program

BACKGROUND

While microsurgical breast reconstruction may require multiple planned operations, unplanned reoperations has not been studied. We sought to investigate unplanned reoperations after microsurgical breast reconstruction.

METHODS

We queried the American College of Surgeons' National Surgical Quality Improvement Program between 2005 and 2018. Current Procedural Terminology code 19364 was used to identify all patients with microsurgical breast reconstruction. Patient demographics, medical comorbidities, preoperative laboratory results, and operative data were analyzed. The primary outcome measure was 30-day unplanned reoperation.

RESULTS

Of 8449 patients meeting inclusion criteria, 1021 required an unplanned reoperation (12.1%). These patients were more likely to be obese, smokers, hypertensive, on steroids preoperatively, needing concomitant mastectomy, and with prolonged operating room time >9 h (p < .05). Multivariable regression model revealed preoperative steroids intake (OR = 1.92, CI 1.09-3.38, p = .03), concomitant mastectomy (OR = 1.45, CI 1.23-1.71, p < .01), and operating room time >9 h (OR = 1.37, CI 1.16-1.62, p < .01) as independent risk factors. Mastectomy was found to be an independent risk factor for early reoperation, that is, ≤2 days (OR = 1.44, CI 1.14-1.82, p < .01), whereas obesity was an independent risk factor for three reoperations (OR = 3.92, CI 1.14-13.46, p = .03).

CONCLUSION

Unplanned reoperations within 30-days after microsurgical breast reconstruction are a significant problem. Mastectomy is an independent risk factor for early reoperation whereas obesity is an independent risk factor for multiple reoperations. Identification of such patients preoperatively may help microsurgeons improve patient safety and quality of care.

A Comparison of Common Plastic Surgery Operations Using the NSQIP and TOPS Databases

Both the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) and the American Society of Plastic Surgeons Tracking Operations and Outcomes for Plastic Surgeons (TOPS) databases track 30-day outcomes.

Methods

Using the 2008-2016 TOPS and NSQIP databases, we compared patient characteristics and postoperative outcomes for 5 common plastic surgery procedures. A weighted TOPS population was used to mirror the NSQIP population in clinical and demographic characteristics to compare postoperative outcomes.

Results

We identified 154,181 cases. Compared with NSQIP patients, TOPS patients were more likely to be younger (47.9 versus 50.0 years), have American Society of Anesthesiologists class I-II (92.1% versus 74.6%), be outpatient (66.0% versus 49.3%), and be smokers (18.7% versus 11.7%). TOPS had extensive missing data: body mass index (40.6%), American Society of Anesthesiologists class (34.9%), diabetes (39.3%), and smoking status (37.2%). NSQIP was missing <1% of all shared categories except race (15.6%). The entire TOPS cohort versus only TOPS patients without missing data had higher rates of dehiscence (5.1% versus 3.5%) and infection (2.1% versus 1.7%). TOPS versus NSQIP patients had higher dehiscence rates (5.1% versus 1.0%) but lower rates of return to the operating room (3.1% versus 6.6%), infection (2.1% versus 3.0%), and medical complications (0.3% versus 2.2%). Nonweighted and weighted TOPS cohorts had similar 30-day outcomes.

Conclusions

NSQIP and TOPS populations are different in characteristics and outcomes, likely due to differences in collection methodology and the types physicians using the databases. The strengths of each dataset can be used together for research and quality improvement.

Reconstruction with Free Flaps of Head and Neck Cancer Defects: A National Cohort Study

The purpose of this study was to determine the perioperative mortality rate, reintervention rate, and total healthcare costs for head and neck cancer patients who underwent free tissue transfer (FTT) in Colombia. The prognostic factors associated with those results were estimated.

Patient specific implants for jawbone reconstruction after tumor resection

In case of benign and malignant tumours affecting the maxillofacial region, the resection of jawbone reflects the standard therapy in more than 5.000 cases per year within the European Union. The resulting large bone defects lead to scarred, mangled facial appearance, loss of mastication and probably speech, requiring aesthetic and functional surgery as a basis for physical and physiological rehabilitation. Although autologous vascularized bone autografts reflect the current golden standard, the portion of bone available for the procedure is limited and subsequent high-dose anti-cancer chemo-/radiotherapy can lead to local tissue necrosis. Autologous vascularized bone from fibular or iliac-crest autografts is current golden standard in jawbone resection post-treatment, however, the portion of transplantable bone is limited and subsequent high-dose anti-cancer chemo-/radiotherapy often results in tissue necrosis Our research focuses on alternative treatment techniques: tissue reconstruction via novel patient-specifically manufactured maxillofacial implant that stimulates bone tissue growth. The planned neoformation of vascularized bone in such implants within the patient's own body as "bioreactor" is the safest approach in tissue engineering. The works described herein included the design of the metallic substrate of the implant with the use of computed tomography basing on real patients scans and then 3D-printing the substrates from the Ti6Al7Nb powder. The metal core was then evaluated in terms of structural characteristic, cytotoxicity and gene expression through the in vitro tests. Further experiments were focused on fabrication of the biocompatible coating for outer surface of the bone implant that would enhance the healing process and accelerate the tissue growth. Functional polymeric granulate dedicated for osteoconductive, osteoinductive and osteogenesis properties were elaborated. Another approach including the coating for the implant surface with two-phase biocompatible layer including polymeric microspheres and hydrogel carrier, which would provide long-time release of bone and cartilage growth factors around the implant were also done. The polymeric granulate containing βTCP improved bone cells growth, but it some modification has to be done in order to improve structural pores to ensure for better osteoconductivity. The biocompatible coating including PVP hydrogel and polymeric microspheres is still in the development process.

High-Risk Plastic Surgery: An Analysis of 108,303 Cases From the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP)

BACKGROUND

There is a lack of large-scale data that examine complications in plastic surgery. A description of baseline rates and patient outcomes allows better understanding of ways to improve patient care and cost-savings for health systems. Herein, we determine the most frequent complications in plastic surgery, identify procedures with high complication rates, and examine predictive risk factors.

METHODS

A retrospective analysis of the 2012 to 2016 American College of Surgeons National Surgical Quality Improvement Program plastic surgery data set was conducted. Complication rates were calculated for the entire cohort and each procedure therein. Microsurgical procedures were analyzed as a subgroup, where multivariate logistic regression models determined the risk factors for surgical site infection (SSI) and related reoperation.

RESULTS

We identified 108 303 patients undergoing a plastic surgery procedure of which 6 264 (5.78%) experienced ≥1 complication. The outcome with the highest incidence was related reoperation (3.31%), followed by SSI (3.11%). Microsurgical cases comprised 6 148 (5.68%) of all cases, and 1211 (19.33%) experienced ≥1 complication. Similar to the entire cohort, the related reoperation (12.83%) and SSI (5.66%) were common complications. Increased operative time was a common independent risk factor predictive of a related reoperation or development of an SSI (P < 001). Of all microsurgeries, 23.3% had an operative time larger than 10 hours which lead to faster increase in reoperation likelihood.

CONCLUSIONS

The complication rate in plastic surgery remains relatively low but is significantly increased for microsurgery. Increased operative time is a common risk factor. Two-team approaches and staged operations could be explored, as a large portion of microsurgeries are vulnerable to increased complications.

Scoping Review of the National Surgical Quality Improvement Program in Plastic Surgery Research

Background

The National Surgical Quality Improvement Program (NSQIP) is a robust, high-quality surgical outcomes database that measures risk-adjusted 30-day outcomes of surgical interventions. The purpose of this scoping review is to describe how the NSQIP is being used in plastic surgery research.

Methods

A comprehensive electronic literature search was completed in PubMed, Embase, MEDLINE, and CINAHL. Two reviewers independently reviewed articles to determine their relevance using predefined inclusion criteria. Articles were included if they utilized NSQIP data to conduct research in a domain of plastic surgery or analyzed surgical procedures completed by plastic surgeons. Extracted information included the domain of plastic surgery, country of origin, journal, and year of publication.

Results

A total of 106 articles met the inclusion criteria. The most common domain of plastic surgery was breast reconstruction representing 35% of the articles. Of the 106 articles, 95% were published within the last 5 years. The Plastic and Reconstructive Surgery journal published most of the (59%) NSQIP-related articles. All of the studies were retrospective. Of note, there were no articles on burns and only one study on trauma as the domain of plastic surgery.

Conclusion

This scoping review describes how NSQIP data are being used to analyze plastic surgery interventions and outcomes in order to guide quality improvement in 106 articles. It demonstrates the utility of NSQIP in the literature, however also identifies some limitations of the program as it applies to plastic surgery.

Prolonged operative duration is associated with complications: a systematic review and meta-analysis

BACKGROUND

The aim of this study was to systematically synthesize the large volume of literature reporting on the association between operative duration and complications across various surgical specialties and procedure types.

METHODS

An electronic search of PubMed, Cochrane Central Register of Controlled Trials, and Cochrane Database of Systematic Reviews from January 2005 to January 2015 was conducted. Sixty-six observational studies met the inclusion criteria.

RESULTS

Pooled analyses showed that the likelihood of complications increased significantly with prolonged operative duration, approximately doubling with operative time thresholds exceeding 2 or more hours. Meta-analyses also demonstrated a 14% increase in the likelihood of complications for every 30 min of additional operating time.

CONCLUSIONS

Prolonged operative time is associated with an increase in the risk of complications. Given the adverse consequences of complications, decreased operative times should be a universal goal for surgeons, hospitals, and policy-makers. Future study is recommended on the evaluation of interventions targeted to reducing operating time.

Quality Indicators: Measurement and Predictors in Head and Neck Cancer Free Flap Patients

Objective

To determine the predictors of length of stay (LOS), readmission within 30 days, and unplanned return to the operating room (OR) within 30 days in head and neck free flap patients.

Study Design

Case series with chart review.

Setting

Tertiary academic cancer hospital.

Subjects and Methods

All head and neck free flap patients at The Ohio State University (OSU, 2006-2012) were assessed. Multivariable logistic regression to assess the impact of patient factors, flap and wound factors, and intraoperative factors on the aforementioned quality metric outcomes.

Results

In total, 515 patients were identified, of whom 66% had oral cavity cancers, 33% had recurrent tumors, and 28% underwent primary radiotherapy. Of the patients, 31.5% had a LOS greater than 9 days, predicted by longer operative time, oral cavity and pharyngeal tumor sites, blood transfusion, diabetes mellitus, and any complication. A total of 12.6% of patients were readmitted within 30 days predicted by absent OSU preoperative assessment clinic attendance and any complication, and 14.8% of patients had an unplanned OR return predicted by advanced age.

Conclusions

When assessing quality metrics, adjustment for the complexity involved in managing patients with head and neck cancer with a high comorbidity index, clean contaminated wounds, and a high degree of primary radiotherapy is important. Patients seen in a preoperative assessment clinic had a lower risk of readmission postoperatively, and this should be recommended for all head and neck free flap patients. Quality improvement projects should focus on predictors and prevention of complications as this was the number one predictor of both increased length of stay and readmission.