Perioperative complications among head and neck surgery patients with COVID‐19

Patients undergoing surgery for head and neck cancer (HNC) have potentially high perioperative complication rates. Recent studies indicate that preoperative COVID‐19 infection poses increased risk for postoperative complications in other fields. However, to date, there has not been data showing the effect of COVID‐19 on complication rates for HNC. Here, a large database was employed to assess if perioperative COVID‐19 increased the risk of perioperative complications among those undergoing HNC surgery.


K E Y W O R D S
COVID-19, head and neck cancer, perioperative complications, perioperative mortality, surgical complications

| INTRODUCTION
Head and neck cancer (HNC) is the sixth most prevalent form of cancer globally, with squamous cell carcinoma (SCC) far and away the most common pathology in the head and neck. 1 While treatment is often multidisciplinary, surgical management, either in the primary or salvage setting, remains a mainstay in head and neck cancer treatment.While surgery represents the preferred treatment option for certain tumor subsites, including salivary, oral cavity and advanced laryngeal cancers, postoperative complications, including hemorrhage, risk for airway compromise, hematomas, and wound dehiscence, can pose a significant risk of morbidity and mortality in the perioperative period. 2 Prior to the onset of the COVID-19 pandemic, the rate of perioperative complications and mortality for HNC patients undergoing surgery was estimated to be 32.6% and 0.3%, respectively. 2In the initial months of the pandemic during the first wave of COVID-19 infection, reports demonstrated that perioperative mortality for all surgeries increased to 23.8% for those infected with the virus. 3Postoperative pulmonary problems were found to be significantly increased in COVID-19 patients undergoing both elective and emergency surgeries, with 50% of COVID-19 patients experiencing postoperative pulmonary dysfunction compared to 8% of non-COVID patients. 3In fact, respiratory compromise was linked to the cause of death in 81% of patients who died in the perioperative period. 3hereas most institutions instituted a moratorium on elective and nontime sensitive surgeries following the declaration of COVID-19 as an international pandemic, in most institutions, time-sensitive oncologic surgeries were allowed to continue given the risk of disease progression and death in the absence of timely, appropriate treatment.At the authors' home institution, a multidisciplinary board was convened to review and approve surgeries as being time-sensitive and appropriate to proceed in the setting of the pandemic.Though routine preoperative testing had become widespread by the summer of 2020, allowing for identification of asymptomatic COVID-19 patients prior to scheduled surgery, postoperative HNC cancer patients, whose hospital stays often last 7 days or longer, may have still been in proximity to COVID-infected patients in the intensive care unit (ICU) and floor settings or have received care from medical staff who were caring for COVID-positive individuals in addition to COVID-negative patients.While the risk of pulmonary dysfunction was not unexpected given the characteristic multifocal pneumonia caused by COVID which proved lethal despite the most advanced interventions including mechanical ventilation, prone positioning, and extracorporeal membranous oxygenation (ECMO), increased risk of additional complications was identified in certain subsets of patients, with nearly 25-fold increased hemorrhage rates and higher rates of blood transfusions in COVID-positive patients undergoing cardiac surgery. 4hile studies in other surgical specialties have demonstrated increased risk of perioperative complications among patients with COVID-19, no broad analysis exists examining the impact of COVID-19 infection on perioperative complication in patients undergoing HNC surgery.While the public health emergency and national state of emergency due to COVID-19 was officially ended on May 11, 2023, new variants of COVID-19 continue to emerge, and the COVID-19 infection is expected to continue to cause infection for many years into the future.In the authors' personal experience, while the latest COVID variants have trended towards being less virulent and more infectious, in the HNC patient population, perioperative COVID-19 infection can still result in significant complications.In this light, this study seeks to perform a retrospective, multi-institutional study comparing perioperative complication rates in patients with and without COVID-19 infection undergoing HNC surgery.

| Study design and participants
Institutional Review Board (IRB) exemption was provided by Georgetown University's IRB.
A retrospective observational study was conducted using the TriNetX Research Network, a federated multiinstitutional network of health data.The database includes deidentified electronic health records (EHRs) with approximately 120 million unique subjects available at the time of analysis.Cohorts were designed using International Classification of Diseases-Clinical Modification (ICD-CM) codes and Current Procedure Terminology (CPT) codes.
T A B L E 1 Boolean operators were employed where items within each category were paired as "or" and each category (COVID-19 codes, malignancy codes, and resection codes) was paired as "and" to other categories.

| Data collection
Final database queries and analyses were performed on February 2023.For all cohorts, ICD and CPT codes were used to identify subjects that received surgical resection for head and neck malignancy.The exposure cohort included those who were COVID-19 positive 7 days before or after surgery compared to controls who were negative during this same time period (Table 1.).Complications were measured from the day of surgery until 30 days postoperatively.

| Statistical analysis
Statistical analyses were performed within the database.The database employs 1:1 propensity score matching using the greedy nearest neighbor algorithm with a caliper width of 0.1 and pooled standard deviation.The output order of rows was randomized to eliminate bias secondary to the nearest neighbor algorithm.Odds ratios (OR) with 95% confidence intervals were calculated for each complication.The Bonferroni correction method was employed to account for multiple comparisons where an initial alpha level of 0.05 was divided by the number of tests performed (0.05/n = 15).Subsequently, p values less than 0.003 were considered statistically significant.Cohorts were matched for sex, age, and race.Since protected health information might be revealed with relatively small cohorts, the database conceals exact patient counts when ≤10 subjects meet the index event for each baseline characteristic and associated complication.Analyses were conducted for matched and unmatched cohorts.
Two hundred eight subjects were identified as COVID-19 positive in the perioperative period and 15 158 were negative.The mean age at time of surgery for COVID-19 positive and negative was 61.4 (SD 14.6) and 62.7 (SD 14.7) years (p = 0.18), respectively.The majority of subjects in each positive and negative cohort were Caucasian, n = 172 (82.7%) and 12 222 (80.6%), p = 0.45, respectively (Table 2).

| DISCUSSION
A large, multi-institutional research network was used to assess if COVID-19-positive patients undergoing HNC surgery are at increased risk for perioperative complications compared to negative controls.Retrospective review using a large database found that patients positive within 7 days before or after surgery were at significantly higher risk for multiple medical and surgical complications in the perioperative period.While a recent observational study conducted at a single institution in South India showed similar results, this is the largest retrospective multi-institutional study demonstrating increased risk of perioperative complication due to COVID-19 infection to the author's best knowledge. 5OVID-19 complications and mortality have largely been associated with pulmonary and cardiovascular pathologies.Interestingly, the pulmonary complications associated with the COVID-19 infection deviate from the typical pathophysiology seen in acute respiratory distress syndrome (ARDS).Characteristically, the manifestation of ARDS begins with diffuse alveolar damage with edema, hyaline membranes, and inflammation with type II pneumocyte hyperplasia, impairing lung compliance and mechanics. 6However, COVID-19 patients were found to have preserved lung compliance and mechanics despite severe hypoxia and prolonged requirements for mechanical ventilation. 6Magro et al. observed distinct histologic findings in COVID-19-positive patient pulmonary tissue with increased deposition of complement components C5b-9, C4d, and MASP2 in vessels and colocalization of complement with the SARS-CoV-2-specific spike glycoprotein, suggesting major injury induced by the virus from over-activation of immune complement pathways (alternative and lectin) directing damage at the microvascular level. 6Invasion of the virus is facilitated by the angiotensin-converting enzyme II (ACE-2) receptor, which is characteristically highly expressed in alveolar epithelial cells and capillary endothelium, further explaining the pathophysiology of virally induced pulmonary damage. 4n array of coagulopathies have been reported in COVID patients, with lab value alterations including thrombocytopenia, prolonged prothrombin (PT), and activated partial thromboplastin time (aPTT). 7The presence of disseminated intravascular coagulation in COVID-19 patients has been reported as a potential cause of FTT failure.As early studies showed improved survival in COVID-19 patients with early initiation of anticoagulant therapy, a retrospective, matched casecontrol study of patients undergoing open-heart cardiac surgery revealed a concerning an approximately 25-fold increase in early and late bleeding complications among perioperative COVID-19-positive patients. 4As arterial or venous thromboembolism represents one of the greatest complications of microvascular free tissue transfer, given the potential for increased free flap complication, several authors even altered their surgical practices during the pandemic, performing more regional tissue transfer when applicable. 8 To avoid observed COVID-associated complications, an institution in the UK even implemented a strict COVID-19 protocol at the peak of the pandemic with a small cohort of patients undergoing head and neck surgery, including rigorous preoperative testing and selfisolation measures prior to the procedure. 9Only two of their patients faced complications, both were COVID-19-negative and therefore these outcomes were not attributed to the virus.Our data does demonstrate increased postoperative free flap complications and failures in our COVID-19-positive HNC cohort, confirming the concern expressed by microvascular surgeons regarding the deleterious effect of COVID on patients undergoing microvascular-free tissue reconstruction.
Evidence of an alarming 4-fold increase in postoperative mortality among COVID-19-positive patients necessitates special attention and medical vigilance. 10Research shows an even greater 6-fold increase in mortality in patients who test positive for COVID-19 within 2 weeks of their procedure (including General Surgery, Neurosurgery, Orthopedics, Gynecology, Thoracic, and Vascular). 10Male sex, higher American Society of Anesthesiologists (ASA) Physical Status Classification score, and smoking history were noted to be strong factors associated with mortality among individuals previously infected with COVID. 10 This data supports our findings showing increased 30-day postoperative mortality in HNC patients with perioperative COVID-19 infection.Another UK study similarly found a remarkable increase in postoperative mortality among patients with hip fracture repair surgery ($40% of patients COVID-19 positive) with an association with male gender and higher age. 11The authors noted predicting factors for

Note:
Odds ratio (OR) with (95% confidence interval).p values less than 0.003 were considered statistically significant.
Abbreviations: CVA, cerebral vascular accident; MI, myocardial infarction.a When <11 subjects meet the outcome event, the platform conceals the exact number of subjects to protect health information.
MARTIN ET AL. mortality, observing increased oxygen requirements and rapid decline seen in chest imaging in the immediate postoperative period as warning signs. 11Altogether, this evidence highlights the critical need for critical need for healthcare professionals to be vigilant in managing postoperative care for COVID-19-positive patients, particularly high risk-individuals and those who test positive within 2 weeks of their operation, in order to reduce fatal risks and improve postoperative outcomes.
The retrospective nature of our study limits overall power of the investigation.Since data was pulled using ICD and CPT codes, patients' exact health status and procedures underwent may have been more nuanced than what the database was able to capture due to the definitions applied to each ICD or CPT code.The CPT code employed for free flap complication was T86.828 which is a global code encompassing a range of complications (e.g., arterial insufficiency of flap, flap necrosis, partial thickness flap loss, and venous thrombosis of flap).Therefore, the presence of COVID-19 infection cannot be attributed to any specific free flap complication.Additional retrospective research may be required to understand which exact complications are associated with the presence of infection.
Furthermore, patients who underwent surgery with a positive COVID-19 test within 7 days of surgery may be characteristically different than patients who were negative.Delaying surgery may not have been feasible for these patients due to disease burden, whereas patients with less disease burden could be delayed until clear of COVID-19.Subsequently, the disease burden might explain the increased risk of postoperative complications observed in positive subjects.For example, a patient with a tumor leading to airway compromise may require urgent surgery despite a recent COVID infection, and the presence of airway compromise by tumor itself may be associated with poorer 30-day outcomes.These possibly confounding variables could not have been detected by the ICD code database search.However, the database receives data from a geographically diverse source of contributors owing to a pool of approximately 120 million subjects from 95+ institutions available for study.Despite this large pool of available subjects, there were not enough COVID-19-positive subjects who underwent HNC surgery to allow for PSM.As COVID-19 remains endemic in the population, future studies might reinvestigate this question when more cases have accrued allowing for PSM.Additionally, with the availability of COVID-19 antivirals, future prospective investigations might explore whether these medications decrease the risk of the complications highlighted here. 12I G U R E 1 Forest plot of association between perioperative COVID-19 infection and risk of 30 days surgical and medical complications.This large populational study suggests HNC patients diagnosed with COVID-19 infection in the perioperative period are at increased risk for death and several perioperative complications, including free tissue transfer failure.This study employed a large, diverse population and is the first study to address this clinical question.Limitations include the retrospective nature of the study and the descriptions inherent to each ICD and CPT code.
[Color figure can be viewed at wileyonlinelibrary.com] Risk of 30 days surgical complications among COVID-19 positive compared to controls.When <11 subjects meet the outcome event, the platform conceals the exact number of subjects to protect health information.Risk of 30 days medical complications among COVID-19 positive compared to controls.
T A B L E 3 a T A B L E 4