Characteristics of Patients Requiring Prolonged Mechanical Ventilation Following Head and Neck Microvascular Reconstructive Procedures

Background: Postoperative failure to wean from ventilator support is associated with poor outcomes. This study characterizes patients requiring prolonged ventilation after head-and-neck reconstructive surgery and their outcomes.Methods: Patients who underwent head-and-neck reconstructive surgery were identified in the National Surgical Quality Improvement Program (NSQIP) database using CPT codes, ICD-10 codes, and free-text search. Hypothesis testing and regression analysis was utilized to compare patient characteristics and peri/postoperative complication rates between those with and without extended ventilation.

NSQIP contained 4661 patients who underwent free flap surgery; 180 experienced prolonged ventilation. Patients with primary oral cavity/sinonasal malignancy, preoperative ventilator dependence, and preoperative open wound/wound infection were more likely to experience prolonged ventilation. Prolonged ventilation patients experienced double the overall hospital stay, and increased rates of wound disruption, reoperation, and mortality. (p <0.05). Conclusion: Prolonged mechanical ventilation can significantly affect outcomes of free flap patients, including longer hospital stays, and higher 30-day mortality. Characteristics of Patients Requiring Prolonged Mechanical V entilation F ollowing Head and Neck Microvascular Reconstructive Procedures

Background: Postoperative failure to wean from ventilator support is associated with poor outcomes. This study characterizes patients requiring prolonged ventilation after head-and-neck reconstructive surgery and their outcomes.

Patients who underwent head-and-neck reconstructive surgery were identified in the National Surgical Quality Improvement Program (NSQIP) database using CPT codes, ICD-10 codes, and free-text search. Hypothesis testing and regression analysis was utilized to compare patient characteristics and peri/postoperative complication rates between those with and without extended ventilation.

NSQIP contained 4661 patients who underwent free flap surgery; 180 experienced prolonged ventilation. Patients with primary oral cavity/sinonasal malignancy, preoperative ventilator dependence, and preoperative open wound/wound infection were more likely to experience prolonged ventilation. Prolonged ventilation patients experienced double the overall hospital stay, and increased rates of wound disruption, reoperation, and mortality. (p <0.05). 1 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary.

Prolonged mechanical ventilation can significantly affect outcomes of free flap patients, in- cluding longer hospital stays, and higher 30-day mortality. Key Points: 1. Prior ventilator dependence, preoperative open wound/infection, and primary oral cavity/sinonasal malignancy were associated with higher rates of prolonged postoperative mechanical ventilation after head-and-neck reconstructive surgery. 2. No significant association was found between diabetes, hypertension, chronic obstructive pulmonary disease, or smoking status and prolonged ventilation. 3. Prolonged ventilation correlated with higher rates of postoperative complications including hemorrhage requiring transfusion, wound complications, and 30-day reoperation. 30-day mortality was also greater in patients who experienced prolonged ventilation after head-and-neck reconstructive surgery. 4. Patients who underwent prolonged ventilation after head-and-neck reconstructive surgery experienced nearly double the length of total hospital stay compared to patients with normal extubation times. 5. Concurrent tracheostomy placement during head-and-neck reconstructive surgery did not reduce rates of prolonged ventilation. 6.

Head-and-neck oncologic operations represent major surgical endeavors that require tumor ablative surgery, as well as reconstructive surgery often with free tissue transfer. These operations are performed in pa- tients who have various preexisting comorbidities including peripheral vascular disease, chronic obstructive pulmonary disease (COPD), dementia, diabetes mellitus, or a history of cerebrovascular accident/transient ischemic attack, complicating both intraoperative management and postoperative recovery 1,2. Occasionally, patients may undergo postoperative mechanical ventilation following their surgery3,4. Prolonged mechanical ventilation has been associated with increased patient morbidity and mortality. Specifically, it has been associated with increased rates of barotrauma, pneumonia, and increased hospital stay 3,5-7. In patients who undergo major head-and-neck surgery, mechanical ventilation has been associated with increased rates of pneumonia, pulmonary embolism, and cardiopulmonary arrest 8. Despite the risks that are known, there remains a gap in the literature regarding the characteristics of patients who undergo prolonged postoperative mechanical ventilation following head-and-neck surgery. Recent litera- ture has begun to shed light on the impacts of prolonged mechanical ventilation. Here, we leverage the large sample size of a national/multi-institutional database to characterize prolonged mechanical ventilation in head-and-neck free tissue transfer. Utilizing data from the National Surgical Quality Improvement Program (NSQIP), this study aims to characterize the patient population requiring extended mechanical ventilation. Furthermore, this study seeks to analyze the thirty-day outcomes of patients who undergo prolonged me- chanical ventilation after head-and-neck surgery, including the incidence of postoperative complications and mortality.

Data Acquisition The NSQIP database was queried to identify patients who underwent various microvascular reconstructive procedures between the years of 2016 and 2020. Patients were initially identified in the NSQIP database by using Current Procedural Terminology (CPT) codes corresponding to free muscle, free myocutaneous, free skin, free fascial, free omental, free jejunal, free osseus or osteocutaneous flaps with microvascular anastomosis. After identifying free flap patients, this group was subsequently filtered to include only patients who under- went specifically head-and-neck microvascular reconstructive surgery. Only patients whose primary surgeon was an otolaryngologist, or who had primary head-and-neck cancer or osteonecrosis as the primary indi- cations for their head-and-neck surgery were included. This stratification was done through employing 2 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. International Classification of Diseases 10 (ICD-10) diagnostic codes as well as free-text searching patient surgical indications for terms including cancer, neoplasm, or osteonecrosis. No specific sites of osteonecrosis (e.g., jaw) were specified in this study protocol. ICD-10 codes that were utilized corresponded to various head-and-neck cancers including tongue, oral cavity, floor of mouth, palate, maxilla, oropharynx, laryngeal, epiglottis, salivary, face skin, facial bone, and scalp cancer. The full list of CPT codes, ICD-10 diagnostic codes, and free text criteria that were implemented in group selection are depicted in Figure 1. Patient Characteristics The head-and-neck patient group was subsequently further separated into a subgroup that experienced prolonged postoperative mechanical ventilation and a subgroup that did not. Prolonged postoperative ven- tilation was defined as patients who remained mechanically ventilated for greater than forty-eight hours following their surgical start time. Hypothesis testing was performed to compare patient characteristics / preoperative factors between the prolonged ventilation and control groups. Patient characteristics that were deemed statistically significant per hypothesis testing were subsequently entered into univariate and enter-variable multivariate logistic regression analysis. Odds ratios (OR) with 95% confidence intervals (CI) were calculated for each patient characteristic. ICD-10 diagnostic codes were used to quantify primary malignancy sites, as well as identify patients who experienced free flap graft failure or complications. CPT codes were used to identify patients who underwent concurrent tracheostomy and total laryngectomy with tracheostomy alongside free flap reconstruction. In addition, CPT codes were used to identify which patients experienced concurrent ablative surgery. The full list of CPT and ICD-10 codes that were utilized for this purpose are depicted in Supplemental Figure 1. Performance of the constructed multivariate regression model was assessed with an Akaike Information Cri- terion, Concordance statistic, and Hosmer-Lemeshow test. Each predictor in the final multivariate regression model was approximated for respective statistical importance through Likelihood Ratio (LR) Chi-Squared testing. Peri/Postoperative Outcomes Hypothesis testing was also used to compare rates of peri and postoperative outcomes between the prolonged ventilation and control groups to evaluate whether prolonged ventilation was associated with increased rates of specific outcomes in head-and-neck free flap patients. Technical Details Figure 2 demonstrates a flowchart highlighting the inclusion criteria and overall patient filtration methodol- ogy. All data analysis was performed in R (version 4.4.2, 10/31/2024). Hypothesis testing involved T-testing for all continuous variables and Chi-Squared testing for all categorical variables. Statistical significance was set at an α level of 0.05. This study was a retrospective database study.

Patient Characteristics A total of 4,661 patients who underwent head-and-neck free flap surgery were included for analysis, with 180 patients identified as part of the prolonged mechanical ventilation group. Patients in the prolonged ventilation group were more often white (74.4% vs. 64.2%, p = 0.005), transferred from another healthcare facility rather than being admitted from home (95.0% vs. 97.7%, p = 0.019). Additionally, the prolonged ventilation group had a higher prevalence of diabetes (20.0% vs. 14.4%, p = 0.040), hypertension (56.1% vs. 46.4%, p = 0.011), smoking within a year of their surgery (37.2% vs. 29.5%, p = 0.027), dyspnea at rest or exertion (p < 0.001), COPD (13.9% vs. 6.7%, p < 0.001), pre- operative ventilator dependence up to 48 hours prior to their surgery (2.2% vs. 0.2%, p < 0.001), and having an open wound/wound infection preoperatively (12.2% vs. 6.4%, p = 0.002). Patients who underwent prolonged ventilation following head-and-neck surgery experienced higher rates of requiring preoperative 3 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. blood products (2.2% vs. 0.5%, p = 0.007). Patients in the prolonged ventilation group had lower rates of undergoing their head-and-neck surgery as an elective operation (90.0% vs. 94.0%, p = 0.029) (Table 1). Free flap patients who had primary malignancies of the oral cavity (50.6% vs. 39.1%, p=0.002) or nasal cavity/sinuses (6.1% vs. 2.4%, p=0.002) experienced higher rates of requiring prolonged postoperative ventilation. Rates of tracheostomy placement among this group of patients with primary oral cavity / sinonasal malignancy were not significantly different between those who experienced prolonged ventilation and those who did not (64.1% vs. 59.3%, p >0.05; 52.3% vs. 36.4%, p >0.05). Patients with cutaneous primary malignancies experienced lower rates of prolonged ventilation (3.9% vs. 11.2%, p=0.002). Patients who had primary malignancies of the oropharynx, larynx, or salivary gland did not experience a significantly different rate of prolonged ventilation compared to control. Patients who underwent concurrent tracheostomy placement alongside free flap surgery or who underwent total laryngectomy with tracheostomy placement did not experience decreased rates of prolonged postoperative ventilation. Additionally, patients who underwent ablative surgery alongside free flap reconstruction, or experienced graft failure / other graft complications did not experience higher rates of prolonged postoperative ventilation (p >0.05) (Table 1). Hematologic and nutritional status markers were also different between groups, with the prolonged ventilation group exhibiting lower hematocrit levels (38.7 vs. 40.0; p = 0.010) and lower serum albumin levels (3.90 vs. 4.00; p = 0.002). (Table 1). All the variables that were deemed statistically significant per hypothesis testing remained statistically significant upon univariate regression testing. Multivariate logistic regression analysis identified previous ventilator dependence up to 48 hours prior to surgery (OR = 4.67, p=0.022, LR = 10.65) and preoperative open wound/wound infection (OR = 2.53, p=0.015, LR = 9.06) as significant predictors of prolonged post- operative mechanical ventilation. Primary malignancies of the oral cavity (OR = 2.06 [1.33-3.24], p=0.001, LR = 17.79) and nasal cavity/sinuses (OR = 3.80 [1.56-8.33], p=0.002, LR = 11.45) were also deemed to be significant predictors of requiring postoperative prolonged ventilation. Conversely, non-white ethnicity (OR = 0.50, p=0.011, LR = 7.99) was deemed as a protective patient characteristic in regard to requiring prolonged mechanical ventilation after head-and-neck free flap surgery (Table 2). Peri and Postoperative Outcomes Postoperative outcomes between the two groups also varied significantly. The prolonged ventilation group had a mean operating time that was 83 minutes longer than the control group (624.7 minutes vs. 540.8 minutes, p < 0.001), and their mean hospital stay was almost twice as long (20.7 days vs. 10.7 days, p < 0.001). Additionally, this group required 0.80 more units of postoperative transfusion products on average (p=0.040) and experienced higher rates of reoperation within 30 days (46.1% vs. 17.0%, p < 0.001). The prolonged mechanical ventilation group also demonstrated a significantly higher postoperative mortality rate (7.2% vs. 1.1%, p < 0.001) and was associated with increased incidences of postoperative complications including bleeding events requiring transfusion (52.8% vs. 22.9%, p < 0.001), pneumonia (35.6% vs. 4.2%, p < 0.001), wound disruption (12.8% vs. 5.4%, p < 0.001), airway complications resulting in unplanned intubation/ventilation within 48 hours of surgery (26.1% vs. 1.2%, p < 0.001), surgical site infection (20.0% vs. 11.5%, p=0.016), pulmonary embolism (2.8% vs. 0.7%, p = 0.003), cardiac arrest/myocardial infarction (18.3% vs. 1.4%, p < 0.001), and sepsis/septic shock (p < 0.001) (Table 3). Thirty-day hospital readmission rates were not significantly greater in the prolonged ventilation group. Figure 3 visually compares incidence rates of various outcomes between both groups.

Major head-and-neck surgeries are complex procedures, often performed on patients with significant pre- existing comorbidities such as COPD, diabetes, heart failure, and hypertension. Postoperatively, certain patients who undergo these surgeries may require mechanical ventilation and experience lengthy hospital stays. Optimizing length of stay and postoperative recovery is of particular importance in the head-and- neck cancer patient population given that these patients often require adjuvant oncologic therapy (e.g., chemotherapy/radiation). Delays in adjuvant therapy are known to be associated with adverse oncologic 4 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. outcomes9,10. Numerous studies and various recovery protocols have been published with the goal of opti- mizing the postoperative course. Prolonged postoperative mechanical ventilation has often been discussed as being associated with poorer outcomes. Here, we characterize these patients with the large sample size of a national database. Our study indicates that certain patient demographics and characteristics are associated with a higher likelihood of requiring prolonged mechanical ventilation. Specifically, we observed that patients who were white, had a previous history of ventilator dependence, presented with an open wound/wound infection preoperatively, and had primary oral cavity or sinonasal malignancies were more likely to experience extended postoperative mechanical ventilation. These findings align with existing literature that identifies similar risk factors for prolonged mechanical ventilation across various surgical disciplines 6,11,12, however our study emphasizes their relevance in the context of major head-and-neck surgeries. Given that major aerodigestive tract surgery can affect the airway, we had also hoped to evaluate whether concurrent tracheostomy has any impact on prolonged mechanical ventilation. It is understood that in certain cases tracheostomy may help facilitate post operative extubation by providing a secure airway. Additionally, tracheostomy rates and tracheostomy dependence are major quality metrics for head-and-neck cancer care. Ultimately, our analysis showed that there was no relationship between concurrent tracheostomy placement and rates of prolonged mechanical ventilation. Specifically, we found that tracheostomy was not associated with any change in rate of prolonged mechanical ventilation regardless of the site of malignancy. While we did attempt to study tracheostomy placement rates, our results could potentially be attributed to inherent

of the NSQIP database. Our analysis is likely not fully accounting for patients who received tracheostomy due to limitations of procedure bundling and multiple CPT code limits. Furthermore, this analysis was unable to assess if patients had pre-existing tracheostomy placement. Ultimately, this study is likely undercounting head-and-neck free flap patients with tracheostomy. Our study did not find any significant association between diabetes, hypertension, COPD, or smoking history and increased rates of prolonged postoperative mechanical ventilation. Previous literature has established that many of these comorbidities tend to correspond with increased patient morbidity and mortality in head-and-neck surgical patients 13-15. This discrepancy could suggest that improved patient selection may be contributing to improved postoperative outcomes. Here, we demonstrate that prolonged mechanical ventilation postoperatively is associated with an increase in patient morbidity and mortality. Prolonged mechanical ventilation was associated with increased rates of postoperative hemorrhage requiring transfusion, wound complications including surgical site infection or wound dehiscence, and an increased 30-day reoperation rate. Additionally, prolonged ventilation was associated with an increase in 30-day mortality. These findings align with recently published literature such as the work published by Mosquera et al. which also discusses increased rates of reoperation and longer overall hospital stay in patients who underwent prolonged postoperative ventilation after head-and-neck reconstructive surgery16. Extended ventilation in the surgical patient also has a significant impact on hospital resource utilization / expenditure. Zilberberg et al. discuss how patients who underwent prolonged mechanical ventilation had an annual cumulative cost of $16 billion in the United States, and composed nearly two-thirds of all mechanical-ventilation related costs17. Our analysis shows how prolonged ventilation head-and-neck patients experienced nearly double the length of hospital stay compared to those who did not undergo prolonged ventilation, which could also be responsible for some of the discussed medical cost disparity. Ultimately, our findings lend support to a conservative approach regarding mechanical ventilation in the postoperative free flap patient. For the first time, we use a national database to study prolonged mechanical ventilation in head-and-neck free flap patients. We use a national database to assess both preoperative/risk factors and evaluate postoperative outcomes, and were able to do so with a sample size of 180 patients which makes our study the largest to date in this field. Previous literature such as the work published by Mosquera et al. also discusses 5 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. characteristics and outcomes associated with head-and-neck patients who undergo prolonged mechanical ventilation, however their analysis is restricted to a single-center database 16. With our study, we hope to highlight trends in head-and-neck patients undergoing prolonged mechanical ventilation on a national scale. Our study does have its limitations. While we controlled for numerous confounding variables, we were limited to the variables that were provided to us from the NSQIP database. Some of the variables that we could not assess for include primary malignancy grading/staging, and specific time to extubation. Additionally, as discussed previously, we were unable to assess for tracheostomies that were not performed as separate procedures, or were performed prior to patients’ free flap surgeries. For completeness sake, we did include free flap patients that underwent total laryngectomy as total laryngectomy involves formalization of a tra- cheostomy, however we are likely still under-capturing the true quantity of free flap patients that received tracheostomy. Another limitation of our study may arise from the large discrepancy in sample size between the prolonged ventilation and control groups which could introduce bias into the study. Although sample sizes did differ, baseline covariates for most patient characteristics that we could assess were well balanced. Future research may focus on prospective studies to validate these findings and explore interventions aimed at reducing prolonged mechanical ventilation in the head-and-neck free flap population. Our study highlights various factors that may help surgeons identify patients who are at increased risk for prolonged mechanical ventilation after head-and-neck reconstructive surgery. We also show how pro- longed ventilation is associated with increased patient morbidity and mortality. Ultimately, our study adds to the growing chorus of data that supports early weaning from mechanical ventilation, especially in the perioperative setting.

1. Pai, K. , et al. The Utility of Comorbidity Indices in Assessing Head and Neck Surgery Outcomes: A Systematic Review.Laryngoscope 132 , 1388-1402 (2022).2. Stordeur, S., et al. Comorbidity in head and neck cancer: Is it associated with therapeutic delay, post-treatment mortality and survival in a population-based study? Oral Oncol 102 , 104561 (2020).3. Chiang, S.H., Ho, M.H., Wu, S.H. & Lin, C.C. Postoperative recovery among head and neck cancer patients receiving microvascular free flap surgery with implementing nurse-protocolized targeted sedation: relationship of use of sedatives and mechanical ventilation to length of ICU stay. Support Care Cancer 31 , 317 (2023).4. Yu, P.K., et al. Postoperative care in an intermediate-level medical unit after head and neck microvascular free flap reconstruction. Laryngoscope Investig Otolaryngol 4 , 39-42 (2019).5. Fadila, M., Rajasurya, V. & Regunath, H. Ventilator Weaning. in StatPearls (Treasure Island (FL), 2024).6. Huang, H.Y., Huang, C.Y. & Li, L.F. Prolonged Mechanical Ventilation: Outcomes and Management. J Clin Med 11 (2022).7. Trudzinski, F.C. , et al. Risk Factors for Prolonged Mechanical Ventilation and Weaning Failure: A Systematic Review. Respiration 101 , 959-969 (2022).8. Manzoor, T., Ahmed, Z., Sheikh, N.A. & Khan, M.M. Pulmonary complications associated with head and neck cancer surgery. J Coll Physicians Surg Pak 17 , 558-561 (2007).9. Dayan, G.S. , et al. Oncologic Significance of Therapeutic Delays in Patients With Oral Cavity Cancer. JAMA Otolaryngology– Head & Neck Surgery 149 (2023/11).10. Treatment delays in oral cavity squamous cell carcinoma and association with survival - PubMed. Head & neck 39 (2017 Apr).11. White, A.C. Long-Term Mechanical Ventilation: Management Strategies.Respiratory Care 57 (2012-06-01).12. S, M., S, B., RR, J. & AK, S. Airway management of patients undergoing oral cancer surgery: a retrospective study - PubMed. European journal of anaesthesiology 22 (2005 Jul).13. Xu, W., Chen, Z. & Zhang, L. Impact of diabetes on the prognosis of patients with oral and oropharyngeal cancer: A meta-analysis. Journal of Diabetes Investigation 15 (2024/08/01).14. Yu, V.X., Long, S. & Tassler, A. Smoking and Head and Neck Cancer. JAMA Otolaryngology–Head & Neck Surgery 149 (2023/05/01).15. Norcliffe-Kaufmann, L. & Palma, J.-A. Blood pressure instability in head and neck cancer survivors. Clinical autonomic research : official journal of the Clinical Autonomic Research Society 30 (2020/08).16. Mosquera, C. , et al. Risk factors for prolonged postoperative mechanical ventilation following head and neck microvascular free flap reconstruction and its association with patient outcomes. International Journal of Oral and Maxillofacial Surgery 0 (2025).17. MD, Z., BH, N., J, W. & AF, S. A Minority of Patients on Mechanical Ventilation Consume Disproportionate Resources: A Retrospective 6 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Cohort Study - PubMed. Chest 159 (2021 May). Pre-operative Patient Characteristic Control Group (n= 4,481) Prolonged V entilation (n=180) Hypothesis T est P-V alue Age (years)* Median [Q1-Q3] 63.0 [55.0-71.0] 63.0 [55.0-71.0] 0.799 Sex** Male: n (%) 2,290 (66.7%) 111 (61.7%) 0.311 Body Mass Index* Median [Q1-Q3] 25.5 [22.2-29.8] 25.0 [21.3-30.1] 0.884 Race** White: n (%) 2,878 (64.2%) 134 (74.4%) 0.005 Ethnicity** Hispanic: n (%) 1,046 (23.3%) 32 (17.8%) 0.083 Received Concurrent Tracheostomy Alongside Free Flap** Yes: n (%) 1,771 (39.5%) 84 (46.7%) 0.055 Underwent Total Laryngectomy** Yes: n (%) 349 (7.8%) 13 (7.2%) 0.781 Primary Cancer Site** Oral Cavity Yes: n (%) 1,753 (39.1%) 91 (50.6%) 0.002 Tracheostomy: 1,124 (64.1%) 54 (59.3%) 0.355 Oropharynx Yes: n (%) 261 (5.8%) 12 (6.7%) 0.637 Tracheostomy: 141 (54.0%) 8 (66.7%) 0.390 Larynx Yes: n (%) 395 (8.8%) 12 (6.7%) 0.317 Tracheostomy: 23 (5.8%) 3 (25.0%) 0.007 Nasal Cavity/Sinuses Yes: n (%) 109 (2.4%) 11 (6.1%) 0.002 Tracheostomy: 57 (52.3%) 4 (36.4%) 0.314 Salivary Gland Yes: n (%) 149 (3.3%) 2 (1.1%) 0.100 Tracheostomy: 23 (15.4%) 0 (0.0%) 0.546 Cutaneous Yes: n (%) 502 (11.2%) 7 (3.9%) 0.002 Tracheostomy: 35 (7.0%) 3 (42.9%) <0.001 Ablative Case** Yes: n (%) 2,828 (63.1%) 120 (66.7%) 0.332 Flap Failure / Complication** Yes: n (%) 11 (0.2%) 1 (0.6%) 0.421 Transfer from other Health Facility** Yes: n (%) 102 (2.3%) 9 (5.0%) 0.019 Elective Surgery** Yes: n (%) 4,209 (94.0%) 162 (90.0%) 0.027 Dependent Functional Status** Yes: n (%) 106 (2.4%) 7 (3.9%) 0.193 Diabetes Mellitus** Yes: n (%) 647 (14.4%) 36 (20.0%) 0.039 Hypertension** Yes: n (%) 2,078 (46.4%) 101 (56.1%) 0.010 Congestive Heart Failure** Yes: n (%) 30 (0.7%) 0 (0.0%) 0.271 Recent/Active Smoker** Yes: n (%) 1,322 (29.5%) 67 (37.2%) 0.026 Dypnea: At Rest** Yes: n (%) 42 (0.9%) 5 (2.8%) <0.001 Dypnea: Moderate Exertion** Yes: n (%) 240 (5.4%) 21 (11.7%) <0.001 Chronic Obstructive Pulmonary Disease** Yes: n (%) 298 (6.7%) 25 (13.9%) <0.001 Preoperative Ventilator Dependence** Yes: n (%) 11 (0.2%) 4 (2.2%) <0.001 Ascites** Yes: n (%) 4 (0.1%) 0 (0.0%) 0.688 Renal Failure** Yes: n (%) 2 (0.0%) 1 (0.6%) 0.008 Dialysis** Yes: n (%) 15 (0.3%) 1 (0.6%) 0.619 Open Wound / Wound Infection** Yes: n (%) 286 (6.4%) 22 (12.2%) 0.002 Immunosuppression** Yes: n (%) 182 (4.1%) 11 (6.1%) 0.176 High Bleed Risk** Yes: n (%) 96 (2.1%) 5 (2.8%) 0.566 Preoperative Transfusion Required** Yes: n (%) 23 (0.5%) 4 (2.2%) 0.003 Sepsis** Yes: n (%) 51 (1.1%) 2 (1.1%) 0.973 Sodium* Median [Q1-Q3] 139 [137-141] 138 [136-140] 0.046 Blood Urea Nitrogen* Median [Q1-Q3] 15 [11-20] 13 [9-19] 0.071 Creatinine* Median [Q1-Q3] 0.85 [0.70-1.02] 0.81 [0.70-1.06] 0.453 Albumin* Median [Q1-Q3] 4.00 [3.60-4.30] 3.90 [3.20-4.20] 0.002 Bilirubin* Median [Q1-Q3] 0.50 [0.40-0.70] 0.40 [0.30-0.60] 0.348 Aspartate Aminotransferase* Median [Q1-Q3] 21.0 [16.8-27.0] 19.0 [15.0-26.0] 0.532 Alkaline Phosphatase* Median [Q1-Q3] 77.0 [63.0-96.0] 80.0 [66.0-100.0] 0.134 7 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Pre-operative Patient Characteristic Control Group (n= 4,481) Prolonged V entilation (n=180) Hypothesis T est P-V alue Leukocytes* Median [Q1-Q3] 7.30 [5.90-9.16] 7.60 [6.46-9.70] 0.307 Hematocrit* Median [Q1-Q3] 40.0 [36.3-43.0] 38.7 [34.5-42.4] 0.010 * = Continuous Variable (T-Testing); ** = Categorical Variable (Chi-Square testing) T able 1 Title: Preoperative Patient Characteristic Comparison T able 1 Legend : Analysis of various preoperative patient characteristics between a patient group that underwent prolonged postoperative ventilation and a group that did not. P values that are bolded indicate statistical significance. Preoperative Patient Characteristic Univariate Analysis OR [95% CI], p-value Multivariate Analysis OR [95% CI], p-value Likelihood Ratio Chi-Squared Statistic, p-value Race: Non-White 0.62 [0.43-0.86], p=0.005 0.50 [0.29-0.84], p=0.011 7.99, p=0.005 Transfer From Other Healthcare Facility 2.26 [1.05-4.30], p=0.022 1.58 [0.62-3.63], p=0.307 - Primary Cancer Site - Oral Cavity 1.59 [1.18-2.15], p=0.002 2.06 [1.33-3.24], p=0.001 17.79, p <0.001 Nasal Cavity / Sinuses 2.61 [1.30-4.73], p=0.003 3.80 [1.56-8.33], p=0.002 11.45, p <0.001 Cutaneous 0.32 [0.14-0.64], p=0.003 0.53 [0.21-1.17], p=0.142 - Elective Surgery 0.57 [0.35-0.97], p=0.029 1.20 [0.63-2.47], p=0.598 - Diabetes Mellitus 1.48 [1.00-2.13], p=0.040 1.08 [0.62-1.82], p=0.766 - Hypertension 1.48 [1.10-2.00], p=0.011 1.17 [0.78-1.76], p=0.441 - Recent/Active Smoker 1.42 [1.04-1.92], p=0.027 1.08 [0.70-1.64], p=0.730 - Dypnea: At Rest 3.25 [1.11-7.59], p=0.014 2.36, [0.70-6.42], p=0.123 - Dypnea: Moderate Exertion 2.39 [1.45-3.75], p <0.001 1.90 [0.95-3.55], p=0.055 - Chronic Obstructive Pulmonary Disease 2.26 [1.43-3.45], p <0.001 1.28 [0.63-2.42], p=0.475 - Preoperative Ventilator Dependence 9.24 [2.54-27.31], p <0.001 4.67 [1.11-16.40], p=0.022 10.65, p=0.001 Renal Failure 12.51 [0.58-131.20], p=0.039 26.40 [1.00-698.44], p=0.024 3.03, p=0.082 Open Wound / Wound Infection 2.04 [1.25-3.17], p=0.002 2.53 [1.43-4.28], p=0.001 9.06, p=0.003 Preoperative Transfusion Required 4.41 [1.28-11.60], p=0.007 2.81 [0.71-9.29], p=0.108 - Serum Sodium 0.96 [0.92-1.00], p=0.046 0.98 [0.92-1.04], p=0.476 - Serum Albumin 0.61 [0.45-0.83], p=0.002 0.77 [0.53-1.14], p=0.188 - Serum Hematocrit 0.96 [0.94-0.99], p=0.010 1.01 [0.97-1.05], p=0.707 - OR = Odds Ratio; CI = Confidence Interval Multivariate Analysis Model: N = 4661, N-Missing = 2386; AIC = 865.8 , C-Statistic = 0.724, Hosmer-Lemeshow Test = 13.36 (p=0.100) T able 2 Title: Patient Characteristic Regression Analysis T able 2 Legend : Results of univariate and enter-variable multivariate analysis for patient characteristics that were deemed statistically significant per hypothesis testing. Odds ratios (OR) with 95% confidence intervals (CI) are also depicted. Likelihood Ratio (LR) Chi-Squared Statistics with corresponding p-values are also provided. The final multivariate analysis model accuracy metrics including Akaike Information Criterion (AIC), Concordance Statistic (C-Statistic), and Hosmer-Lemeshow Test results are provided on the bottom row. P-values that are bolded indicate statistical significance. Peri/Postoperative Complication Control Group (n= 4,481) Prolonged V entilation (n=180) Hypothesis T est P-V alue Hospital Length of Stay (days)* Mean (SD) 10.68 (7.24) 20.65 (11.40) 30 Days** Yes: n (%) 121 (2.7%) 34 (18.9%) <0.001 Unplanned Readmission** Yes: n (%) 447 (10.0%) 13 (7.2%) 0.225 Reason: Surgical Site Infection: Yes: n (%) 2 (0.4%) - Wound Disruption: Yes: n (%) 12 (2.6%) - 8 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Peri/Postoperative Complication Control Group (n= 4,481) Prolonged V entilation (n=180) Hypothesis T est P-V alue Cardiopulmonary Events (e.g., pulmonary embolism, CVA) Yes: n (%) 1 (0.2%) - Sepsis / SIRS Yes: n (%) (0.8%) - Pneumonia Yes: n (%) 1 (0.2%) - Bleeding Events Yes: n (%) 6 (1.3%) - Respiratory Events (e.g., hypoxemia, respiratory failure, pulmonary collapse, pneumothorax, pulmonary edema, pleural effusion) Yes: n (%) 22 (4.8%) - Reoperation** Yes: n (%) 761 (17.0%) 83 (46.1%) <0.001 Perioperative Transfusion** Yes: n (%) 1,028 (22.9%) 95 (52.8%) <0.001 Total Transfusion Amount (Units)* Mean (SD) 2.26 (1.68) 3.11 (2.13) 0.040 30-Day Mortality** Yes: n (%) 50 (1.1%) 13 (7.2%) <0.001 Home Discharge** Yes: n (%) 3,828 (85.4%) 114 (63.3%) <0.001 Surgical Site Infection** Yes: n (%) 508 (11.5%) 36 (20.0%) 0.016 Wound Disruption** Yes: n (%) 244 (5.4%) 23 (12.8%) <0.001 Pneumonia ** Yes: n (%) 186 (4.2%) 64 (35.6%) <0.001 Unplanned Intubation/Ventilation** Yes: n (%) 54 (1.2%) 47 (26.1%) <0.001 Pulmonary Embolism** Yes: n (%) 33 (0.7%) 5 (2.8%) 0.003 Deep Venous Thrombosis** Yes: n (%) 61 (1.4%) 10 (5.6%) <0.001 Renal Failure** Yes: n (%) 2 (0.0%) 3 (1.7%) <0.001 Urinary Tract Infection** Yes: n (%) 50 (1.1%) 2 (1.1%) 0.995 Clostridium Dificile Occurrence** Yes: n (%) 57 (1.3%) 9 (5.0%) <0.001 Stroke** Yes: n (%) 24 (0.5%) 3 (1.7%) 0.050 Cardiac Arrest / Myocardial Infarction** Yes: n (%) 63 (1.4%) 33 (18.3%) <0.001 Sepsis** Yes: n (%) 129 (2.9%) 14 (7.8%) <0.001 Septic Shock** Yes: n (%) 20 (0.4%) 22 (12.2%) <0.001 Operation Time (min.)* Mean (SD) 540.8 (196.3) 624.7 (218.8) <0.001 * = Continuous Variable (T-Testing); ** = Categorical Variable (Chi-Square testing) T able 3 Title: Peri and Postoperative Complication Incidence Rates T able 3 Legend : Analysis of perioperative/postoperative complication rates between a patient group that underwent prolonged postoperative ventilation and a group that did not. P-values that are bolded indicate statistical significance. 9 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Figure 1 Title: Search Criteria for Identifying Free Flap Patients Figure 1 Legend : A comprehensive list of Current Procedural Terminology (CPT) codes, International Classification of Disease (ICD-10) codes, and free-text search criteria that were utilized to identify the patient group who underwent head and neck microvascular reconstructive surgery. 10 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Figure 2 Title : Study Design Flowchart Figure 2 Legend : A flowchart depicting the inclusion criteria that was utilized to identify the final patient group who underwent head and neck microvascular reconstructive surgery. 11 Posted on 12 Sep 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.175767133.38284058/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Figure 3 Title: Postoperative Complication Rate Comparison after Head and Neck Reconstructive Surgery Figure 3 Legend: Bar graph depicting a comparison of statistically significant (p<0.05) peri/postoperative complication rates between patients who underwent prolonged ventilation (Orange) and those who did not (Blue). For the full list of postoperative complications with p-values, see Table 3. 12