2018-19 Section 7-Neoplastic and Inflammatory Diseases of the Head and Neck eBook

Otolaryngology–Head and Neck Surgery 158(5)

For each liter increase in crystalloid given intraoperatively, odds of wound-healing complications increased multiplicatively by 1.21 (95% CI, 1.05-1.38). Those with sites classified as oral cavity (including maxilla) and larynx/hypopharynx/oropharynx had greater odds of wound-healing complications compared to those with nonmucosal subsites with an odds ratio of 19.1 (95% CI, 2.5-144.0) and 24.2 (95% CI, 3.1-190.5), respectively. Predictors of Wound Infection Complications The univariable and multivariable logistic regression predic- tors of wound infection complications are presented in Table 6 . Multivariable analysis demonstrated that age ( P \ .01), history of diabetes ( P \ .01), and malnutrition ( P = .04) were all significant predictors of developing wound infection complications. For each year increase in age, odds of wound infection complications decreased multiplicatively by 0.97 (95% CI, 0.95-0.99). Those with a history of diabetes had greater odds of a wound infection complication compared to those without a history of diabetes with an odds ratio of 2.47 (95% CI, 1.28-4.77). Those with a history of malnutrition had greater odds of a wound infection complication with an odds ratio of 1.89 (95% CI, 1.03-3.50). Discussion This study comprehensively analyzes all complications in patients who have received free flap reconstruction for head and neck defects at our institution, thus comprising the larg- est cohort in the current literature. It demonstrates a high overall complication rate (54%), which seems to be pre- dicted by prolonged OR time, comorbidity index, and muco- sal tumor sites. Wound-healing complications were less prevalent (15%) but were similarly predicted by prolonged OR time and mucosal subsites, as well as increased crystal- loid administration. Last, wound infection complications (12%) were predicted by advanced age, diabetes mellitus, and malnutrition. The overall complication rate reported in our study (54%; n = 515) is slightly lower than that reported in the only other study in the literature assessing all complications in this population (64%; n = 192). 6 McMahon et al 6 demon- strated that wound and pulmonary complications accounted for the majority, which is not surprising given the high inci- dence of smoking, COPD, and complex clean contaminated wounds in an irradiated/chemoirradiated surgical bed. Length of OR time, an important independent predictor of complications, can always be improved by increasing effi- ciency and working concurrently in 2 or sometimes 3 teams. This should be further studied and could certainly be a qual- ity improvement target using lean process design. However, despite assessing for and excluding colinear variables in our multivariable analysis, it may be that prolonged OR time is still a surrogate for case complexity (through residual unmeasured confounding), which would not surprisingly be associated with increased complication rates. Similarly, a defect including mucosal subsites was associated with higher complications, given that many of these patients

Table 2. Patient Comorbidity Status (n = 515).

Characteristic

Value

5.0 6 2.3

Charlson comorbidity index (age adjusted), mean 6 SD

Smoking history, No. (%) Never

113 (21.9) 200 (38.8) 202 (39.2) 194 (37.7) 84 (16.3) 122 (23.7) 94 (18.3) 259 (50.3) 43 (8.4) 27 (5.2) 84 (16.3) 69 (13.4) 103 (20.0)

Yes—past

Yes—current

Alcoholism at time of operation, No. (%)

Diabetes mellitus, No. (%)

Coronary artery disease, No. (%) Peripheral vascular disease, No. (%)

Chronic obstructive pulmonary disease, No. (%)

Chronic renal failure, No. (%)

Hepatic disease, No. (%)

Obesity, No. (%)

Hypothyroidism, No. (%) Malnutrition, No. (%)

Table 3. Peri- and Postoperative Complication (Outcome) Summary. a

Complication

No. (%)

Predictors of Wound-Healing Complications The univariable and multivariable logistic regression predic- tors of wound-healing complications are presented in Table 5 . Multivariable analysis demonstrated that increased opera- tive level ( P \ .01), more intraoperative crystalloid given ( P \ .01), and tumor site (oral cavity and larynx/hypophar- ynx; P \ .01) were all significant predictors of developing wound-healing complications. Again, a conversion of OR time from minutes to hours was done to allow for easier interpretation of the results. For each hour increase in OR time, odds of wound-healing complications increased multiplicatively by 1.24 (95% CI, 1.09-1.41). Similarly, a conversion of crystalloid given from milliliters to liters was done to allow for easier interpretation of the results. 278 (54.0) Wound-healing complication (fistula/dehiscence) 77 (15.0) Wound infection (cellulitis/abscess) 60 (11.7) Flap loss 18 (3.5) Medical complications Myocardial infarction 8 (1.6) Cerebrovascular event 4 (0.8) Pneumonia 38 (7.4) Delirium tremens (alcohol withdrawal) 33 (6.4) Clostridium difficile colitis 7 (1.4) Death 8 (1.6) a In bold are the primary and secondary outcomes of the study. Any complication

195