Methods: Fifty five PE reconstructions were performed following resection or stenosis, with or without laryngectomy. 15 variables relevant to functional outcomes and postoperative morbidity were recorded. The Functional Outcome Swallowing Score (FOSS), 0-5 system measured swallowing (0=normal swallowing). Uni- and multivariate statistical models were used for outcome analysis. Complication rates were recorded.

Results: 74% of resections included total laryngectomy and 81% of reconstructions were performed primarily. 64% of patients received prior treatment and 80% underwent neck dissection. The mean best swallowing score was 1.8, achieved at a median interval of 2 (0.5-9.5) months. The ultimate swallowing score was 2.4, reached at an interval of 7 (1-84) months from reconstruction. Voice was by larynx in 22%, tracheoesophageal puncture in 35% and electrolarynx in 43%. Fistulae occurred in 15%. Significant associations existed between poorer best swallow scores (2-5) and fasciocutaneous flaps (odds ratio {O.R.}=5.7) comorbidity (O.R.=5.4) and use of saliva bypass tubes (BT) (O.R.=10.5) in univariate analyses. In multivariate models poorer ultimate swallowing scores correlated significantly with presence of comorbidity (O.R.=5.6) and use of BT (O.R.=4.1). No variables predicted voicing method. Complications were significantly associated with prior treatment (O.R.=8.4) neck dissections (O.R.=9.6) and primary reconstruction (O.R.=7.8) with multivariate analysis.

Conclusions: Surgeons may wish to limit use of BT, be cautious in predicting ultimate swallowing results for patients with comorbidity and expect higher neck wound complications/fistulae with salvage surgery, primary reconstruction and neck dissections.