Predicting postlaryngectomy voice outcome in an era of primary tracheoesophageal fistulization: a retrospective evaluation.

OBJECTIVES

Not all laryngectomees appear to have the same potential to develop functional spoken communication. Our goal was to evaluate voice outcome in different functional subgroups of laryngectomees and to identify physical and demographic factors associated with success and failure to achieve functional spoken communication.

DESIGN

Retrospective chart review.

SETTING

Surgery was performed at a tertiary care hospital. Multidisciplinary follow-up was conducted at weekly head and neck clinics held at the associated regional cancer centre. Voice rehabilitation procedures took place in both settings.

PATIENTS

Sixty-four consecutive patients who had undergone total laryngectomy during the era of primary tracheoesophageal fistulization (TEF) at this facility. Patients were subdivided into four groups according to whether they had undergone primary TEF, or whether this had been contraindicated by locoregional factors of TEF candidacy/performance status, or both.

INTERVENTIONS

Primary TEF was performed whenever technically feasible and traditional TEF candidacy criteria were met. Voice rehabilitation procedures were initiated prior to discharge.

OUTCOME MEASURES

A judgement of voice outcome was assigned based on documentation on at least one of three patient treatment records by a physician or speech-language pathologist that a patient had demonstrated functional spoken communication within the clinical setting.

RESULTS

Forty-five of 64 patients (70%) achieved functional spoken communication. Six laryngectomized subgroups were ultimately identified and characterized. Voice outcome varied considerably between these subgroups. Prelaryngectomy communication status and age emerged as predictors of voice outcome.

CONCLUSIONS

Voice outcome is related to several factors present prior to or at laryngectomy. Different combinations of such factors create various postlaryngectomy recovery streams for which voice outcome may be predicted more specifically.