The prospective experience of a maxillofacial surgeon with the percutaneous endoscopic gastrostomy technique.

Insertion of a percutaneous endoscopic gastrostomy (PEG) was attempted on 225 occasions, mainly for oral malignancy. Seventy-five percent (169/225) were inserted at the time of definitive surgery. There were significant incidental findings during 5% (11/225). The rate of successful insertion was 97% (219/225). The incidence of minor complications was 12% (26/225) and major complications 3% (7/225). There was no procedure-related mortality. The 30-day mortality rate, including those with terminal malignant disease, was 6% (14/225). An increased risk of death was associated with age of 65 years and over (P=0.004). The median PEG duration was 337 (SE 31) days. Duration was significantly longer for stage T3-4 tumours (P=0.028), N1 or greater neck disease (P=0.034), following surgery with radiotherapy when compared to surgery alone (P<0.001), particularly glossectomy (P=0.038) and maxillectomy procedures (P=0.003), after two separate surgical procedures and radiotherapy (P=0.046) and following a composite bone resection (P=0.031), or radiotherapy alone when compared to surgery alone (P=0.003). There was no relationship to the type of flap used for reconstruction. Four patients have a long-term PEG. Only two patients did not use the PEG. The early insertion of a PEG in all patients undergoing free or pedicled flap reconstruction appears to be appropriate. The PEG procedure may be safely performed by an appropriately trained maxillofacial surgeon.