TUBEFEEDING AND MORTALITY IN CHILDREN WITH SEVERE DISABILITIES AND MENTAL-RETARDATION

Objective. To study the contribution of tubefeeding to mortality for c hildren with severe disabilities and mental retardation. Previous rese arch has suggested an association between tubefeeding and mortality. H owever, risk has never been determined using population-based data or defined in regard to patient variables. Methods. Retrospective analysi s of a comprehensive statewide data set comprised of 4921 children wit h severe disabilities and mental retardation living in community and c ongregate care settings. The outcome measure was mortality; primary st udy variables included the presence of a feeding tube, measures of fun ctional independence, type of residence, and medical comorbidity. Resu lts. There were four findings. First, the use of a feeding tube was as sociated with virtually every disability. Second, when no study variab les were controlled, statistically significant differences in mortalit y rates were noted between children who were tubefed and those who wer e not. The relative risk of mortality associated with use of a feeding tube was 2.1. Third, the use of a feeding tube was associated with a reduction in relative risk of mortality in children with tracheostomy (relative risk of mortality: .55). However, this association did not a chieve statistical significance. Fourth, when study variables were con trolled in a multivariate analysis, feeding tube use was associated wi th no identifiable increase in mortality among children with very seve re disabilities, but was associated with an approximated doubled morta lity rate among those with less severe disabilities. Conclusions. We h ypothesize that the increased mortality associated with tubefeeding ma y be attributable to a differential increase in pulmonary disease seco ndary to overly vigorous nutritional maintenance and subsequent aspira tion after tube placement. For children with tracheostomy this risk ma y be reduced. If tracheostomy proves to be associated with a relativel y more favorable outcome for tubefeeding, we hypothesize that it would reflect the benefits of tracheostomy in allowing access to the airway for suctioning and ventilation. Given the observed higher mortality r ates among the less severely disabled children who are tubefed and the substantial costs associated with tubefeeding, a prospective, control led study may be clinically indicated, ethically justifiable, and econ omically warranted.