Do sociodemographic and clinical characteristics affect mortality rates in people with intellectual disability and dysphagia who have a percutaneous endoscopic gastrostomy? A cohort study between 2000 and 2022

Abstract

BackgroundPeople with intellectual disability frequently have eating, drinking and swallowing difficulties (dysphagia) and are at greater risk of premature mortality, particularly from aspiration and respiratory infections. The insertion of a percutaneous endoscopic gastrostomy (PEG), as part of a multidisciplinary management plan, may help to maintain and improve nutrition. This study included people with intellectual disability who had ever had a PEG inserted and who presented to the specialist Speech and Language Therapy team in one Greater Manchester borough between 2000 and 2022 and assessed the risk of death and sociodemographic and clinical factors associated with this.MethodsThis was a cohort study using secondary data. Kaplan–Meier estimates were used to construct the probability of mortality curves. Cox proportional hazards were used to compare death rates in the different sociodemographic and clinical characteristic groups and were presented as hazard ratios and 95% confidence intervals.FindingsOf 42 people included in the study, 18 (43%) died from the point of PEG insertion to the end of the study (December 2022). The median (interquartile range) time to death from PEG insertion was 5 (2–10) years, with four people (10%) dying in the first year. Aspiration pneumonia and unspecified pneumonia were responsible for seven (39%) deaths. No significant associations were found between sociodemographic and clinical factors and risk of death.ConclusionIn persons with intellectual disability and a PEG who were followed up between 2000 and 2022, the mortality was around 40% with deaths occurring in the first year and respiratory conditions being an important cause. The lack of association with sociodemographic and clinical characteristics may have been due to a limited sample size. Further research is needed with larger samples and more variables, including quality of life data, to help understand and improve clinical practice in this area.