Invasive Home Mechanical Ventilation: 10-Year Experience of a Pediatric Home Care Service

摘要

BACKGROUND: Children dependent on invasive home mechanical ventilation (HMV) represent a growing population worldwide. The objective of this study was to assess the experience of 10 years of medical assistance given to pediatric patients on continuous invasive HMV at a Brazilian Home Care Service (HCS), specifically patient characteristics and predictors of outcome (ie, hospital readmission, death, and location of death). METHODS: Medical records for children on invasive HMV at the HCS between 2007 and 2016 were evaluated to collect the following data: age at admission to HCS, sex, principal diagnosis, length of hospital admission and home care period, number and cause of hospital read-missions, number of procedures, death and location of death. The odds ratio was used to understand the likelihood of death for each diagnosis, hospital readmission, and admission age, using a binary logistic regression model. RESULTS: Twenty-seven children were evaluated. The most prevalent diagnosis was cerebral palsy (37.0%). The mean duration of home care was higher than the mean hospital length of stay (955.0 +/- 4.6 d versus 341.0 +/- 0.5 d, respectively). First hospital readmission mean was at 392.6 +/- 548.9 d, and the main cause was respiratory tract infection (45.9%), especially tracheitis. Of the total number of deaths (13), 76.9% occurred in hospital units. There was no statistically significant result observed for greater odds of death for any of the diagnoses and admission age on HCS. However, children who had a hospital readmission < 6 months after hospital discharge presented 10% greater chance of death (P = .02). CONCLUSIONS: The most prevalent diagnosis of children on continuous invasive HMV was cerebral palsy. The main cause of hospital readmission was respiratory tract infection, especially tracheitis. Having the first hospital readmission at < 6 months after discharge was shown to be a risk factor associated with mortality.