Objective To investigate the temporal association between sensorineural hearing loss (SNHL) and clinical variables prior to and during extracorporeal membrane oxygenation (ECMO).
Methods ECMO survivors treated for neonatal respiratory failure at Children's Hospital Los Angeles were identified between 1987 and 1991. Developmental follow-up, including a hearing evaluation, was available in 48 patients and is the focus of this study. To minimize bias, one investigator (M.M.) reviewed the clinical data and was blinded to the outcome of audiological testing. Relevant clinical variables were collected. Patients were categorized into normal hearing (NH) and SNHL groups. Continuous variables were tested by t-test. For categorical and continuous variables recorded into ordinal measures, the distribution or proportion of patients between groups were compared by Fisher's exact test. Cox proportional-hazard regression was used to take into account the variable length of follow-up and assess the independent effect of each risk factor for SNHL.
Results During the study period, 126 neonates survived ECMO. Complete follow-up information was available from 48/126(38%) infants. Eight of the 48 patients (17%) were found to have SNHL. There were no differences between the SNHL group and NH group in regards to birth weight, gender, Apgar scores at 1 minute, primary diagnosis, age at ECMO initiation, or type of ECMO support (VA vs VV). A significant difference between SNHL and NH groups was found in median 5-minute Apgar scores (6 vs 8, p = .02), mean ECMO duration (191 vs 137 hr, p = .04), proportion of infants with pre-ECMO CO2 < 3O mm Hg (100% vs 61%, p = .04), pre-ECMO use of Lasix (75% vs 12%, p = .03), mean total dose of muscle relaxant (6.4 ± 16.9 vs 0.52 ± 0.8 mg/kg, p < .01). Factors associated with proportional-hazards regression analysis showing a significant association with the occurrence of SHNL were pre-ECMO clinical seizure (hazard ratio of 13.3) and duration of ECMO (hazard ratio of 1.01).
Conclusion In infants who undergo ECMO therapy, clinical seizure activity prior to ECMO and the duration of ECMO therapy are significantly associated with SNHL. We speculate that these results indicate an important independent contribution of pre-ECMO CNS injury in the pathogenesis of SNHL.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.