Objectives: To determine the relationship between receipt of aerosolized ribavirin and the hospital course of high-risk infants and children with respiratory syncytial virus (RSV) lower respiratory infection (LRI).
Methods: The 1993-1994 Pediatric Investigators Collaborative Network on Infections in Canada (PICNIC) RSV database consists of prospectively enrolled children with acute RSV LRI, admitted to nine Canadian pediatric tertiary care centers. After excluding cases with compromised immunity and/or nosocomial infection, subsets with any congenital heart disease (CHD), chronic lung disease (CLD), age </=6 weeks (INFANT), gestation </=36 weeks (PREM), or severe disease within 48 hours of admission as shown by an oxygen saturation </=90% or an FiO2 requirement of >.35 (EARLY HYPOXIA) were studied in two ways. First, each risk group subset was analyzed separately to assess the association between ribavirin receipt and measures of disease severity including duration of intensive care, mechanical ventilation, hypoxia and RSV-attributable hospital stay. Secondly, ribavirin was added as an independent variable to a previously described multiple regression model for RSV-attributable length of hospital stay and two mutually exclusive subsets were analyzed: 1) previously healthy patients with >/=1 of: INFANT, PREM, or EARLY HYPOXIA; 2) patients with CHD and/or CLD.
Results: Between January 1993 and June 1994, 1425 community-acquired hospitalized cases of RSV LRI were entered into the RSV database. Among these 750 (52.6%) fit into one or more of the defined subsets including 97 CHD, 134 CLD, 213 INFANT, 211 PREM, and 463 EARLY HYPOXIA. The proportion ventilated in each group was 20.6%, 20.9%, 15.5%, 15.2%, and 13.3%, respectively. Across the subsets ribavirin use ranged from 36% to 57% of ventilated patients and 6% to 39% of nonventilated patients. For nonventilated patients in each subset the median RSV-attributable hospital length of stay (RSV-LOS) was 2 to 3 days longer for ribavirin recipients and the duration of hypoxia was significantly increased. Duration of intensive care unit (ICU) stay was also increased for all ribavirin-treated subgroups except those with CHD. In contrast, for ventilated patients, ribavirin therapy was not significantly associated with any of the outcome measures regardless of risk group. In the multiple regression model, ribavirin was significantly associated with a prolonged RSV-LOS both for children with CHD and/or CLD as well as for those whose only risk factors included INFANT, PREM, and/or EARLY HYPOXIA.
Conclusions: These data raise further doubts about the clinical effectiveness of ribavirin in infants and children with risk factors for severe disease. Selection bias, with ribavirin used for sicker children, may have influenced outcome. Nevertheless the long durations of hospitalization, ICU, ventilation, and oxygen supplementation in nonventilated ribavirin recipients stress the need for further randomized trials to assess its efficacy.