Does Ribavirin Impact on the Hospital Course of Children With Respiratory Syncytial Virus (RSV) Infection? An Analysis Using the Pediatric Investigators Collaborative Network on Infections in Canada (PICNIC) RSV Database Barbara J. Law, MD*; Elaine E. L. Wang, MDCM‡; Noni MacDonald, MD§; Jane McDonald, MDi; Simon Dobson, MD¶; Francois Boucher, MD#; Joanne Langley, MD**; Joan Robinson, MD‡‡; Ian Mitchell, MD§§; and Derek Stephens MSc‡ ABSTRACT. Objectives. To determine the relation- 20.6%, 20.9%, 15.5%, 15.2%, and 13.3%, respectively. ship between receipt of aerosolized ribavirin and the Across the subsets ribavirin use ranged from 36% to 57% hospital course of high-risk infants and children with of ventilated patients and 6% to 39% of nonventilated respiratory syncytial virus (RSV) lower respiratory infec- patients. For nonventilated patients in each subset the tion (LRI). median RSV-attributable hospital length of stay (RSV- Methods. The 1993–1994 Pediatric Investigators Col- LOS) was 2 to 3 days longer for ribavirin recipients and laborative Network on Infections in Canada (PICNIC) the duration of hypoxia was significantly increased. Du- RSV database consists of prospectively enrolled children ration of intensive care unit (ICU) stay was also increased with acute RSV LRI, admitted to nine Canadian pediatric for all ribavirin-treated subgroups except those with tertiary care centers. After excluding cases with compro- CHD. In contrast, for ventilated patients, ribavirin ther- mised immunity and/or nosocomial infection, subsets apy was not significantly associated with any of the with any congenital heart disease (CHD), chronic lung outcome measures regardless of risk group. In the mul- disease (CLD), age 6 weeks (INFANT), gestation 36 tiple regression model, ribavirin was significantly asso- weeks (PREM), or severe disease within 48 hours of ciated with a prolonged RSV-LOS both for children with admission as shown by an oxygen saturation 90% or an CHD and/or CLD as well as for those whose only risk FiO2 requirement of >.35 (EARLY HYPOXIA) were stud- factors included INFANT, PREM, and/or EARLY HY- ied in two ways. First, each risk group subset was ana- POXIA. lyzed separately to assess the association between riba- Conclusions. These data raise further doubts about virin receipt and measures of disease severity including the clinical effectiveness of ribavirin in infants and chil- duration of intensive care, mechanical ventilation, hy- dren with risk factors for severe disease. Selection bias, poxia and RSV-attributable hospital stay. Secondly, riba- with ribavirin used for sicker children, may have influ- virin was added as an independent variable to a previ- enced outcome. Nevertheless the long durations of hos- ously described multiple regression model for RSV- pitalization, ICU, ventilation, and oxygen supplementa- attributable length of hospital stay and two mutually tion in nonventilated ribavirin recipients stress the need exclusive subsets were analyzed: 1) previously healthy for further randomized trials to assess its efficacy. patients with 1 of: INFANT, PREM, or EARLY HY- Pediatrics 1997;99(3). URL: http://www.pediatrics.org/cgi/ POXIA; 2) patients with CHD and/or CLD. content/full/99/3/e7; ribavirin, respiratory syncytial virus, Results. Between January 1993 and June 1994, 1425 pneumonia, bronchiolitis. 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 ABBREVIATIONS. RSV, respiratory syncytial virus; LRI, lower CHD, 134 CLD, 213 INFANT, 211 PREM, and 463 EARLY respiratory infection; ICU, intensive care unit; CHD, congenital # HYPOXIA. The proportion ventilated in each group was heart disease; CLD, chronic lung disease; PREM, gestation 36 weeks; INFANT, postnatal age #6 weeks; EARLY HYPOXIA, oxygen saturation #90%. From the *Winnipeg Children’s Hospital and University of Manitoba, Win- nipeg, Manitoba; the ‡Hospital for Sick Children and University of Toronto, Respiratory syncytial virus (RSV) is the major Toronto, Ontario; the §Children’s Hospital of Eastern Ontario and Univer- cause of lower respiratory infection (LRI) and hospi- sity of Ottawa, Ottawa, Ontario; the iMontreal Children’s Hospital, and McGill University, Montreal, Quebec; ¶British Columbia Children’s Hos- talization among infants and toddlers in North 1 pital and University of British Columbia, Vancouver, British Columbia; America. Nearly all children are infected by age 2 #Centre Hospitalier de l’Universite Laval, Quebec City, Quebec; **Izaak years and 1% to 2% of those infected require hospi- Walton Killam Hospital and Dalhousie University, Halifax, Nova Scotia; talization.2 Among those admitted to hospital with ‡‡Children’s Health Centre and the University of Alberta, Edmonton, Al- no apparent risk factors for severe disease, 4% to 15% berta; and §§Alberta Children’s Hospital, and University of Calgary, Cal- gary, Alberta. are admitted to the intensive care unit (ICU), 1% to 3–7 This work was presented in part at the Annual Society for Pediatric Re- 5% require assisted ventilation, and ,1% die. In search Meeting, in San Diego, CA (May 1995). contrast, among children with underlying heart/ Received for publication Feb 23, 1996; accepted May 10, 1996. lung disease, prematurity (gestation #36 weeks) and Reprint requests to (B.J.L.) Department of Medical Microbiology, Room 530, # 730 William Ave, Winnipeg, Manitoba, Canada, R3E 0W3. young age ( 6 weeks) the corresponding figures for PEDIATRICS (ISSN 0031 4005). Copyright © 1997 by the American Acad- ICU, ventilation, and mortality range from 10% to emy of Pediatrics. 40%, 8% to 27%, and up to 10%, respectively.3–7 http://www.pediatrics.org/cgi/content/full/99/3/Downloaded from www.aappublications.org/newse7 by guestPEDIATRICS on September 26, Vol. 2021 99 No. 3 March 1997 1of6 Ribavirin, first approved in the United States in observation relative to the RSV illness. Additional days spent in 1986, is the only licensed antiviral therapy for RSV hospital for investigation of other problems, elective surgery, or delayed discharge due to social or transportation problems were infection. In 1993 the American Academy of Pediat- not counted. Because the data were not normally distributed, the rics (AAP) recommended that ribavirin should be Wilcoxon test was used throughout to compare outcome duration. given to selected infants and children at risk for or x2 or Fisher’s exact tests of proportion were used to compare already manifesting severe disease as well as for all distribution of demographic factors, coexisting risk factors, and ventilated patients.8 Although several randomized indicators of disease severity among the ribavirin and no-ribavirin therapy groups. trials suggested that ribavirin was efficacious among Receipt of ribavirin was also added as an independent variable nonventilated patients,9–14 the use of the drug re- to a previously developed multiple regression model for predict- mained controversial, due to the small numbers ing RSV-attributable length of hospitalization.18 Separate analyses studied, and concern over the validity, generalizabil- were performed for the CHD/CLD subsets and the INFANT/ PREM/EARLY HYPOXIA subsets. The other independent vari- ity, and clinical relevance of the outcome measures ables included in the model for both analyses were INFANT, used. For ventilated patients, controversy existed as PREM, EARLY HYPOXIA, aboriginal racial background, history well, with one trial supporting a beneficial effect for of apnea or respiratory arrest before or at the time of admission to ribavirin15 while another failed to show a differ- hospital, consolidation on the admitting chest radiograph, and ence.16 Large increases in the cost of ribavirin since hospital center. For analysis of children with prior heart or lung disease CHD and CLD were included as independent variables. 1993 raised further questions about its relative cost- benefit. In Canada over 90% of RSV-related hospital RESULTS costs are due to daily bed charges.17 With the daily cost of ribavirin at approximately $1500 (Canadian Study Groups funds), therapy would have to lead to a substantial Between January 1, 1993 and June 6 1994, 2116 reduction in the duration of hospital stay to justify children were hospitalized with acute RSV LRI at the the costs of ribavirin. Accordingly, the AAP recom- participating PICNIC study centers. Of these 1516 mendations for patients who should receive ribavirin (72%) were enrolled in the prospective RSV study. were used to select cases enrolled in a prospective Excluded from this analysis were 91 cases of hospi- study of children hospitalized for RSV LRI and out- tal-acquired RSV infection. Among the remaining comes were examined to determine the relationship 1425 community-acquired cases of RSV LRI, previ- between ribavirin therapy and length of hospitaliza- ously recognized chronic disease affected 220 (15%). tion. From these, the 173 cases with underlying cardiopul- monary disease and no known immunodeficiency METHODS were chosen for analysis, including 97 CHD and 134 The Pediatric Investigators Collaborative Network on Infec- CLD. Among the CHD group, 49 (50%) had a left to tions in Canada (PICNIC) RSV Database includes prospectively right shunt and 20 (21%) had pulmonary hyperten- collected data on demographic factors, disease severity, and daily sion. Among the CLD group 88 (66%) had a history management in hospital of 1516 infants and children with acute of