Letters to the Editor

To the Editor;

We are a group of paediatric health care professionals (community and hospital general paediatrics, nursing, neonatology, paediatric respirology, paediatric infectious disease and health economics) who wish to express concerns about the position statement published in the August/September issue of the Journal.

While rigorously developed guidelines play a pivotal role in establishing standards of practice, poorly developed guidelines may cause harm. As reviewed by Haroon et al (1), a well-conducted review process is important in avoiding potential bias, and guaranteeing validity and feasibility for practice. In this case, key stakeholders external to the guideline group, such as respirologists and cardiologists, appear to have been omitted. Of more concern, it is unclear what criteria were used in searching for and evaluating the evidence behind the position statement and which methodology was employed to reach consensus, in addition, there was an absence of grading for each final recommendation.

Specifically, Table 1 purports to summarize results of the efficacy of palivizumab in randomized controlled trials (RCTs). This table is incomplete both from the placebo controlled (2) or combined palivizumab-placebo trials and the comparative palivizumab-motavizumab trials perspectives. In the Cochrane review (3), seven RCTs were identified, confirming the efficacy of palivizumab in reducing the incidence of serious respiratory syncytial virus (RSV) disease in children with chronic lung disease (CLD), congenital heart disease (CHD) and those born preterm <35 weeks gestational age (wGA).

The recommendation for palivizumab targets patients who are oxygen dependent at 36 wGA or medically dependent in the first year, and only oxygen dependent in the second year. This reflects a narrow view of the limitations of using oxygen dependency alone as a diagnostic criterion for CLD (4), the pathophysiology of lung function in healthy preterms and those with CLD and the impact of RSV infection, and the rates of RSV-related hospitalization (RSVH), which are 12- to 18-fold higher in this cohort compared with healthy infants up to two years of age (5). The mean RSVH rate for patients with CLD across the 10 observational studies portrayed in the statement, albeit a methodologically flawed comparison, are approximately similar (16.8 first season; 14.7 first or second season; 13.9 first and second season). Moreover, the incidence of CLD overall has remained relatively unchanged because of its multifactorial etiology (6) and, following RSVH, both preterm and CLD infants have two- to threefold higher rates of intensive care unit (ICU) admission and ventilation (7,8). Although the number needed to treat (NNT) for CLD in the Impact trial was 20 (9), studies with larger sample sizes in ‘real world’ experience indicate a NNT range of eight to 10 (10–12). Late preterm infants incur significant respiratory morbidities (13) and are at high risk for RSVH (14,15). Three robust models targeting RSV prophylaxis cost effectively for these infants were overlooked (16–18) in favour of non-treatment, and strongly merits reconsideration because the estimated NNT across all three models equals 13. Overall, the guidelines for RSV prophylaxis restricted to infants <29 completed wGA without CLD is a serious step back to the era before 1998, and may result in significant burden of illness in untreated infants with higher ICU admissions and need for mechanical ventilation (29 to 32 wGA [n=243]: 49%, 25%; 33 to 34 wGA [n=279]: 41%, 19%; 35 wGA [n=187] 32%, 14%, respectively) (19).

The references in support of RSV prophylaxis for one versus two seasons for infants with CHD is limited to one RCT (20), which showed a declining RSVH incidence from 12% to 4% between one and two years, but >50% palivizumab efficacy for infants <6 months of age and those one to two years of age. The recommendation discounts both the severity of fully uncorrected, unstable CHD, the fourfold higher RSVH rates in these infants compared with healthy babies (21) and the significant weighted mean mortality rate of 5.2% with RSV infection (22). We have recently demonstrated that the hazard ratio for RSVH between the first and second RSV season for treated CHD infants is similar (HR 1.4 [95% CI 0.5 to 4.0]; P=0.52) (23) and prophylaxis for selected, cardiology sanctioned, high-risk infants for two seasons requires revisitation.

In regard to the “Cost-effectiveness of palivizumab”, we were unable to determine the threshold used to determine cost effectiveness. The Canadian threshold of up to $100,000 per quality-adjusted life year (QALY) is considered to be a good use of resources (24). Hence, we cannot assess whether the statement that palivizumab “is unlikely to be cost effective in children with prematurity, CLD or CHD…” is accurate. Also, the perspective is not stated, for clarity see Guidelines for the Evaluation of Health Technologies: Canada, 3rd edition, 2006 (25).

In a publication produced by the Canadian Paediatric Society, which describes itself as “a national advocacy association that promotes the health needs of children and youth”, we would expect a societal perspective rather than a payer perspective and, as such, consideration of the family costs of admission to hospital with RSV infection. Our preceding comments should not be interpreted to mean that we are not sensitive to the need to examine costs. However, the position statement starts with the assumption that the first priority is to “save drug costs”. As advocates for children, we suggest that the first stage should be to determine the best course of action to benefit children based on an unbiased review of the evidence, and then examine the funding implications. The position statement appears to have acted in the reverse order.

The position statement concludes that “people serving on this panel (Provincial Annual Reviews) should not have conflicts of interest, including research funding and participation in a speakers bureau or financial links with the pharmaceutical firm that makes palivizumab”. This simplistic approach would exclude many knowledgeable individuals on RSV infection, including members of the committees who developed and reviewed the document. Overall, this position statement overturns strong evidence supporting current practice for RSV prophylaxis in Canada and, if adopted, will lead to unnecessary infections in already vulnerable children.