Article Text

RSV prevention
1. SANJEEV DESHPANDE, Consultant Neonatologist
1. Royal Shrewsbury Hospital
2. Mytton Oak Road, Shrewsbury, UK
3. e-mail: deshpande{at}which.net
1. R M NICHOLL
1. on behalf of the Evidence Based Neonatal Medicine Journal Club
2. Appraisers: C Cane, E Costales, U Chovodath, S Goldring, I Halberstat, S Hughes, S Lindsay, S Salgia, S Shekhar
3. Neonatal Unit, North West London Hospitals NHS Trust (Northwick Park Campus)
4. Harrow HA1 3UJ, UK

## Statistics from Altmetric.com

Editor,—I recently saw the UK guidance for the use of synagis (Palivizumab) brought out by the manufacturer and contributed to (?endorsed) by a number of UK experts. My interpretation of the IMpact-RSV (respiratory syncytial virus) study1appears to be at variance to the conclusions drawn and the recommendations of the “guidance”.

To me, in the above mentioned study,1 compared to placebo, Palivizumab: (a) did not (the trial lacked the power to demonstrate) reduce death or need for mechanical ventilation; (b) reduce RSV related hospitalisation by 55%—the primary outcome of interest.

Based on (b), the “guidance” recommends Palivizumab for premature infants of < 36 weeks' gestation who are < 6 months old at the start of RSV season as well as those with bronchopulmonary dysplasia < 2 years old who have required medical treatment within six months of the RSV season beginning.

Looking at the data in another way, one would need to treat 16 suggested “at risk” babies (20 for babies with bronchopulmonary dysplasia) to prevent one RSV related hospitalisation. Is it worth it in clinical or economic costs?

Clinical—There was no evidence that RSV illness was less severe among hospitalised Palivizumab recipients than among hospitalised placebo recipients. (The paper provides figures for secondary efficacy end points as per 100 treated children. I suppose they are clinically more relevant than for per 100 hospitalised children.) Although statistically insignificant, both the RSV related deaths occurred in the Palivizumab recipients. Three per cent of placebo recipients (but 15 of 53 hospitalised placebo recipients) and 1.3% of Palivizumab recipients (but 13 of 48 hospitalised Palivizumab recipients) had RSV intensive care unit admissions. Only 0.2% (one of 500 treated or 53 hospitalised) placebo recipients and 0.7% (seven of 1002 treated or 48 hospitalised) Palivizumab recipients required mechanical ventilation. Similar conclusions can be drawn for total ventilator days. In other words, reduction in hospitalisation rates does not necessarily equate with reduction in severity of illness.

Economic—As 16 at risk infants need to be treated to prevent one hospitalisation, and as the infants would receive five monthly injections of the antibody costing between £424 (US$680) and £706 (US$1130) per injection (depending on their weight), the cost of preventing one hospitalisation from RSV related illness is at least £32 000 (US£51 000). I do not know the latest figures for a paediatric inpatient stay but analogies from neonatal “special care” category days would suggest that a six to eight day admission (calculated figures from the IMpact-RSV trial) should not cost more than £1800 to £2500 (US$2880–4000). So I see very little clinical benefit to those “at risk” children who receive Palivizumab and get RSV illness nor do I see any economic advantage to prevention of RSV related hospitalisations. Indeed both the PREVENT2 and Impact-RSV1studies show much lower rates of hospitalisation among the placebo infants than for historic data, which has been attributed among other things to “extensive education”. The valid conclusion then should be that education is the most important tool for prophylaxis against RSV hospitalisation. The “guidance” suggests drawing conclusions from local RSV hospitalisation rates but these are affected by geographic and social factors as much as clinical ones. Moreover, would local figures have any meaningful value if a large multinational study fails conclusively to prove clinical or economical benefits of the treatment? I do not dispute that there will be an occasional baby with such severe lung disease that any help, however proved, would be welcome. But a generalised preventive measure such as this will be difficult to justify. Have I got it all wrong? ## References Editor,—Palivizumab, a monoclonal respiratory syncytial virus (RSV) antibody that may reduce the need for hospital admission because of infection with RSV has recently received a licence for use in the UK. Guidelines for its use have already been published in the USA,1-1 and there are early indications that some paediatricians in this country plan to use this product.1-2 The USA guidelines are based on the results of a multicentre, randomised, double blind, placebo controlled trial (n = 1502) of Palivizumab1-3 conducted at 139 centres in the USA, Canada, and the UK. Treatment or placebo was administered by intramuscular injection at monthly intervals for five months during the RSV season, to infants considered at risk for hospital admission (the primary end point). This paper ends by stating that “prophylaxis with this monoclonal antibody results in a significant (55%) reduction in RSV hospitalization in children at high risk for severe RSV infection”. We have appraised the evidence on which this and other conclusions are drawn in the paper and feel it important to bring our comments to the attention of UK colleagues who may be considering this treatment as we approach the start of another RSV season. • The study group was heterogeneous in terms of baseline risk for acquiring RSV. Some were less than 35 weeks' gestation but had not been ventilated, others were of extreme low birthweight and had had severe bronchopulmonary dysplasia (BPD). Whether the study group was truly “high risk” could be debated. Only 1% of the entire study population required intensive care. Mortality was similar in both placebo (1%) and Palivizumab groups (0.4%). • The baseline risk for the primary outcome (hospital admission) was just 10.6% in the placebo control group. This is important as the reduction in hospital admission rates are expressed as relative risk reduction (RRR) with 95% confidence intervals (CI) and p values rather than absolute risk reduction (ARR), and number needed to treat (NNT)1-4 with 95% CI.1-5 • The use of RRR in the paper flatters the results, as table 1-1 illustrates. Data in bold have been extracted by us from the paper. Expenditure data is based on the number of babies needed to treat in each group to prevent one extra hospital admission and assuming the baby weighs 3 kg. Dose regimen is 15 mg/kg per dose for five doses; the cost of a 50 mg ampoule is £424 (US$680).

Table 1-1

RSV antibody: clinically useful measures of effect

While prevention of hospital admission is clearly desirable, many may feel the cost to be prohibitive. A UK clinical trial of Palivizumab in infants at high risk of needing intensive care because of RSV (for example, babies discharged from neonatal units and in home oxygen at the start of winter) might be more informative.

## References

1. 1-1.
2. 1-2.
3. 1-3.
4. 1-4.
5. 1-5.
View Abstract

## Request Permissions

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.