We documented very high hospitalization incidence in infants with HsCHD in Hong Kong with prolonged RSV circulation. This hospitalization incidence was higher than that previously documented in Hong Kong infants < 6 months of age who did not have congenital heart disease.[1] The incidence we reported is at least comparable if not higher than that previously reported in patients with HsCHD.[6, 9, 10, 18] We also documented that infants with HsCHD in Hong Kong were hospitalized for RSV infection in nearly every month of the year.
Palivizumab has been demonstrated to be a reasonably effective immunoprophylaxis in preventing RSVh in patients with HsCHD in North American and European countries where seasonal clustering of RSV is often evident.[12, 19–21] This had resulted in the recommendation from the AAP on a 5 consecutive dose of palivizumab immunoprophylaxis during RSV seasons for patients with HsCHD below 1 year of age.[13] However, such regimen has not been uniformly successful, with persistence of significant RSVh risk being reported despite Palivizumab prophylaxis for patients with HsCHD in countries with a distinct RSV season.[22] The applicability of the AAP recommendation is even more questionable in subtropical and tropical areas that often lack a defined RSV season.[23–25] Limited studies have tried to elucidate the efficacy of alternative Palivizumab regimens in such countries. Chiu et al from Taiwan reported the use of a novel palivizumab prophylaxis protocol, involving 6 monthly injections commencing right after diagnosis of HsCHD has been made, and concluded that to be effective in preventing RSVh.[16] However, the study was limited by its low baseline RSVh rate of 4.8% only from a historical cohort as compared with 11% of our cohort which is more in line with other published data.[6, 12, 26]
Our 3-year study with a range RSV circulation activity provided a better estimation of the effectiveness of palivizumab prophylaxis in areas without a sharp RSV season. The NNT to prevent 1 RSVh could vary from 35 to as high as 166 in a three-year period, with a very high NNT in 2014 when the RSV activity was the lowest with RSVh rate of 4% only. This year to year variation of incidence has actually been well reported in subtropical and tropical areas, and could be related to various meteorological factors.[27–30] The overall high NNT and high cost in preventing RSVh with palivizumab could be explained by the year round circulation of RSV activity resulting in a significant proportion of RSVh occurring outside the prophylaxis period.
The high cost of palivizumab is yet another important matter of concern. Rigorous effort has been instigated in addressing the cost-effectiveness of its use in preventing RSVh in patients with HsCHD.[26, 31–40] Unfortunately, only a few studies were focused on this group of patients alone in areas with a distinct RSV season, and none of which were population-based.[26, 31, 33, 40, 41] These studies were mainly based on either hypothetical patient cohorts or decision-analysis models using ICER/Quality Adjusted Life year (QALY) to assess cost-effectiveness with an exceedingly wide range reported from €9,259 to US$114,337/QALY.[33, 40, 41] Despite a very favorable ICER/QALY was reported by Nuijten et al, a scenario analysis excluding mortality from the model was reported to raise the ICER/QALY significantly from €9,259 to €123,439/QALY.[41] Judging from the fact that the RSV related mortality in this group of patients reported from the REGAL series covers a wide range from 0 to 3.3%, we think using ICER/QALY in hypothetical model may result in significant inaccuracy. Since there was no reported RSV related mortality in our study cohort, we preferred to assess the cost-effectiveness with ICER/HAP, which was probably more appropriate for medical systems with low to minimal RSV related mortality. Furthermore, since the protection against RSVh rendered by palivizumab was not 100% as reported by Feltes et al[12], unlike many of the aforementioned studies, we have also included in our calculation the cost of RSVh as a result of palivizumab failure. From our results, not only was the cost of US$85,937/HAP (based on 5-dose regimen) much higher than that reported by Rackhman et al[26], our data also pointed out an important effect of the year-to-year variation in RSV activity on cost-effectiveness with ICER/HAP showing an almost 7-fold differences comparing years with lowest and highest RSV activity with a similar number of infants with newly diagnosed HsCHD each year in the population. Judging from our data, albeit a lack of consensus for an acceptable threshold for ICER/HAP, a Palivizumab program is not cost effective in preventing RSVh in patients with HsCHD. On the other hand, other strategies may prove to be more cost effective in reducing RSVh. In our cohort, the mean waiting time for HsCHD patient with heart failure to undergo surgical intervention was 4.3 months. A reduction in this waiting time, and hence the cumulative time at risk of RSVh, can potentially reduce the burden of RSVh. Development of RSV vaccine is certainly another option and numerous vaccine candidates are currently in the pipeline.[42] Once licensed, such vaccines may replace Palivizumab in preventing RSVh in high risk patients including those with HsCHD.
Our study had several limitations. Firstly, the assessment of efficacy and cost-effectiveness of Palivizumab was based on assumptions of a relative risk reduction in RSVh from published data in regions with distinct RSV seasons rather than from local prospective data with round the year RSV activity. To compensate for the difference in RSV activity, we have factored into the time of infection and considered those infected outside the hypothetical program as prophylaxis failure. Secondly, tangible cost like administrative cost of Palivizumab and non-tangible costs like cares’ absence from work were not included in the calculation. However, by only considering drug cost, our data reflected the minimal cost of giving Palivizumab. Finally, since the patient cohort was identified from a database, we might have inadvertently omitted patients who might qualify for Palivizumab if recruited prospectively. An example being patients presented with heart failure and an initial intent for surgery, but in the end did not require surgical intervention. In a prospective study, these patients would be eligible for Palivizumab injection. For that matter, a post-hoc analysis was conducted for all patients classified as mild CHD with RSVh. Amongst this group there were only 2 RSVh and both of which happened beyond 6 months after the diagnosis of CHD was made, making them unpreventable by Palivizumab even if given. Hence including these patients with “borderline” indication would only worsening the ICER/HAP making it even less cost-effective.
Strengths of this study included the data obtained were robust. Firstly, we are the only major pediatric cardiology center that served nearly all pediatric patients with congenital heart disease in Hong Kong. In addition, RSV testing was routine in children with acute respiratory infections and was PCR-based across all hospitals in Hong Kong. Moreover, we have also elected to obtain data over the span of 3 years with wide variation of RSV circulation burden enabling us to assess scenario with a range of RSV activity levels. Lastly, each HsCHD case was intensively reviewed by a pediatric cardiologist, not only relying on diagnostic codes from the database, ensuring correct classification of the patients’ cardiac status.