Cost-effectiveness of palivizumab compared to no prophylaxis in term infants residing in the Canadian Arctic

Re: Cost-effectiveness of palivizumab compared to no prophylaxis in term infants residing in the Canadian Arctic CMAJ Oct 2016

Anna Banerji MPH MD, Kaspar Ng BPharm MD, Theo J. Moraes MD PhD, Vladimir Panzov MD, Joan Robinson MD, Bonita E. Lee MD

We read with interest the recent cost-effectiveness analysis by Banerji et al. (1) in CMAJ Open. There is no disputing the fact that lower respiratory tract infections (LRTIs) are a very important child health issue in northern Canada and one that we are faced with regularly as clinicians and public health practitioners working in Nunavut, and the north. We do, however, have several concerns with this manuscript that we feel are important to highlight.

As the basis of the cost effectiveness analysis an assumed effectiveness of palivizumab of 96% was used. This is the highest point estimate for effectiveness for palivizumab of which we are aware. This estimate is based on the finding that during the 2009-10 seasons 5 of the 10 infants identified by the investigators as eligible for palivizumab but who did not receive prophylaxis were admitted to hospital with RSV associated LRTI (2). All of these children were born prematurely and 2 of the five had other significant co-infections (parainfluenza and Bordetella pertussis). In fact 8 of these 10 children were only identified because of the study procedure, which was to specifically review all admissions to hospital due to LRTI. This procedure (identifying non-prophylaxed children from those admitted with LRTI) would therefore likely overestimate RSV positive admission rate among non-prophylaxed children. In contrast, a recent Cochrane collaboration systematic review (3) estimated the risk ratio of 0.49 for admission to hospital due to RSV for those provided with palivizumab prophylaxis vs. placebo. Most of the children included in the meta-analysis were born prematurely. Studies looking at other indications for palivizumab in children born at term (e.g. congenital cardiac disease or cystic fibrosis) have generally shown lower estimates of effectiveness (4). To therefore assume that term Inuit babies (who unlike premature infants, receive significant transplacental maternal RSV antibodies) would have such significant added protection from palivizumab would not seem supported by the data provided from the study.

The cost estimates are also based on only a single year of data (2009 season). This particular year had a discrete season, which therefore favours a lower cost estimate for palivizumab prophylaxis. It is well known that the seasonality for various infectious diseases in arctic regions are subject to significant variability. RSV has been a reportable infection since 2001 in Nunavut and government of Nunavut surveillance data has shown that there are frequently prolonged seasons with lower numbers of cases (e.g. March to October 2008 and March to August 2016). There can also be dramatic variation in the total number of RSV cases diagnosed on an annual basis (e.g. 2009 had the highest number of reported cases at 168 since surveillance began, whereas 2008 only had a total of 87 cases). This extreme variability (generally not seen in southern centres) draws into significant question an analysis based solely on one year of data. The cost analysis was also performed using relatively artificial regional sub-analyses. For example one region of Nunavut where the sub-analysis showed favourability for the intervention was Kitikmeot (population 6,010 in 2011 census). It is completely unclear why there are differences in admission rates seen in various regions (and whether these are in fact stable over time). These may reflect health system differences (e.g. single health centre referral patterns) that may be very transient and not any underlying innate persistent vulnerability of the children residing in these regions. In fact there is data from Inuit communities in Greenland (5) which has shown rates of admission for LRTI in young children to be much lower. The Banerji et al. cost effectiveness analysis also elected to exclude Iqaluit (population 6699 in 2011) from the analysis presumably because the admission rates were also too low to justify such a costly intervention. Of note, even in Banerji et al.'s companion descriptive paper, the vast majority of admissions to hospital were not due to RSV alone. Other infectious agents are obviously important contributors LRTI in these regions. These would all suggest that investments in bolstering the ability of individual communities to prevent infections and manage children with LRTI caused by ALL infectious agents (not only RSV) would be a more effective and sustainable use of resources.

The accompanying descriptive CMAJ Open article by Banerji et al. clearly states that the funding that supported that work was received from the palivizumab manufacturer. However for Banerji et al.'s palivizumab effectiveness and economic analyses (1,2) the funding source is not specified. It certainly would appear that neither of these analyses would have been possible without the larger industry funded descriptive study. A separate cost effectiveness analysis involving some of the same investigators (5) does explicitly mention industry funding for the research. We feel that it is very important to clarify whether this research was supported by industry funding, particularly given the implications for this marginalized and potentially vulnerable population. We recognise the large need for more research on this topic and would welcome this, conducted by investigators independent of industry sponsorship.

In addition to LRTIs, Inuit children have also unfortunately been shown to have very high rates of other infections (e.g. tuberculosis, invasive bacterial infections, ear infections, rotavirus, cryptosporidiosis, etc.), and double the preterm birth rates of the rest of Canada. They also have poor socioeconomic indicators including poverty, extremely high rates of over-crowded housing, and perhaps most concerning, among the highest rates of food insecurity in the developed world. As practitioners in these communities we would be the first to admit that these are complex, multifaceted issues and solutions unfortunately are not simple in this resource limited setting. We ask the authors - in a territory where 69% of children live in food insecure households (7) - a large proportion going full days without meals - what is the likely validity of an intervention that would cost $6500 or more per child, targets a single virus, and has uncertain effectiveness? There are many other contributors to overall health (including respiratory health) where significant resources are needed - e.g. targeting maternal and child nutritional status, improved breastfeeding rates, smoking cessation, etc. Funding and resources are obviously limited and perhaps would be better directed to these areas that would help improve overall maternal/child health and not simply target a single pathogen.

Health care spending in this region is quite high in large part related to its vast geography and sparse population. This however does not mean that we should be encouraged to medicalize an entire population of children in infancy through use of expensive and unproven (in this population) monthly synthetic antibody injections. We certainly welcome further research into solutions for what we recognize is a real problem, but given the data provided to date it would not appear to us that universal palivizumab use is a viable answer for Nunavut or any of its regions.

Sincerely,

Amber Miners MD, FRCPC Chief of Pediatrics Qikiqtani General Hospital Iqaluit, Nunavut [email protected]

W. Alexander (Sandy) Macdonald MD Territorial Chief of Staff Department of Health, Government of Nunavut, [email protected]

Maureen Baikie MD, FRCPC Public Health Advisor Past Chief Medical Officer of Health, Government of Nunavut [email protected]

David Goldfarb MD, FRCPC Consultant Pediatrician Kimmirut, Nunavut Medical Microbiologist, BC Children's Hospital [email protected]

References: 1.Banerji, A., et al., Cost-effectiveness of palivizumab compared to no prophylaxis in term infants residing in the Canadian Arctic. CMAJ Open, 2016(4): p. E623-633.

2.Banerji A, Panzov V, Young M, et al. The real-life effectiveness of palivizumab for reducing hospital admissions for respiratory syncytial virus in infants residing in Nunavut. Can Respir J 2014;21:185-9.

3.Andabaka T, Nickerson JW, Rojas-Reyes MX, et al. Monoclonal antibody for reducing the risk of respiratory syncytial virus infection in children. Cochrane Database Syst Rev. 2013 Apr 30;(4):CD006602.

4.Ralston SL, Lieberthal AS, Meissner HC, et al. Clinical practice guideline: the diagnosis, management, and prevention of bronchiolitis. Pediatrics. 2014 Nov;134(5):e1474-502.

5.Koch A, S?rensen P, Hom?e P, et al. Population-based study of acute respiratory infections in children, Greenland. Emerg Infect Dis. 2002 Jun;8(6):586-93.

6.Tam DY, Banerji A, Paes BA, et al. The cost effectiveness of palivizumab in term Inuit infants in the Eastern Canadian Arctic. J Med Econ. 2009;12(4):361-70.

7.Rosol R, Huet C, Wood M, Lennie C, Osborne G, Egeland GM. Prevalence of affirmative responses to questions of food insecurity: International Polar Year Inuit Health Survey, 2007-2008. Int J Circumpolar Health 2011;70(5):488-497.

Conflict of Interest:

None declared