Cost analysis of standard Sanger sequencing versus next generation sequencing in the ICONIC study

Abstract Background HIV and hepatitis C virus (HCV) are a major cause of morbidity and mortality, and both viruses contain high genomic variation. To date, viral gene sequencing has been handled by standard Sanger sequencing (SSS) for the detection of specific drug-resistance determinants for HIV and HCV. However, SSS-derived information is very limited. By contrast, full-length viral gene sequences when linked to clinical data might influence the monitoring of drug resistance to optimally guide treatment, identify sources of viral transmissions within health-care settings, and track emerging epidemics. The ICONIC (Infection Response through Virus Genomics) study aims to introduce a novel method called next generation sequencing (NGS) within UK health-care settings, testing potential implementation in routine practice. With use of samples within established diagnostic laboratory workflows, NGS has the potential to produce higher informational content than SSS and report in a timely fashion. However, economic evidence for this emerging method is scarce. We aimed to use the examples of HIV and HCV to compare the cost of NGS versus SSS. Methods We performed a bottom-up cost analysis using published, genomic-testing, costing templates to estimate the mean cost per sample for SSS and NGS methods over a 1 year period at a major London hospital. Data on resource use associated with genomic testing were based on estimates from individual sample data, routinely collected from a UK population. Findings With SSS, mean cost per sample, including operating costs, was £178 for HCV (2080 samples) and £79 for HIV (520). Mean cost per sample with NGS was £119 (2207 samples), including operating costs, generating a cost saving of £59 for HCV and a surplus of £40 for HIV. Although this method is still research based and prices vary widely, our results demonstrated a broad NGS and SSS cost equivalence. Interpretation NGS is data rich and could be integrated in emerging stratified patient treatments. The mean cost per sample for the two methods should be similar and provide added-value information. A number of costing toolkits should be designed to address the appropriate pricing, including health-care consultation and operating costs when considering NGS efficiency and cost-effectiveness. The mean cost per sample for NGS as part of routine health care requires further exploration. Funding This project is funded by the Health Innovation Challenge Fund, a parallel partnership between the Wellcome Trust and the Department of Health (ref HICF-T5-344). FRB received funding for this study from the National Institute for Health Research (NIHR) Biomedical Research Centre, and the UCLH/UCL Biomedical Research Centre funded this NIHR Health Informatics Collaborative study.