Cost-Effectiveness of Universal Duffy Testing at 9-12 Months: An Effort Towards Equity in Interpreting Absolute Neutrophil Counts (ANC)

Introduction: The Duffy null phenotype is found in 10% of people in the United States (US). This variant results in lower ANC without increased risk of infection. There have been recent large-scale efforts to develop Duffy null specific ANC reference ranges across the US, but limited efforts to obtain universal Duffy testing. We assessed the cost-effectiveness of adding universal testing for Duffy status to recommended testing at 9-12 months old.

Methods: We developed a Markov cohort model to evaluate the cost-effectiveness of 1) universal testing by serological testing for Duffy status versus 2) no testing (status quo) in the 2022 American birth cohort of 3.67 million children. Epidemiologic age- and sex-adjusted baselines for births and deaths were obtained. Duffy cohort characteristics (including rates of ANC<2000 by Duffy status) were drawn directly from a US-based cross-sectional study. Our analysis was performed from a US health system perspective, at a lifetime time-horizon and across all accepted willingness-to-pay (WTP) thresholds. As there are no neutropenia diagnosis guidelines, evaluation for both basic and full workup for asymptomatic neutropenia was informed by expert opinion as the model base-case and scenario analyses, respectively. Basic work-up included referral to hematology and anti-neutrophil antibodies, Duffy testing, immunoglobulins, and fecal elastase or pancreatic amylase. Full work-up included basic testing plus nutritional, inflammatory, autoimmune, and infectious work-up as well as congenital neutropenia genetic testing and bone marrow biopsy. Costs were informed from the Centers for Medicare & Medicaid Services, utilities were set on an age- and sex-adjusted general US population baseline, conservatively assuming that there is no quality-of-life improvement associated with Duffy testing. The primary outcome was the incremental net monetary benefit (iNMB) for universal testing vs no testing, with the secondary outcome being a threshold sensitivity analyses for minimum Duffy prevalence (for testing to be cost-effective). We performed deterministic sensitivity analysis varying all parameters the greater of their adjudicated 95% confidence intervals or ±20% and concluded with probabilistic sensitivity analyses that captured uncertainty in all parameters simultaneously over 10,000 Monte Carlo iterations.

Results: In the base-case, the discounted, lifetime costs for universal testing strategy vs status quo in the 2022 American birth cohort accrued $577.98 billion and $577.99 billion, respectively, accruing 98.6 million in discounted quality-adjusted life-years with both strategies, with an iNMB of $15 million [95% credible interval $5 million - $24 million] in favor of no testing. No testing was the cost-effective strategy in 100% of 10,000 Monte Carlo iterations. In scenario analysis of the full diagnostic work-up these costs were $578.13 billion and $578.26 billion, respectively, accruing 98.6 million, with an iNMB of $133 million [95% CI $96 million - $174 million] in favor of testing. Here testing was the cost-effective strategy in 100% of 10,000 Monte Carlo iterations. Threshold sensitivity analyses showed that the minimum Duffy prevalence for cost-effective population testing is 15% in the base-case and 3% in scenario analysis. Only variation in Duffy-null prevalence could alter the iNMB to favor the alternative strategy in the base-case and scenario analysis.

Conclusion: Universal Duffy testing is a novel approach to avoid intervention to mitigate patient harm and healthcare costs. The minimum threshold Duffy null prevalence for cost-effective US birth cohort testing ranges from 15% to 3% under basic and full neutropenia work-up, respectively. Duffy null prevalence in the US is ~10% and increases to >50% among patients who self-identify as Black or Middle Eastern. Next steps include developing detailed consensus on neutropenia testing and tiered (rather than basic vs full) testing. Also, this model only addresses Duffy testing and neutropenia evaluations and does not account for impact on disease or treatment outcomes, such as chemotherapy dose reductions secondary to neutropenia or discontinuation of other critical medications (i.e., azathioprine, clozapine). Quantitative consideration of the differential outcomes of these neoplastic, autoimmune, and/or psychiatric comorbidities due to unknown Duffy status is necessary.