Articles www.thelancet.com/psychiatry Vol 1 June 2014 55 Monitoring white blood cell count in adult patients with schizophrenia who are taking clozapine: a cost-effectiveness analysis François R Girardin*, Antoine Poncet*, Marc Blondon, Victoria Rollason, Nathalie Vernaz, Yves Chalandon, Pierre Dayer, Christophe Combescure Summary Background Long-term monitoring of white blood cell count is compulsory in patients taking clozapine, although the incidence of drug-induced agranulocytosis is lower than previously expected. The cost-effectiveness of this monitoring is unknown. We aimed to assess the cost-effectiveness of various strategies to monitor white blood cell count in adult patients with schizophrenia taking clozapine. Methods We assessed the cost-effectiveness of four strategies for monitoring white blood cell count (national strategies used in the UK, USA, and European countries, and a hypothetical 8-week strategy) compared with that of no monitoring. We used a semi-Markov model to do the cost–utility analysis from a health-care perspective with a 3-year time horizon, assuming a probability of 0·7% that a patient would develop agranulocytosis. Clinical and resource parameters were based on data from national registries of patients treated with clozapine, study cohorts, and a pharmacovigilance database; we derived estimates of health-related quality of life and mortality from the scientific literature. We assessed model uncertainty, including time horizon, with one-way and probabilistic sensitivity analyses. Findings Compared with no monitoring, all four monitoring strategies increased quality-adjusted survival by less than 1 day per patient; more than 5000 patients would need to be monitored to avoid one death. The incremental cost- effectiveness ratios (ICERs) were at least US$970 000 per quality-adjusted life-year gained for all four strategies compared with no monitoring. The ICERs were highest in the strategies with highest frequencies and longest durations of monitoring. The results remained robust in the one-way and probabilistic sensitivity analyses, suggesting that no monitoring had the highest probability of being cost effective. Interpretation Existing strategies for monitoring white blood cell count in patients taking clozapine, based on divergent national requirements, do not seem to be cost effective. This finding should be taken into account by public health authorities and policy makers in the revision of guidance for clozapine prescription. Funding University Hospitals of Geneva. Introduction Schizophrenia is a chronic relapsing illness and affects about 1% of the population in early adulthood. Findings from cost-of-illness studies 1 suggest that schizophrenia and related disorders are among the most debilitating diseases and have very high resource consumption. Patients with schizophrenia benefit from maintenance treatment with antipsychotic drugs, with reductions in relapses, readmission to hospital, and suicide. 2 Compared with other neuroleptic drugs, clozapine is an atypical antipsychotic drug with higher overall efficacy, better quality of life outcomes, 3 and few extrapyramidal side- effects, 4 meaning that it can even be prescribed for Parkinson’s disease psychosis. 5 In Europe and North America, about 25% of patients with schizophrenia are treated with clozapine. 6 After the worldwide withdrawal of clozapine in 1975 because of a cluster of eight fatal cases of agranulocytosis, 7 the drug was reintroduced onto the market in 1990 as a third-line treatment for patients with refractory psychosis resistant to other antipsychotic drugs. 8 Its prescription was associated with stringent conditions, such as a registry-based prescribing system and a requirement for long-term monitoring of white blood cell count. Late-onset agranulocytosis seems to be quite rare: 9 the cumulative incidence of agranulocytosis (0·38–0·80% up to a 3-year time-horizon) and related case-fatality (2·7–4·2%) in large monitoring registries 10–12 were lower than previously expected. This finding led to a relaxation of the frequency of required monitoring in many countries, but also to divergent recommendations in the USA, UK, and European countries, where long-term monitoring is still compulsory. The cost-effectiveness of monitoring white blood cell counts for patients taking clozapine remains largely unknown. Results from one benchmark case analysis 13 with strong assumptions and no probabilistic approach suggested that long-term monitoring is not cost effective, even in conservative scenarios with an agranulocytosis mortality rate of up to 20%. To investigate the efficiency of divergent national regulations, we did an economic assessment of continuous monitoring of white blood cell count and measured the cost-effectiveness of four monitoring strategies compared with no monitoring in Lancet Psychiatry 2014; 1: 55–62 Published Online May 29, 2014 http://dx.doi.org/10.1016/ S2215-0366(14)70245-7 See Comment page 8 *Contributed equally Department of Anesthesiology, Intensive Care, and Clinical Pharmacology (F R Girardin MD, V Rollason PhD, Prof P Dayer MD), Department of Health and Community Medicine (A Poncet MSc, C Combescure PhD), Department of Pharmacy (N Vernaz PhD), and Department of Medical Specialties (M Blondon MD, Prof Y Chalandon MD), University Hospitals and University of Geneva, Geneva, Switzerland; and Centre for Health Economics, University of York, York, UK (F R Girardin) Correspondence to: Dr François R Girardin, Division of Psychopharmacology, Clinical Pharmacology, and Toxicology, University Hospitals of Geneva, 1205 Geneva, Switzerland francois.girardin@hcuge.ch