Polycythemia vera (PV) and essential thrombocythemia (ET) are two common types of myeloproliferative disorders (MPD). The prevalence of PV and ET in the United States (US) has not been well documented. Recent breakthroughs in the molecular etiology of these disorders and the accelerated development of targeted pharmacotherapeutics to treat the MPD have significantly increased the need to accurately define the affected population. In the present study, we obtained detailed demographic and health claims data from major commercial insurance payers in Connecticut and the Center for Medicare and Medicaid Services to estimate the prevalence of PV and ET. Health claims data from one payer and the actual diagnoses made by physicians who submitted claims with MPD-related ICD-9 codes were utilized to develop claim-based statistical algorithms to predict the probability that an individual with claims suggestive of MPD truly has PV or ET. Specifically, logistic regression was used to develop the algorithms, and area under the receiver operating characteristics curve (AUC) was used as the measure of goodness-of-fit for each model. Different models were fit, and the model with the highest AUC was selected. For PV, the best-fitting model included age as a continuous variable and the frequency of two PV-related ICD-9 codes (238.4 and 289.6 combined) as a continuous variable, and the AUC was 0.95. For ET, the best-fitting model included age as a continuous variable and the frequency of code 289.9 as a categorical variable (=0, =1, or ≥2), and the AUC was 0.72. For both models, the addition of gender or the frequency of the non-specific code 238.7 did not improve AUC. Subsequently, the algorithms were applied to health claims from multiple payers to estimate the number of PV and ET patients in Connecticut. The total number of Connecticut residents included in the study was close to 2,900,000, which represented about 83.2% of the estimated population of the entire state in July 2003. As of 2003, the age-standardized prevalence of PV and ET in Connecticut was 22 per 100,000 and 24 per 100,000, respectively. Applying the age-specific prevalence of PV and ET to the entire US population resulted in an estimated total of 65,243 patients with PV and 71,078 patients with ET in the US in 2003. This study is the first to assess the prevalence of PV and ET in a very large US population. Given the large number of individuals afflicted with these diseases and the fact that demographic changes alone will further increase the burden of these diseases in the foreseeable future, it is imperative to conduct more systematic research into the etiology and treatment of PV and ET.

Author notes

Disclosure: No relevant conflicts of interest to declare.

Sign in via your Institution