Beta-Thalassemia Treatment and Complications in Two Large US Insured Population Databases

Beta(β)-thalassemia is a rare disease, with limited data available on the typical clinical course and outcomes of patients in the US. There are currently no nationwide surveillance efforts to monitor the prevalence or changes in outcomes among those affected, or to inform the development of standards of care, policy, or advocacy efforts. Most information on β-thalassemia treatment in the US comes from studies involved with the Thalassemia Clinical Research Network, which ended data collection in 2012.  Large administrative claims databases have been used to study numerous diseases, but insufficient information on β-thalassemia diagnosis codes has precluded their use for β-thalassemia research until recently. In 2011, new International Classification of Disease-9 (ICD9) codes for thalassemia subtypes were introduced, allowing analyses of various thalassemia populations in diverse settings across the US.  The purpose of this study was to describe the prevalence of β-thalassemia diagnoses among pediatric and adult patients in two large administrative databases covering broad US geographic areas, and to examine treatment patterns and health outcomes in these samples of β-thalassemia patients.

We analyzed data from two large insurance claims databases: OptumInsight Life Sciences Inc. Clinformatics™ DataMart MultiPlan and Truven Health MarketScan® Research commercial insurance database.  Each of these databases includes over 90 million commercially-insured patients in the US, with a greater representation of the Northeast region in Clinformatics™ and of the South in MarketScan®.  Primary analysis focused on medical and pharmaceutical claims between 2011 and 2014. Patients with β-thalassemia were defined as probable cases, with ≥3 diagnosis codes for β-thalassemia (ICD9 282.44) or HbE/β-thalassemia (ICD9 282.47). Transfusion-dependence was defined as ≥ 8 transfusion procedure codes in a 12-month period. For comparison to general population, 5 controls were selected per case, matched on age, sex, and follow-up time. Complications were defined by relevant ICD9 or procedure codes, and categorized using the Clinical Classification System developed by the Agency for Health Research and Quality.

A total of 880 patients (707 adults, 183 patients under age 18 y) with β-thalassemia were observed from 2011 onwards, of which 304 were from the Clinformatics™ database and 576 were from the MarketScan® database.  Iron overload was diagnosed in 21-22% of all cases and upwards of 68% of transfusion-dependent cases.  The most common oral chelating agent was deferasirox (69.8% of transfusion-dependent patients); deferoxamine claims occurred in 28.9% and deferiprone in 12.8% of patients with frequent transfusions. Diseases of the heart and endocrine system were common (Table).  

Percentage with Select Complications, by Database
	Clinformatics™		MarketScan®
	Total (N=304)	Transfusion-Dependent* (n=25)	Total (N=576)	Transfusion-Dependent (n=110)
Transfusion	24.7	100	28.8	100
Iron overload	21.1	32.0	22.4	68.2
Spleen disorder	10.5	4.0	9.9	14.6
Diseases of the heart	36.5	52.0	30.7	46.4
Liver disease	16.1	4.0	12.7	15.5
Osteoporosis	8.6	24.0	6.8	15.5
Nutritional, endocrine, & metabolic	42.8	76.0	40.3	80.0
Anxiety disorders	9.9	8.0	10.2	8.2

    *Additional transfusion-dependent cases may have been present among the 304 cases but excluded because bundling of codes obscured the frequency of transfusions.

Narcotic analgesics were used by 25% of β-thalassemia patients in the outpatient setting, and anti-anxiety or antidepressants medications were used by 16-17% of patients. 

In a 12-month period, β-thalassemia cases had significantly higher occurrence of numerous diagnoses, including viral and bacterial infections, neoplasms, lower gastrointestinal disorders, and contraceptive/procreative management, compared to controls.

Thus, using a vast amount of data from relatively large samples of β-thalassemia patients, we observed significant comorbidities that may be overlooked in small clinical studies in which such broad data collection is not feasible. This research provides unique information on the prevalence of various outcomes in β-thalassemia and over time may provide robust estimates of incidence and healthcare utilization patterns. Such insight on the long-term clinical course of β-thalassemia patients can inform clinical studies aiming to address unmet medical needs.