1 Introduction
Hemoglobinopathies are among the most prevalent clinical genetic disorders worldwide [
1‐
3]. Although hemoglobinopathies are rare in Germany [
4], global migration patterns have resulted in a higher national prevalence [
5‐
7] and increased healthcare burden in Germany [
6]. Sickle cell disease (SCD) and β-thalassemia are hemoglobinopathies of particular concern, as they require lifelong management and are associated with reduced life expectancy [
8,
9]. SCD and β-thalassemia are hereditary monogenic hemoglobinopathies arising from mutations in the gene that encodes for β-globin [
10,
11]. These mutations result in the production of hemoglobin that is marked by a dysfunctional sickled phenotype in SCD [
9,
12], or profoundly reduced or absent synthesis of β-globin and excessive synthesis of unbound α-globin in β-thalassemia [
11,
13].
Both hemoglobinopathies are associated with significant, hallmark clinical complications [
14,
15]. In SCD, deoxygenated sickled hemoglobin polymerizes, obstructing blood vessels and leading to vaso-occlusive crises (VOCs), which are unpredictable painful events associated with profound morbidity and early mortality [
9,
16,
17]. Similarly, β-thalassemia leads to common clinical consequences such as ineffective erythropoiesis, hemochromatosis (iron overload), and chronic anemia. Managing transfusion-dependent β-thalassemia (TDT) requires individuals to receive regular red blood cell transfusions (RBCTs) to survive and iron chelation therapy to alleviate clinical symptoms and prevent further complications associated with RBCT-related iron overload [
11,
13]. For these reasons, the management of both disorders is associated with significant healthcare resource utilization (HCRU) [
14,
18,
19].
Recent claims-based analyses have highlighted the burden associated with managing SCD and β-thalassemia among these growing patient populations in Germany. With ongoing national efforts in Germany to optimize the management of SCD in the past decade [
5], the incidence of acute chest syndrome has decreased and use of hydroxyurea (HU) has increased [
5], trends that likely reflect the recent changes in clinical practice and adoption of treatment guidelines that favor the use of HU for the treatment of SCD [
5]. In one study (
n = 623), 32 patients with β-thalassemia required ≥ 1 RBCT to manage their disease and nearly one-third of these patients (28.1%) received ≥ 8 RBCTs over the 12-month follow-up period [
20]. Furthermore, the proportion of patients reporting reduced earning capacity was five times higher among patients requiring ≥ 1 RBCT than patients with β-thalassemia (15.6% versus 3.1%) [
20].
Data describing the spectrum of clinical complications and economic burden associated with SCD with recurrent VOCs and TDT in Germany are limited, especially among older patients and those with more severe disease. Studies published in other countries, however, have reported increased clinical complications and HCRU among this patient group [
14,
19,
21]. Given the significant impacts associated with VOCs among patients with SCD and the ongoing management needed for patients with TDT, the primary aim of this retrospective, real-world claims analysis was to describe the current clinical burden and healthcare system impact associated with these two hemoglobinopathies in Germany.
2 Methods
2.1 Study Design and Data Source
This observational, retrospective cohort study utilized the Betriebskrankenkassen (BKKs) German Sickness Fund Database to identify eligible patients during the overall study period (1 January 2010 to 31 December 2019). The BKKs German Sickness Fund Database is a secondary anonymized health insurance database that includes healthcare claims data for approximately 5.3 million insured patients in Germany (or 6‒7% of the German statutory health insurance [SHI] population) that has been utilized in previously in research [
22‐
25]. Data housed in the BKKs represent the entire German population that is insured by statutory health insurance with respect to common demographics such as age, gender, and federal states and health insurance funds expenditures. These data include patient demographics; inpatient and outpatient diagnoses coded by the International Classification of Disease, Tenth Revision, German Modification (ICD-10-GM); outpatient dispensing codes using the World Health Organization Anatomical Therapeutic Chemical (ATC) classification system; and inpatient and outpatient care.
Patient data were anonymized prior to transfer into the study analysis software system, in compliance with German data protection regulations. Analyses were performed by the Team Gesundheit GmbH, Essen, Germany. All required study approvals were secured prior to the start of the study.
2.2 Study Population
Patients with a diagnosis of SCD or β-thalassemia in the BKKs database between 1 January 2010 and 31 December 2018 were identified.
2.2.1 Patients with SCD
Patients with SCD with recurrent VOCs were included if they had experienced ≥ 2 VOCs (defined as SCD with crisis, priapism, or acute chest syndrome) per year in any two consecutive years and had ≥ 12 months of data available before and after the index date (defined as the date of the second VOC in the second of two consecutive years). The index date for patients with SCD with recurrent VOCs was the second VOC in the second consecutive year to index patients immediately when they achieved the recurrent VOC definition; this is consistent with previously published literature [
14]. Patients with SCD with recurrent VOCs were excluded if there was evidence of hereditary persistence of fetal hemoglobin (HPFH) or hematopoietic stem cell transplantation (HSCT) at any time in their medical records.
2.2.2 Patients with TDT
Patients with β-thalassemia were included if they had ≥ 8 RBCT claims in any 12-month period and ≥ 12 months of data available after the index date (defined as the date of their first RBCT claim). The index date for patients with TDT being the first transfusion is because patients with TDT generally begin requiring regular transfusions as early as 2 years of age; the index date definition is aligned with previously published literature [
26‐
28]. RBCT events were identified via national procedural codes and were considered distinct if they had occurred at least 1 day apart from each other. Patients with TDT were excluded if at any time in their medical records they had a diagnosis of SCD or evidence of HPFH, HSCT, or α-thalassemia, unless the diagnosis code was preceded by the diagnosis code for β-thalassemia.
2.2.3 Matched Controls
Each patient with SCD with recurrent VOCs and each patient with TDT was independently matched to five controls who did not have SCD, TDT, or any other blood disorder, or who met other exclusion criteria outlined above; one to five matching was utilized given the likely limited sample size for cases and is consistent with other published studies. Matched controls were selected from the general population of Germany in the BKKs by random sampling, and exact matching was based on age (in years) at the index date, sex, and region of residence. In addition, a record of healthcare use (i.e., consultation, drug dispensation, biological tests) around the index date (± 1 month) was required for the matched patients with SCD/TDT [
29]. Controls had a similar index date as the patients with SCD with recurrent VOCs or TDT with whom they were matched.
Both patient groups were followed for ≥ 12 months from the index date to a censoring event, defined as either patient death, patient deregistration due to the patient ending their insurance coverage, or the end of the study period (31 December 2019), whichever came first.
2.3 Study Outcomes
Patient demographics were assessed at baseline or the index date. HCRU outcomes (hospitalizations, outpatient visits, and outpatient prescriptions) were assessed during follow-up. Clinical complications were assessed using ICD-10 codes during follow-up; clinical complications were selected on the basis of common conditions in people with SCD or TDT and clinician input [
30,
31].
2.4 Statistical Analysis
Descriptive analyses were conducted to describe demographics, clinical complications, and HCRU among patients with SCD with recurrent VOCs, patients with TDT, and matched controls. Categorical variables are presented as number of patients (or events) and proportion. Outcomes with patient numbers < 5 were masked (i.e., noted as “–”) to safeguard patient confidentially. Continuous variables are presented as mean and standard deviation (SD). Acute complication rates were calculated using the number of events as the numerator and number of years during follow-up (person-time) as the denominator (per patient per year [PPPY]). The proportion of individuals with a chronic complication was calculated over the follow-up period. Comparative analyses were conducted for HCRU outcomes between patients and matched controls. Z tests were used to assess between-group differences, with p < 0.05 denoting statistical significance.
Subgroup analyses were conducted for HCRU outcomes on the basis of the annualized rate of VOCs (< 2 and ≥ 2 PPPY) among patients with SCD with recurrent VOCs and the annualized rate of RBCTs (< 8, 8–12, 12 to < 16, and ≥ 16 PPPY) among patients with TDT.
4 Discussion
To our knowledge, this retrospective, real-world claims analysis is the first to use SHI databases to describe clinical complications and HCRU among patients with severe hemoglobinopathies, including SCD with recurrent VOCs and TDT, in Germany. Patients with SCD and patients with TDT experienced substantial clinical complications and frequent VOCs and RBCTs, respectively, which largely contributed to their increased HCRU. This study used different definitions to identify patients with severe forms of SCD (SCD with ≥ 2 VOCs [SCD with crisis, priapism, or acute chest syndrome] per year for two consecutive years) [
32] and TDT (β-thalassemia with ≥ 8 RBCTs in a 12-month period) [
33], highlighting the significant burden of severe illness among these patient populations.
The prevalence of clinical complications observed in this study was higher than that observed in previous studies of SCD in Germany [
5,
34]. This finding may reflect the difference in the severity of disease examined (i.e., the broad SCD population [
5] versus patients with SCD with recurrent VOCs); the different claims databases used to identify patients with SCD and consequent study population demographics (i.e., our study captured an older SCD patient population with a mean age of 40.1 years using the BKKs database); the different SCD definitions used (i.e., SCD diagnoses by experts and/or two independent physicians in Kunz et al. [
5] versus ICD-10-GM codes in our study); and the different study methodologies used [
34]. For example, one previous German study of patients with SCD used the Allgemeine Ortskrankenkassen (AOK) database, which included general local SHIs, and noted that the largest proportion of patients with SCD was under 30 years of age [
5], while another registry-based German study was limited to data from pediatric centers [
34]. Although these previously published studies provide much-needed insight into the increasing burden of SCD in Germany, there is limited information on the more severe phenotypes of SCD (i.e., recurrent VOCs); our study’s focus on this more severe subgroup adds to the evidence on substantial burden associated with SCD.
The degree of clinical complications observed in this study may indicate patient populations that have lived with severe hemoglobinopathies for several decades and the progressive nature of these hemoglobinopathies [
13,
19]. Acute and chronic clinical complications were prevalent and affected all major organ systems (i.e., cardiopulmonary, endocrine, mental health, vision, and urinary), reinforcing our understanding of the widespread, lifelong burden associated with managing these hemoglobinopathies [
13,
19,
35]. Although others studies of SCD or TDT in Germany have reported fewer complications in these patient groups [
35], rates of clinical complications observed in this study were consistent with those observed among older patients (aged > 15 years) with TDT in Germany [
35]. The prevalence of malignancy was also notable among patients with TDT (44.1%) in this study, and although this finding may indicate that the risk of malignancy increases with age [
36,
37], it may also indicate that this patient population has an increased risk of malignancy, which would represent a novel finding in TDT and calls for additional prospective studies on this important issue. These findings highlight the need for appropriate management of enduring clinical burden across the entire age continuum and the importance of treating these disorders as early as possible in the disease course before significant disease progression.
Findings from this study are consistent with those of previous studies in other countries that reported increased HCRU among patients with severe hemoglobinopathies [
15,
18,
19,
38‐
40]. HCRU was significantly higher among patients with SCD with recurrent VOCs and patients with TDT than among matched controls, especially the mean numbers of hospitalizations PPPY, which were 6- and 11-fold higher among patients with SCD with recurrent VOCs and patients with TDT, respectively. Additionally, increased outpatient visits and prescriptions were observed among patients with SCD with recurrent VOCs and were even higher among patients with TDT, who had nearly 60 mean outpatient visits PPPY. Similar to previous reports in Europe and North America, HCRU increased with the number of annual VOCs [
14,
38] and RBCTs [
19], an increase that likely reflects the greater economic burden among these patient groups [
15,
38]. Results from this study may be indicative of current practices in disease management and ongoing national efforts to improve the management of SCD in Germany [
5,
41]; however, these results underscore the substantial burden of illness in patients with severe hemoglobinopathies and highlight a potential opportunity to introduce similar efforts for β-thalassemia. While other studies have described the importance of diagnosing SCD early in the disease course [
5] and the recent implementation of newborn screening for SCD [
41], this study and other studies in Germany [
4,
5,
20,
35] reinforce that hemoglobinopathies are a prevalent health problem in Germany [
4,
6] and that novel therapies are urgently needed, especially among patients who may be under-represented and/or may require lifelong monitoring.
Additionally, recent studies in Germany have noted the importance of using consistent diagnosis criteria for SCD. In one study, use of the “diagnosis by experts” definition (documented diagnosis with ICD-10 codes from hematologist or hospital outpatient clinic, or as the principal diagnosis preceding a hospital admission) versus the “two independent diagnoses” definition captured different proportions of children and adolescents with SCD < 20 years of age [
5]. Although this study provides important context for the current SCD landscape in Germany, its broad focus on patients with SCD does not capture the nuanced burden of illness among individuals with more severe phenotypes (i.e., SCD with recurrent VOCs), and its use of multiple diagnostic criteria suggests that specific subsets of patients may be under-represented when different criteria are applied.
This study had several limitations. First, our use of the BKKs German Sickness Fund Database may have limited the overall generalizability to all patients with SCD or TDT residing in and recently immigrating to Germany. The ages of patients with SCD with recurrent VOCs and patients with TDT in our study were higher than in other reports in Germany; therefore, the patient populations included in our study may not represent all younger patients with these hemoglobinopathies. However, findings from this study uncovered increased clinical complications in an older patient cohort. Although the prevalence of malignancies for patients with TDT was reanalyzed and confirmed, the unexpected nature of these findings warrants additional studies using other data sources, especially in older patients with TDT. Given that this study utilizes ICD-10 codes within a claims database to capture complications, there is some inherent risk of potential misclassification of complications. In addition, the database did not include a specific category for emergency room visits, but there were classifications of hospitalizations (i.e., ambulatory [hospitalizations lasting ≤ 1 day]). While it is likely that the ambulatory hospitalizations might capture some of the emergency department visits, there may be significant HCRU by patients with SCD or TDT not captured. As there is currently no standard definition for SCD with recurrent VOCs, our ability to directly compare this study to other studies was limited. Although the majority of VOCs in the follow-up period were due to pain crisis and our use of a broader definition for VOC (i.e., SCD with crisis, priapism, or acute chest syndrome) was in line with current clinical trials investigating gene therapy for SCD [
33], direct comparisons with studies using varying definitions were limited. In addition, there is no validated algorithm for diagnosis of β-thalassemia in claims and the positive predictive values for ICD-9/10 codes are unknown. However, our use of the current definition requiring ≥ 8 RBCTs in any 12-month period likely resulted in high specificity for identifying patients with TDT. Lastly, the small sample size of this study has limited our ability to conduct subgroup analyses.
Acknowledgements
Medical writing and editing support were provided under the guidance of the authors by Jenifer Li, MSc, and Nicholas Strange, BA, of Complete HealthVizion, IPG Health Medical Communications, Chicago, IL, USA, funded by Vertex Pharmaceuticals Incorporated.