Hamostaseologie
DOI: 10.1055/a-2558-9886
Original Article

Epidemiology, Health Care Resource Utilization, and Costs of Hemophilia A and B by Treatment Regimen: A Retrospective Analysis of German Claims Data from 2016 to 2021

Dominik Obermüller
1   InGef - Institut für angewandte Gesundheitsforschung Berlin GmbH, Berlin, Germany
,
Karin Berger
2   Department of Medicine III, University Hospital, Ludwig-Maximilians-University of Munich, Munich, Germany
3   Bavarian Cancer Research Center (BZKF), Munich, Germany
4   German Cancer Consortium (DKTK), Partner Site Munich, Munich, Germany
,
Robert Klamroth
5   Zentrum für Gefäßmedizin, Klinik für Innere Medizin – Angiologie und Hämostaseologie, Vivantes Klinikum im Friedrichshain, Berlin, Germany
,
Maria Kleppisch
6   Pfizer Pharma GmbH, Berlin, Germany
,
Stephan Rauchensteiner
6   Pfizer Pharma GmbH, Berlin, Germany
,
Ines Ecke
6   Pfizer Pharma GmbH, Berlin, Germany
,
Sandra Hermann
6   Pfizer Pharma GmbH, Berlin, Germany
,
Dorota Pawlowska-Phelan
1   InGef - Institut für angewandte Gesundheitsforschung Berlin GmbH, Berlin, Germany
,
Dennis Häckl
7   Wig2 GmbH, Leipzig, Sachsen, Germany
,
6   Pfizer Pharma GmbH, Berlin, Germany
› Institutsangaben

Funding This study was funded by Pfizer Pharma GmbH. The funding organization provided financial support for the design, data collection, analysis, and interpretation of the study data.
 

Abstract

Introduction

In Germany, hemophilia patients with a severe bleeding phenotype receive lifelong prophylactic treatment with intravenous concentrated factor VIII (FVIII) or IX (FIX) to prevent bleeding events. To assess the economic value of emerging treatment options, studies describing the economic burden of hemophilia under the current standard of care in Germany are needed.

Aim

This study classified hemophilia A (HA) and B (HB) patients by treatment regimen in administrative claims data to examine the real-world economic burden of Hemophilia in Germany from 2016 to 2021.

Methods

Hemophilia patients were identified in InGef statutory health insurance claims data via ICD-10 codes D66 (HA) and D67 (HB) in combination with one or more claims for hemophilia-related medication.

Each patient's factor regimen was classified as either indicative of a severe phenotype needing prophylaxis or a non-severe phenotype treated on demand using a classification threshold of 100,000 International Units (IU) FVIII/year (HA) and 80,000 IU FIX/year (HB). Health care resource utilization and cost outcomes were captured for each study year.

Results

Male prevalence per 100,000 ranged from 6.39 to 7.81 (HA) and 1.26 to 1.89 (HB), with 43 to 53% (HA) and 40 to 56% (HB) categorized as severe. In 2021, mean (standard deviation [SD]) per-patient medication costs were €321,987 (€157,915) in the severe treatment group versus €43,487 (€92,821) in the non-severe group for HA and €289,411 (€132,400) versus €19,253 (€23,655) for HB.

Conclusion

The results demonstrate the high economic burden of severe HA and HB in Germany, driven by the need for continuous factor replacement therapy, and give an estimate of treatment costs based on a real-world therapy mix.


Introduction

Hemophilia is a rare, X-recessive, congenital bleeding disorder characterized by a deficient endogenous production of blood clotting factor VIII (FVIII) in hemophilia A (HA) and factor IX (FIX) in hemophilia B (HB). The disease occurs predominantly in males, while females may exhibit symptoms of mostly mild hemophilia. In Germany, the incidence is approximately 1 per 5,000 for male births for HA and 1 per 25,000 male births for HB,[1] and the prevalence is estimated at 10.4 and 1.77 per 100,000 males for HA and HB, respectively.[2] [3] The life expectancy of patients with hemophilia has markedly increased and now approaches that of the general population in high-income countries due to improved management and treatment.[4] [5]

The standard of care in Germany for patients with hemophilia with a severe bleeding phenotype is a lifelong prophylactic treatment with intravenous concentrated FVIII or FIX to prevent bleeding events; prophylaxis is also recommended for patients with moderate hemophilia experiencing regular bleeds.[6] In addition, newer treatment options are available with the aim to alleviate the limitations of conventional factor replacement therapy, notably the need for frequent intravenous injections and inhibitor development, through different mechanisms: the extension of the circulating half-life of recombinant factors, non-factor treatments such as emicizumab, and the induction of the production of endogenous clotting factors with gene therapy.[7]

Due to the required lifetime treatment and management of adverse effects, hemophilia is a very costly disease. Costs are almost entirely attributed to direct medical costs for factor replacement,[8] [9] [10] [11] [12] [13] with indirect costs found neglectable despite the significant burden on patients and family.[14]

To assess the economic value of emerging treatment options, studies describing the economic burden of hemophilia under the current standard of care in Germany are needed. Current data on epidemiology, health care resource utilization (HCRU), and costs in the German hemophilia population are limited and challenging to obtain due to the rarity of the condition. Here, we present a structured approach classifying hemophilia patients by treatment regimen in statutory health insurances (SHIs) administrative claims data to examine real-world patient characteristics, treatment patterns, HCRU, and costs among treated patients with HA and HB in Germany from 2016 to 2021.


Methods

This retrospective study used anonymized SHI claims data from January 1, 2015, to December 31, 2021, from the InGef research database, which includes longitudinal claims data of approximately 4.8 million subjects from approximately 50 SHIs (mainly company or guild health insurances) in Germany. The database is representative of the total German population with regard to age and sex.[15] All analyses were conducted in a cross-sectional design where patients were examined per calendar year.

Patient Selection

Patients were required to have continuous insurance coverage during the respective analysis year (until December 31 or date of death) and the baseline period (i.e., the year prior to the analysis year).

Patients had to meet one the following diagnosis criteria for inclusion from January 1, 2016, to December 31, 2021, one or more main or secondary inpatient diagnosis of congenital hemophilia (ICD-10; HA: D66; HB: D67); OR ≥2 verified outpatient HA/HB diagnoses in two quarters of the respective study year; OR ≥1 verified outpatient HA/HB diagnosis, and one or more prescription of a medication for HA/HB treatment (see [Supplementary Tables S1]–[S4] [available in the online version only]).

German ICD-10 does not contain codes for hemophilia severity grades. To allow the classification of the patients by weight-dependent factor consumption regimens, the overall study population was further restricted to all patients aged 12 years or older.


Classification by Factor Consumption

Factor prescriptions in each study year were used to classify each patient's factor regimen type as either indicative of a severe phenotype predominantly requiring prophylactic treatment or a non-severe phenotype predominantly treated on demand, in adaptation from.[16] Patient-specific yearly factor consumption was determined by summing the total factor quantity (International Units; IU) from each factor prescription observed during the 12-month period. The calculated yearly factor consumption was compared with a prespecified classification threshold. The classification thresholds were determined based on the low end of continuous prophylaxis dose range with standard half-life factor products (SHL; 20–40 IU/kg), the recommended dosing frequency (every 2–3 days for HA and every 3–4 days for HB),[6] [17] and an estimated low end of body weight distribution in a 12+ male population of 43 kg.[18] This resulted in a threshold of 100,000 IU for HA and 80,000 IU for HB. For HB, recognizing widespread use of extended half-life factor products (EHL), the threshold was halved to 40,000 IU if the patient had received at least one prescription of EHL in the respective study year, to adjust for reduced dosing frequency from twice weekly to once every 7 to 10 days for FIX.[17] If the calculated yearly factor consumption exceeded the threshold, the patient was designated as “severe.”

Patients with a yearly factor consumption that is less than or equal to the threshold were classified as “non-severe”; this includes patients with no factor prescription in the study year but one or more prescriptions of desmopressin (HA only) or one or more prescriptions of tranexamic acid.

Patients with current evidence of inhibitors (one or more prescription of a bypassing agent [HA only] or one or more prescription of emicizumab in combination with an ICD-10-GM diagnosis D68.31 [hemorrhagic disorder due to factor VIII antibodies]) were excluded from classification by factor consumption and were not included in the analyses of severe or non-severe patients.

The three treatment groups, “severe,” “non-severe,” and “inhibitors,” combined formed the overall study population, consisting of all treated HA/HB patients with one or more prescriptions of a medication for HA/HB treatment (see [Supplementary Tables S1]–[S4] [available in the online version only]).


Health Care Resource Utilization and Costs

HCRU and cost outcomes were captured for each study year. HCRU outcomes were the frequency of all-cause hospitalizations, physician visits, and hemophilia-related medication. Yearly factor consumption was determined by summing the total factor quantity (IU) from each factor prescription observed during the 12-month period. The number of inpatient days per patient was calculated by dividing the sum of inpatient days by the number of patients in the respective study year.

Health care cost outcomes included all-cause inpatient and outpatient costs, costs for remedies and aids, and medication costs. To determine medication costs, the net costs from an SHI perspective (the net amount of the effective expenditure of the SHI = pharmacy sales price minus discounts; all inclusive of value added tax [VAT]) were extracted from the claims data on the day the drug was issued.


Statistical Analysis

Summary statistics for patient characteristics, HCRU, and cost outcomes included means and standard deviations (SDs), median, interquartile range (IQR) for continuous variables, and counts and percentages for categorical variables. Statistical comparisons between severe and non-severe subgroups (medication costs, number of inpatient days) were conducted using t-tests. Patient numbers were projected to the total German population using official population statistics from the German Federal Bureau of Statistics.[19]

Cell sizes of less than five were suppressed to protect study subjects from potential re-identification and privacy considerations; where an analysis could not be made due to a cell being suppressed, this was indicated (non-applicable; NA).



Results

Hemophilia A

Patient Characteristics

Overall, 125 to 151 treated HA patients were identified in the database based on the inclusion and exclusion criteria in each study year from 2016 to 2021 ([Table 1]). The mean age in our HA study population ranged from 34 to 37 years. Over 80% of our study population were adults, with 15.6 to 19.8% adolescents aged 12 to 17 years ([Table 2]). A total of 83.2 to 87.0% of the study population was male. The mean age of the severe group was slightly lower (30–32 years), with up to 30.2% adolescents and over 98% males.

Table 1

Patient flow in InGef Database—hemophilia A

Selection step 

2016 

2017 

2018 

2019 

2020 

2021 

1) Persons in the InGef database 

4,171,377

4,201,273

4,278,020

4,284,818

4,282,793

4,321,057

2) Persons fully insured in respective baseline and analysis years 

3,990,293

4,000,856

3,946,476

3,911,267

3,896,796

3,721,474

3) Diagnosis M2Q D66

856

909

905

964

1,019

993

4) Age 12 yrs or older

784

836

833

901

950

924

5a) Treatment group “Severe”

55

58

62

66

71

63

5b) Treatment group “Non-Severe”

72

74

68

57

74

75

5c) Treatment group “Inhibitors”

<5

<5

<5

<5

6

<5

5) Overall treated study population

129

135

131

125

151

142

Abbreviations: M2Q, minimum two quarters; yrs, years.


Table 2

Patient characteristics of hemophilia A study population and treatment groups

2016

2017

2018

2019

2020

2021

N (percentage of overall)

Overall

129 (100)

135 (100)

131 (100)

125 (100)

151 (100)

142 (100)

Severe

55 (42.6)

58 (43.0)

62 (47.3)

66 (52.8)

71 (47.0)

63 (44.4)

Non-Severe

72 (55.8)

74 (54.8)

68 (51.9)

57 (45.6)

74 (49.0)

75 (52.8)

Inhibitors

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

6 (4.0)

<5 (NA)

Age in years; mean ± SD

Overall

35 ± 17

35 ± 17

37 ± 18

34 ± 18

36 ± 18

35 ± 18

Severe

32 ± 15

31 ± 13

31 ± 14

30 ± 14

31 ± 14

30 ± 16

Non-Severe

38 ± 19

38 ± 19

42 ± 20

39 ± 20

41 ± 20

38 ± 19

Inhibitors

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

37 ± 17

<5 (NA)

Male, n (percentage of treatment group)

Overall

112 (86.8)

115 (85.2)

114 (87.0)

104 (83.2)

126 (83.4)

123 (86.6)

Severe

54 (98.2)

57 (98.3)

61 (98.4)

66 (100)

71 (100)

62 (98.4)

Non-Severe

57 (79.2)

56 (75.7)

52 (76.5)

37 (64.9)

52 (70.3)

57 (76.0)

Inhibitors

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

Adults, n (percentage of treatment group)

Overall

106 (82.2)

114 (84.4)

105 (80.2)

101 (80.8)

125 (82.8)

119 (83.8)

Severe

43 (78.2)

46 (79.3)

47 (75.8)

49 (74.2)

54 (76.1)

44 (69.8)

Non-Severe

61 (84.7)

64 (68.5)

55 (80.9)

48 (84.2)

63 (85.1)

69 (92.0)

Inhibitors

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

Abbreviations: NA, not available; SD, standard deviation.



Prevalence

The projection of patient numbers in the database to the German population yielded estimates of 2,748 to 3,337 treated HA patients aged 12 years or older in Germany per year. Male prevalence ranged from 6.39 to 7.81 per 100,000 ([Fig. 1A]). Forty-three percent to 53% of patients were in the severe treatment group; less than 5% of patients had current evidence of inhibitors.

Zoom
Fig. 1 Extrapolated number of cases and prevalence of hemophilia in the German population per 100,000 males. (A) Patients with hemophilia A; (B) patients with hemophilia B.

Health Care Resource Utilization

The proportion of HA patients hospitalized for any cause per year was stable throughout the study period from 2016 to 2021 at approximately 60% ([Table 3]). Throughout the study period, the severe group spent more days in hospital than the non-severe group with differences in means reaching statistical significance in most study years ([Fig. 2A]).

Zoom
Fig. 2 Number of inpatient days per patient by subgroup. (A) Patients with hemophilia A; (B) patients with hemophilia B. Blue dots: mean; Red dots: maximum; Boxes: IQR; fences: ±1.5 IQR. ** for p < 0.01. IQR, interquartile range; n.s., not statistically significant.
Table 3

Health care resource utilization—hemophilia A

Type of HCRU

2016

2017

2018

2019

2020

2021

Overall

All-cause hospitalizations, n (%)

80 (62.0)

135 (63.7)

131 (62.6)

125 (68)

151 (66.2)

142 (65.5)

All-cause outpatient visits, n (%)

129 (100)

135 (100)

131 (100)

124 (99.2)

150 (99.3)

142 (100)

Hemophilia-related medication, n (%)

Factor replacement therapy: Any

102 (79.1)

107 (79.3)

102 (77.9)

97 (77.6)

117 (77.5)

110 (77.5)

● pdSHL

23 (17.8)

26 (19.3)

28 (21.4)

24 (19.2)

30 (19.9)

25 (17.6)

● recSHL

78 (60.5)

79 (58.5)

71 (54.2)

56 (44.8)

61 (40.4)

51 (35.9)

● Any SHL

101 (78.3)

103 (76.3)

98 (74.8)

80 (64.0)

90 (59.6)

76 (53.5)

● recEHL

<5 (NA)

<5 (NA)

10 (7.6)

26 (20.8)

28 (18.5)

34 (23.9)

Emicizumab

0 (0)

0 (0)

<5 (NA)

<5 (NA)

8 (5.3)

12 (8.5)

Desmopressin

13 (10.1)

12 (8.9)

13 (9.9)

12 (9.6)

9 (6.0)

6 (4.2)

Tranexamic acid

26 (20.2)

33 (24.4)

29 (22.1)

34 (27.2)

44 (29.1)

37 (26.1)

Severe

All-cause hospitalizations, n (%)

34 (61.8)

36 (62.1)

40 (64.5)

47 (71.2)

44 (62.0)

41 (65.1)

All-cause outpatient visits, n (%)

55 (100)

58 (100)

62 (100)

65 (98.5)

71 (100)

63 (100)

Hemophilia-related medication, n (%)

Factor replacement therapy: Any

55 (100)

58 (100)

62 (100)

66 (100)

69 (97.2)

59 (93)

● pdSHL

9 (16.4)

11 (19.0)

11 (17.7)

14 (21.2)

16 (22.5)

13 (20.6)

● recSHL

46 (83.6)

47 (81.0)

48 (77.4)

38 (57.6)

36 (50.7)

27 (42.9)

● Any SHL

55 (100)

56 (96.6)

58 (93.6)

52 (78.8)

51 (71.8)

40 (63.5)

● recEHL

<5 (NA)

<5 (NA)

10 (16.1)

22 (33.3)

23 (32.4)

22 (34.9)

Emicizumab

0 (0)

0 (0)

0 (0)

<5 (NA)

6 (8.5)

11 (17.5)

Desmopressin

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

Tranexamic acid

<5 (NA)

<5 (NA)

<5 (NA)

8 (12.1)

7 (9.9)

<5 (NA)

Non-Severe

All-cause hospitalizations, n (%)

45 (62.5)

48 (64.9)

41 (60.3)

36 (63.2)

50 (67.6)

49 (65.3)

All-cause outpatient visits, n (%)

72 (100)

74 (100)

68 (100)

57 (100)

73 (98.7)

75 (100)

Hemophilia-related medication, n (%)

Factor replacement therapy: Any

46 (63.9)

46 (62.2)

39 (57.4)

30 (52.6)

44 (59.5)

47 (62.7)

● pdSHL

14 (19.4)

14 (18.9)

16 (23.5)

10 (17.5)

14 (18.9)

12 (16)

● recSHL

32 (44.4)

32 (43.2)

23 (33.8)

18 (31.6)

25 (33.8)

24 (32.0)

● Any SHL

46 (63.9)

46 (62.2)

39 (57.4)

28 (49.1)

39 (52.7)

36 (48.0)

● recEHL

0 (0)

0 (0)

0 (0)

<5 (NA)

5 (6.8)

12 (16.0)

Emicizumab

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

Desmopressin

13 (18.1)

12 (16.2)

13 (19.1)

12 (21.1)

9 (12.2)

6 (8.0)

Tranexamic acid

22 (30.6)

28 (37.8)

26 (38.2)

25 (43.9)

34 (46.0)

33 (44.0)

Abbreviations: HCRU, health care resource utilization; NA, not available; pdSHL, plasma-derived standard half-life factor; recEHL, recombinant extended half-life factor; recSHL, recombinant standard half-life factor.


Plasma-derived SHL (pdSHL) use remained stable, with around 20% of all HA patients treated with pdSHL, both in the severe and non-severe treatment groups ([Table 3]). Recombinant SHL (recSHL) use declined from 60.5% in 2016 to 35.9% in 2021 and was partly replaced by recEHL. Less than 40% of severe HA patients were treated with EHL in 2021, and less than 20% were treated in the non-severe treatment group. Non-factor replacement was available and used from 2018 onward, with 17.5% of severe patients treated with emizicumab in 2021 ([Table 3]).

Mean per-patient per year (PPPY) factor consumption in the non-severe treatment group ranged from 19,000 IU to 32,000 IU (SHL) and 17,000 IU to 43,000 IU (EHL), indicative of a patient group with on-demand therapy ([Table 4]). In the severe treatment group, SHL consumption declined from 325,000 IU in 2016 to 282,000 IU in 2021, and an increase in mean PPPY EHL consumption from 195,000 IU in 2018 to 372,000 IU in 2021.

Table 4

Yearly factor consumption in 1,000 IU per hemophilia A patient by product category and treatment group

2016

2017

2018

2019

2020

2021

Overall

SHL

n = 101

n = 103

n = 98

n = 80

n = 90

n = 76

192 ± 211

208 ± 238

198 ± 213

209 ± 200

166 ± 195

157 ± 195

120 (30–290)

120 (20–300)

120 (18–320)

180 (32–323)

101 (12–297)

75 (10–245)

2; 870

0.5; 1,272

1; 960

0.5; 860

0.75; 964

0.5; 724

EHL

<5 (NA)

<5 (NA)

n = 10

n = 26

n = 28

n = 34

195 ± 92

221 ± 182

292 ± 193

246 ± 244

208 (133–250)

195 (62–308)

300 (125–360)

180 (24–360)

40; 320

4; 700

6; 793

2; 840

Severe

SHL

n = 55

n = 56

n = 58

n = 52

n = 51

n = 40

325 ± 206

348 ± 233

315 ± 202

307 ± 182

277 ± 196

282 ± 198

280 (180–420)

300 (182–407)

260 (175–360)

273 (180–364)

238 (131–964)

221 (163–370)

30; 870

110; 1,272

30; 960

20; 860

10; 964

4; 724

EHL

<5 (NA)

<5 (NA)

n = 10

n = 22

n = 23

n = 22

195 ± 92

259 ± 173

346 ± 168

372 ± 216

208 (133–250)

223 (152–314)

333 (280–390)

341 (190–553)

40; 320

30; 700

95; 390

60; 840

Non-Severe

SHL

n = 46

n = 46

n = 39

n = 28

n = 39

n = 36

32 ± 28

29 ± 30

20 ± 21

26 ± 29

21 ± 25

19 ± 24

26 (6–60)

20 (5–38)

12 (6–24)

12 (5–43)

10 (6–26)

11 (5–20)

2; 83

0.5; 100

1; 100

0.5; 100

0.75; 100

0.5; 90

EHL

n = 0

n = 0

n = 0

<5 (NA)

n = 5

n = 12

0

0

0

43 ± 32

17 ± 18

0

0

0

30 (30–60)

8 (4–24)

0

0

0

6; 90

2; 54

Abbreviations: EHL, extended half-life factor products; IQR, interquartile range; Max, maximum; Min, minimum; NA, not available; SD, standard deviation; SHL, standard half-life factor products.


n, number of patients with prescriptions; mean ± SD; median (IQR); Min, Max.



Costs

Mean PPPY all-cause treatment costs for the HA cohort remained stable over the study period. They were consistently and statistically significantly higher in the severe than in the non-severe group amounting to over €300,000 in the severe group, and below €50,000 in the non-severe group ([Fig. 3A]). Outpatient costs, aids and remedies' costs, and inpatient costs are comparable between severe and non-severe groups. Differences in treatment costs are driven by the medication costs, constituting over 97% of total costs in the severe group and consistently and statistically significantly higher in the severe than in the non-severe group. In 2021, mean (SD) PPPY medication costs were €321,987 (€157,915) versus €43,487 (€92,821). [Figure 4A] shows a shift in the distribution of medication costs over the study period, with increasing numbers of patients with high mean PPPY medication costs (>€100,000) in the non-severe group in later study years.

Zoom
Fig. 3 Mean per-patient health care costs per treatment group and cost category (in €). (A) Patients with hemophilia A; (B) patients with hemophilia B. *** for p < 0.001.
Zoom
Fig. 4 Distribution of Mean per-patient medication costs, per treatment group. (A) Patients with hemophilia A; (B) patients with hemophilia B. Blue dots: mean; Red dots: maximum; Boxes: IQR; Fences: ±1.5 IQR. *** for p < 0.001. IQR, interquartile range.


Hemophilia B

Patient Characteristics

Overall, 26 to 35 treated HB patients were identified in the database based on the inclusion and exclusion criteria in each study year from 2016 to 2021 ([Table 5]). The mean age in our HB study population ranged from 36 to 40 years. Over 80% of our study population were adults, with 16.1 to 20.0% adolescents aged 12 to 17 years ([Table 6]). A total of 80.7 to 90.0% of the study population was male. The mean age of the severe group was lower (27–31 years), with up to 41.7% adolescents and over 97% males on average.

Table 5

Patient flow in InGef database—hemophilia B

Selection step 

2016 

2017 

2018 

2019 

2020 

2021 

1) Persons in the Ingef database 

4,171,377 

4,201,273 

4,278,020 

4,284,818 

4,282,793 

4,321,057 

2) Persons fully insured in respective baseline and analysis years 

3,990,293 

4,000,856 

3,946,476 

3,911,267 

3,896,796 

3,721,474 

3) Diagnosis M2Q D67

149

151

153

150

149

141

4) Age 12 yrs or older

134

134

137

138

132

128

5a) Treatment group “Severe”

12

12

13

12

17

19

5b) Treatment group “Non-Severe”

18

15

13

17

13

15

5c) Treatment group “Inhibitors”

0

0

0

0

<5

<5

5) Overall treated study population

30

27

26

29

31

35

Abbreviations: M2Q, minimum two quarters; yrs, years.


Table 6

Patient characteristics of hemophilia B study population and treatment groups

2016

2017

2018

2019

2020

2021

N (percentage of overall)

Overall

30 (100)

27 (100)

26 (100)

29 (100)

31 (100)

35 (100)

Severe

12 (40)

12 (44.4)

13 (50)

12 (41.3)

17 (54.8)

19 (54.3)

Non-Severe

18 (60)

15 (55.6)

13 (50)

17 (58.6)

13 (41.0)

15 (42.9)

Inhibitors

0 (0)

0 (0)

0 (0)

0 (0)

<5 (NA)

<5 (NA)

Age in years, mean ± SD

Overall

37 ± 22

37 ± 19

37 ± 19

36 ± 18

40 ± 20

37 ± 17

Severe

27 ± 18

29 ± 12

27 ± 12

28 ± 13

31 ± 16

27 ± 12

Non-Severe

44 ± 22

43 ± 22

46 ± 20

42 ± 19

53 ± 18

49 ± 18

Inhibitors

NA

NA

NA

NA

NA

NA

Male, n (percentage of treatment group)

Overall

27 (90.0)

23 (85.2)

21 (80.8)

24 (82.8)

25 (80.7)

30 (85.7)

Severe

12 (100)

11 (91.7)

13 (100)

12 (100)

15 (88.2)

19 (100)

Non-Severe

15 (83.3)

12 (80.0)

8 (61.5)

12 (70.6)

9 (69.2)

10 (66.7)

Inhibitors

0 (0)

0 (0)

0 (0)

0 (0)

<5 (NA)

<5 (NA)

Adults, n (percentage of treatment group)

Overall

24 (80.0)

22 (81.4)

21 (80.8)

24 (82.8)

26 (83.9)

31 (88.6)

Severe

7 (58.3)

10 (83.3)

9 (69.2)

9 (75)

13 (76.5)

15 (78.9)

Non-Severe

17 (94.4)

12 (80)

12 (92.3)

15 (88.2)

12 (92.3)

15 (100)

Inhibitors

0 (0)

0 (0)

0 (0)

0 (0)

<5 (NA)

<5 (NA)

Abbreviations: NA, not available, SD, standard deviation.



Prevalence

The projection of patient numbers in the database to the German population yielded estimates of 566 to 802 treated HB patients aged 12 years or older in Germany per year. Male prevalence ranged from 1.26 to 1.89 per 100,000 ([Fig. 1B]). Forty percent to 56% of patients were in the severe treatment group; less than 5% of patients had current evidence of inhibitors in 2020 and 2021.


Health Care Resource Utilization

The proportion of HB patients hospitalized for any cause per year was stable throughout the study period from 2016 to 2021 at approximately 75% ([Table 7]). Over the study period, patients in the severe treatment group stayed more days in hospital on average than the non-severe group but without reaching statistical significance in the differences. The number of inpatient days per patient declined in the severe treatment group, resulting in comparable mean, median, and IQR in both groups in 2020 and 2021 ([Fig. 2B]).

Table 7

Health care resource utilization—hemophilia B

Type of HCRU

2016

2017

2018

2019

2020

2021

Overall

All-cause hospitalizations, n (%)

23 (76.7)

22 (81.5)

19 (73.1)

22 (75.9)

21 (67.7)

26 (74.3)

All-cause outpatient visits, n (%)

30 (100)

27 (100)

26 (100)

29 (100)

31 (100)

35 (100)

Hemophilia-related medication, n (%)

Factor replacement therapy: Any

26 (86.7)

24 (88.9)

24 (92.3)

25 (86.2)

24 (77.4)

30 (85.7)

● pdSHL

13 (43.3)

10 (37.0)

7 (26.9)

5 (17.2)

<5 (NA)

5 (14.3)

● recSHL

12 (40.0)

8 (29.6)

8 (30.8)

8 (27.6)

<5 (NA)

5 (14.3)

● Any SHL

24 (80.0)

18 (66.7)

15 (57.7)

13 (44.8)

8 (25.8)

10 (28.6)

● recEHL

5 (16.7)

7 (25.9)

12 (46.2)

15 (51.7)

17 (54.8)

21 (60.0)

Tranexamic acid

8 (26.7)

6 (22.2)

<5 (NA)

8 (27.6)

11 (35.5)

10 (28.6)

Severe

All-cause hospitalizations, n (%)

10 (83.3)

11 (91.7)

11 (84.6)

9 (75.0)

12 (70.6)

16 (84.2)

All-cause outpatient visits, n (%)

12 (100)

12 (100)

13 (100)

12 (100)

17 (100)

19 (100)

Hemophilia-related medication, n (%)

Factor replacement therapy: Any

12 (100)

12 (100)

13 (100)

12 (100)

17 (100)

19 (100)

● pdSHL

6 (50.0)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

● recSHL

6 (50.0)

6 (50.0)

<5 (NA)

<5 (NA)

<5 (NA)

0 (0.0)

● Any SHL

11 (91.7)

9 (75.0)

5 (38.5)

<5 (NA)

<5 (NA)

<5 (NA)

● recEHL

<5 (NA)

<5 (NA)

11 (84.6)

12 (100.0)

14 (82.4)

18 (94.7)

Tranexamic acid

<5 (NA)

0 (0)

0 (0)

<5 (NA)

<5 (NA)

<5 (NA)

Non-Severe

All-cause hospitalizations, n (%)

13 (72.2)

11 (73.3)

8 (61.5)

13 (76.5)

8 (61.5)

9 (60.0)

All-cause outpatient visits, n (%)

18 (100)

15 (100)

13 (100)

17 (100)

13 (100)

15 (100)

Hemophilia-related medication, n (%)

Factor replacement therapy: Any

14 (77.8)

12 (80)

11 (84.6)

13 (76.5)

7 (53.9)

11 (73.3)

● pdSHL

7 (38.9)

7 (46.7)

5 (38.5)

<5 (NA)

<5 (NA)

<5 (NA)

● recSHL

6 (33.3)

<5 (NA)

5 (38.5)

6 (35.3)

<5 (NA)

5 (33.3)

● Any SHL

13 (72.2)

9 (60.0)

10 (76.9)

10 (58.8)

<5 (NA)

8 (53.3)

● recEHL

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

Tranexamic acid

6 (33.3)

6 (40.0)

<5 (NA)

7 (41.2)

7 (53.9)

9 (60.0)

Abbreviations: HCRU, health care resource utilization; n, number of patients with prescriptions/visit; NA, not available; pdSHL, plasma-derived standard half-life factor; recEHL, recombinant extended half-life factor; recSHL, recombinant standard half-life factor.


Overall, EHL use increased from 16.7% in 2016 to 60.0%; SHL use decreased from 80.0% in 2016 to 28.6% in 2021 ([Table 7]). The switch to EHL factor products was particularly prominent in the severe treatment group: In 2016, 91.7% of the patients in this group were treated with SHL; in 2021, 94.7% received EHL products. Mean PPPY SHL consumption in the severe group was 150,000 IU and 200,000 IU in 2016 and 2017 and dropped to 90,000 IU in 2018 ([Table 8]). This decrease coincides with the increasing uptake of EHL products and is therefore likely due to mid-year treatment switches. The mean PPPY EHL consumption ranged from 126,000 IU to 165,000 IU, a dosage indicative of a patient group on prophylactic treatment.

Table 8

Yearly factor consumption in 1,000 IU per hemophilia B patient by product category and treatment group

2016

2017

2018

2019

2020

2021

Overall

SHL

n = 24

n = 18

n = 15

n = 13

n = 8

n = 10

86 ± 88

116 ± 98

50 ± 73

33 ± 32

111 ± 181

42 ± 63

65 (29–110)

85 (37–205)

29 (5–62)

18 (6–60)

49 (13–95)

18 (12–28)

2; 360

3; 273

1; 291

2; 100

10, 548

4; 210

EHL

n = 5

n = 7

n = 12

n = 15

n = 17

n = 21

37 ± 25

66 ± 56

124 ± 79

105 ± 58

137 ± 96

120 ± 72

30 (30–60)

55 (40–72)

105 (83–176)

100 (92–135)

115 (76–240)

100 (89–180)

2; 62

2; 180

8; 265

12; 248

2; 300

3; 300

Severe

SHL

n = 11

n = 9

n = 5

<5 (NA)

<5 (NA)

<5 (NA)

150 ± 94

200 ± 61

90 ± 117

110 (86–186)

208 (170–240)

40 (24–90)

60; 360

90; 273

5; 291

EHL

<5 (NA)

<5 (NA)

n = 11

n = 12

n = 14

n = 18

135 ± 73

126 ± 45

165 ± 82

137 ± 63

110 (90–178)

105 (96–140)

132 (101–246)

114 (90–180)

42; 265

90; 248

70; 300

83; 300

Non-Severe

SHL

n = 13

n = 9

n = 10

n = 10

<5 (NA)

n = 8

31 ± 27

33 ± 28

30 ± 29

28 ± 28

25 ± 24

30 (6–52)

36 (5–48)

22 (5–55)

16 (7–51)

18 (12–27)

2; 80

3; 80

1; 80

2; 73

4; 80

EHL

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

<5 (NA)

Abbreviations: EHL, extended half-life factor products; IQR, interquartile range; Max, maximum; Min, minimum; NA, not available; SD, standard deviation; SHL, standard half-life factor products.


n, number of patients with prescriptions; mean ± SD; median (IQR); Min, Max.


In the non-severe group, a slight decline in SHL use can be observed, with more than half of the patients still treated with SHL in 2021. Fewer than five patients in this group received EHL. The mean PPPY SHL consumption remained stable around 30,000 IU, indicative of a patient group on on-demand therapy.


Costs

Mean PPPY all-cause health care costs for the HB-treated cohort increased between 2016 and 2018 and stabilized at around €300,000 as of 2018 in the severe group, whereas costs stayed below €55,000 throughout the study period in the non-severe group ([Fig. 3B]).

Outpatient costs, aids and remedies' costs, and inpatient costs are comparable between severe and non-severe groups. Differences in treatment costs are driven by the medication costs, constituting 87 to 99% of total costs in the severe group and consistently and statistically significantly higher in the severe than in the non-severe group. In 2021, mean (SD) PPPY medication costs were €289,411 (€132,400) versus €19,253 (€23,655). Medication costs stayed below €100,000 for all patients in the non-severe group in all study years ([Fig. 4B]).



Limitations

Although the analysis dataset obtained from the InGef database covers approximately 4.8 million insured members of SHIs all over Germany, representativeness for the entire population can only be guaranteed with regard to age and sex.

Overall, sample sizes are small, particularly for HB, limiting several analyses due to data protection considerations and increasing the imprecision of our study results. The classification threshold of patients by yearly factor consumption was based on the assumed low end of average body weight across the study population without individual threshold adjustment based on the patient's age and sex.

The classification threshold for HA patients did not include an adjustment for reduced dosing frequency in patients using EHL, potentially leading to misclassification of severe patients in the non-severe group and an overestimation of HCRU and costs in non-severe HA in the later study years.

Diagnoses codes D66 and D67 are used, potentially incorrectly, in high numbers of patients without hemophilia-specific medication ([Table 1]). German ICD-10 does not code hemophilia severity grades.

Inhibitor patients who had achieved partial tolerance and used high factor doses without bypassing agents could potentially have been misclassified in the severe group.

This claims-database approach remains methodologically challenging. Robust registry data are an important source to address and confirm our results.

Analyses on HCRU and costs were done based on all-cause outcomes and, therefore, include non-hemophilia-related burdens. Reasons for hospitalization were not analyzed. Costs were not inflation-adjusted.



Discussion

HB male prevalence estimates of 1.26 to 1.89 were comparable to the prevalence reported by Stonebraker et al. of 1.77 per 100,000 males in Germany.[2] The projected total numbers in 2020 and 2021 of 684 and 802 HB patients aged 12+ were in-line with 2020/2021 German Hemophilia Registry (DHR) reports of 860 and 926 patients (all ages), and 634 and 659 adults. For HA, male prevalence estimates of 6.39 to 7.81 were lower than the previous prevalence reported of 10.4 per 100,000 males in Germany.[3] Projected total numbers of HA patients aged 12+ in 2020 and 2021 of 3,337 and 3,272 were also slightly lower than numbers reported by DHR of 4,518 and 4,873 patients (all ages), and 3,335 and 3,567 adults.[1] [20]

The mean age of our HA and HB cohorts was in-line with previous reports.[9] [13] [21] The small number of patients with current evidence of inhibitors was as expected in an HA/HB population. Female patients were predominantly found in the non-severe treatment groups, which is compatible with the milder phenotype observed in female HA/HB carriers.[17] [21] Our range of HA patients in the severe treatment group (42.8–53.3%) and the range for the HB severe treatment group (40–54.8%) are broadly in-line with severity proportions reported in DHR.[1] [20] It should be noted that our classification by intensity of treatment might also classify patients with moderate hemophilia into the severe treatment group. Patients who did not receive any hemophilia-specific medication in the study year were excluded from our analysis, whereas DHR includes approximately 30% of patients that are not factor-treated and approximately 14% subclinical patients.

Our data show that EHL are now the mainstay treatment in severe HB patients, with over 95% treated with EHL, whereas pdSHL and recSHL remain in use for prophylactic treatment in a high percentage of HA patients. Widespread uptake of EHL in severe HB patients over the study period coincides with a marked decline in days spent in hospital in this group. No similar decline could be observed in HA patients as would be expected with a pandemic effect. The decline in hospital stays could therefore be indicative of an improved disease management achieved over the course of this time period.

In 2021, HA patients in the severe treatment group incurred €321,987 (€157,915) in mean (SD) PPPY medication costs, while HB patients incurred €289,411 (€132,400). O'Hara et al. (2017) estimated per-patient costs for clotting factor replacement therapy at €312,157 in 2014 based on questionnaires among 194 adult German patients with severe HA and HB.[12] The cross-sectional CHESS II study reported mean PPPY FIX treatment costs of €393,263 for adult, severe HB patients on prophylaxis from Germany (n = 10).[9] This higher estimate is based on only adult, male, and severe patients, while our severe subgroup also included adolescents, females, and potentially moderate HB patients. A recent German SHI claims data analysis found lower total medication costs incurred per severe HA patient with €255,000 and per severe HB patient with €242,000 in 2021[22]; this sample, however, also included on-demand patients with fewer than 183 Defined Daily Dose (DDD) of factor products/year.


Conclusion

This analysis provides a comprehensive overview of real-world treatment patterns, HCRU, and costs among persons with HA and B in Germany from 2016 to 2021, stratified by patients' factor regimen type (severe/non-severe type).

Despite a small sample size and the methodological challenge of classification by treatment type in claims data, results on epidemiology, patient characteristics, factor consumption, and costs overall and by treatment type are broadly in-line with literature and recent reports and demonstrate a solid and plausible differentiation between both treatment groups.

The results demonstrate the high economic burden in severe HA and B patients in Germany, driven by the need for continuous factor replacement therapy and give an estimate of treatment costs based on a real-world therapy mix.



Conflict of Interest

MK, SH, and AK are employees of Pfizer Pharma in Germany and may hold Pfizer stocks.

SR, and IE were employees of Pfizer Pharma in Germany at the time of study, but not at the time of publication.

DH is an employee of WIG2, which is a paid consultant of Pfizer Pharma GmbH for designing the study, conducting the analyses, and interpreting the results.

DO and D-PP are employees of InGef GmbH, which acted as a subcontractor and received funding from WIG2 GmbH for designing the study, conducting the analyses, and interpreting the results.

RK has received honoraria and/or funds for research from Bayer, Biotest, Biomarin, CSL Behring, Grifols, Kedrion, LFB, Novo Nordisk, Octapharma, Pfizer, Roche/Chugai, Sanofi, SOBI, and Takeda.

KB has no conflict of interest to declare.

Authors' Contributions

D.O., K.B., R.K., M.K., S.R., I.E., S.H., D.H., and AK designed the study. D.O. and D-P.P. conducted the data analysis. D.O., K.B., R.K., S.R., I.E., S.H., and A.K. interpreted the results. A.K. drafted the manuscript with critical review and revisions from all authors. All authors reviewed and approved the final manuscript.


Supplementary Material


Address for correspondence

Dr. Agnes Kisser
Pfizer Pharma GmbH
Friedrichstrasse 110, 10117 Berlin
Germany   

Publikationsverlauf

Eingereicht: 26. November 2024

Angenommen: 31. März 2025

Artikel online veröffentlicht:
14. Juli 2025

© 2025. Thieme. All rights reserved.

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany


Zoom
Fig. 1 Extrapolated number of cases and prevalence of hemophilia in the German population per 100,000 males. (A) Patients with hemophilia A; (B) patients with hemophilia B.
Zoom
Fig. 2 Number of inpatient days per patient by subgroup. (A) Patients with hemophilia A; (B) patients with hemophilia B. Blue dots: mean; Red dots: maximum; Boxes: IQR; fences: ±1.5 IQR. ** for p < 0.01. IQR, interquartile range; n.s., not statistically significant.
Zoom
Fig. 3 Mean per-patient health care costs per treatment group and cost category (in €). (A) Patients with hemophilia A; (B) patients with hemophilia B. *** for p < 0.001.
Zoom
Fig. 4 Distribution of Mean per-patient medication costs, per treatment group. (A) Patients with hemophilia A; (B) patients with hemophilia B. Blue dots: mean; Red dots: maximum; Boxes: IQR; Fences: ±1.5 IQR. *** for p < 0.001. IQR, interquartile range.