Summary
Congenital thrombotic thrombocytopenic purpura (cTTP) is an ultra rare haematological disorder. This study aimed to estimate the clinical burden, healthcare resource use (HCRU) and associated costs of cTTP in England using primary and secondary care data. A retrospective cohort study was undertaken using the Clinical Practice Research Datalink (Aurum) linked to Hospital Episode Statistics data. The study period was from 01 January 2000 to 31 December 2019. Patients with cTTP were identified using an algorithm and matched 1:2 to patients without thrombotic thrombocytopenic purpura (TTP) (non‐TTP cohort) to contextualise the findings. A total of 36 cTTP patients were identified. The mean follow‐up for cTTP patients was 4.4 and 5.7 years in primary and secondary care, respectively, and 4.9 and 5.8 years for the non‐TTP cohort. Of the 36 patients, 38.9% experienced an acute episode and 19.4% experienced an organ damage event during follow‐up. HCRU (including inpatient admission, outpatient appointments and primary care appointments) was higher in the cTTP cohort relative to the non‐TTP cohort. The average yearly cost of healthcare for a cTTP patient was £6155, relative to £858 in the non‐TTP cohort. This study demonstrates the high clinical and economic burden of cTTP and provides evidence to inform healthcare planning.
Keywords: Clinical Practice Research Datalink, congenital thrombotic thrombocytopenic purpura, health economic burden, healthcare resource use, real‐world evidence
INTRODUCTION
Thrombotic thrombocytopenic purpura (TTP) is a rare, life‐threatening haematological disease that may be genetically inherited (congenital TTP [cTTP]) or acquired (immune‐mediated TTP [iTTP]). cTTP is an ultra rare disease with an incidence of approximately one in a million, 1 accounting for <5% of all TTP cases. 2 , 3 , 4
cTTP is caused by an absence or severe deficiency of A Disintegrin And Metalloproteinase with ThromboSpondin type 1 motif, member 13 (ADAMTS13), an enzyme involved in the regulation of blood clotting. 2 cTTP can lead to very low platelet levels, destruction of red blood cells and organ damage due to inadequate blood supply. 2 Organ damage can include chronic kidney disease, cerebral infarction and cardiac hypofunction. 2 , 5 , 6 In a UK TTP registry study of 73 patients with cTTP presenting from 2003 to 2018, the most common symptoms and events reported in untreated patients at presentation or following diagnosis were recurrent headaches and non‐hemiplegic migraines (44%), stroke (19%), lethargy (18%), abdominal pain (14%) and transient ischaemic attack (8%). 7
cTTP also involves relapsing, acute episodes, which are considered medical emergencies requiring prompt diagnosis and management. 2 , 8 Acute TTP episodes are defined by clinical and biological criteria such as multivisceral ischaemic symptoms, microangiopathic haemolytic anaemia and severe thrombocytopenia. 2
The British Society for Haematology guidelines recommend consideration of plasma prophylactic treatment for all cTTP patients to replenish ADAMTS13. 8 This is associated with frequent hospital visits and high healthcare resource use (HCRU).
Due to its rarity and the challenges of capturing real‐world outcomes, data on cTTP remain limited. This study aimed to evaluate the clinical burden, HCRU and associated costs of treating patients with cTTP in England and assess how this clinical and economic burden differs from the general population. Linked primary and secondary care data for a defined cohort with cTTP and a matched cohort without TTP (non‐TTP cohort) were used to better understand the impact of cTTP on patients and the healthcare system in England.
METHODS
Study design
This study employed a retrospective, matched cohort design, drawing on healthcare data from primary and secondary sources across England. Data were obtained from the Clinical Practice Research Datalink (CPRD) Aurum (May 2022 release) linked to Hospital Episode Statistics (HES). CPRD collects data from a network of UK General Practitioner (GP) practices and is broadly representative of the UK population in terms of age, sex and ethnicity. 9
The study period was from 01 January 2000 to 31 December 2019 (Figure 1). Patients with TTP were identified by relevant diagnosis codes in their primary/secondary care records between 01 January 2010 and 31 December 2019 (the identification period). A patient's first TTP code within the identification period was their index date, with the same index date used for corresponding non‐TTP controls. In primary care, follow‐up was defined as index date to registration end date or death (whichever occurred first). Some patients had no follow‐up in primary care, that is, if the first TTP diagnosis was identified in secondary care and the patient was no longer actively registered at a participating CPRD practice. In secondary care, follow‐up was defined as index date to end of study or death (whichever occurred first).
FIGURE 1.
Study design overview. cTTP, congenital thrombotic thrombocytopenic purpura; ICD, International Classification of Diseases; iTTP, immune‐mediated thrombotic thrombocytopenic purpura; TTP, thrombotic thrombocytopenic purpura. [Colour figure can be viewed at wileyonlinelibrary.com]
This study was approved by the CPRD Research Data Governance process (protocol 22_001897). CPRD holds an overarching ethics approval for observational research using anonymised data.
Population
The population of interest was patients with cTTP. There were no codes available specifically to identify patients with cTTP and iTTP. Therefore, all patients with a TTP diagnosis code (Table S1) were included in the TTP cohort and an algorithm was developed to identify those with cTTP (based on relevant medcodes, International Classification of Diseases [ICD]‐10 codes or Office of Population Censuses and Surveys [OPCS] codes). Specifically, they must have met at least one of the following criteria to be included in the study:
A genetic disease/carrier.
Received an infusion of coagulation factor.
Received >1 plasma infusion.
Furthermore, patients were excluded from the cTTP cohort if they met any of the following criteria:
Received more plasma exchanges than plasma infusions.
Undergone a splenectomy.
Received ≥3 rituximab prescriptions.
The cTTP algorithm was based on UK treatment guidelines and the best available data, given that ADAMTS13 assays were not available and was validated by a clinical expert (WL, a UK Consultant Haematologist with experience diagnosing and treating cTTP).
Exact matching was used in a 1:2 ratio to match patients with cTTP to those without a TTP diagnosis, based on year of birth, gender and primary care practice, to reduce the likelihood that differences in outcomes are due to these factors. Patients with cTTP without two matches were excluded from the study.
Deduplication
In CPRD, a person can have two or more unique patient identifiers (IDs). This occurs when a patient leaves a GP practice and joins another participating GP practice (immediately or at a later date). Registration dates in CPRD are used to ensure that duplicated data are not included in analyses. However, this can reduce cohort size. To maximise the use of HES data, a deduplication algorithm was developed (see Supporting Information for further details).
Duplication of data can also occur when one practice is absorbed by another practice that also contributes (or goes on to contribute) to the CPRD Aurum database. CPRD recommend excluding practices listed in the CPRD Aurum Data Specification v3.1. 10 Therefore, data from practices identified by CPRD were excluded.
Outcomes
The clinical and demographic baseline characteristics were reported for the cTTP and matched non‐TTP cohorts based on the look‐back period for each individual. The non‐TTP cohort was used to contextualise the findings for patients with cTTP. All outcomes were assessed during follow‐up. The outcomes of interest were the occurrence of acute TTP episodes, organ damage and all‐cause HCRU and associated costs (overall and relating to an acute TTP episode). Further detail is provided in Table 1.
TABLE 1.
Outcomes.
| Outcomes | |
|---|---|
| Baseline characteristics |
|
| Acute TTP episodes |
|
| Organ damage |
|
| HCRU and associated costs |
|
| HCRU related to an acute TTP episode | Rate of HCRU within 14 days from the end of an acute TTP episode for:
|
Abbreviations: A&E, accident and emergency; BPL 8Y, Bio Products Laboratory Plasma‐derived factor VIII/von Willebrand factor; cTTP, congenital thrombotic thrombocytopenic purpura; HCRU, healthcare resource use; ICU, intensive care unit; TTP, thrombotic thrombocytopenic purpura.
Organ damage events were identified during hospital admissions using ICD‐10 codes (Table S2). An acute TTP episode was defined as a non‐elective, not day‐case hospital admission with one or more of the following:
A TTP diagnosis as the primary reason for admission.
A thrombocytopenia diagnosis as the primary reason for admission.
A thromboembolic event (diagnosis or relevant procedure).
Use of plasma exchange, plasma infusion or administration of coagulation factor, namely, intermediate purity plasma‐derived factor VIII/von Willebrand factor concentrates containing ADAMTS13 (Bio Products Laboratory Plasma‐Derived Factor VIII/von Willebrand Factor [BPL 8Y]).
For acute TTP episodes including only one admission, the episode length was defined as the duration of the admission. For acute TTP episodes made up of more than one admission, the episode duration was defined as time from first admission to final discharge.
Primary care costs were obtained from the Unit Costs of Health and Social Care 2023 manual. 11 All secondary care costs (inpatient, outpatient and accident and emergency department [A&E]) were obtained from the National Health Service (NHS) Payment Scheme 24/25 Pay Award Prices. 12 Cost estimates were based on complete cases only. The specific cost approaches are presented in Table S3.
Statistical analysis
This study was descriptive with no formal statistical comparison between groups. Continuous variables were summarised using mean, standard deviation (SD), median, quartile 1 (Q1), quartile 3 (Q3), minimum and maximum. Categorical variables were summarised using frequency counts and percentages. The rate of an event was calculated by dividing the number of observed events by the total number of person‐years. Based on Public Health England guidance, 13 Byar's method was used to calculate 95% confidence intervals (CIs) for all rates. 14 The average yearly cost of healthcare was estimated by multiplying HCRU rates by the median HCRU cost. Missing data, where appropriate, were considered a separate category and described using frequency counts and percentages.
For compliance with CPRD requirements, 15 any covariate resulting in cell counts <5 was excluded from the study to preserve patient anonymity.
Analyses were performed using R version 4.4.3.
RESULTS
Identification of subjects
A total of 1855 patients with TTP were identified within the identification period. After applying the inclusion/exclusion criteria, deduplication algorithm and ensuring sufficient matches, 1190 patients with TTP remained. Using the TTP subtype algorithm, 36 people (3.0%) were defined as having cTTP. These 36 patients with cTTP were matched with 72 non‐TTP controls (Figure 2).
FIGURE 2.
Flow diagram for study inclusion/exclusion criteria. APC, admitted patient care; CPRD, Clinical Practice Research Datalink; cTTP, congenital thrombotic thrombocytopenic purpura; GP, general practitioner; HES, Hospital Episode Statistics; iTTP, immune‐mediated thrombotic thrombocytopenic purpura; TTP, thrombotic thrombocytopenic purpura.
The mean follow‐up time for the cTTP cohort was 4.4 years (SD: 3.5) and 5.7 years (SD: 2.9) in primary and secondary care respectively. The mean follow‐up time for the non‐TTP cohort was 4.9 years (SD: 2.8) and 5.8 years (SD: 2.8) for primary and secondary care respectively.
Baseline characteristics
Baseline characteristics are presented in Table 2. Both the cTTP cohort and their matched non‐TTP cohorts had the same median age of 35.5 (Q1, Q3: 26.8, 45.2) years, and 86.1% were female. In the cTTP cohort, 52.8% had previously experienced prespecified organ damage at baseline, relative to 8.3% in the non‐TTP cohort. In the cTTP cohort, 24 patients (66.7%) had a previous TTP diagnosis and 25 patients (69.4%) had at least one previous TTP treatment recorded.
TABLE 2.
Baseline characteristics for the cTTP cohort and non‐TTP matched cohort.
| Baseline characteristic | cTTP, n = 36 | Non‐TTP (cTTP matched), n = 72 |
|---|---|---|
| Age, mean (SD) | 39.4 (17.2) | 39.4 (17.1) |
| Age, median (IQR) | 35.5 (18.5) | 35.5 (18.5) |
| Sex, n (%) | ||
| Male | 5 (13.9) | 10 (13.9) |
| Female | 31 (86.1) | 62 (86.1) |
| Ethnicity, n (%) | ||
| White | 30 (83.3) | 54 (75.0) |
| Other a | 6 (16.7) | 6 (8.3) |
| Unknown/missing | 0 (0.0) | 12 (16.7) |
| Organ damage | ||
| Previous organ damage, n (%) | 19 (52.8) | 6 (8.3) |
| Previous organ damage by type, n (%) | ||
| Cerebral | 8 (22.2) | NR |
| Cardiac | 5 (13.9) | 5 (6.9) |
| Liver | NR | NR |
| Renal | 9 (25.0) | NR |
| Number of organ damage types, mean (SD) | 0.6 (0.7) | 0.1 (0.5) |
| Number of organ damage events, mean (SD) | 2.1 (3.8) | 0.4 (1.8) |
| TTP treatments | ||
| Previous TTP diagnosis (based on diagnosis codes), n (%) | 24 (66.7) | NA |
| Previous TTP treatment, n (%) | 25 (69.4) | |
| Previous TTP treatment by type, n (%) | ||
| Plasma exchange | 7 (28.0) | |
| Plasma infusion | 22 (88.0) | |
| Immunosuppression | 0 (0.0) | |
| BPL 8Y | NR | |
Abbreviations: BPL 8Y, Bio Products Laboratory Plasma derived factor VIII/von Willebrand factor; cTTP, congenital thrombotic thrombocytopenic purpura; HCRU, healthcare resource use; IQR, interquartile range; NA, not applicable; NR, not reported; SD, standard deviation; TTP, thrombotic thrombocytopenic purpura.
Due to low counts for individual ethnic groups, the remaining groups have been combined. Outcomes are not reported when count <5.
Acute TTP episodes
In the cTTP cohort, 14/36 (38.9%) patients experienced an acute episode during the follow‐up period (up to 10 years). There were 27 acute TTP episodes observed, resulting in a rate of 0.13 per person‐year (95% CI: 0.09, 0.19). In 10 (37.0%) of the acute TTP episodes, patients were admitted from A&E. In 7 (25.9%) of the acute episodes, patients experienced an episode of care in the intensive care unit (ICU) during their admission. The mean number of acute TTP episodes was 1.43 (SD: 1.62) in those experiencing an organ damage event during follow‐up (n = 7), compared with 0.59 (SD: 1.21) in those without organ damage during follow‐up (n = 29).
A summary of acute TTP episodes in patients with cTTP and HCRU within 14 days is shown in Table 3. The mean duration of an acute TTP episode was 13 days (SD: 20). Within 14 days of an acute TTP episode, the median number of outpatient appointments, inpatient admissions and primary care appointments was 2 (Q1, Q3: 1, 3), 0 (Q1, Q3: 0, 1) and 0 (Q1, Q3: 0, 0), respectively. The median cost of treating an acute TTP episode was £13 872 (Q1, Q3: £13 872, £17 231).
TABLE 3.
HCRU following an acute TTP episode for patients with cTTP.
| Summary of acute TTP episodes | n | Min | Max | Mean (SD) | Median (Q1, Q3) |
|---|---|---|---|---|---|
| Length of acute TTP episode, days | 27 | 1 | 77 | 13 (20) | 6 (3, 12) |
| Time between acute TTP episodes for patients with two or more acute TTP episodes | 13 | 29 | 3059 | 1015 (1045) | 544 (166, 1910) |
| HCRU within 14 days of an acute TTP episode a | Events | Min | Max | Mean (SD) | Median (Q1, Q3) |
|---|---|---|---|---|---|
| Primary care appointments | 9 | 0.0 | 3 | 0.3 (0.8) | 0 (0, 0) |
| Inpatient admissions | 18 | 0.0 | 3 | 0.7 (0.9) | 0 (0, 1) |
| Outpatient appointments | 53 | 0.0 | 5 | 2.0 (1.3) | 2 (1, 3) |
| Cost of acute TTP episode, £ | 15 | 5995 | 31 103 | 15 846 (6726) | 13 872 (13 872, 17 231) |
Abbreviations: A&E, accident and emergency; cTTP, congenital thrombotic thrombocytopenic purpura; HCRU, healthcare resource use; Q1, quartile 1; Q3, quartile 3; SD, standard deviation; TTP, thrombotic thrombocytopenic purpura.
A&E attendance not reported due to low counts (frequencies for outcomes <5).
Organ damage during follow‐up
In the cTTP cohort, 7/36 (19.4%) patients experienced at least one organ damage event during the follow‐up period (up to 10 years). Most patients (29/36) did not experience an organ damage event. The maximum number of events experienced by a patient was 2. Of those who experienced an organ damage event, the median age was 62 years (Q1, Q3: 44.5, 67.5). In the cTTP cohort, the rate of organ damage events per person‐year was 0.05 (95% CI: 0.02–0.09). In those who experienced at least one organ damage event, the rate of organ damage events was 0.25 (95% CI: 0.12–0.47). Organ damage outcomes were not reported by organ damage subtype or for the non‐TTP cohort because fewer than five people had an event.
HCRU and associated costs
The rates of HCRU are reported in Table 4. The rate of primary care appointments was 1.42 times higher in the cTTP cohort relative to the matched non‐TTP cohort (5.76 [95% CI: 5.24, 6.31] vs. 4.06 [95% CI: 3.86, 4.28] per person‐year). The rate of outpatient appointments was 4.01 times higher in the cTTP cohort relative to the matched non‐TTP cohort (7.90 [95% CI: 7.52, 8.29] vs. 1.97 [95% CI: 1.83, 2.11] per person‐year). The rate of inpatient admissions was 26.76 times higher in the cTTP cohort relative to the matched non‐TTP cohort (8.83 [95% CI: 8.43, 9.25] vs. 0.33 [95% CI: 0.28, 0.39] per person‐year). The rate of A&E attendances was 1.87 times higher in the cTTP cohort relative to the matched non‐TTP cohort (0.58 [95% CI: 0.48, 0.69] vs. 0.31 [95% CI: 0.26, 0.37] per person‐year).
TABLE 4.
HCRU rates in primary and secondary care for the cTTP and matched non‐TTP cohorts.
| HCRU rates per person‐year | cTTP, n = 36 | Non‐TTP (cTTP matched), n = 72 | ||||
|---|---|---|---|---|---|---|
| Events | Total FU a | Rate (95% CI) | Events | Total FU a | Rate (95% CI) | |
| Primary care appointments | ||||||
| All | 455 | 79 | 5.76 (5.24–6.31) | 1437 | 354 | 4.06 (3.86–4.28) |
| GP | 386 | 79 | 4.89 (4.41–5.40) | 1222 | 354 | 3.45 (3.26–3.65) |
| Nurse | 69 | 79 | 0.87 (0.68–1.11) | 215 | 354 | 0.61 (0.53–0.69) |
| Inpatient admissions | ||||||
| All | 1801 | 204 | 8.83 (8.43–9.25) | 137 | 414 | 0.33 (0.28–0.39) |
| Outpatient appointments | ||||||
| All | 1611 | 204 | 7.90 (7.52–8.29) | 814 | 414 | 1.97 (1.83–2.11) |
| A&E attendances | ||||||
| All | 118 | 204 | 0.58 (0.48–0.69) | 129 | 414 | 0.31 (0.26–0.37) |
Abbreviations: A&E, accident and emergency; CI, confidence interval; cTTP, congenital thrombotic thrombocytopenic purpura; FU, follow‐up; GP, general practitioner; HCRU, healthcare resource use; SD, standard deviation; TTP, thrombotic thrombocytopenic purpura.
FU time in years.
In the cTTP cohort, the proportions of HCRU related to TTP (based on diagnosis codes) were 43.6% (n = 786), 5.9% (n = 27) and 2.5% (n = 40) for inpatient admissions, primary care appointments and outpatient appointments respectively.
A summary of the costs associated with HCRU events for the cTTP and matched non‐TTP cohorts is shown in Table 5. The average yearly cost of healthcare for a patient with cTTP was £6155, relative to £858 in the matched non‐TTP cohort. The average yearly cost of healthcare for a patient with cTTP consists of the cost of inpatient admissions (75.6%), outpatient appointments (18.6%), primary care (4.0%) and A&E attendances (1.8%).
TABLE 5.
Summary of HCRU costs per event for cTTP and matched non‐TTP cohorts (GBP, £).
| HCRU costs per event | Total number of events | Min (£, GBP) | Max (£, GBP) | Mean (SD) (£, GBP) | Median (Q1, Q3) (£, GBP) |
|---|---|---|---|---|---|
| Primary care appointments | |||||
| cTTP | 455 | 9 | 49 | 36 (17) | 49 (20, 49) |
| Non‐TTP | 1437 | 9 | 49 | 37 (16) | 49 (20, 49) |
| Inpatient admissions | |||||
| cTTP | 432 | 291 | 20 191 | 1338 (3204) | 527 (425, 527) |
| Non‐TTP | 58 | 199 | 12 383 | 2311 (2369) | 1030 (746, 3923) |
| Outpatient appointments | |||||
| cTTP | 1172 | 63 | 415 | 148 (49) | 145 (145, 145) |
| Non‐TTP | 425 | 63 | 371 | 139 (60) | 145 (86, 176) |
| A&E attendances | |||||
| cTTP | 117 | 94 | 459 | 201 (90) | 188 (141, 247) |
| Non‐TTP | 97 | 94 | 459 | 187 (81) | 188 (141, 206) |
Abbreviations: A&E, accident and emergency; cTTP, congenital thrombotic thrombocytopenic purpura; GBP, Great British Pounds; HCRU, healthcare resource use; Q1, quartile 1; Q3, quartile 3; SD, standard deviation; TTP, thrombotic thrombocytopenic purpura.
DISCUSSION
To the authors' knowledge, this is the first published study that evaluates the clinical burden, HCRU and associated costs of treating patients with cTTP in England, relative to a matched non‐TTP cohort. This study provides a comprehensive analysis using linked data from CPRD and HES.
A total of 36 patients with cTTP in England were identified and included in the study. A study of 1041 patients with TTP from the UK TTP Registry found that 73 patients (7.0%) in the UK had cTTP based on genetic testing. 7 Differences in the number of patients with cTTP identified may be due to the wider recruitment period in the UK TTP Registry study from 2003 to 2018 and differing methods of patient identification. Additionally, the UK TTP Registry study aimed to cover all TTP centres across the United Kingdom but recruitment was by consent, meaning the proportion of patients with cTTP may not be directly comparable with the current study. The current study identified cTTP cases with a clinician‐verified algorithm, since genetic testing or other diagnostic information on TTP type was not available in the data. However, the criteria used for the algorithm increase the confidence that patients in the cTTP cohort truly have cTTP, rather than iTTP.
There is a clear clinical burden for patients with cTTP, with 38.9% of patients experiencing at least one acute TTP episode during the follow‐up period. The rate of acute TTP episodes per person‐year was 0.13. This was lower than a similar study using the International Hereditary TTP Registry that reported an annual incidence rate of acute episodes of 0.35 per person‐year. 16 However, the difference in the observed rates may be due to differences in acute TTP definitions and the data available to identify these. For acute TTP episodes, a large proportion of admissions from A&E (37.0%) and episodes of care in ICU (26.0%) suggest a high emergency demand. The median cost associated with an acute TTP episode of £13 872 further underscores the high burden of cTTP on the healthcare system.
About 53% of patients with cTTP had a history of organ damage at baseline and 19.4% of patients with cTTP experienced an organ damage event during the follow‐up period. This is similar to the proportion of patients with cTTP experiencing organ damage reported by Sakai and Matsumoto who conducted a questionnaire study in a Japanese cTTP cohort of 55 patients and reported that 16 patients (29.1%) developed organ damage. 5 Differences may relate to varying lengths of follow‐up between studies, which were not reported by Sakai and Matsumoto.
There were higher HCRU rates in the cTTP cohort in relation to the non‐TTP cohort. The greatest difference in HCRU rate in the cTTP cohort, relative to the non‐TTP cohort, was for inpatient admissions. This may have been due to the treatment needs of patients with cTTP, given that almost half of inpatient admissions were directly related to TTP (43.6%). Overall, cTTP has an average yearly healthcare cost of £6155, relative to £858 for the non‐TTP cohort, demonstrating a considerable economic burden associated with cTTP in England.
Strengths and limitations
CPRD and HES are particularly valuable for studying rare diseases due to their large and diverse patient populations. 9 , 17 , 18 , 19 A strength of this study was the ability to link CPRD and HES databases to estimate multiple HCRU components in primary and secondary care. This enables a more complete and accurate estimation of HCRU across different settings as well as for acute clinical outcomes in the secondary care sector. Another strength of using these databases is that analysis using real‐world data reflects actual patient experiences, making the findings more generalisable to the broader population.
The data used in this study may present incomplete follow‐up for the population, as encounters at facilities outside the CPRD and HES networks are not captured. 20 As a result, the estimates presented in this manuscript are considered conservative. Coding errors, under‐coding and the inclusion of diagnoses for rule‐out purposes could have further impacted the accuracy of the data. However, it was assumed that the events of interest were reliably captured due to their clinical severity. Healthcare Resource Group (HRG) codes are used as proxies for secondary care costs, but these codes represent the reimbursement costs for the hospital rather than the actual cost of care and are also susceptible to human error.
The Coronavirus 2019 (COVID‐19) pandemic impacted healthcare data due to changes in HCRU. To accurately analyse trends and draw reliable conclusions, a pre‐pandemic study period was used; however, this means that the findings reflect the treatment landscape up to 2019. Another limitation is the difficulty in distinguishing between iTTP and cTTP due to the lack of specific diagnostic codes. The algorithm developed to identify patients with cTTP was validated by a clinical expert to ensure robustness in the absence of genetic cTTP confirmation in CPRD/HES data. However, misclassification is possible, particularly given that clinical recording of treatment codes is susceptible to human error. Furthermore, infusion of coagulation factor is typically administered at home by the patient/carer and, therefore, may not be recorded in CPRD/HES datasets.
CONCLUSION
This study highlights the considerable clinical and economic burden of cTTP. Almost 40% of patients experienced an acute TTP episode during the follow‐up period (up to 10 years), with over 25% of acute TTP episodes involving a stay in ICU, highlighting that the risk of this medical emergency remains high despite the availability of plasma‐based therapies. HCRU was markedly higher in the cTTP cohort relative to the non‐TTP cohort, resulting in considerably higher average annual healthcare costs. These results provide valuable insights for healthcare decision makers and may support future healthcare planning, the development of clinical guidelines and policy initiatives aimed at improving the management of rare diseases such as cTTP.
AUTHOR CONTRIBUTIONS
Erin Barker: Conceptualization; writing – original draft; writing – review and editing; formal analysis; methodology. Will Lester: Writing – review and editing; conceptualization. Monica Garrett: Project administration; writing – original draft; writing – review and editing; formal analysis; conceptualization; methodology. Heather Riley: Formal analysis; writing – review and editing; methodology; conceptualization. Amie Padhiar: Conceptualization; methodology; writing – review and editing. Annie Berkley: Conceptualization; methodology; writing – review and editing. Oliver Heard: Conceptualization; methodology; writing – review and editing. Sudhakar Manne: Conceptualization; data curation; methodology; writing – review and editing. Umesh Doobaree: Conceptualization; methodology; data curation; writing – review and editing.
FUNDING INFORMATION
This work was commissioned and funded by Takeda UK Ltd. Takeda UK Ltd. provided insight into the manuscript and the decision to publish results.
CONFLICT OF INTEREST STATEMENT
AP, AB, SM, UD and OH were employees of Takeda at the time of the study. EB, HR and MG are employees of York Health Economics Consortium, who received consultancy fees from Takeda UK Ltd. WL received consultancy fees from Takeda and Sanofi.
ETHICS STATEMENT
This study was approved via the CPRD RDG process (Protocol 22_001897). The CPRD holds ethics approval for observational studies that use only anonymised data.
PATIENT CONSENT STATEMENT
Patient consent was not required for this study. The data utilised were obtained from CPRD, which contains anonymised data collected during routine primary and secondary care. Access to the CPRD data was granted following a successful application via the CPRD RDG protocol (Protocol 22_001897).
PERMISSION TO REPRODUCE MATERIAL FROM OTHER SOURCES
The authors confirm that all data presented in this manuscript are original and have not been previously published elsewhere.
Supporting information
Table S1. TTP diagnosis codes.
Table S2. Organ damage ICD‐10 codes.
Table S3. Specific approaches taken for estimating secondary care costs.
ACKNOWLEDGEMENTS
The authors would like to acknowledge Sarah Campbell‐Hill, Niki Lam, Kunali Ghelani (employees of Takeda at the time of the study) and Karina Watts and Stuart Mealing (employees of York Health Economics Consortium at the time of the study) for their contributions to this study/manuscript. The authors would also like to acknowledge Kunali Ghelani (employee of Takeda at the time of the study) for their work on developing the deduplication algorithm with UD and SM.
DATA AVAILABILITY STATEMENT
This study uses data from the CPRD obtained under licence from the UK Medicines and Healthcare products Regulatory Agency. The data are collected as part of routine health appointments in primary and secondary care. The data are subject to a licence agreement and cannot be shared publicly. Secondary researchers can request access via the CPRD Research Data Governance (RDG) process.
REFERENCES
- 1. Dekimpe C, Roose E, Sakai K, Tersteeg C, De Meyer SF, Vanhoorelbeke K. Toward gene therapy for congenital thrombotic thrombocytopenic purpura. J Thromb Haemost. 2023;21(5):1090–1099. [DOI] [PubMed] [Google Scholar]
- 2. Joly BS, Coppo P, Veyradier A. Thrombotic thrombocytopenic purpura. Blood. 2017;129(21):2836–2846. [DOI] [PubMed] [Google Scholar]
- 3. Mariotte E, Azoulay E, Galicier L, Rondeau E, Zouiti F, Boisseau P, et al. Epidemiology and pathophysiology of adulthood‐onset thrombotic microangiopathy with severe ADAMTS13 deficiency (thrombotic thrombocytopenic purpura): a cross‐sectional analysis of the French National Registry for Thrombotic Microangiopathy. Lancet Haematol. 2016;3(5):e237–e245. [DOI] [PubMed] [Google Scholar]
- 4. Joly BS, Stepanian A, Leblanc T, Hajage D, Chambost H, Harambat J, et al. Child‐onset and adolescent‐onset acquired thrombotic thrombocytopenic purpura with severe ADAMTS13 deficiency: a cohort study of the French National Registry for Thrombotic Microangiopathy. Lancet Haematol. 2016;3(11):e537–e546. [DOI] [PubMed] [Google Scholar]
- 5. Sakai K, Matsumoto M. Clinical manifestations, current and future therapy, and long‐term outcomes in congenital thrombotic thrombocytopenic purpura. J Clin Med. 2023;12(10):3365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Scully M, Yarranton H, Liesner R, Cavenagh J, Hunt B, Benjamin S, et al. Regional UK TTP registry: correlation with laboratory ADAMTS 13 analysis and clinical features. Br J Haematol. 2008;142(5):819–826. [DOI] [PubMed] [Google Scholar]
- 7. Alwan F, Vendramin C, Liesner R, Clark A, Lester W, Dutt T, et al. Characterization and treatment of congenital thrombotic thrombocytopenic purpura. Blood. 2019;133(15):1644–1651. [DOI] [PubMed] [Google Scholar]
- 8. Scully M, Rayment R, Clark A, Westwood JP, Cranfield T, Gooding R, et al. A British Society for Haematology guideline: diagnosis and management of thrombotic thrombocytopenic purpura and thrombotic microangiopathies. Br J Haematol. 2023;203(4):546–563. [DOI] [PubMed] [Google Scholar]
- 9. Herrett E, Gallagher AM, Bhaskaran K, Forbes H, Mathur R, Van Staa T, et al. Data resource profile: clinical practice research datalink (CPRD). Int J Epidemiol. 2015;44(3):827–836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. CPRD . CPRD Aurum data specification. 2023. [cited 15 Jul 2025]. Available from: https://www.cprd.com/sites/default/files/2023‐09/CPRD%20Aurum%20Data%20Specification%20v3.1.pdf
- 11. Jones K, Weatherly H, Birch S, Castelli A, Chalkley M, Dargan A, et al. Unit costs of health and social care 2023 manual. UK: Kent; 2024. Available from: https://kar.kent.ac.uk/105685/ [Google Scholar]
- 12. NHS England . 2023‐25 NHS payment scheme (amended). 2023. [cited 24 Mar 2025]. Available from: https://www.england.nhs.uk/publication/2023‐25‐nhs‐payment‐scheme/
- 13. Public Health England . Technical guide confidence intervals. Available from: https://fingertips.phe.org.uk/documents/PHDS%20Guidance%20‐%20Confidence%20Intervals.pdf
- 14. Breslow NE, Day NE. Statistical methods in cancer research. Volume II—the design and analysis of cohort studies. IARC Sci Publ. 1987;82:1–406. [PubMed] [Google Scholar]
- 15. Medicines & Healthcare Products Regulatory Agency . CPRD safe outputs guidance. London: Medicines & Healthcare Products Regulatory Agency; 2024. Available from: https://www.cprd.com/sites/default/files/2024‐10/CPRD%20Safe%20Outputs%20Guidance.pdf [Google Scholar]
- 16. Tarasco E, Bütikofer L, Friedman KD, George JN, Hrachovinova I, Knöbl PN, et al. Annual incidence and severity of acute episodes in hereditary thrombotic thrombocytopenic purpura. Blood. 2021;137(25):3563–3575. [DOI] [PubMed] [Google Scholar]
- 17. Pimenta JM, Irving M, Cheung M, Mazzeo L, Landis S, Mukherjee S. Higher rates of non‐skeletal complications and greater healthcare needs in achondroplasia compared to the general UK population: a matched cohort study using the CPRD database. Orphanet J Rare Dis. 2023;18(1):211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kingswood JC, Nasuti P, Patel K, Myland M, Siva V, Gray E. The economic burden of tuberous sclerosis complex in UK patients with renal manifestations: a retrospective cohort study in the clinical practice research datalink (CPRD). J Med Econ. 2016;19(12):1116–1126. [DOI] [PubMed] [Google Scholar]
- 19. Hawley S, Shaw NJ, Delmestri A, Prieto‐Alhambra D, Cooper C, Pinedo‐Villanueva R, et al. Higher prevalence of non‐skeletal comorbidity related to X‐linked hypophosphataemia: a UK parallel cohort study using CPRD. Rheumatology. 2021;60(9):4055–4062. [DOI] [PubMed] [Google Scholar]
- 20. Clinical Practice Research Datalink (CPRD) . What are the limitations of CPRD data? 2024. [cited Feb 2025]. Available from: https://www.cprd.com/introduction‐cprd#What%20are%20the%20limitations%20of%20CPRD%20data?
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. TTP diagnosis codes.
Table S2. Organ damage ICD‐10 codes.
Table S3. Specific approaches taken for estimating secondary care costs.
Data Availability Statement
This study uses data from the CPRD obtained under licence from the UK Medicines and Healthcare products Regulatory Agency. The data are collected as part of routine health appointments in primary and secondary care. The data are subject to a licence agreement and cannot be shared publicly. Secondary researchers can request access via the CPRD Research Data Governance (RDG) process.