Real-world Diagnosis and Testing of Acquired Hemophilia A in Japan: A Web-based Physician Survey

Abstract

Objective

Acquired hemophilia A (AHA) is a rare and often severe bleeding disorder caused by autoantibodies against coagulation factor VIII (FVIII). Both the prompt diagnosis and treatment of patients with AHA are essential. Data on Japanese AHA diagnostic delay are lacking; therefore, this study aimed to determine the time to AHA diagnosis in Japan, as well as to identify factors influencing the time from presentation to diagnosis.

Methods

This web-based, prospective, observational study surveyed physicians in Japan who had diagnosed or treated patients with AHA. The time from bleeding onset to diagnosis was the primary endpoint, with diagnostic testing barriers as the secondary endpoints.

Results

Of the 322 physicians who accessed the survey, 112 were included in the study. The time from bleeding onset to referral and from referral to diagnosis was ≥8 days in 31.3% (n=35) and 23.2% (n=26) of cases, respectively. The tests most commonly conducted in-hospital were activated partial thromboplastin time (aPTT) and aPTT cross-mixing tests (85.7% and 73.2%, respectively), for which results typically took ＜1 day to obtain. FVIII activity and FVIII inhibitor tests were more often outsourced, and when performed in-hospital, the results took ≥2 days in over 25% of cases. Across the tests surveyed, 41.7-76.2% of physicians noted difficulty in interpreting test results as a reason for AHA diagnosis delay.

Conclusion

More efficient diagnostic procedures and enhanced physician education and awareness are needed to shorten AHA diagnosis times in Japan.

Introduction

Acquired hemophilia A (AHA) is a bleeding disorder caused by the development of autoantibodies, also known as inhibitors, to coagulation factor VIII (FVIII), leading to the inhibition of FVIII function (1). Considered a very rare disease, studies in Europe have estimated the incidence of AHA to be 1.3-1.5 individuals per million per year (2-4). In 2022, 296 cases of AHA were reported in Japan (5). AHA can be severe and life-threatening, with a reported mortality of 15-42% (6,7). In Japan, two studies reported mortality rates of 13.8% (over 17 months) and 25% (over 3 years) in patients with AHA (8,9).

AHA manifests as spontaneous bleeding at sites throughout the body, with patients typically presenting with subcutaneous, gastrointestinal, muscle, or, more rarely, intracranial hemorrhage, often without a history of bleeding (3,10-12). Laboratory tests of patients with AHA often show an isolated and prolonged activated partial thromboplastin time (aPTT), reduced FVIII activity (FVIII:C), and the presence of autoantibodies against FVIII, as detected by Bethesda assay or enzyme-linked immunosorbent assay (ELISA) (12). To distinguish coagulation factor deficiency from the presence of inhibitors, aPTT cross-mixing tests may be conducted if FVIII:C testing is not immediately available. FVIII inhibitors in AHA are time- and temperature-dependent, so aPTT results should be compared immediately following the mixture of control and patient plasma and after a 2-hour incubation period (12). The Japanese guidelines for the management of AHA published in 2017 suggest performing cross-mixing tests with various mixing ratios of patient plasma and normal plasma to differentiate lupus anticoagulant from AHA based on differences in the shape of the aPTT graph (13). The differential diagnoses of low FVIII:C include von Willebrand disease and acquired von Willebrand syndrome (12).

Diagnostic delays, inadequate treatment, and complications from invasive procedures for controlling hemorrhages all contribute to the severity of bleeding in patients with AHA (14). A delayed diagnosis of AHA significantly affects the time from bleeding onset to treatment initiation (4); therefore, timely diagnosis is essential to improve patient outcomes. However, barriers to the timely and accurate diagnosis of AHA remain. Owing to the rarity of the disease, AHA is often unrecognized or misdiagnosed (15), especially as patients often first present to physicians who do not have prior experience with the disease (16). Additionally, although FVIII:C and FVIII inhibitor tests are essential for the definitive diagnosis of AHA (13), our understanding is that only a few medical centers in Japan can perform these tests in-house. Other medical centers must rely on outsourcing the tests, increasing the time to diagnosis. In comparison, the aPTT cross-mixing test does not provide a definitive diagnosis of AHA, but can be conveniently conducted at the majority of Japanese facilities and is useful for the detection of inhibitors against intrinsic coagulation factors (13).

In Japan, no studies on AHA diagnostic delays or barriers to diagnosis have so far been conducted. This survey of physicians in Japan who have treated patients with AHA aimed to determine the time to AHA diagnosis, as well as to identify factors that influence the time from presentation to diagnosis.

Materials and Methods

Study design

This web-based, prospective, cross-sectional, observational study was conducted between April 1 and April 12, 2024 (UMIN Clinical Trials Registry ID: UMIN000053897). Physicians completed an online survey on their real-world experience in treating patients with AHA, including time to diagnosis, time to treatment, and issues experienced in diagnosing AHA. The survey questions are presented in Supplementary material 1. The study was conducted in accordance with the principles of the Declaration of Helsinki, Ethical Guidelines for Medical and Health Research Involving Human Subjects (17), and all other applicable laws and regulations. All study documents were approved by the Japan Physicians Association Institutional Review Board on March 11, 2024. Physicians were recruited from a physician panel owned by CareNet using a dedicated web system, and were compensated for their participation. Informed consent was obtained from all physicians before survey initiation. Collection of informed consent, registration, and survey participation were all conducted on the web system. Personal identification details, such as name, address, and telephone number, were stored securely by CareNet, and persons outside CareNet did not have access to these details. Physicians were anonymized using central registration numbers for identification and anonymized data were provided for analysis.

Participants

Physicians were included if they belonged to the hematology, emergency medicine, general internal medicine, general medicine, or pediatrics departments and had diagnosed or treated patients with AHA in the 3 years before giving their consent. These departments were chosen as those most likely to be involved in diagnosing and treating patients with AHA. Physicians were excluded if they, at the time of giving their consent, had less than 2 years of clinical experience (excluding residency); were working in facilities other than clinics, university hospitals, general hospitals, or national and public hospitals; were not treating outpatients or inpatients; or required legal representatives.

Endpoints

The primary endpoint was the time from the onset of bleeding to AHA diagnosis. Specifically, physicians were asked to consider the most recent patient with AHA that they had encountered, and to report the time taken for this patient to be referred to their department once bleeding had started. They were then asked about the time taken from the patient's initial visit to their department to the diagnosis of AHA.

The secondary endpoints included details of tests conducted to diagnose AHA (including the types of tests conducted, whether the tests were performed in-hospital or outsourced, and the time taken to receive test results) and the challenges faced by physicians in the diagnosis and testing of AHA (including which tests caused delays or difficulties in diagnosing AHA and why, and the referral process from other departments).

Sample size and statistical methods

As this was not a confirmatory study, the sample size was not determined based on statistical calculations. The target number of participants was 100, considering the rarity of AHA. The dataset of participant responses used for analysis was developed by Macromill CareNet using QuickCross, an original software program of Macromill Carenet Statistical analysis was performed by EviPRO. In the event of abnormal responses or data that may need to be excluded, data handling decisions were made by the principal investigator, and the excluded data were treated as missing data. Missing values were not imputed. Demographic information, baseline characteristics, and primary and secondary endpoints are described as summary statistics of categorical variables. Spearman's rank correlation coefficients, 95% confidence intervals (CIs), and p values were calculated for exploratory correlation analyses.

Results

Physician disposition and characteristics

In total, 322 physicians accessed the survey portal, of whom 210 were excluded (refusal of consent form, n=64; did not meet the inclusion criteria, n=122; met the exclusion criteria, n=24) (Fig. 1). The remaining 112 physicians passed the screening, completed the survey, and were included in the analysis population. No physicians or their associated data were excluded during the study. Most physicians were in the departments of hematology (n=71, 63.4%), general internal medicine (n=22, 19.6%), or emergency medicine (n=8, 7.1%). In total, 31 physicians (27.7%) belonged to university hospitals, 23 (20.5%) to national and public hospitals, 46 (41.1%) to other general hospitals, and 12 (10.7%) to clinics (Fig. 2).

Figure 1.

Participant disposition.

Figure 2.

Medical department and facility characteristics of the physicians (n=112).

Time from onset of bleeding to diagnosis of AHA

For the most recent patient with AHA encountered by physicians, the most common length of time from the onset of bleeding to referral to their department was 2-7 days (n=39, 34.8%), followed by 8 days to less than 1 month (n=26, 23.2%), and 1 day (n=16, 14.3%) (Fig. 3a). Some patients took much longer to be referred; eight (7.1%) were referred after 1 month to less than 6 months, and one (0.9%) was referred after 6 months or more.

Figure 3.

Length of time from bleeding onset to referral (a) and from referral to diagnosis (b) (n=112). ^aPatient seen directly in own department.

The most common duration from referral to diagnosis was 2-7 days (n=46, 41.1%) (Fig. 3b). Regarding patients for whom diagnosis took longer, 17 (15.2%) took 8 days to less than 1 month, eight (7.1%) took 1 month to less than 6 months, and one (0.9%) took more than 6 months. Fifteen patients (13.4%) were diagnosed in 1 day and 11 (9.8%) were diagnosed within 24 hours. When stratified by the physicians' department, 22/71 patients (31.0%) referred to the hematology department were diagnosed within 1 day after referral; this proportion was lower in other departments [emergency medicine: 2/8 (25.0%), general internal medicine: 2/22 (9.1%), general medicine: 0/6 (0%), pediatrics: 0/5 (0%)].

Referral process from other departments

Most physicians encountered patients with AHA through referrals from other departments, with the most frequent (chosen by ≥15% of physicians) being general internal medicine (n=40, 35.7%), emergency medicine (n=30, 26.8%), general medicine (n=23, 20.5%), and dermatology (n=18, 16.1%) (Supplementary material 2). In seven (6.3%) cases, the patients presented to the physicians' department directly without a referral (Fig. 3a, b). The patients were most commonly referred from other general hospitals (n=51/106, 48.1%), followed by clinics (n=37/106, 34.9%), university hospitals (n=23/106, 21.7%), and national and public hospitals (n=16/106, 15.1%) (Supplementary material 3).

Nature of tests conducted for the diagnosis of AHA

The details regarding the type of tests conducted for diagnosing AHA, as well as whether they were performed in-hospital at the physicians' facility or outsourced, are shown in Fig. 4. The tests most often conducted at the physicians' facility were the aPTT (85.7% in-hospital, 10.7% outsourced) and aPTT cross-mixing tests (73.2% in-hospital, 17.9% outsourced). The other AHA diagnostic testing options listed in the survey [FVIII:C; FVIII antigen level; FVIII inhibitor titer (Bethesda method); anti-FVIII binding antibody titer (ELISA); von Willebrand factor (VWF): Ristocetin Cofactor (RCo), and VWF antigen level] were more commonly outsourced (13.4-28.6% in-hospital, 64.3-75.0% outsourced).

Figure 4.

Type and location (in-hospital or outsourced) of tests conducted for AHA diagnosis. AHA: acquired hemophilia A, aPTT: activated partial thromboplastin time, ELISA: enzyme-linked immunosorbent assay, FVIII: coagulation factor VIII, FVIII:C: FVIII activity level, FIX: coagulation factor IX, RCo: ristocetin cofactor, VWF: von Willebrand factor

Time required for in-hospital tests

For tests performed in-hospital at the physicians' facility, most physicians reported that it took 1 day or less to obtain the results from aPTT (n=88/96, 91.7%) and aPTT cross-mixing tests (n=75/82, 91.5%). Other listed tests typically required a longer time between conducting the test and receiving the results, with results taking 2 days or more in 28.2-66.7% of cases (Fig. 5).

Figure 5.

Time taken to receive results from in-hospital tests. aPTT: activated partial thromboplastin time, ELISA: enzyme-linked immunosorbent assay, FVIII: coagulation factor VIII, FVIII:C: FVIII activity level, RCo: ristocetin cofactor, VWF: von Willebrand factor

Diagnostic tests as the cause of delays in AHA diagnosis

When physicians were asked which tests caused difficulties in diagnosing AHA, the most common answers were aPTT cross-mixing tests (n=35/111, 31.5%), FVIII inhibitor titer tests (Bethesda method; n=32/111, 28.8%), and FVIII:C tests (n=30/111, 27.0%) (Fig. 6). There was a significant positive correlation between the time from referral to diagnosis and difficulties in diagnosis because of the anti-FVIII binding antibody titer test [ELISA; correlation coefficient (95% CI): 0.21 (0.01, 0.39), p=0.035, Table]. Thirty-eight of 111 (34.2%) physicians did not experience any AHA diagnosis difficulties because of diagnostic testing. A significant negative correlation was observed between the time from referral to diagnosis and the absence of difficulties due to diagnostic testing [correlation coefficient (95% CI): -0.23 (-0.41, -0.03), p=0.023, Table].

Figure 6.

Difficulties in the diagnosis of AHA caused by testing (n=111). Multiple answers were allowed. AHA: acquired hemophilia A, aPTT: activated partial thromboplastin time, ELISA: enzyme-linked immunosorbent assay, FVIII: coagulation factor VIII, FVIII:C: FVIII activity level, RCo: ristocetin cofactor, VWF: von Willebrand factor

Table.

Correlation between the Time Taken from Referral to Diagnosis and Difficulty in Diagnosing AHA Using Laboratory Tests (n=98).

Variable	Spearman correlation coefficient (95% CI)	p value
Experience with difficulty in diagnosing AHA using specific tests
aPTT	0.13 (-0.07, 0.32)	0.21
aPTT cross-mixing tests	0.11 (-0.10, 0.30)	0.30
FVIII activity	0.10 (-0.11, 0.29)	0.35
FVIII antigen level	0.13 (-0.07, 0.32)	0.20
FVIII inhibitor titer (Bethesda method)	0.11 (-0.09, 0.30)	0.30
Anti-FVIII binding antibody titer (ELISA)	0.21 (0.01, 0.39)	0.035
VWF:RCo	0.08 (-0.13, 0.27)	0.46
VWF antigen level	0.12 (-0.08, 0.31)	0.23
None of the above applies	-0.23 (-0.41, -0.03)	0.023

AHA: acquired hemophilia A, aPTT: activated partial thromboplastin time, CI: confidence interval, ELISA: enzyme-linked immunosorbent assay, FVIII: coagulation factor VIII, RCo: ristocetin cofactor, VWF: von Willebrand factor

Reasons for difficulty with diagnostic tests

To clarify why specific tests may have led to delays in the diagnosis of AHA, physicians were asked to explain the reasons for the difficulties caused by the identified tests. ‘I was lost in the interpretation of the results’ was chosen as the main reason leading to diagnostic difficulties for aPTT (n=16/21, 76.2%), aPTT cross-mixing (n=26/35, 74.3%), FVIII:C (n=20/30, 66.7%), and FVIII antigen level (n=14/22, 63.6%) tests (Fig. 7). For the other tests [FVIII inhibitor titer (Bethesda method), anti-FVIII binding antibody titer (ELISA), VWF:RCo, and VWF antigen level], the reasons for difficulty were approximately equally split between ‘I was lost in interpretation of the results’ and ‘It took a long time to get the results’.

Figure 7.

Reasons for difficulty in making an AHA diagnosis, by test type. Multiple answers were allowed. AHA: acquired hemophilia A, aPTT: activated partial thromboplastin time, ELISA: enzyme-linked immunosorbent assay, FVIII: coagulation factor VIII, FVIII:C: FVIII activity level, RCo: ristocetin cofactor, VWF: von Willebrand factor

Physicians who selected ‘I was lost in the interpretation of the results’ as the main reason for diagnostic difficulties were asked a follow-up question to identify specific reasons for their confusion in result interpretation (Fig. 8). The reasons for the uncertainty in interpretation varied according to the test. For the aPTT cross-mixing test, the most common reason identified was the difficulty in distinguishing between lupus anticoagulant and inhibitor patterns. For FVIII:C, anti-FVIII binding antibody titer (ELISA), VWF:RCo, and VWF antigen level tests, the most common reason for diagnosis difficulty was that the physician did not know the clinically significant abnormal value.

Figure 8.

Reason for difficulty in interpreting the test results, by test type. Multiple answers were allowed. aPTT: activated partial thromboplastin time, ELISA: enzyme-linked immunosorbent assay, FVIII: coagulation factor VIII, FVIII:C: FVIII activity level, LA: lupus anticoagulant, RCo: ristocetin cofactor, VWF: von Willebrand factor

Discussion

This study, which was the first large-scale survey of its kind in Japan, surveyed 112 physicians who had diagnosed or treated patients with AHA to determine the time to AHA diagnosis and identify issues within the referral and diagnostic testing processes. The most common time between bleeding onset and referral to the surveyed physicians' department was 2-7 days (34.8%). However, 31.2% of patients were referred after 8 days or more, suggesting that some physicians at the patient's first point of contact had challenges with initially suspecting AHA. Once patients were referred to the surveyed physicians' department, 41.1% were diagnosed within 2-7 days, while 23.2% of cases required more than 8 days, further indicating that diagnosing patients with AHA was sometimes challenging for some physicians.

According to a study conducted using European Acquired Haemophilia Registry data from 2003 to 2008, 37.2% (174/467) of patients with AHA were diagnosed within 1 day of bleeding onset (4). Although this result was not divided into ‘from onset to referral’ and ‘from referral to diagnosis,’ comparison with our results suggests that it takes more time to diagnose AHA in Japan than in Europe. Previously, the exact time frame from symptom onset to diagnosis of AHA in Japan was unclear. However, the findings of this web-based study have demonstrated that there are substantial delays in AHA diagnosis in Japan, which is concerning given that patients with AHA remain at risk of life-threatening bleeding, even if they are not actively experiencing an acute bleeding episode. AHA is more common in elderly individuals and can cause severe bleeding symptoms, leading to significant impairment in patients' activities of daily living within a short period. Therefore, a prompt diagnosis is essential to ensure the timely initiation of appropriate treatment (18).

Most patients were referred from general internal medicine, emergency medicine, or general medicine. General internal medicine is a common and often large primary care department, and the high number of referrals from this department aligns with the likelihood of patients first interacting with primary care physicians when experiencing AHA symptoms. The high number of referrals from the emergency department is indicative of the often severe bleeding symptoms of AHA, with patients experiencing severe anemia, hemorrhagic shock, and immobilization caused by severe pain. Delays in the referral process was identified as a key reason for delays in diagnosing AHA in Japan, with only 6.3% of patients visiting the surveyed physicians' departments without referral. Additionally, when referred to the emergency, general internal medicine, or general medicine departments, patients were diagnosed with AHA within 1 day in only 25.0%, 9.1%, and 0% of cases, respectively, which was lower than in patients referred to the hematology department (31.0%). Increasing disease awareness of AHA in first-touch departments (such as among primary care physicians) is therefore of key importance. The need for greater education of AHA among physicians has also been recognized more broadly, with the Zürich Haemophilia Forum, an expert panel of European hemophilia specialists, suggesting ‘improving initial diagnosis of AHA in all settings’ as the first principle of AHA care (19). Some suggestions for improving awareness and diagnosis of AHA outside of specialist centers include the sharing of case studies of patients with AHA who presented to first touch departments, including information regarding clinical presentation, bleeding phenotype, and differences to congenital hemophilia, and the distribution of simple, visual guidelines to aid diagnosis. We conducted another survey to investigate awareness of AHA among physicians in Japan (UMIN Clinical Trials Registry ID: UMIN000053895), the results of which will be reported separately.

The majority of the aPTT and aPTT cross-mixing tests were performed in-hospital, and most of these in-hospital tests were completed in 1 day or less. In comparison, other tests, such as FVIII:C and FVIII inhibitor tests, were mainly outsourced. The turnaround time for FVIII:C and FVIII inhibitor tests is 2-4 days when outsourced to a private Japanese laboratory company (20). When the same tests were performed in-hospital, the results took 4 days or more in 6.3% (FVIII:C) and 14.3% (FVIII inhibitor) of cases. It should be noted that the definitions for ‘in-hospital’ and ‘outsourcing’ were not specifically defined in the survey, so it is possible that the in-hospital versus outsourcing rates, and their associated turnaround times, were influenced by participants' differing interpretations of what constitutes ‘outsourcing.’ For example, some participants may consider the ‘outsourcing’ option to include outsourcing to another hospital with better resources or to another department within the same hospital. Nevertheless, these results indicate overall that the percentage of tests conducted in-hospital, and associated in-hospital turnaround times, can still be improved. These results suggest that in Japan, challenges to conducting the diagnostic tests in-hospital may contribute to delays in AHA diagnosis.

More than a quarter of the physicians reported difficulties (e.g., with interpretation of results) with aPTT cross-mixing, FVIII:C, and FVIII inhibitor tests. For the aPTT cross-mixing test, many physicians reported difficulty in distinguishing between lupus anticoagulant and inhibitor patterns. The challenges associated with the aPTT cross-mixing test include its qualitative rather than quantitative nature, atypical curves on the cross-mixing graph for some patients with AHA who have high inhibitor titers (21), and the fact that some patients develop AHA and lupus anticoagulant simultaneously, which influences the curve (22). Thus, the results of the aPTT cross-mixing test require a high level of expertise for interpretation. For the FVIII:C and FVIII inhibitor titer (Bethesda method) tests, which are required for the definitive diagnosis of AHA, physicians noted difficulties in identifying clinically significant values from these tests, difficulties in identifying the cause of abnormal values, and complications arising from the need to consider the anticoagulant medication being taken by the patient. The correlation between FVIII:C in patients with AHA and bleeding severity or global coagulation potential has been reported to be low (23), and some patients with AHA still show FVIII:C over 20% (23,24). Additionally, lupus anticoagulant, which causes anti-phospholipid syndrome, can affect FVIII:C and FVIII inhibitor titers measured by the one-stage clotting assay and cause pseudo-positive FVIII inhibitors and pseudo-FVIII deficiencies (25), contributing to the difficulty in interpretation. Recommended solutions to mitigate these issues include facilitating the widespread use of anti-FVIII binding antibody titer testing by ELISA, as well as measurement of FVIII:C and FVIII inhibitors using plasma samples at a high dilution to eliminate the lupus anticoagulant effect (13). Additionally, physicians should be encouraged to measure FVIII:C using both the one-stage clotting and chromogenic substrate assays. However, to most reliably eliminate a potential lupus anticoagulant effect, lupus anticoagulant testing should be performed in accordance with the International Society on Thrombosis and Haemostasis lupus anticoagulant detection and interpretation guidelines (26).

Interestingly, a positive correlation (0.21) was observed between the time to diagnosis and difficulties with the anti-FVIII binding antibody (ELISA) test. In Japan, the anti-FVIII binding antibody test is not common because the test fee is not reimbursed by the national medical insurance scheme; therefore, it is possible that the test may not be widely recognized by physicians.

The aPTT test was identified as a reason for diagnostic difficulties by 18.9% of physicians, with uncertainty about the cause of the abnormal values and the need to consider the influence of the patient's anticoagulation medication given as reasons for difficulties in interpreting the aPTT results. Some aPTT reagents are not very sensitive to deficiencies in clotting factors (27), and mild decreases in FVIII:C may not be reflected by a prolonged aPTT. Measurements of aPTT may also be prolonged by anticoagulants (direct oral anticoagulants, warfarin, etc.), which complicates the interpretation of aPTT results and may require physicians to wait for the washout of anticoagulants, leading to a delayed diagnosis. Ultimately, physicians who did not have difficulties with any of the tests were able to diagnose patients with AHA more quickly, indicating the necessity of educational activities to improve physicians' abilities to interpret laboratory tests associated with AHA.

This study had several limitations. Because physicians were recruited from a physician panel owned by CareNet, there was a risk of some selection bias. The physicians who responded to the survey may not be representative of all physicians in the target departments who have diagnosed or treated patients with AHA in the past 3 years in Japan. Owing to the web-based nature of the survey, it was also not possible to confirm that all physicians fully understood the meaning of the survey questions or were familiar with all the diagnostic tests mentioned before responding, which is reflected in the common reasons given by participants for why they had difficulties interpreting test results. Participants also may not all have had similar interpretations of the survey questions. Additionally, as the survey asked about past experiences, the answers may have been affected by some recall bias. However, limited clinical data are available for AHA because it is a very rare disease, so such survey studies are considered to be valuable in this context. Furthermore, because of the rarity of AHA, the sample size of participating physicians was limited, especially for some questions and subgroup analyses. Finally, this survey was conducted among physicians working in Japan only; therefore, the real-world findings of the study may not be applicable to AHA diagnosis conditions in other countries.

Conclusion

More than 7 years have passed since the revised guidelines for the management of AHA were published in 2017 (13). Given that the data for this study were collected in 2024, it reflects the state of AHA diagnostic experiences in Japan following the implementation of the guidelines. This study identified that the diagnosis of AHA is often delayed in Japan, and is the first study to evaluate the extent of and reasons behind these delays. Delays in receiving diagnostic test results and difficulties in their interpretation were identified as barriers to diagnosis. Some countermeasures to shorten the time from bleeding onset to AHA diagnosis may be required, such as speeding up the process of obtaining diagnostic test results and providing education and guidance on their interpretation.

Author’s disclosure of potential Conflicts of Interest (COI).

Yoshinobu Seki: Honoraria, Chugai Pharmaceutical, Takeda Pharmaceuticals, and Sanofi. Madoka Go: Employee at Takeda Pharmaceutical. Masahiro Ieko: Honoraria, Takeda Pharmaceuticals.

This work was supported by Takeda Pharmaceutical.

Supplementary Material

Supplementary Table 1.

Questions and Response Options.

Supplementary Figure 1.

Original department of patient referral (n=112)

Supplementary Figure 2.

Original hospital or facility of patient referral (n=106)

Funding Statement

The authors would like to thank the study team members of Takeda Pharmaceutical, Tokyo, Japan, for their support in conducting the study. Data collection and management were provided by Macromill Carenet and statistical analysis was provided by EviPRO. Medical writing support was provided by Jennifer Hung Ph. D. and Sarah Graham Ph. D. of Oxford PharmaGenesis, Melbourne, Australia, and funded by Takeda Pharmaceutical in accordance with the Good Publication Practice (GPP 2022) guidelines (www.ismpp.org/gpp-2022).