Direct transfer of data of people with haemophilia from the Thai Haemophilia Treatment Centre Registry to the World Bleeding Disorders Registry of the World Federation of Hemophilia.

Abstract

Most haemophilia treatment centres (HTCs) have their own registry. In 2016, the Thai Society of Haematology (TSH) initiated the 'Haemostasis Registry' consisting of two separate parts, 'Hereditary Bleeding Disorders' and 'Thrombosis'. Both parts of this registry were developed and ready for data entry in 2020. The 'Hereditary Bleeding Disorders Registry (HBDR)' collects data from patients with hereditary bleeding disorders caused by altered endothelium, platelet and coagulation, including haemophilia A and B. Similarly, the World Federation of Hemophilia (WFH) has established a global database on people with haemophilia (PWH) called the 'World Bleeding Disorders Registry (WBDR)' in 2018.1, 2 Both HBDR and WBDR are privacy-protected web-based data entry systems using a computerised-generated identification for each patient. Only registered medical personnel can assess their own patients’ data. An accessible patient registry will strengthen the capacity to diagnose, treat and care for PWH. To avoid duplicate workload of medical personnel interested in entering data in both registries, an electronic connection of the HBDR and the WBDR was created between January and December 2020 with the approval of the Ethics Committee of the Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand (COA.MURA 2019/604). Written informed consent must be obtained from PWH and parents for enrolling in both HBDR and WBDR. The computerised programme of MySQL (Version 5.7)3, 4 and phpMyAdmin (Version 4.9.7)5, 6 were selected for the transferring process. MySQL is the world's most popular open-source database, and is fast, reliable, scalable and easy to use. It allowed us to define the data in tables, update and query data, and access data in the database. The phpMyAdmin, a web application, has also been used to handle the administration of MySQL over the web, and is also an open-source tool written in PHP. A well-planned database set up should be precise and cover the essential issues. Importantly, all data should be available and useful, providing an easy way for clinicians and medical personnel to follow the clinical progress of their patients. These output tools should be conceived similarly to the design of the registry so any errors either too much or too little data will be evident from the start. Moreover, always improving patient care by adding quality of life assessment tools for PWH is important, which should be planned when entering data. Think big for possible services that will be added for better patient care.7 Data from 26 relational tables in the HBDR's database were transferred to 11 Excel tables of minimal and extended dataset8 as shown in Table 1. The minimal dataset included 'baseline data', including basic demographic data, 'baseline visit', 'baseline hospitalisation' and 'baseline treatment' capturing retrospective clinical events, hospitalisation and replacement therapy experienced in the previous 6 months before the registration date, respectively; and 'follow-up visit', 'follow-up hospitalisation' and 'follow-up treatment' recording prospective clinical events, hospitalisation and replacement therapy at each follow-up visit tracking the registration date, respectively. Additional extended datasets included inhibitor, musculoskeletal assessment and mortality. However, some constraints required appropriate solutions. First, the platform of WBDR requires the entry of retrospective data for the 6 months before the registration date, and then ongoing data were entered at each of the follow-up visit for the time between visits. This resulted in a staggered visit schedule which was not based on a calendar year. During the evaluation process, we determined that the HBDR data would be entered each calendar year, with no overlapping years permitted, rather than on a rolling basis at each follow-up visit. This was deemed to be helpful for the summation of episodes of hospitalisation, number of days of hospitalisation, number of bleeds, number of joint bleeds as well as the amount of replacement therapy within one calendar year for the individual annual report. Second was the factor replacement therapy. The WBDR requires the name of the treatment product, amount, unit and intake of the administered factor concentrate, and the database automatically calculated the total used factor concentrate to treat one episode of bleeding. On the other hand, HBDR was developed to collect the total amount of the used factor concentrate to treat one bleeding episode, including on demand, prophylaxis and immune tolerance induction. To minimise data entry errors, the HBDR transfers the total amount of factor used to the WBDR. Third was the terminology used in the registries. The words used to define data fields and outcomes in various registries differed based on cultural and linguistic differences in countries. As such, some of the terminology used in the HBDR differed from the WBDR, but the concepts behind the terminology were sufficiently similar that they could be matched between the two registries. For example, one of the 'reasons to stop treatment' for episodic treatment is defined as 'good result' in the HBDR but is defined as 'on demand treatment complete' in the WBDR. These two terms have the same meaning and can be matched when linking the databases. Fourth was the security of data and potential for duplication of PWH in the WBDR. The WBDR uses a secure one-way hashing algorithm to generate unique patient identifiers (UPI), which ensures no duplication of PWH. This process involves the input of four immutable patient identifiers, which are not stored in the registry but are used to generate the UPI. Typically, the participating HTC will input these pieces of data into the WBDR to generate the UPI. For this project, the selection of the four immutable pieces of data were modified to include computer-created PWH identifiers from the HBDR preceded by the word 'THAI' so that each identification is unique to each PWH from Thailand, for instance, THAI-20200600030. Upon the completeness of the data transfer designation, a virtual conference among HTCs in Thailand was organized with the subsequent training of data entering for the physicians and medical personnel. An easy-to-understand instruction was prepared in Thai and distributed to the participating HTCs. HTCs have the alternative of opting out by simply not pressing the button 'Connect to WBDR'. This may result in some HTCs not participating in data transfer; however, and importantly, it also has the advantage that a data transfer from the HBDR to the WBDR will not be abandoned in the event that not all HTCs agree to participate. To transfer selected data from the HBDR to the WBDR, two query screens were added for each user at the HBDR to complete. First, 'Would you like to connect to the WDBR?' and if ’Yes’, the second screen would ask: 'Does your department/hospital receive ethics approval from a Faculty Ethics Committee?' If 'No', the user will be asked to obtain ethics approval before connecting to the WBDR. If 'Yes', the user will be able to transfer the data to the WBDR by clicking a button called 'Connect to WBDR'. The selected data of PWH are captured from the HBDR and transferred to the WBDR. Only age at first bleed required an additional access from the database of the HBDR. An expansion of this project, Phase 2, is currently being evaluated. Phase 2 will include the direct transfer of select data from the HBDR directly to the WFH Annual Global Survey (AGS). Currently the National Member Organization in Thailand compiles all required data from the HTCs in Thailand and sends it to the WFH on an annual basis. Obtaining the data directly from the HBDR to the WFH AGS on an annual basis will be reliable, accurate and time-saving. In fact, data entering for the registry constituted extra efforts by the medical personnel who already had routine workload. The benefit of improving patient care as a whole and for individuals should outweigh this constraint. Moreover, the TSH has an arrangement with the haematology team in each hospital for more patients being registered. Although 52 HTCs are located at the university, regional and provincial hospitals in Thailand, eight of 10 HTCs at the university hospitals actively participate in the HBDR and WBDR. In November 2021, a total of 300 PWH of 559 patients with hereditary bleeding disorders, registered to the HBDR, had already been transferred to the WBDR. To our knowledge, this is the first report illustrating the possibility of transferring minimal and extended datasets of PWH from an HTC to an existing local registry, and to the WBDR. This method of having each HTC be responsible for sending the data themselves, provides each HTC the ability to decide to participate or not. The extent of transferred data depends upon the appropriate data selection and accessibility of the computerised programme. Advanced technology should ease the difficulty and constraints of data management. It should be smart, friendly and capable of assisting the routine workload. In conclusion, the data of PWH from the existing local registry can be directly transferred to the WBDR using computerised programmes namely MySQL and phpMyAdmin. This method can be adapted to other existing registries in both economically developed and less developed countries. The real-world data of PWH for the WBDR will be useful for the evidence-based advocacy initiatives to serve PWH globally. The authors would like to thank the patients and families participating in the HBDR and WBDR, and gratefully acknowledge that this study was partially funded by a grant from Grifols. The authors state that they have no interests which might be perceived as posing a conflict or bias. A.C. designed and conducted the study with data analysis and wrote the manuscript; W.Y. designed the computerised programme and wrote the manuscript; N.S. designed the data collection of the HBDR; R.N. designed the data collection of both HBDR and WBDR; T.R. conducted the transfer of HBDR to WBDR. All authors approved the manuscript. The data that the findings of this report are available from the corresponding author upon reasonable request.