In the UK at the time the survey was conducted there were two SACs (in London and Sheffield); in German there were nine (in Lunbeck, Munchen, Tubingen, Bonn, Essen, Aachen, Berlin, Dusseldorf and Magdeburg); and, in Italy there were eleven (in Florence, Milan, Messina, Naples, Rome (2 centres), Siena, Turin, Pisa, Genova and Bologna). We conducted a survey of patients and families affected by ataxia in these three countries, collecting data on their use of health services and whether or not they attended a SAC for treatment. We also collected data on socio-demographic characteristics and other factors relating to the diagnosis and management of the condition. The survey questionnaire was developed with input from patient group representatives, a specialist ataxia neurologist, a specialist ataxia nurse, a health economist and representatives of pharmaceutical companies involved in ataxia research. The questionnaire mainly contained close-ended questions with defined response categories. Participants were provided with an information sheet about the survey, which included the purpose of the study, the organisations involved, and assurances around anonymity and aggregation of data for reporting purposes. The first part of the questionnaire explained the context of the study and the medical terms used. Participants were then asked a series of questions about their socio-demographic characteristics, including the type of ataxia they were affected by. The next section of the survey asked about experiences of diagnosis. This was followed by a series of questions about health service contacts of different types, and about whether or not the person affected by ataxia attended a SAC for some or all of their treatments, and if so, their experiences of this. Participants were then asked a series of questions about the treatment they received and their satisfaction with this. Finally, they were asked questions about their symptoms and the management of these. See Appendix 1 for the UK version of the survey. This version was revised and updated after being distributed in the UK, before being translated and distributed in Germany and Italy.

Participation in the survey was open to all patients with ataxia (or carers of patients with ataxia, who could act as proxy respondents), who were aged 16 years old or more. The ataxia patient associations in each country (Ataxia UK in the UK; Deutsche Heredo Ataxie Gesellschaft (DHAG) and Ataxie Forderverein e.V. in Germany; Associazione Italiana per la lotta alle Sindromi Atassiche (AISA) in Italy) publicised the surveys to potential participants via newsletters, social media, website and events. In the UK the survey was mainly distributed online via Ataxia UK’s mailing list, website, magazine and social media channels. In Germany and Italy the survey was publicised via the patients associations and also via clinicians working at SACs. In the UK the survey was submitted for ethical approval via the Integrated Research Application System (IRAS; reference 252,966) and received approval by the Cambridge Research Ethics Committee (REC; reference 19/EE/0030). As part of this process all materials related to the survey, including the patient information sheets and final questionnaire, were validated by a clinical expert and an informal review panel including ataxia patients, nominated by Ataxia UK. For the two other countries, ethical approval for an anonymised survey was not needed. The survey was ‘live’ from March to May 2019 in the UK, from February to October 2020 in Germany, and from May to September 2021 in Italy; note the different time periods.

Responses were removed for all respondents who did not provide informed consent or who did not provide positive responses to the three screening questions contained within survey (questions 3, 4 and 5; see Appendix 1). Where the respondent gave clear contradictory responses, the responses to those questions were removed from the analysis. Incomplete surveys were not removed from the analysis, as respondents chose to answer some questions and to skip others. The database used for cleaning and analysis was anonymised.

Participants were asked to record the number of health care contacts they had received in the preceding 6 (in the case of the UK) or 12 (in Germany and Italy) months specifically to manage their ataxia. Participants were asked to record the number of health care contacts with the general practitioner (GP), hospital outpatient clinic visits with a neurologist (not at a SAC), SAC visits, hospital inpatient stays, accident and emergency department visits, physiotherapy appointments, speech and language therapy appointments, occupational therapy appointments, and other consultant specialist visits (e.g. an ophthalmologist, an ear, nose and throat specialist, urologist, gastroenterologist and others). Figures for the UK were multiplied by two to give 12-month estimates, commensurate with the other two counties [11]. The wording in the UK version of the questionnaire could have been misunderstood, as respondents were asked to report the number visits to the specialist ataxia centre, and this may have been interpreted by respondents to mean the number of visits to see a specialist about their ataxia at the hospital where the SAC was, irrespective of the type of specialist seen or whether or not the visit was at a SAC. We know that according to routine practice in the UK people with ataxia who attended a SAC are invited to visit the centre once per year. Therefore, all UK respondents who reported attending the SAC were calculated to have one visit to the SAC each year and any additional visits that were reported were included in the economic analysis as other consultant specialist. Unit costs for each type of contact were obtained from local providers or published sources [12, 13]. All costs were calculated in 2021 €; 2021 unit costs for the UK were reported in UK£ and converted into € using GDP Purchasing Power Parities; unit costs for Germany and Italy were reported in €; prices were converted into 2021 values using consumer price indices for each country. Given the small numbers of respondents, and the variation in the numbers of respondents to each resource use question, we multiplied the unit cost by the mean volume of resource use for each type of resource use and summed these across all types of resource use to calculate mean annual treatment costs per patient per annum in each of the three countries.

Participants were also asked whether they were currently attending or had ever attended a SAC. Participants who reported they had never been to a SAC were grouped in the “non-SAC” group; those who reported they were currently attending a SAC for some or all of their treatment were grouped in the “SAC” group. A third group comprising people who used to go to a SAC for some or all of their treatment but no longer attend were not included in the analysis; the number of respondents in this group was small, and it was unclear whether their resource use and costs would be affected by their previous contact with the SAC.

We compared average resource use for each contact type, and costs, over a 12-month period, stratifying patients by whether they were currently attending a SAC or had never attended a SAC. We tested for significant differences in mean resource use and costs between the SAC/non-SAC groups in each country using unadjusted and adjusted ordinary least squares regression analysis, the latter controlling for age, sex, number of symptoms experienced as a result of ataxia, and whether or not the patient had comorbidities. We also tested for significant differences in mean contacts and costs between countries separately for non-SAC and SAC groups using adjusted ordinary least squares regression analysis, controlling for age, sex, number of symptoms experienced as a result of ataxia, and whether or not the patient had comorbidities.