The DiGAReal registry was approved by the Institutional Review Board of the Medical Faculty of the University of Erlangen-Nürnberg, Germany (Reg No. 22-113-1-B). All procedures followed relevant guidelines and regulations, including the Declaration of Helsinki. Informed consent was obtained from all individual respondents included in the study. Eligible participants included adults (aged ≥ 18 years) presenting to a German rheumatology department who had received a prescription for one of 12 eligible DiGAs (Cara Care, Deprexis, Esysta, HelloBetter-Schmerz, HelloBetter Stress, Kaia Rückenschmerzen, Nichtraucherhelden, Oviva, Selfapy, Somnio, Vivira, Zanadio). This DiGA pre-selection had the consent of all rheumatologists from the participating study centers, based on a discussion of previous DiGA experiences and results of a previous monocentric study [13]. These patients were informed about the registry by the local rheumatology staff and provided written informed consent prior to participation. Upon consent, each patient was assigned a unique registry code, which was used to complete electronic questionnaires at baseline (T0) and at the 3-month follow-up (T1). Demographic data, including age, sex, and disease, were entered into electronic case report forms by local staff using the same registry identification code.

All outcomes were self-reported by patients through electronic questionnaires. In accordance with DigiG requirements, these outcomes included DiGA usage, patient-perceived effectiveness, usability, and satisfaction. Both generic and DiGA-specific metrics were assessed to capture a comprehensive understanding of patient experiences and the impact of DiGA on their health.

The initial T0 questionnaire queried patient’s education status, and their ability (yes/no) to independently and actively participate in virtual meetings (Zoom, Teams, Google Meet). Patients were asked “How motivated are you to tackle the DiGA indication/symptoms?” on a 0–10 numeric rating scale (0 = not at all motivated; 10 = very motivated).

To enable an overarching comparison of DiGA effectiveness, patients completed a core common set of assessments, including the Patient Global Impression of Change (PGIC) questionnaire at T1 by answering “How have your symptoms changed since you started using the DiGA prescribed for your condition?” using one of seven answer options ranging from “very much worse” to”very much improved”. Additionally, at baseline patients answered an adapted version of the PGIC questionnaire: “How well do you think you will be doing with the DiGA compared to now, with regard to your respective therapy goal?” with the original seven answer options. A 3-month time frame (T1) was selected as it aligns with the typical prescription duration of DiGAs,

The German version of the 13-item Patient Activation Measure (PAM®; Phreesia, Inc., Wilmington, DL, USA) [18] was administered at T0 and T1 to evaluate patients' self-management capabilities. Using the official PAM® scoring algorithm the total raw score was converted to a standardized scale ranging from 0 to 100; this score was then categorized into four levels of patient activation, indicating varying degrees of engagement in self-care. The levels, as defined by Hibbard et al. [18], are: (1) recognizing the importance of an active role; (2) gaining confidence and knowledge to take action; (3) actively taking action; and (4) sustaining healthy behaviors even when facing setbacks. An individual score change of 3 to 4 points is considered to be clinically significant, marking the difference between engaging and not engaging in self-care behaviors [19, 20].

Patient health literacy was measured at T0 and T1 using the German 16-item short form of the European Health Literacy Questionnaire (HLS-EU-Q16-GER) [21]. Following standard procedures [22], individual responses were binarized by combining the two positive response categories (“very easy” and “fairly easy,” scored as 1) and the two negative response categories (“fairly difficult” and “very difficult,” scored as 0). The general health literacy score was then calculated as the sum of the 16 binary items. Scores were categorized into three levels: “sufficient” (score 13–16), “problematic” (score 9–12), and “inadequate” (score 1–8).

Additionally, patients completed four 0–10 numeric rating scales at T0 and T1, evaluating current overall health status (0 = worst imaginable health status, 10 = best imaginable health status); level of exhaustion (fatigue) in the past week (0 = no exhaustion, 10 = worst imaginable level of exhaustion); level of pain in the past week (0 = no pain, 10 = worst imaginable pain); and level of disease activity of the primary rheumatic disease in the past week (0 = no disease activity, 10 = very high disease activity).

To assess the indication-specific DiGA effectiveness, each patient completed electronic patient reported outcomes (ePROs), which had been used as the primary outcomes in the respective prior DiGA approval studies (see Electronic Supplementary Material [ESM] Table S1).

At T1, patients reported their average DiGA usage, with frequency options ranging from “not at all” to “daily.” Additionally, the extent of use was queried with options ranging from “never really used it” to “completed the whole program.” The validated 6-item German Telehealth Usability and Utility Short Questionnaire (TUUSQ) [23], which is an adapted, shortened and translated version of the Telehealth Usability Questionnaire (TUQ) [24], was utilized to evaluate DiGA healthcare utility, usability and satisfaction at T1. Patients rated the individual items using on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

Statistical analyses were performed using Microsoft Excel 2019 (Microsoft Corp., Redmond, WA, USA) and GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). A threshold of p < 0.05 was considered to be statistically significant. Descriptive statistics for patient comparisons were presented as the median and interquartile range (IQR, defined as the 25th and 75th percentiles) for continuous variables, and as absolute numbers (n) and percentages for categorical variables. Differences between groups were assessed using the Mann–Whitney U-test for non-parametric data.