SBStefan Bilz

Stefan Bilz

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Stefan Bilz has a total of 482 co-authors. Here are the top co-authors by number of shared publications:

  • Michael Brändle

    Internal Medicine & Endocrinology/Diabetes, Kantonsspital St.Gallen, Switzerland

    8 shared publications
  • Drahomir Aujesky

    Department of General Internal Medicine, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland

    6 shared publications
  • Jonas Rutishauser

    Internal Medicine, Kantonsspital Baden, Switzerland

    6 shared publications
  • Michael Brändle

    Division of Endocrinology and Diabetes, Cantonal Hospital St.Gallen, HOCH Health Ostschweiz

    6 shared publications
  • Pascal Tribolet

    Medical University Department, Division of General Internal and Emergency Medicine, Kantonsspital Aarau, Aarau, Switzerland; Department of Health Professions, Bern University of Applied Sciences, Bern, Switzerland

    6 shared publications
  • Philipp Schuetz

    Medical University Department, Division of General Internal and Emergency Medicine, Kantonsspital Aarau, Aarau, Switzerland; Medical Faculty of the University of Basel, Basel, Switzerland

    6 shared publications

Latest Publications

  • Open Access

    Reminder Strategies to Improve Meal-Logging Adherence: Protocol for a Microrandomized Trial

    by Magdalena Fuchs, Victoria Brügger, Qiuhan Jin, Benjamin Wirth, Stefan Bilz, Florian von Wangenheim, Michael Brändle, Tobias Kowatsch, Mia Jovanova

    Background: Accurate measurement of lifestyle factors is central to understanding how daily behaviors act as risk factors or protective buffers to noncommunicable diseases. While wearable devices enable passive monitoring of physical activity or sleep, nutritional intake still depends on active participant input, such as manual dietary logs or image-based recordings. Adherence to such logging tasks often declines rapidly, impacting data completeness and clinical utility. Theory-based reminders drawing on loss-framing or logging consistency feedback (ie, tracking streaks) can improve adherence to lifestyle data collection, but the effects of these strategies on repeated dietary logging remain unclear. Objective: The study aims to examine the effects of 2 theory-driven reminder strategies—loss-framing and logging consistency—on adherence to repeated, image-based meal logging. Methods: We conducted a microrandomized trial embedded within a 4-week observational lifestyle phenotyping study in Switzerland (N=200, age ≥45 y, BMI ≥25 kg/m²). Participants photographed their meals at each mealtime (breakfast, lunch, and dinner) using a mobile app over 28 days. A decision point was scheduled prior to participant-defined habitual mealtimes. Participants were randomly assigned with equal probability to (1) a reminder emphasizing loss of a daily financial reward for not logging (“loss-framing”), (2) a reminder providing feedback on recent logging consistency (“logging consistency”), or (3) a neutral reminder (“active control”). The proximal outcome is whether the participant logs a meal within 2 hours of receiving a reminder. Participants earned a daily financial reward contingent on meal-logging completion. To estimate intervention effects, we will use marginal excursion effect models for binary outcomes, adjusting time-varying covariates (eg, day in study) and baseline covariates (eg, age and gender). Ethics approval for this study was granted by the Cantonal Ethics Committee of Eastern Switzerland (BASEC ID 2025‐00972). Results: Enrollment began in November 2025, and the study was initiated on December 8, 2025. As of August 2026, 130 participants have been recruited, 120 have been enrolled, and 76 participants have completed the study, with an anticipated completion date of January 2027. Data analysis has not yet begun, and study results are expected to be published in Q2 2027. Conclusions: By clarifying the proximal effects of loss-framed and consistency-based reminders, findings will inform the design of future digital health studies to improve meal-logging adherence in daily life. This work contributes to the development of scalable, theory-driven reminder strategies for enhancing dietary data quality in observational and interventional research.


  • Open Access

    Reminder Strategies to Improve Meal-Logging Adherence: Micro-Randomized Trial Protocol (Preprint)

    by Magdalena Fuchs, Victoria Brügger, Qiuhan Jin, Benjamin Wirth, Stefan Bilz, Florian von Wangenheim, Michael Brändle, Tobias Kowatsch, Mia Jovanova

    BACKGROUND Accurate measurement of lifestyle factors is central to understanding how daily behaviors act as risk factors or protective buffers to non-communicable diseases. While wearable devices enable passive monitoring of physical activity or sleep, nutritional intake still depends on active participant input such as manual dietary logs or image-based recordings. Adherence to such logging tasks often declines rapidly, impacting data completeness and clinical utility. Theory-based reminders drawing on loss-framing or logging consistency feedback (i.e., tracking streaks) can improve adherence to lifestyle data collection, but effects of these strategies on repeated dietary logging remain unclear. OBJECTIVE To examine examines the effects of two theory-driven reminder strategies, loss-framing and logging consistency, on adherence to repeated, image-based meal-logging. METHODS We employ a micro-randomized trial (MRT) embedded within a 4-week observational lifestyle phenotyping study in Switzerland (N=200, age ≥45, BMI ≥25 kg/m²). Participants photograph their meals at each mealtime (breakfast, lunch, dinner) using a mobile app over 28 days. A decision point is scheduled prior to participant-defined habitual mealtimes. Participants are randomly assigned with equal probability to: (1) a reminder emphasizing loss of a daily financial reward for not logging ("loss-framing"), (2) a reminder providing feedback on recent logging consistency ("logging consistency"), or (3) a neutral reminder ("active control"). The proximal outcome is whether the participant logs a meal within two hours of receiving a reminder. Participants earn a daily financial reward contingent on meal logging completion. To estimate intervention effects, we will use marginal excursion effect models for binary outcomes, adjusting time-varying covariates (e.g., day in study) and baseline covariates (e.g., age, gender). Ethics approval for this study was granted by the Cantonal Ethics Committee of Eastern Switzerland (BASEC ID: 2025-00972). RESULTS Enrollment began in November 2025, and the study was initiated on December 8, 2025, with an anticipated completion date of January 2027. CONCLUSIONS By clarifying the proximal effects of loss-framed and consistency-based reminders, findings will inform the design of future digital health studies to improve meal-logging adherence in daily life. This work contributes to the development of scalable, theory-driven reminder strategies for enhancing dietary data quality in observational and interventional research. CLINICALTRIAL ClinicalTrials.gov NCT07555262; https://clinicaltrials.gov/study/NCT07555262 ClinicalTrials.gov NCT07373418; https://clinicaltrials.gov/ct2/show/NCT07373418

  • Open Access

    GLOW UP Study: Protocol for an Observational Digital Biomarker Study for Prediabetes Screening and Digital Phenotyping

    by Victoria Bruegger, Magdalena Fuchs, Qiuhan Jin, Benjamin Wirth, Stefan Bilz, Michael Braendle, Tobias Kowatsch, Mia Jovanova

    Introduction: Prediabetes, a key precursor to type 2 diabetes, is highly prevalent and underdiagnosed, particularly among adults aged ≥45 years with elevated body mass index (BMI). Early detection is critical because lifestyle interventions can delay or prevent progression to type 2 diabetes. The Glow Up (GLucose Observation and Wearable Use for Prevention) study aims to (1) test the feasibility of a digital biomarker for prediabetes screening using wearable– and smartphone-derived lifestyle factors (e.g., sleep, physical activity, and nutrition patterns), in daily life and to (2) characterize individual– and metabolic subgroup-level variability in lifestyle factors and glycemic control. Specifically, we aim to examine how lifestyle factors relate to diabetes risk and identify personalized predictors of early metabolic dysregulation. Methods and analysis: Glow Up is a prospective, single-center, observational case-control study conducted in Switzerland. Adults (N=200) aged ≥45 years with BMI ≥25 kg/m² will be recruited, including n=100 individuals with prediabetes and n=100 age– and sex-matched case-control normoglycemic controls. Participants will undergo four weeks of continuous monitoring using a blinded continuous glucose monitor (CGM), and commercial– and medical-grade wearables; e.g. capturing physical activity, sleep, and physiological markers (heart rate variability, heart rate and skin temperature); in addition to completing daily image-based meal logs, using smartphones. Glycated haemoglobin (HbA1c), fasting plasma glucose (FPG) and body anthropometrics will be collected at baseline and follow-up, four weeks apart. Primary outcomes include HbA1c and FPG, measured at approximately four week follow-up. Secondary outcomes include CGM metrics, lifestyle profiling (sleep, physical activity, stress, and nutrition), and adherence to image-based meal logging. Ethics and dissemination: The study has received ethics approval from the Ethics Committee of Eastern Switzerland (BASEC ID.: 2025-00972). Results will be published in international peer-reviewed journals and at national and international conferences as posters, presentations, and articles. Summaries will be provided to the funders and personalized reports to participants. Trial registration number: NCT07373418 Article Summary: Strengths and limitations of this study: Glow Up presents the first study in Switzerland to investigate glycaemic and lifestyle profiling among individuals with normoglycaemia and prediabetes in an observational, free-living setting. To our knowledge, Glow Up is the first dataset to combine continuous glucose monitoring, commercial– and medical-grade wearable sensors, smartphone-based questionnaires, image-based dietary logging under free-living conditions, and clinical biomarkers, HbA1c and FPG; allowing for rich digital phenotyping. Inclusion of baseline and follow-up clinical biomarkers (HbA1c and FPG) provides clinically validated ground truth measures, strengthening outcome validity and anchoring digital and wearable-derived markers to established clinical, diagnostic standards. The absence of direct measures of insulin resistance, lipid metabolism, and gut microbiome compositions may limit more mechanistic investigation of glucose dysregulation patterns and metabolic sub phenotyping. The four-week follow-up period may limit inference on longer-term glycaemic trajectories and progression to T2D. Findings may not be generalisable beyond at-risk adults living in Switzerland who meet the study eligibility criteria.


  • Open Access

    Early health technology assessment of digital diabetes screening in Switzerland: cost-effectiveness and budget impact analyses

    by Wasu Mekniran, Victoria Bruegger, Magdalena Fuchs, Qiuhan Jin, Benjamin Wirth, Stefan Bilz, Michael Braendle, Elgar Fleisch, Tobias Kowatsch, Mia Jovanova

    Objectives: Digital biomarkers offer scalable screening for type 2 diabetes, yet adoption is stalled by uncertainty regarding economic viability. This study evaluates the cost-effectiveness and budget impact of digital screening compared to opportunistic screening from a Swiss payer perspective. Methods: A probabilistic Markov cohort model was developed to simulate at-risk Swiss adults (age ≥45, BMI ≥25 kg/m²) over a 40-year horizon. The model incorporates a digital attrition parameter, inputs derived from Swiss-specific sources (e.g., the CoLaus study and FSO life tables), and statutory tariffs. Costs and outcomes were discounted at 3.0%. Results: In the deterministic base-case, digital screening yielded an incremental cost-effectiveness ratio of CHF 2,912 per quality-adjusted life-year gained. Probabilistic sensitivity analysis indicated a 93.2% probability of cost-effectiveness at the CHF 50,000 threshold. The budget impact analysis estimated a Year 1 gross investment budget of CHF 27 million to identify prevalent cases, followed by long-term savings from averted complications. Conclusions: Digital screening can be highly cost-effective in Switzerland. While the required Year 1 gross investment poses a liquidity challenge, reimbursement via pathway-oriented models under the Swiss tariff could align incentives with long-term complication avoidance.


  • Open Access

    Correction to: Safety and Effectiveness of Glargine 300 U/ml After Switching from Basal Insulins in Patients with Type 1 Diabetes: COMET-T Study

    by Stefan Gölz, Julia K. Mader, Stefan Bilz, Julia Kenzler, Thomas Danne

  • Open Access

    Safety and Effectiveness of Glargine 300 U/ml After Switching from Basal Insulins in Patients with Type 1 Diabetes: COMET-T Study

    by Stefan Gölz, Julia K. Mader, Stefan Bilz, Julia Kenzler, Thomas Danne

    Introduction Appropriate glycemic control is paramount for people with type 1 diabetes (PwT1D) by the effective delivery of exogenous insulin. However, glycemic variability and the risk of severe hypoglycemia must be reliably controlled. Methods COMET-T is a prospective, multicenter, observational study conducted in Germany, Austria, and Switzerland during 2021–2022 to assess the effectiveness and safety of insulin glargine 300 U/ml (Gla-300) after switching from other basal insulins. Out of 135 PwT1D, data of 94 patients were analyzed. The primary endpoint was the change in time in range (TIR) approximately 12 and 24 weeks after switching to Gla-300. Secondary endpoints were: change in HbA1c, fasting plasma glucose (FPG), coefficient of variation (CV%) of plasma glucose, body weight (BW) and insulin dose. Results Patients had mean age of 48.6 ± 16.5 years, included 39.4% males and had 18.2 ± 15.5 years T1D duration. From baseline (54.3%), TIR changed at week 12 (mean change 0.3% [± 14.3]; p = 0.8383) and at week 24 (+ 4.5% [± 14.9], p = 0.078). At week 24, TIR significantly increased in patients with body mass index > 30 kg/m2 (8.4% [± 12.8] p = 0.0057) and patients who previously received insulin detemir (10.5%; [± 12.93]; p = 0.0005). At week 24, there was a significant reduction in the HbA1c value (8.1 ± 0.6% vs. 7.7 ± 0.9%; p < 0.001), a reduction in the CV% of plasma glucose (36.1 ± 12.4% vs. 32.8 ± 9.6%, p = 0.056), and increase in bolus insulin dose (26.5 ± 16.3 vs. 27.9 ± 16.6 U/day; p = 0.042). FPG, BW, and basal insulin doses were not significantly changed. Conclusions Although switching to Gla-300 in poorly controlled PwT1D did not significantly reduce TIR, it significantly decreased HbA1c values and glycemic variability without changes in BW and basal insulin dose.


  • Open Access

    Tracking Real-World Physical Activity in Chronic Obstructive Pulmonary Disease Over One Year: Results from a Monocentric, Prospective, Observational Cohort Study

    by Maximilian Boesch, Florent Baty, Stefan Bilz, Martin H Brutsche, Frank Rassouli

    Introduction: Lung function constraints and comorbidities such as coronary heart disease, sarcopenia, and mood disorders make chronic obstructive pulmonary disease (COPD) patients avoid physical activity (PA). However, PA represents an important pillar of COPD management and is explicitly recommended by professional associations to enhance physical functioning and positively modulate disease progression. Methods: In this monocentric, prospective, observational feasibility study, it was our primary objective to investigate the association between PA and the evolution of the COPD assessment test (CAT) and the occurrence of acute exacerbations of COPD (AECOPD), respectively. To this end, we equipped 42 COPD patients with an activity tracking wearable and telemonitored their daily PA levels over one year using a dedicated web-based interface. Patients additionally provided weekly CAT scores using the same telehealth platform and came in for 3 study visits to assess functional parameters and biochemical markers related to nutrition and inflammation. Results: A principal study finding was that PA was inversely associated with CAT score (drop of 0.21 points associated with an increase of 1000 daily steps, p = 0.004), and that the 50% of patients with higher PA levels showed less CAT score progression over time (0.42 points per year) than the 50% of patients with lower PA levels (3.26 points per year) (p < 0.001). In addition, higher PA levels were significantly associated with a lower likelihood of experiencing a moderate-to-severe AECOPD (31% risk reduction associated with an increase of 1000 daily steps, p = 0.0097). Discussion: Our study demonstrates the relevance of PA for key COPD outcome metrics in a real-world setting and underpins the importance of PA for COPD self-management in everyday life. Our study paves the way for future intervention trials to prospectively identify medically relevant PA thresholds and establish training recommendations for different patient subgroups.


  • Umstellung anderer Basalinsuline (BI) auf Glargin 300 E/ml (Gla-300) bei Patienten mit kontinuierlicher Glukosemessung (CGM) und unzureichend kontrolliertem Typ-1-Diabetes (T1D): Ergebnisse einer prospektiven, multinationalen Beobachtungsstudie (COMET-T)

    by Stefan Gölz, Julia K. Mader, Stefan Bilz, Daniel Heinze, Julia Müller, Thomas Danne

    Eine hohe Zeit im Zielbereich (TIR; 70 – 180 mg/dl) ist für T1D-Patienten von entscheidender Bedeutung. Dieser stehen jedoch Glukosevariabilität und das Risiko schwerer Hypoglykämien gegenüber. In COMET-T wurden Wirksamkeit und Sicherheit der Umstellung anderer BI auf Gla-300 innerhalb einer Basal-Bolus-Therapie (BBT) im klinischen Alltag bei CGM-überwachten T1D-Erwachsenen ausgewertet. Primärziel war die Veränderung der TIR 12 Wochen (W12) und 24 Wochen (W24) nach Behandlungsbeginn mit Gla-300. Sekundärziele: Änderungen beim HbA1c, Nüchternblutzucker (NBZ), Variationskoeffizient der Blutglukose (VK%), Körpergewicht (KG) und BI-Dosis. Insgesamt wurden 94 Teilnehmer (Alter: 48,6±16,5 Jahre, männlich: 39,4%, T1D-Dauer: 18,2±15,5 Jahre) analysiert. Die TIR unterschied sich bei W12 oder W24 nicht wesentlich vom Ausgangswert (BL). Im Zeitverlauf zeigte sich ein Trend zu einem Anstieg der TIR (BL: 54,3±18,4% vs. W24: 58,8±19,9%, p=0,078) und einem verringerten VK% (BL: 36,1±12,4% vs. W24: 32,8±9,6%, p=0,056). Die Umstellung auf Gla-300 was assoziiert mit einer signifikanten Senkung des HbA1c-Wertes (BL: 8,1±0,6% vs. W24: 7,7±0,9%; p<0,001) und einer Erhöhung der Bolusinsulindosis (BL: 26,5±16,3 E/Tag vs. W24: 27,9±16,6 E/Tag; p=0,042), ohne signifikante Änderung des NBZ, des KG oder der BI-Dosis. Zusammenfassend führte die Umstellung des BI auf Gla-300 innerhalb einer BBT bei sub-optimal eingestellten T1D-Patienten im klinischen Alltag zu einer signifikanten Senkung des HbA1c-Wertes mit einem Trend zu einer TIR-Erhöhung, einer VK%-Senkung ohne Änderung der BI-Dosis. Unterstützt von Sanofi-Aventis Deutschland GmbH. Article published online: 02 May 2023 © 2023. Thieme. All rights reserved. Georg Thieme Verlag Rüdigerstraße 14, 70469 Stuttgart, Germany

  • Prospective validation of five malnutrition screening and assessment instruments among medical inpatients: Secondary analysis of a randomized clinical trial

    by Lena Stalder, Nina Kaegi-Braun, Carla Gressies, Claudia Gregoriano, Pascal Tribolet, Dileep N. Lobo, Filomena Gomes, Claus Hoess, Vojtech Pavlicek, Stefan Bilz, et al.