Health data space nodes for privacy-preserving linkage of medical data to support collaborative secondary analyses.

Introduction: The potential for secondary use of health data to improve healthcare is currently not fully exploited. Health data is largely kept in isolated data silos and key infrastructure to aggregate these silos into standardized bodies of knowledge is underdeveloped. We describe the development, implementation, and evaluation of a federated infrastructure to facilitate versatile secondary use of health data based on Health Data Space nodes. Materials and methods: Our proposed nodes are self-contained units that digest data through an extract-transform-load framework that pseudonymizes and links data with privacy-preserving record linkage and harmonizes into a common data model (OMOP CDM). To support collaborative analyses a multi-level feature store is also implemented. A feasibility experiment was conducted to test the infrastructures potential for machine learning operations and deployment of other apps (e.g., visualization). Nodes can be operated in a network at different levels of sharing according to the level of trust within the network. Results: In a proof-of-concept study, a privacy-preserving registry for heart failure patients has been implemented as a real-world showcase for Health Data Space nodes at the highest trust level, linking multiple data sources including (a) electronical medical records from hospitals, (b) patient data from a telemonitoring system, and (c) data from Austria's national register of deaths. The registry is deployed at the tirol kliniken, a hospital carrier in the Austrian state of Tyrol, and currently includes 5,004 patients, with over 2.9 million measurements, over 574,000 observations, more than 63,000 clinical free text notes, and in total over 5.2 million data points. Data curation and harmonization processes are executed semi-automatically at each individual node according to data sharing policies to ensure data sovereignty, scalability, and privacy. As a feasibility test, a natural language processing model for classification of clinical notes was deployed and tested. Discussion: The presented Health Data Space node infrastructure has proven to be practicable in a real-world implementation in a live and productive registry for heart failure. The present work was inspired by the European Health Data Space initiative and its spirit to interconnect health data silos for versatile secondary use of health data.

SAINTENS: Self-Attention and Intersample Attention Transformer for Digital Biomarker Development Using Tabular Healthcare Real World Data.

BACKGROUND: Deep learning currently struggles with tabular data, but it can benefit from multimodal learning. SAINT is a deep learning model for tabular data on which we base our presented developments. OBJECTIVES: In this study, we introduce SAINTENS as a new deep learning method, specifically for the in healthcare predominant tabular real world data. METHODS: For this purpose, we compare SAINTENS with SAINT and the State of the Art Machine Learning methods for tabular data. We use tabular data from geriatrics to predict four different targets (dysphagia, pressure ulcers, decompensated heart failure and delirium). We determine the relevant feature sets and train the models on these sets. RESULTS: Both SAINTENS and SAINT models are at least on the same performance level as the current State of the Art (Gradient Boosting Decision Trees). CONCLUSION: In combination with multimodal learning SAINTENS and SAINT may be used on real world data comprising tabular, text and image data, for discovery and development of new digital biomarkers.

Digital algorithm-guided insulin therapy in home healthcare for elderly persons with type 2 diabetes: A proof-of-concept study.

GlucoTab@MobileCare, a digital workflow and decision support system with integrated basal and basal-plus insulin algorithm was investigated for user acceptance, safety and efficacy in persons with type 2 diabetes receiving home health care by nurses. During a three months study nine participants (five female, age 77 ± 10 years, HbA1c 60 ± 13 mmol/mol (study start) vs. 57 ± 12 mmol/mol (study end) received basal or basal-plus insulin therapy as suggested by the digital system. In total 95% of all suggested tasks (blood glucose (BG) measurements, insulin dose calculations, insulin injections) were performed according to the digital system. Mean morning BG was 171 ± 68 mg/dL in the first study month vs. 145 ± 35 mg/dL in the last study month, indicating a reduced glycemic variability of 33 mg/dL (standard deviation). No hypoglycemic episode < 54 mg/dL occurred. User's adherence was high and the digital system supported a safe and effective treatment. Larger scale studies are needed to confirm findings under routine care. German Clinical Trials Register ID: DRKS00015059.

Efficient and safe glycaemic control with basal-bolus insulin therapy during fasting periods in hospitalized patients with type 2 diabetes using decision support technology: A post hoc analysis.

AIM: To evaluate the efficacy and safety of basal-bolus insulin therapy in managing glycaemia during fasting periods in hospitalized patients with type 2 diabetes. MATERIALS AND METHODS: We performed a post hoc analysis of two prospective, uncontrolled interventional trials that applied electronic decision support system-guided basal-bolus (meal-related and correction) insulin therapy. We searched for fasting periods (invasive or diagnostic procedures, medical condition) during inpatient stays. In a mixed model analysis, patients' glucose levels and insulin doses on days with regular food intake were compared with days with fasting periods. RESULTS: Out of 249 patients, 115 patients (33.9% female, age 68.3 ± 10.3 years, diabetes duration 15.1 ± 10.9 years, body mass index 30.1 ± 5.4 kg/m2 , HbA1c 69 ± 20 mmol/mol) had 194 days with fasting periods. Mean daily blood glucose (BG) was lower (modelled difference [ModDiff]: -0.5 ± 0.2 mmol/L, P = .006), and the proportion of glucose values within the target range (3.9-10.0 mmol/L) increased on days with fasting periods compared with days with regular food intake (ModDiff: +0.06 ± 0.02, P = .005). Glycaemic control on fasting days was driven by a reduction in daily bolus insulin doses (ModDiff: -11.0 ± 0.9 IU, P < .001), while basal insulin was similar (ModDiff: -1.1 ± 0.6 IU, P = .082) compared with non-fasting days. Regarding hypoglycaemic events (BG < 3.9 mmol/L), there was no difference between fasting and non-fasting days (χ2 0.9% vs. 1.7%, P = .174). CONCLUSIONS: When using well-titrated basal-bolus insulin therapy in hospitalized patients with type 2 diabetes, the basal insulin dose does not require adjustment during fasting periods to achieve safe glycaemic control, provided meal-related bolus insulin is omitted and correction bolus insulin is tailored to glucose levels.

Diabetes Management According to Health Status in Older Adults with Type 2 Diabetes Staying in Geriatric Care Facilities.

BACKGROUND: About 25% of adults >70 years suffer from type 2 diabetes. Due to the heterogeneity of the geriatric population, guidelines emphasize the need to individualize glycemic goals and simplify treatment strategies with the main focus of avoiding hypoglycemia. The aim of this study was to assess glycemic control in patients with type 2 diabetes in geriatric care facilities based on their individual health status. METHODS: 170 medical records of older adults with type 2 diabetes in geriatric care facilities were retrospectively assessed (64.7% female, age 80 ± 9 years; glycated hemoglobin 6.8% ± 3.6% [51 ± 16 mmol/mol]; body mass index 27.9 ± 5.8 kg/m2). Based on the individual health status, patients were allocated to three groups (healthy n = 27, complex n = 86, and poor n = 57). RESULTS: The overall blood glucose (BG) value was highest in the poor health group with 188 ± 47 mg/dL (poor) vs 167 ± 42 mg/dL (complex) vs 150 ± 34 mg/dL (healthy). BG values of 1.6% (poor) vs 2.8% (complex) vs 1.5% (healthy) of patients were below 90 mg/dL. 36.8% (poor) vs 23.4% (complex) vs 18.5% (healthy) of patients received insulin as the main diabetes therapy, but of these only 14.3% (poor) vs 20% (complex) vs 40% (healthy) were treated with basal insulin. CONCLUSIONS: Overall, BG values were higher in the poor and complex health group. There were a few low BG values in all groups. Although recommended by international guidelines, basal insulin therapy with its low complexity and low hypoglycemic risk is still underused, especially in the poor health group. Therefore, simplification of diabetes therapy should be considered further.

Safe and Sufficient Glycemic Control by Using a Digital Clinical Decision Support System for Patients With Type 2 Diabetes in a Routine Setting on General Hospital Wards.

The aim was to investigate the applicability of a clinical decision support system in a real-world inpatient setting for patients with type 2 diabetes on general hospital wards.A total of 150 patients with type 2 diabetes requiring subcutaneous insulin therapy were treated with basal-bolus insulin therapy guided by a decision support system (GlucoTab) providing automated workflow tasks and suggestions for insulin dosing to health care professionals.By using the system, a mean daily blood glucose (BG) of 159 ± 32 mg/dL was achieved. 68.8% of measurements were in the target range (70 to <180 mg/dL). The percentage of BG values <40, <70, and ≥300 mg/dL was 0.02%, 2.2%, and 2.3%, respectively. Health care professionals' adherence to suggested insulin doses and workflow tasks was high (>93% and 91%, respectively).The decision support system facilitates safe and efficacious inpatient diabetes care by standardizing treatment workflow and providing decision support for basal-bolus insulin dosing.

Electronic Diabetes Management System Replaces Paper Insulin Chart: Improved Quality in Diabetes Inpatient Care Processes Due to Digitalization.

BACKGROUND: GlucoTab, an electronic diabetes management system (eDMS), supports healthcare professionals (HCPs) in inpatient blood glucose (BG) management at point-of-care and was implemented for the first time under routine conditions in a regional hospital to replace the paper insulin chart. METHOD: To investigate quality of the eDMS for inpatients with type 2 diabetes mellitus a monocentric retrospective before-after evaluation was conducted. We compared documentation possibilities by assessing a blank paper chart vs the eDMS user interface. Further quality aspects were compared by assessing filled-in paper charts (n = 106) vs filled-in eDMS documentation (n = 241). HCPs (n = 59) were interviewed regarding eDMS satisfaction. RESULTS: The eDMS represented an improvement of documentation possibilities by offering a more structured and comprehensive user interface compared to the blank paper chart. The number of good diabetes days averaged to a median value of four days in both groups (paper chart: 4.38 [0-7] vs eDMS: 4.38 [0-7] days). Median daily BG was 170 (117-297) mg/dL vs 168 (86-286) mg/dL and median fasting BG was 152 (95-285) mg/dL vs 145 (69-333) mg/dL, and 0.1% vs 0.4% BG values <54 mg/dL were documented. Diabetes documentation quality improved when using eDMS, for example, documentation of ordered BG measurement frequency (1% vs 100%) and ordered BG targets (0% vs 100%). HCPs stated that by using eDMS errors could be prevented (74%), and digital support of work processes was completed (77%). Time saving was noted by 8 out of 11 HCPs and estimated at 10-15 minutes per patient day by two HCPs. CONCLUSIONS: The eDMS completely replaced the paper chart, showed comparable glycemic control, was positively accepted by HCPs, and is suitable for inpatient diabetes management.

Study protocol for assessing the user acceptance, safety and efficacy of a tablet-based workflow and decision support system with incorporated basal insulin algorithm for glycaemic management in participants with type 2 diabetes receiving home health care: A single-centre, open-label, uncontrolled proof-of-concept study.

INTRODUCTION: Diabetes management can be especially complex for older adults who receive health care at home. Thus, international guidelines recommend basal-insulin regimens due to simpler handling and low hypoglycaemia risk. A basal-insulin algorithm (including basal-plus) was developed to also include participant's health status and subsequently implemented into a tablet-based workflow and decision support system, GlucoTab@MobileCare. This study protocol describes a proof-of-concept study to investigate user acceptance, safety and efficacy of the GlucoTab@MobileCare system in participants receiving home health care. METHODS: The open-label, single-centre, uncontrolled study will recruit a maximum of ten participants with insulin treated type-2-diabetes (age ≥18 years) who receive home health care. During a three month study period participants will receive basal- or basal-plus-insulin therapy once daily as suggested by the GlucoTab@MobileCare system. Statistical analysis will be conducted on an intention-to-treat basis. The primary endpoint is the percentage of tasks (BG measurements, insulin dose calculations, insulin injections) that were performed according to GlucoTab@MobileCare suggestions relative to the total of suggested tasks. Secondary endpoints include user acceptance, safety and efficacy parameters. The study was approved by the ethics committee and regulatory authorities. Before obtaining written informed consent, all participants will receive oral and written information about all aspects of the study. Results will be published in a peer-reviewed journal and at diabetes and geriatric conferences. DISCUSSION: Potential implications may be improved quality and safety of basal-insulin therapy in older adults as well as support for health-care-providers in daily routine including evidence-based knowledge. TRIAL REGISTRATION: German Clinical Trials Register (DRKS00015059).

GlucoTab-guided insulin therapy using insulin glargine U300 enables glycaemic control with low risk of hypoglycaemia in hospitalized patients with type 2 diabetes.

AIMS: To investigate efficacy, safety and usability of the GlucoTab system for glycaemic management using insulin glargine U300 in non-critically ill hospitalized patients with type 2 diabetes (T2D). MATERIALS AND METHODS: In this open, non-controlled single-arm pilot study, glycaemic control at the general ward of a tertiary care hospital was guided by a mobile decision support system (GlucoTab) for basal-bolus insulin dosing using the novel basal insulin analogue insulin glargine U300 for the first time. Glycaemic control was surveilled with capillary glucose measurements and continuous glucose monitoring (CGM). The primary endpoint was efficacy of glycaemic management, defined as the percentage of blood glucose measurements within the target range of 3.9 to 7.8 mmol/L. RESULTS: A total of 30 patients with T2D (12 female; age, 67 ± 11 years; HbA1c, 70 ± 26 mmol/mol; BMI, 31.8 ± 5.6 kg/m2 ; length of study, 8.5 ± 4.5 days) were included. In total, 894 capillary glucose values and 49 846 data points of CGM were available, of which 56.1% of all measured capillary glucose values and 54.3% of CGM values were within the target area (3.9-7.8 mmol/L). Overall capillary mean glucose was 8.5 ± 1.2 and 8.4 ± 1.2 mmol/L assessed by CGM. Time within glucose target improved continuously during the course of treatment, while time within hypoglycaemia (<3.9 mmol/L) decreased substantially. The GlucoTab-suggested total daily dose was accepted by staff in 97.3% of situations. CONCLUSIONS: Treatment with GlucoTab using insulin glargine U300 in hospitalized patients with T2D is effective and safe.

Development of a Protocol for Automated Glucose Measurement Transmission Used in Clinical Decision Support Systems Based on the Continua Design Guidelines.

BACKGROUND: A fast and accurate data transmission from glucose meter to clinical decision support systems (CDSSs) is crucial for the management of type 2 diabetes mellitus since almost all therapeutic interventions are derived from glucose measurements. OBJECTIVES: Aim was to develop a prototype of an automated glucose measurement transmission protocol based on the Continua Design Guidelines and to embed the protocol into a CDSS used by healthcare professionals. METHODS: A literature and market research was performed to analyze the state-of-the-art and thereupon develop, integrate and validate an automated glucose measurement transmission protocol in an iterative process. RESULTS: Findings from literature and market research guided towards the development of a standardized glucose measurement transmission protocol using a middleware. The interface description to communicate with the glucose meter was illustrated and embedded into a CDSS. CONCLUSION: A prototype of an interoperable transmission of glucose measurements was developed and implemented in a CDSS presenting a promising way to reduce medication errors and improve user satisfaction.

Feasibility and Design of an Electronic Surgical Safety Checklist in a Teaching Hospital: A User-Based Approach.

BACKGROUND: The Surgical Safety Checklist (SSC) is routinely used in operating rooms (OR) but its acceptance is low. One promising way to improve acceptance of the SSC and thus quality of patient care is digitalization. OBJECTIVE: To investigate how a digitalization of the SSC could be implemented in a teaching hospital. Based on the identified user requirements we designed a first user interface (UI). METHOD: We performed a literature review, identified user perceptions and requirements during 12 interviews including a standardized questionnaire in surgical departments at the University Hospital Graz (Austria). Subsequently a first prototype of a UI was designed. RESULTS: Seven different approaches for digital SSC were identified in literature. Our interviews showed that 90% of the participants had a positive attitude towards a digitalization of SSC. The most favoured version of a digitalized SSC was a tablet-based client-server system with integration in the EHR and projection on an OR monitor. CONCLUSION: Digitalization of the SSC is requested by medical and nursing personnel. Based on the identified user requirements we designed a process oriented UI of a digital SSC.

A Mobile Computerized Decision Support System to Prevent Hypoglycemia in Hospitalized Patients With Type 2 Diabetes Mellitus.

BACKGROUND: Diabetes management requires complex and interdisciplinary cooperation of health care professionals (HCPs). To support this complex process, IT-support is recommended by clinical guidelines. The aim of this article is to report on results from a clinical feasibility study testing the prototype of a mobile, tablet-based client-server system for computerized decision and workflow support (GlucoTab®) and to discuss its impact on hypoglycemia prevention. METHODS: The system was tested in a monocentric, open, noncontrolled intervention study in 30 patients with type 2 diabetes mellitus (T2DM). The system supports HCPs in performing a basal-bolus insulin therapy. Diabetes therapy, adverse events, software errors and user feedback were documented. Safety, efficacy and user acceptance of the system were investigated. RESULTS: Only 1.3% of blood glucose (BG) measurements were <70 mg/dl and only 2.6% were >300 mg/dl. The availability of the system (97.3%) and the rate of treatment activities documented with the system (>93.5%) were high. Only few suggestions from the system were overruled by the users (>95.7% adherence). Evaluation of the 3 anonymous questionnaires showed that confidence in the system increased over time. The majority of users believed that treatment errors could be prevented by using this system. CONCLUSIONS: Data from our feasibility study show a significant reduction of hypoglycemia by implementing a computerized system for workflow and decision support for diabetes management, compared to a paper-based process. The system was well accepted by HCPs, which is shown in the user acceptance analysis and that users adhered to the insulin dose suggestions made by the system.

Impact of errors in paper-based and computerized diabetes management with decision support for hospitalized patients with type 2 diabetes. A post-hoc analysis of a before and after study.

OBJECTIVE: Most preventable adverse drug events and medication errors occur during medication ordering. Medication order entry and clinical decision support are available on paper or as computerized systems. In this post-hoc analysis we investigated frequency and clinical impact of blood glucose (BG) documentation- and user-related calculation errors as well as workflow deviations in diabetes management. We aimed to compare a paper-based protocol to a computerized medication management system combined with clinical workflow and decision support. METHODS: Seventy-nine hospitalized patients with type 2 diabetes mellitus were treated with an algorithm driven basal-bolus insulin regimen. BG measurements, which were the basis for insulin dose calculations, were manually entered either into the paper-based workflow protocol (PaperG: 37 patients) or into GlucoTab(®)-a mobile tablet PC based system (CompG: 42 patients). We used BG values from the laboratory information system as a reference. A workflow simulator was used to determine user calculation errors as well as workflow deviations and to estimate the effect of errors on insulin doses. The clinical impact of insulin dosing errors and workflow deviations on hypo- and hyperglycemia was investigated. RESULTS: The BG documentation error rate was similar for PaperG (4.9%) and CompG group (4.0%). In PaperG group, 11.1% of manual insulin dose calculations were erroneous and the odds ratio (OR) of a hypoglycemic event following an insulin dosing error was 3.1 (95% CI: 1.4-6.8). The number of BG values influenced by insulin dosing errors was eightfold higher than in the CompG group. In the CompG group, workflow deviations occurred in 5.0% of the tasks which led to an increased likelihood of hyperglycemia, OR 2.2 (95% CI: 1.1-4.6). DISCUSSION: Manual insulin dose calculations were the major source of error and had a particularly strong influence on hypoglycemia. By using GlucoTab(®), user calculation errors were entirely excluded. The immediate availability and automated handling of BG values from medical devices directly at the point of care has a high potential to reduce errors. Computerized systems facilitate the safe use of more complex insulin dosing algorithms without compromising usability. In CompG group, missed or delayed tasks had a significant effect on hyperglycemia, while in PaperG group insufficient precision of documentation times limited analysis. The use of old BG measurements was clinically less relevant. CONCLUSION: Insulin dosing errors and workflow deviations led to measurable changes in clinical outcome. Diabetes management systems including decision support should address nurses as well as physicians in a computerized way. Our analysis shows that such systems reduce the frequency of errors and therefore decrease the probability of hypo- and hyperglycemia.

Standardized Glycemic Management with a Computerized Workflow and Decision Support System for Hospitalized Patients with Type 2 Diabetes on Different Wards.

BACKGROUND: This study investigated the efficacy, safety, and usability of standardized glycemic management by a computerized decision support system for non-critically ill hospitalized patients with type 2 diabetes on four different wards. MATERIALS AND METHODS: In this open, noncontrolled intervention study, glycemic management of 99 patients with type 2 diabetes (62% acute admissions; 41 females; age, 67±11 years; hemoglobin A1c, 65±21 mmol/mol; body mass index, 30.4±6.5 kg/m(2)) on clinical wards (Cardiology, Endocrinology, Nephrology, Plastic Surgery) of a tertiary-care hospital was guided by GlucoTab(®) (Joanneum Research GmbH [Graz, Austria] and Medical University of Graz [Graz, Austria]), a mobile decision support system providing automated workflow support and suggestions for insulin dosing to nurses and physicians. RESULTS: Adherence to insulin dosing suggestions was high (96.5% bolus, 96.7% basal). The primary outcome measure, percentage of blood glucose (BG) measurements in the range of 70-140 mg/dL, occurred in 50.2±22.2% of all measurements. The overall mean BG level was 154±35 mg/dL. BG measurements in the ranges of 60-70 mg/dL, 40-60 mg/dL, and <40 mg/dL occurred in 1.4%, 0.5%, and 0.0% of all measurements, respectively. A regression analysis showed that acute admission to the Cardiology Ward (+30 mg/dL) and preexisting home insulin therapy (+26 mg/dL) had the strongest impact on mean BG. Acute admission to other wards had minor effects (+4 mg/dL). Ninety-one percent of the healthcare professionals felt confident with GlucoTab, and 89% believed in its practicality and 80% in its ability to prevent medication errors. CONCLUSIONS: An efficacious, safe, and user-accepted implementation of GlucoTab was demonstrated. However, for optimized personalized patient care, further algorithm modifications are required.

Taking a Closer Look--Continuous Glucose Monitoring in Non-Critically Ill Hospitalized Patients with Type 2 Diabetes Mellitus Under Basal-Bolus Insulin Therapy.

BACKGROUND: Inpatient glucose management is based on four daily capillary blood glucose (BG) measurements. The aim was to test the capability of continuous glucose monitoring (CGM) for assessing the clinical impact and safety of basal-bolus insulin therapy in non-critically ill hospitalized patients with type 2 diabetes mellitus (T2DM). MATERIALS AND METHODS: Eighty-four patients with T2DM (age, 68±10 years; glycosylated hemoglobin, 72±28 mmol/mol; body mass index, 31±7 kg/m(2)) were treated with basal-bolus insulin. CGM was performed with the iPro(®)2 system (Medtronic MiniMed, Northridge, CA) and calibrated retrospectively. RESULTS: A remarkable consistency between CGM and BG measurements and therapy improvement was shown over the study period of 501 patient-days. The number of CGM and BG measurements (CGM/BG) in the range from 3.9-10 mmol/L increased from 67.7%/67.2% (on Day 1) to 77.5%/78.6% (on the last day) (P<0.04). The number of low glycemic episodes (3.3 to <3.9 mmol/L) during nighttime detected by CGM was 15-fold higher, and the number of episodes >13.9 mmol/L detected by CGM during night was 12.5-fold higher than the values from the BG measurements. Ninety-nine percent of data points were in the clinically accurate or acceptable Clarke Error Grid Zones A+B, and the relative numbers of correctly identified episodes of <3.9 and >13.9 mmol/L detected by CGM (sensitivity) were 47.3% and 81.5%, respectively. CONCLUSIONS: Our data exhibit a good agreement between overall CGM and BG measurements, but there were a high number of missed hypo- and hyperglycemic episodes with BG measurements, particularly during nighttime. Overall assessment of glycemic control using CGM is feasible, whereas the use of CGM for individualized therapy decisions needs further improvement.