Background: Individuals with type 2 diabetes (T2D) are at increased risk of developing cardiovascular disease (CVD) which necessitates monitoring of risk factors and appropriate pharmacotherapy. This study aimed to identify factors predicting emergency department visits, hospitalizations, and mortality among T2D patients after being newly diagnosed with CVD.

Methods: In a retrospective observational study conducted in Region Halland, individuals aged > 40 years with T2D diagnosed between 2011 and 2019, and a new diagnosis of CVD between 2016 and 2019, were followed for one year from the date of CVD diagnosis. The first encounter for CVD diagnosis was categorized as inpatient-, outpatient-, primary-, or emergency department care. Follow-up included laboratory tests, blood pressure, pharmacotherapies, and healthcare utilization. Hazard ratios (HR) in two Cox regression analyses determined relative risks for emergency visits/hospitalization and mortality, adjusting for age, sex, glucose regulation, lipid levels, kidney function, blood pressure, pharmacotherapy, and healthcare utilization.

Results: The study included a total of 1759 T2D individuals who received a new CVD diagnosis, with 67% diagnosed during inpatient care. The average hospitalization stay was 6.5 days, and primary care follow-up averaged 10.1 visits. Patients with CVD diagnosed in primary care had a HR 0.52 (confidence interval [CI] 0.35–0.77) for emergency department visits/hospitalization, but age had a HR 1.02 (CI 1.00-1.03). Pharmacotherapy with insulin, DPP4-inhibitors, aldosterone antagonists, and beta-blockers had a raised HR. Highest mortality risk was observed when CVD was diagnosed inpatient care, systolic blood pressure < 100 mm Hg and elevated HbA1c. Age had a HR 1.05 (CI 1.03–1.08), eGFR < 30 ml/min HR 1.46 (CI 1.01–2.11), and LDL-Cholesterol > 2,5 h 1.46 (CI 1.01–2.11) and associated with increased mortality risk. Pharmacotherapy with metformin had a HR 0.41 (CI 0.28–0.62), statins a HR 0.39 (CI 0.27–0.57), and a primary care follow-up < 30 days a HR 0.53 (CI 0.37–0.77) and associated with lower mortality risk.

Conclusions: T2D individuals who had a new diagnosis of CVD were predominantly diagnosed when hospitalized, while follow-up typically occurred in primary care. Identifying factors that predict risks of mortality and hospitalization should be a focus of follow-up care, underscoring the critical role of primary care in the effective management of T2D and CVD. © The Author(s) 2024.

In the latent bandit problem, the learner has access to reward distributions and – for the non-stationary variant – transition models of the environment. The reward distributions are conditioned on the arm and unknown latent states. The goal is to use the reward history to identify the latent state, allowing for the optimal choice of arms in the future. The latent bandit setting lends itself to many practical applications, such as recommender and decision support systems, where rich data allows the offline estimation of environment models with online learning remaining a critical component. Previous solutions in this setting always choose the highest reward arm according to the agent’s beliefs about the state, not explicitly considering the value of information-gathering arms. Such information-gathering arms do not necessarily provide the highest reward, thus may never be chosen by an agent that chooses the highest reward arms at all times.

In this paper, we present a method for information-gathering in latent bandits. Given particular reward structures and transition matrices, we show that choosing the best arm given the agent’s beliefs about the states incurs higher regret. Furthermore, we show that by choosing arms carefully, we obtain an improved estimation of the state distribution, and thus lower the cumulative regret through better arm choices in the future. Through theoretical analysis we show that the proposed method retains the sub-linear regret rate of previous methods while having much better problem dependent constants. We evaluate our method on both synthetic and real-world data sets, showing significant improvement in regret over state-of-the-art methods. © 2022 The Author(s). 

This thesis presents work conducted in the domain of sequential decision-making in general and Bandit problems in particular, tackling challenges from a practical and theoretical perspective, framed in the contexts of mobile Health. The early stages of this work have been conducted in the context of the project ``improving Medication Adherence through Person-Centred Care and Adaptive Interventions'' (iMedA) which aims to provide personalized adaptive interventions to hypertensive patients, supporting them in managing their medication regimen. The focus lies on inadequate medication adherence (MA), a pervasive issue where patients do not take their medication as instructed by their physician. The selection of individuals for intervention through secondary database analysis on Electronic Health Records (EHRs) was a key challenge and is addressed through in-depth analysis of common adherence measures, development of prediction models for MA, and discussions on limitations of such approaches for analyzing MA. Providing personalized adaptive interventions is framed in several bandit settings and addresses the challenge of delivering relevant interventions in environments where contextual information is unreliable and full of noise. Furthermore, the need for good initial policies is explored and improved in the latent-bandits setting, utilizing prior collected data to optimal selection the best intervention at every decision point. As the final concluding work, this thesis elaborates on the need for privacy and explores different privatization techniques in the form of noise-additive strategies using a realistic recommendation scenario.         

The contributions of the thesis can be summarised as follows: (1) Highlighting the issues encountered in measuring MA through secondary database analysis and providing recommendations to address these issues, (2) Investigating machine learning models developed using EHRs for MA prediction and extraction of common refilling patterns through EHRs, (3) formal problem definition for a novel contextual bandit setting with context uncertainty commonly encountered in Mobile Health and development of an algorithm designed for such environments. (4) Algorithmic improvements, equipping the agent with information-gathering capabilities for active action selection in the latent bandit setting, and (5) exploring important privacy aspects using a realistic recommender scenario.   

Background: The development of diabetes technology is rapid and requires education and resources to be successfully implemented in diabetes care management.

Method: In an observational study, we evaluated the use of advanced diabetes technology, resource utilization, and glycemic control. The study population was 725 individuals with type 1 diabetes (T1D) living in Region Halland, Sweden. The study cohort was followed for 7 years between 2013 and 2019.

Results: Children aged 0 to 17 years were associated with significantly better glucose control than young adults aged 18 to 25 years. The mean HbA1c in children and young adults was 53 mmol/mol (7.0%) compared to 61 mmol/mol (7.7%) (P <.0001), respectively. Comorbidities such as attention deficit hyperactivity disorder (ADHD), autism, and coelic disease were associated with higher HbA1c. All groups, regardless of age and comorbidity, showed a positive effect on glucose control after visiting a dietitian or psychologist. Differences were found between the age groups in terms of more use of advanced diabetes technology and more frequent visits to a physician in children compared to young adults.

Conclusions: More frequent visits to physicians, and a visit to dietitians, and psychologists were associated with improved glucose control in individuals with T1D 0 to 25 years. Increased resources, including access to more advanced technologies, may be required in young adults with T1D. © 2022 Diabetes Technology Society.

Medical and technological advancement in the last century has led to the unprecedented increase of the populace's quality of life and lifespan. As a result, an ever-increasing number of people live with chronic health conditions that require long-term treatment, resulting in increased healthcare costs and managerial burden to the healthcare provider. This increase in complexity can lead to ineffective decision-making and reduce care quality for the individual while increasing costs. One promising direction to tackle these issues is the active involvement of the patient in managing their care. Particularly for chronic diseases, where ongoing support is often required, patients must understand their illness and be empowered to manage their care. With the advent of smart devices such as smartphones, it is easier than ever to provide personalised digital interventions to patients, help them manage their treatment in their daily lives, and raise awareness about their illness. If such new approaches are to succeed, scalability is necessary, and solutions are needed that can act autonomously without costly human intervention. Furthermore, solutions should exhibit adaptability to the changing circumstances of an individual patient's health, needs and goals. Through the ongoing digitisation of healthcare, we are presented with the unique opportunity to develop cost-effective and scalable solutions through Artificial Intelligence (AI).

This thesis presents work that we conducted as part of the project improving Medication Adherence through Person-Centered Care and Adaptive Interventions (iMedA) that aims to provide personalised adaptive interventions to hypertensive patients, supporting them in managing their medication regiment. The focus lies on inadequate medication adherence (MA), a pervasive issue where patients do not take their medication as instructed by their physician. The selection of individuals for intervention through secondary database analysis on Electronic Health Records (EHRs) was a key challenge and is addressed through in-depth analysis of common adherence measures, development of prediction models for MA and discussions on limitations of such approaches for analysing MA. Furthermore, providing personalised adaptive interventions is framed in the contextual bandit setting and addresses the challenge of delivering relevant interventions in environments where contextual information is significantly corrupted.       

The contributions of the thesis can be summarised as follows: (1) Highlighting the issues encountered in measuring MA through secondary database analysis and providing recommendations to address these issues, (2) Investigating machine learning models developed using EHRs for MA prediction and extraction of common refilling patterns through EHRs and (3) formal problem definition for a novel contextual bandit setting with context uncertainty commonly encountered in Mobile Health and development of an algorithm designed for such environments.  

Background: There is a strong need to improve medication adherence (MA) for individuals with hypertension in order to reduce long-term hospitalization costs. We believe this can be achieved through an artificial intelligence agent that helps the patient in understanding key individual adherence risk factors and designing an appropriate intervention plan. The incidence of hypertension in Sweden is estimated at approximately 27%. Although blood pressure control has increased in Sweden, barely half of the treated patients achieved adequate blood pressure levels. It is a major risk factor for coronary heart disease and stroke as well as heart failure. MA is a key factor for good clinical outcomes in persons with hypertension.

Objective: The overall aim of this study is to design, develop, test, and evaluate an adaptive digital intervention called iMedA, delivered via a mobile app to improve MA, self-care management, and blood pressure control for persons with hypertension.

Methods: The study design is an interrupted time series. We will collect data on a daily basis, 14 days before, during 6 months of delivering digital interventions through the mobile app, and 14 days after. The effect will be analyzed using segmented regression analysis. The participants will be recruited in Region Halland, Sweden. The design of the digital interventions follows the just-in-time adaptive intervention framework. The primary (distal) outcome is MA, and the secondary outcome is blood pressure. The design of the digital intervention is developed based on a needs assessment process including a systematic review, focus group interviews, and a pilot study, before conducting the longitudinal interrupted time series study.

Results: The focus groups of persons with hypertension have been conducted to perform the needs assessment in a Swedish context. The design and development of digital interventions are in progress, and the interventions are planned to be ready in November 2020. Then, the 2-week pilot study for usability evaluation will start, and the interrupted time series study, which we plan to start in February 2021, will follow it.

Conclusions: We hypothesize that iMedA will improve medication adherence and self-care management. This study could illustrate how self-care management tools can be an additional (digital) treatment support to a clinical one without increasing burden on health care staff. © Kobra Etminani, Carina Göransson, Alexander Galozy, Margaretha Norell Pejner, Sławomir Nowaczyk.

Driver monitoring is expected to contribute greatly to safety in nascent smart cities, also in complex, mixed-traffic scenarios with autonomous vehicles (AVs), vulnerable road users (VRUs), and manually driven vehicles. Until now, one focus has been on detecting bio signals during the relatively short time when a person is inside a vehicle; but, life outside of the vehicle can also affect driving. For example, loneliness, depression, and sleep-deprivation, which might be difficult to detect in time, can increase the risk of accidents–raising possibilities for new and alternative intervention strategies. Thus, the current conceptual paper explores one idea for how continuous care could be provided to improve drivers’ mental states; in particular, the idea of a “robot” that could positively affect a driver’s health through interactions supported by social media mining on Facebook. A speculative design approach is used to present some potential challenges and solutions in regard to a robot’s interaction strategy, user modeling, and ethics. For example, to address how to generate appropriate robot activities and mitigate the risk of damage to the driver, a hybrid neuro-symbolic recognition strategy leveraging stereotypical and self-disclosed information is described. Thereby, the aim of this conceptual paper is to navigate through some “memories” of one possible future, toward stimulating ideation and discussion within the increasingly vital area of safety in smart cities Copyright © 2021

Background and purpose: Patients’ adherence to medication is a complex, multidimensional phenomenon. Dispensation data and electronic health records are used to approximate medication-taking through refill adherence. In-depth discussions on the adverse effects of data quality and computational differences are rare. The purpose of this article is to evaluate the impact of common pitfalls when computing medication adherence using electronic health records.

Procedures: We point out common pitfalls associated with the data and operationalization of adherence measures. We provide operational definitions of refill adherence and conduct experiments to determine the effect of the pitfalls on adherence estimations. We performed statistical significance testing on the impact of common pitfalls using a baseline scenario as reference.

Findings: Slight changes in definition can significantly skew refill adherence estimates. Pickup patterns cause significant disagreement between measures and the commonly used proportion of days covered. Common data related issues had a small but statistically significant (p < 0.05) impact on population-level and significant effect on individual cases.

Conclusion: Data-related issues encountered in real-world administrative databases, which affect various operational definitions of refill adherence differently, can significantly skew refill adherence values, leading to false conclusions about adherence, particularly when estimating adherence for individuals. © 2020 The Authors. Published by Elsevier B.V. 

Background and purpose

Low adherence to medication in chronic disease patients leads to increased morbidity, mortality, and healthcare costs. The widespread adoption of electronic prescription and dispensation records allows a more comprehensive overview of medication utilization. In combination with electronic health records (EHR), such data provides new opportunities for identifying patients at risk of nonadherence and provide more targeted and effective interventions. The purpose of this article is to study the predictability of medication adherence for a cohort of hypertensive patients, focusing on healthcare utilization factors under various predictive scenarios. Furthermore, we discover common proportion of days covered patterns (PDC-patterns) for patients with index prescriptions and simulate medication-taking behaviours that might explain observed patterns.

Procedures

We predict refill adherence focusing on factors of healthcare utilization, such as visits, prescription information and demographics of patient and prescriber. We train models with machine learning algorithms, using four different data splits: stratified random, patient, temporal forward prediction with and without index patients. We extract frequent, two-year long PDC-patterns using K-means clustering and investigate five simple models of medication-taking that can generate such PDC-patterns.

Findings

Model performance varies between data splits (AUC test set: 0.77–0.89). Including historical information increases the performance slightly in most cases (approx. 1–2% absolute AUC uplift). Models show low predictive performance (AUC test set: 0.56–0.66) on index-prescriptions and patients with sudden drops in PDC (Recall: 0.58–0.63). We find 21 distinct two-year PDC-patterns, ranging from good adherence to intermittent gaps and early discontinuation in the first or second year. Simulations show that observed PDC-patterns can only be explained by specific medication consumption behaviours.

Conclusions

Prediction models developed using EHR exhibit bias towards patients with high healthcare utilization. Even though actual medication-taking is not observable, consumption patterns may not be as arbitrary, provided that medication refilling and consumption is linked.  © 2020 The Authors. Published by Elsevier Inc.