Name : Khushi Dipak Solanki
Impact of Electronic Health Records on patient outcomes
1) Electronic health records (EHR) are increasingly used in health research, but bias in EHR-derived disease status can lead to inaccurate results.This paper proposes new strategies to address misclassification and selection bias in EHR-based association studies.Three novel likelihood-based bias correction strategies are proposed to address misclassification.Calibration and inverse probability weighting methods are extended to address selection bias.New strategies are proposed to address both biases simultaneously.Valid standard error estimators are derived, and software implementation is provided.The methods are applied to data from The Michigan Genomics Initiative.The proposed strategies aim to improve the accuracy and reliability of EHR-based research findings.
2) Tobacco use negatively impacts cancer treatment outcomes, but many healthcare providers fail to support patients in quitting due to time constraints and lack of knowledge.Patient Reported Outcomes (PRO) measurement can help address these barriers.A study implemented an automated PRO tobacco use screener and referral system via an electronic health record (EHR) patient portal.Preliminary results showed that 3.6% of patients screened positive for recent cigarette smoking and 49.7% of those patients engaged in treatment.The PRO/MyChart system may improve smoker identification and increase treatment offerings.Longer-term evaluation is needed to draw conclusions about effectiveness.PRO measurement via the EHR patient portal may play a crucial role in comprehensive tobacco cessation treatment.
3) Electronic Health Records (EHRs) aim to improve healthcare quality and efficiency.Research on EHRs lacks generalizability and a theory-driven approach.This study examines the impact of EHR implementation on quality management and financial performance of hospitals.Data from 210 hospitals in Texas was analyzed.Results show a positive relationship between EHR implementation and quality of care.EHR implementation also moderates the relationship between quality management and financial performance.This study provides a comprehensive view of EHR’s impact on healthcare quality and financial performance.The findings have implications for hospital administrators, practitioners, and policymakers.
4) Cardiac dysrhythmia (CD) is a common heart rhythm disorder affecting millions of Americans.It can be managed with medications, behaviour change, or cardiac procedures. Electronic health records (EHR) contain valuable information for predicting CD. This study applied a deep learning model (LSTM) to time-series EHR data to predict CD. The model explored the contribution of modifiable cardiovascular risk factors to CD development. The analysis used EHR data from clinics across the US to characterize cardiovascular health and CD outcomes. The study evaluated the association between time-series cardiovascular health and CD diagnoses. The results showed that CD can be predicted using EHR data. This approach has the potential to improve early detection and prevention of CD. The study demonstrates the effectiveness of deep learning algorithms in predicting CD using EHR data.
5) Implementations of eHealth technologies are underway globally, with significant investment. However, the scientific basis for claims of improved quality and safety remains unclear. A systematic review of systematic reviews was conducted to assess the impact of eHealth on healthcare quality and safety.53 systematic reviews were identified, but found to be of substandard quality. The reviews were categorized into three areas: data management, clinical decision support, and remote care. Despite policymaker support, there was little empirical evidence to support claims of improved quality and safety. Best practice guidelines for eHealth development and deployment are lacking. Future eHealth technologies should be evaluated against comprehensive measures throughout their life cycle.
6) Overactive bladder (OAB) is a common condition characterized by urinary incontinence, urgency, and frequency. Antimuscarinic therapies are often used as first-line treatment, while mirabegron is used as second-line therapy. Mirabegron has been shown to be effective and safe, but may increase blood pressure. This study aims to compare cardiovascular risk profiles and outcomes between OAB patients initiating antimuscarinics vs. mirabegron. The study will use electronic health record (EHR) and claims data to assess cardiovascular risk profiles and outcomes. A key challenge is addressing residual bias in observational data due to unmeasured confounders. The study will use statistical methodology to mitigate residual bias and compare cardiovascular outcomes between treatment groups. The goal is to provide a real-world assessment of cardiovascular risk in OAB patients.
7) Electronic healthcare data (EHD) is used to estimate medication adherence, impacting health outcomes and cost-effectiveness. Standardization and transparency of data processing are concerns, and open-source algorithms can facilitate high-quality evidence. Adherer, a package for R, supports researchers in computing EHD-based adherence estimates and visualizing medication histories. Adherer implements functions consistent with current adherence guidelines and definitions. The package facilitates transparent decision-making and data sharing. Adhere R allows researchers to perform complete analyses in R, from importing data to modeling relationships. The package promotes an open approach to science and replicability of processes and results. Adhere R also enables insights into temporal adherence patterns for individual patients.
8) This study examines the impact of electronic medical records (EMRs) on patient safety indicators (PSIs). Results show a positive effect of EMRs on PSIs, primarily through decision support.EMRs with decision support are more effective in reducing PSIs for less complicated cases. This finding suggests that previous studies showing negligible impacts of EMRs may not apply in all settings. The study’s results highlight the importance of decision support in EMRs for improving patient safety. The findings have implications for healthcare providers and policymakers considering EMR adoption. Overall, the study provides evidence of the benefits of EMRs in improving patient safety.
9) This study examines whether hospitals should source electronic health records (EHR) systems from a single vendor or multiple vendors. The study proposes a framework to define healthcare value as the effective use of clinical resources to improve patient outcomes. A moderated-mediation model is used to explore the impact of EHR sourcing strategies on healthcare value. The study finds that single-sourced EHR systems lead to greater health information sharing and improved healthcare value. Hospital-physician practice integration moderates the impact of single sourcing on health information sharing and value. Tighter integration between hospitals and physician practices can create greater value when aligned with EHR sourcing strategies. The findings provide a roadmap for practitioners and policymakers to improve hospital performance under value-based payment reform. The study highlights the importance of EHR sourcing strategies in achieving high-value care.
10) Shared decision making (SDM) involves sharing evidence between patients and providers to make informed decisions. This study examines how electronic health records (EHRs) can improve SDM.A systematic search of PubMed yielded 1369 articles, with 5 studies meeting the inclusion criteria. These studies showed that EHRs can improve clinical outcomes, lifestyle behaviours, and reduce decisional conflict. EHRs can support providers during all steps of SDM, but few EHRs have integrated SDM.Even fewer evaluations of these EHRs exist, highlighting the need for further research. The study concludes that EHRs have potential to support SDM, but this potential has yet to be fully exploited.
11) Electronic Health Records (EHRs) are increasingly utilized in health research; however, biases such as misclassification and selection bias can lead to inaccurate findings. To address these issues, recent studies have introduced novel strategies. One study proposes three likelihood-based bias correction methods to tackle misclassification and extends calibration and inverse probability weighting techniques to mitigate selection bias. Valid standard error estimators are derived, and software implementations are provided. These methods aim to enhance the accuracy and reliability of EHR-based research outcomes. Another study developed an automated Patient Reported Outcomes (PRO) tobacco use screener and referral system via an EHR patient portal. Preliminary results indicated that 3.6% of patients screened positive for recent cigarette smoking, and 49.7% of these patients engaged in treatment. This system may improve the identification of smokers and increase treatment offerings, though longer-term evaluations are necessary to assess its effectiveness fully. A comprehensive study examined the impact of EHR implementation on hospital quality management and financial performance. Data from 210 Texas hospitals revealed a positive relationship between EHR implementation and quality of care. Additionally, EHR implementation moderated the relationship between quality management and financial performance, highlighting EHR’s role in enhancing healthcare quality and financial outcomes. In the realm of cardiac care, a study applied a deep learning model (LSTM) to time-series EHR data to predict cardiac dysrhythmia (CD). The analysis utilized EHR data from clinics across the US to evaluate the association between cardiovascular health metrics and CD diagnoses, demonstrating the potential of deep learning algorithms in early detection and prevention of CD. Despite global investments in eHealth technologies, a systematic review found limited empirical evidence supporting claims of improved healthcare quality and safety. The review emphasized the need for best practice guidelines and comprehensive evaluations throughout the lifecycle of eHealth technologies to ensure their effectiveness. Regarding overactive bladder (OAB) treatment, a study is underway to compare cardiovascular risk profiles and outcomes between patients initiating antimuscarinics versus mirabegron. Using EHR and claims data, the study aims to provide real-world assessments of cardiovascular risks associated with these treatments, employing statistical methods to address potential biases. To facilitate medication adherence research, the open-source package Adhere R supports the computation of electronic healthcare data-based adherence estimates and visualization of medication histories. Adhere R implements functions consistent with current adherence guidelines, promoting transparency and replicability in research. In terms of patient safety, a study examined the impact of electronic medical records (EMRs) on patient safety indicators (PSIs). The findings indicated that EMRs, particularly those with decision support systems, positively affect PSIs, especially for less complicated cases, underscoring the importance of decision support in EMRs for improving patient safety. A study also explored whether hospitals should source EHR systems from a single vendor or multiple vendors. The research found that single-sourced EHR systems lead to greater health information sharing and improved healthcare value, with hospital-physician practice integration moderating this impact. These insights provide guidance for practitioners and policymakers aiming to enhance hospital performance under value-based payment reforms. Lastly, shared decision-making (SDM) involves collaborative evidence-sharing between patients and providers. A systematic search of studies revealed that while EHRs can support providers in all steps of SDM, few EHRs have integrated SDM features, highlighting the need for further research to fully harness EHRs’ potential in supporting SDM.
References :
1) Aixia Guo & Sakima Smith & Yosef M Khan & James R Langabeer II & Randi E Foraker et al (2021): Application of a time-series deep learning model to predict cardiac dysrhythmias in electronic health records
2) Alyce Mei-Shiuan Kuo & Berry Thavalathil & Glyn Elwyn & Zsuzsanna Nemeth & Stuti Dang et al (2018): The Promise of Electronic Health Records to Promote Shared Decision Making: A Narrative Review and a Look Ahead
3) Austyn Snowden & Hildegard Kolb et al (2017): Two years of unintended consequences: introducing an electronic health record system in a hospice in Scotland
4) Indranil R. Bardhan & Chenzhang Bao & Sezgin Ayabakan et al(2023): Value Implications of Sourcing Electronic Health Records: The Role of Physician Practice Integration
5) Ingram, M. & Doubleday, K. & Bell, M.L. & Lohr, A. & Murrieta, L. & Velasco, M. & Blackburn, J. & Sabo, S. & De Zapien, J.G. & Carvajal, S.C. et al(2017): Community health worker impact on chronic disease outcomes within primary care examined using electronic health records
6) Julia R. May & Elizabeth Klass & Kristina Davis & Timothy Pearman & Steven Rittmeyer & Sheetal Kircher & Brian Hitsman et al (2020): Leveraging Patient Reported Outcomes Measurement via the Electronic Health Record to Connect Patients with Cancer to Smoking Cessation Treatment
7) Lauren J. Beesley & Bhramar Mukherjee et al(2022): Statistical inference for association studies using electronic health records: handling both selection bias and outcome misclassification
8) Malhan, Amit & Pavur, Robert & Pelton, Lou & Manuj, Ila et al(2022): Role of Electronic Healthcare Record Adoption in Enhancing the Relationship between Quality Measures and Hospital Financial Performance
9) Seth Freedman & Haizhen Lin & Jeffrey Prince et al(2015): Information Technology and Patient Health: Analysing Outcomes, Populations, and Mechanisms
10) Sukhpreet Kaur & Rajinder Kaur & Rashmi Aggarwal et al(2019): E-health and its Impact on Indian Health Care: An Analysis