Browsing by Subject "Electronic Health Records"
- Results Per Page
- Sort Options
Item Open Access A Framework to Support the Sharing and Reuse of Computable Phenotype Definitions Across Health Care Delivery and Clinical Research Applications.(EGEMS (Wash DC), 2016) Richesson, RL; Smerek, MM; Cameron, CBINTRODUCTION: The ability to reproducibly identify clinically equivalent patient populations is critical to the vision of learning health care systems that implement and evaluate evidence-based treatments. The use of common or semantically equivalent phenotype definitions across research and health care use cases will support this aim. Currently, there is no single consolidated repository for computable phenotype definitions, making it difficult to find all definitions that already exist, and also hindering the sharing of definitions between user groups. METHOD: Drawing from our experience in an academic medical center that supports a number of multisite research projects and quality improvement studies, we articulate a framework that will support the sharing of phenotype definitions across research and health care use cases, and highlight gaps and areas that need attention and collaborative solutions. FRAMEWORK: An infrastructure for re-using computable phenotype definitions and sharing experience across health care delivery and clinical research applications includes: access to a collection of existing phenotype definitions, information to evaluate their appropriateness for particular applications, a knowledge base of implementation guidance, supporting tools that are user-friendly and intuitive, and a willingness to use them. NEXT STEPS: We encourage prospective researchers and health administrators to re-use existing EHR-based condition definitions where appropriate and share their results with others to support a national culture of learning health care. There are a number of federally funded resources to support these activities, and research sponsors should encourage their use.Item Open Access A Novel Use of Social Network Analysis and Routinely Collected Data to Uncover Care Coordination Processes for Patients with Heart Failure(2021) Wei, SijiaEffective patient care transitions require consideration of the patient’s social and clinical contexts, yet how these factors relate to the processes in care coordination remains poorly described. This dissertation aimed to describe provider networks and clinical care and social contexts involved during longitudinal care transitions across settings. The overall purpose of this dissertation is to uncover the longitudinal patterns of utilization and relational processes needed for effective care coordination in transitional care, so we can redesign interventions that focus on informational and relationship networks to improve interaction patterns and system performance for people living with heart failure (HF) as they undergo transitions across settings and over time. This dissertation was a retrospective exploratory study. Chapter 2 is an integrative review examining coordination processes in transitional care interventions for older adults with HF by integrating a social network analysis framework. We subsequently selected a cohort of patients aged 18 years or older (n = 1269) with an initial hospitalization for HF at Duke University Health System between January 1, 2016 and December 31, 2018 based on encounter, sociodemographic, and clinical data extracted from electronic health records (EHR). In Chapter 3, a latent growth trajectory analysis was used to identify distinct subgroups of patients based on the frequency of outpatient, as well as emergency department (ED) and inpatient encounters 1 year before and 1 year after the index hospitalization; multinomial logistic regression was then used to evaluate how outpatient utilization was related to acute care utilization. Based on findings (described in Chapter 3), we purposively sampled 11 patients from the Chapter 3 cohort for a second empirical study (described in Chapter 4) with a mixed-methods sequential explanatory design. These 11 patients had a full spectrum of experience in socioeconomic disadvantages based on three strata (race, insurance, and Area Deprivation Index), but they had similar levels of comorbidity and average severity of illness and displayed the same change in the severity of illness during the study period. We used quantitative and qualitative data available from clinical notes in the EHR, and integrated results from quantitative and qualitative analysis to better understand the social and clinical context and social structure essential for care coordination. High variability in transitional care is likely because care coordination processes are highly relational. The relational structure of transitional care interventions varied from triadic to complex network structures. Use of a network analysis framework helped to uncover relational structures and processes underlying transitional care to inform intervention development. Chapter 3 revealed that high heterogeneity exists in patients’ utilization patterns. A small subgroup of high users utilized a substantial amount of the resources. Patients with high outpatient utilization had more than 4 times the likelihood of also having high acute care utilization, and change in the severity of illness had the highest level of significance and strongest magnitude of effect on influencing high acute care utilization. Chapter 4 demonstrated the feasibility of using clinical notes and social network analysis (SNA) to assess the provider networks for patients with HF in care transitions. People who were experiencing more socioeconomic disadvantages and social instability were less likely to have densely connected provider teams and providers who were central and influential in the system network. Lacking consistent and reciprocal relationships with outpatient provider teams, especially primary care provider and cardiology teams, was precedent to poor care management and coordination. Turbulence in care transition can result from sources other than transitioning between settings. This dissertation demonstrated the (a) importance of understanding relational processes and structure during patients’ utilization of acute and outpatient care services and (b) potential to capture structural inequalities that may influence the efficiency of care coordination and health outcomes for patients with HF.
Item Open Access A Systematic Framework to Rapidly Obtain Data on Patients with Cancer and COVID-19: CCC19 Governance, Protocol, and Quality Assurance.(Cancer cell, 2020-12) COVID-19 and Cancer Consortium. Electronic address: jeremy.warner@vumc.org; COVID-19 and Cancer ConsortiumWhen the COVID-19 pandemic began, formal frameworks to collect data about affected patients were lacking. The COVID-19 and Cancer Consortium (CCC19) was formed to collect granular data on patients with cancer and COVID-19 at scale and as rapidly as possible. CCC19 has grown from five initial institutions to 125 institutions with >400 collaborators. More than 5,000 cases with complete baseline data have been accrued. Future directions include increased electronic health record integration for direct data ingestion, expansion to additional domestic and international sites, more intentional patient involvement, and granular analyses of still-unanswered questions related to cancer subtypes and treatments.Item Open Access An Atlas of Genetic Variation Linking Pathogen-Induced Cellular Traits to Human Disease.(Cell host & microbe, 2018-08) Wang, Liuyang; Pittman, Kelly J; Barker, Jeffrey R; Salinas, Raul E; Stanaway, Ian B; Williams, Graham D; Carroll, Robert J; Balmat, Tom; Ingham, Andy; Gopalakrishnan, Anusha M; Gibbs, Kyle D; Antonia, Alejandro L; eMERGE Network; Heitman, Joseph; Lee, Soo Chan; Jarvik, Gail P; Denny, Joshua C; Horner, Stacy M; DeLong, Mark R; Valdivia, Raphael H; Crosslin, David R; Ko, Dennis CPathogens have been a strong driving force for natural selection. Therefore, understanding how human genetic differences impact infection-related cellular traits can mechanistically link genetic variation to disease susceptibility. Here we report the Hi-HOST Phenome Project (H2P2): a catalog of cellular genome-wide association studies (GWAS) comprising 79 infection-related phenotypes in response to 8 pathogens in 528 lymphoblastoid cell lines. Seventeen loci surpass genome-wide significance for infection-associated phenotypes ranging from pathogen replication to cytokine production. We combined H2P2 with clinical association data from patients to identify a SNP near CXCL10 as a risk factor for inflammatory bowel disease. A SNP in the transcriptional repressor ZBTB20 demonstrated pleiotropy, likely through suppression of multiple target genes, and was associated with viral hepatitis. These data are available on a web portal to facilitate interpreting human genome variation through the lens of cell biology and should serve as a rich resource for the research community.Item Open Access Applying active learning to high-throughput phenotyping algorithms for electronic health records data.(Journal of the American Medical Informatics Association : JAMIA, 2013-12) Chen, Yukun; Carroll, Robert J; Hinz, Eugenia R McPeek; Shah, Anushi; Eyler, Anne E; Denny, Joshua C; Xu, HuaObjectives
Generalizable, high-throughput phenotyping methods based on supervised machine learning (ML) algorithms could significantly accelerate the use of electronic health records data for clinical and translational research. However, they often require large numbers of annotated samples, which are costly and time-consuming to review. We investigated the use of active learning (AL) in ML-based phenotyping algorithms.Methods
We integrated an uncertainty sampling AL approach with support vector machines-based phenotyping algorithms and evaluated its performance using three annotated disease cohorts including rheumatoid arthritis (RA), colorectal cancer (CRC), and venous thromboembolism (VTE). We investigated performance using two types of feature sets: unrefined features, which contained at least all clinical concepts extracted from notes and billing codes; and a smaller set of refined features selected by domain experts. The performance of the AL was compared with a passive learning (PL) approach based on random sampling.Results
Our evaluation showed that AL outperformed PL on three phenotyping tasks. When unrefined features were used in the RA and CRC tasks, AL reduced the number of annotated samples required to achieve an area under the curve (AUC) score of 0.95 by 68% and 23%, respectively. AL also achieved a reduction of 68% for VTE with an optimal AUC of 0.70 using refined features. As expected, refined features improved the performance of phenotyping classifiers and required fewer annotated samples.Conclusions
This study demonstrated that AL can be useful in ML-based phenotyping methods. Moreover, AL and feature engineering based on domain knowledge could be combined to develop efficient and generalizable phenotyping methods.Item Open Access Blood pressure level impacts risk of death among HIV seropositive adults in Kenya: a retrospective analysis of electronic health records.(BMC Infect Dis, 2014-05-22) Bloomfield, Gerald S; Hogan, Joseph W; Keter, Alfred; Holland, Thomas L; Sang, Edwin; Kimaiyo, Sylvester; Velazquez, Eric JBACKGROUND: Mortality among people with human immunodeficiency virus (HIV) infection is increasingly due to non-communicable causes. This has been observed mostly in developed countries and the routine care of HIV infected individuals has now expanded to include attention to cardiovascular risk factors. Cardiovascular risk factors such as high blood pressure are often overlooked among HIV seropositive (+) individuals in sub-Saharan Africa. We aimed to determine the effect of blood pressure on mortality among HIV+ adults in Kenya. METHODS: We performed a retrospective analysis of electronic medical records of a large HIV treatment program in western Kenya between 2005 and 2010. All included individuals were HIV+. We excluded participants with AIDS, who were <16 or >80 years old, or had data out of acceptable ranges. Missing data for key covariates was addressed by inverse probability weighting. Primary outcome measures were crude mortality rate and mortality hazard ratio (HR) using Cox proportional hazards models adjusted for potential confounders including HIV stage. RESULTS: There were 49,475 (74% women) HIV+ individuals who met inclusion and exclusion criteria. Mortality rates for men and women were 3.8 and 1.8/100 person-years, respectively, and highest among those with the lowest blood pressures. Low blood pressure was associated with the highest mortality incidence rate (IR) (systolic <100 mmHg IR 5.2 [4.8-5.7]; diastolic <60 mmHg IR 9.2 [8.3-10.2]). Mortality rate among men with high systolic blood pressure without advanced HIV (3.0, 95% CI: 1.6-5.5) was higher than men with normal systolic blood pressure (1.1, 95% CI: 0.7-1.7). In weighted proportional hazards regression models, men without advanced HIV disease and systolic blood pressure ≥140 mmHg carried a higher mortality risk than normotensive men (HR: 2.39, 95% CI: 0.94-6.08). CONCLUSIONS: Although there has been little attention paid to high blood pressure among HIV+ Africans, we show that blood pressure level among HIV+ patients in Kenya is related to mortality. Low blood pressure carries the highest mortality risk. High systolic blood pressure is associated with mortality among patients whose disease is not advanced. Further investigation is needed into the cause of death for such patients.Item Open Access Bridging the integration gap between patient-generated blood glucose data and electronic health records.(Journal of the American Medical Informatics Association : JAMIA, 2019-07) Lewinski, Allison A; Drake, Connor; Shaw, Ryan J; Jackson, George L; Bosworth, Hayden B; Oakes, Megan; Gonzales, Sarah; Jelesoff, Nicole E; Crowley, Matthew JTelemedicine can facilitate population health management by extending the reach of providers to efficiently care for high-risk, high-utilization populations. However, for telemedicine to be maximally useful, data collected using telemedicine technologies must be reliable and readily available to healthcare providers. To address current gaps in integration of patient-generated health data into the electronic health record (EHR), we examined 2 patient-facing platforms, Epic MyChart and Apple HealthKit, both of which facilitated the uploading of blood glucose data into the EHR as part of a diabetes telemedicine intervention. All patients were offered use of the MyChart platform; we subsequently invited a purposive sample of patients who used the MyChart platform effectively (n = 5) to also use the Apple HealthKit platform. Patients reported both platforms helped with diabetes self-management, and providers appreciated the convenience of the processes for obtaining patient data. Providers stated that the EHR data presentation format for Apple HealthKit was challenging to interpret; however, they also valued the greater perceived accuracy the Apple HealthKit data. Our findings indicate that patient-facing platforms can feasibly facilitate transmission of patient-generated health data into the EHR and support telemedicine-based care.Item Open Access Can we understand population healthcare needs using electronic medical records?(Singapore medical journal, 2019-09) Chong, Jia Loon; Low, Lian Leng; Chan, Darren Yak Leong; Shen, Yuzeng; Thin, Thiri Naing; Ong, Marcus Eng Hock; Matchar, David BruceIntroduction
The identification of population-level healthcare needs using hospital electronic medical records (EMRs) is a promising approach for the evaluation and development of tailored healthcare services. Population segmentation based on healthcare needs may be possible using information on health and social service needs from EMRs. However, it is currently unknown if EMRs from restructured hospitals in Singapore provide information of sufficient quality for this purpose. We compared the inter-rater reliability between a population segment that was assigned prospectively and one that was assigned retrospectively based on EMR review.Methods
200 non-critical patients aged ≥ 55 years were prospectively evaluated by clinicians for their healthcare needs in the emergency department at Singapore General Hospital, Singapore. Trained clinician raters with no prior knowledge of these patients subsequently accessed the EMR up to the prospective rating date. A similar healthcare needs evaluation was conducted using the EMR. The inter-rater reliability between the two rating sets was evaluated using Cohen's Kappa and the incidence of missing information was tabulated.Results
The inter-rater reliability for the medical 'global impression' rating was 0.37 for doctors and 0.35 for nurses. The inter-rater reliability for the same variable, retrospectively rated by two doctors, was 0.75. Variables with a higher incidence of missing EMR information such as 'social support in case of need' and 'patient activation' had poorer inter-rater reliability.Conclusion
Pre-existing EMR systems may not capture sufficient information for reliable determination of healthcare needs. Thus, we should consider integrating policy-relevant healthcare need variables into EMRs.Item Open Access Clinical factors associated with persistently poor diabetes control in the Veterans Health Administration: A nationwide cohort study.(PloS one, 2019-01) Alexopoulos, Anastasia-Stefania; Jackson, George L; Edelman, David; Smith, Valerie A; Berkowitz, Theodore SZ; Woolson, Sandra L; Bosworth, Hayden B; Crowley, Matthew JObjective
Patients with persistent poorly-controlled diabetes mellitus (PPDM) despite engagement in clinic-based care are at particularly high risk for diabetes complications and costs. Understanding this population's demographics, comorbidities and care utilization could guide strategies to address PPDM. We characterized factors associated with PPDM in a large sample of Veterans with type 2 diabetes.Methods
We identified a cohort of Veterans with medically treated type 2 diabetes, who received Veterans Health Administration primary care during fiscal years 2012 and 2013. PPDM was defined by hemoglobin A1c levels uniformly >8.5% during fiscal year (FY) 2012, despite engagement with care during this period. We used FY 2012 demographic, comorbidity and medication data to describe PPDM in relation to better-controlled diabetes patients and created multivariable models to examine associations between clinical factors and PPDM. We also constructed multivariable models to explore the association between PPDM and FY 2013 care utilization.Results
In our cohort of diabetes patients (n = 435,820), 12% met criteria for PPDM. Patients with PPDM were younger than better-controlled patients, less often married, and more often Black/African-American and Hispanic or Latino/Latina. Of included comorbidities, only retinopathy (OR 1.68, 95% confidence interval (CI): 1.63,1.73) and nephropathy (OR 1.26, 95% CI: 1.19,1.34) demonstrated clinically significant associations with PPDM. Complex insulin regimens such as premixed (OR 10.80, 95% CI: 10.11,11.54) and prandial-containing regimens (OR 18.74, 95% CI: 17.73,19.81) were strongly associated with PPDM. Patients with PPDM had higher care utilization, particularly endocrinology care (RR 3.56, 95% CI: 3.47,3.66); although only 26.4% of patients saw endocrinology overall.Conclusion
PPDM is strongly associated with complex diabetes regimens, although heterogeneity in care utilization exists. While there is evidence of underutilization, inadequacy of available care may also contribute to PPDM. Our findings should inform tailored approaches to meet the needs of PPDM, who are among the highest-risk, highest-cost patients with diabetes.Item Open Access Designing risk prediction models for ambulatory no-shows across different specialties and clinics.(Journal of the American Medical Informatics Association : JAMIA, 2018-08) Ding, Xiruo; Gellad, Ziad F; Mather, Chad; Barth, Pamela; Poon, Eric G; Newman, Mark; Goldstein, Benjamin AObjective:As available data increases, so does the opportunity to develop risk scores on more refined patient populations. In this paper we assessed the ability to derive a risk score for a patient no-showing to a clinic visit. Methods:Using data from 2 264 235 outpatient appointments we assessed the performance of models built across 14 different specialties and 55 clinics. We used regularized logistic regression models to fit and assess models built on the health system, specialty, and clinic levels. We evaluated fits based on their discrimination and calibration. Results:Overall, the results suggest that a relatively robust risk score for patient no-shows could be derived with an average C-statistic of 0.83 across clinic level models and strong calibration. Moreover, the clinic specific models, even with lower training set sizes, often performed better than the more general models. Examination of the individual models showed that risk factors had different degrees of predictability across the different specialties. Implementation of optimal modeling strategies would lead to capturing an additional 4819 no-shows per-year. Conclusion:Overall, this work highlights both the opportunity for and the importance of leveraging the available electronic health record data to develop more refined risk models.Item Open Access Development and validation of an electronic health records-based opioid use disorder algorithm by expert clinical adjudication among patients with prescribed opioids.(Pharmacoepidemiology and drug safety, 2023-05) Ranapurwala, Shabbar I; Alam, Ishrat Z; Pence, Brian W; Carey, Timothy S; Christensen, Sean; Clark, Marshall; Chelminski, Paul R; Wu, Li-Tzy; Greenblatt, Lawrence H; Korte, Jeffrey E; Wolfson, Mark; Douglas, Heather E; Bowlby, Lynn A; Capata, Michael; Marshall, Stephen WBackground
In the US, over 200 lives are lost from opioid overdoses each day. Accurate and prompt diagnosis of opioid use disorders (OUD) may help prevent overdose deaths. However, international classification of disease (ICD) codes for OUD are known to underestimate prevalence, and their specificity and sensitivity are unknown. We developed and validated algorithms to identify OUD in electronic health records (EHR) and examined the validity of OUD ICD codes.Methods
Through four iterations, we developed EHR-based OUD identification algorithms among patients who were prescribed opioids from 2014 to 2017. The algorithms and OUD ICD codes were validated against 169 independent "gold standard" EHR chart reviews conducted by an expert adjudication panel across four healthcare systems. After using 2014-2020 EHR for validating iteration 1, the experts were advised to use 2014-2017 EHR thereafter.Results
Of the 169 EHR charts, 81 (48%) were reviewed by more than one expert and exhibited 85% expert agreement. The experts identified 54 OUD cases. The experts endorsed all 11 OUD criteria from the Diagnostic and Statistical Manual of Mental Disorders-5, including craving (72%), tolerance (65%), withdrawal (56%), and recurrent use in physically hazardous conditions (50%). The OUD ICD codes had 10% sensitivity and 99% specificity, underscoring large underestimation. In comparison our algorithm identified OUD with 23% sensitivity and 98% specificity.Conclusions and relevance
This is the first study to estimate the validity of OUD ICD codes and develop validated EHR-based OUD identification algorithms. This work will inform future research on early intervention and prevention of OUD.Item Open Access Do diabetic veterans use the Internet? Self-reported usage, skills, and interest in using My HealtheVet Web portal.(Telemed J E Health, 2010-06) Cho, Alex H; Arar, Nedal H; Edelman, David E; Hartwell, Patricia H; Oddone, Eugene Z; Yancy, William SOBJECTIVE: The Veterans Health Administration has developed My HealtheVet (MHV), a Web-based portal that links veterans to their care in the veteran affairs (VA) system. The objective of this study was to measure diabetic veterans' access to and use of the Internet, and their interest in using MHV to help manage their diabetes. MATERIALS AND METHODS: Cross-sectional mailed survey of 201 patients with type 2 diabetes and hemoglobin A(1c) > 8.0% receiving primary care at any of five primary care clinic sites affiliated with a VA tertiary care facility. Main measures included Internet usage, access, and attitudes; computer skills; interest in using the Internet; awareness of and attitudes toward MHV; demographics; and socioeconomic status. RESULTS: A majority of respondents reported having access to the Internet at home. Nearly half of all respondents had searched online for information about diabetes, including some who did not have home Internet access. More than a third obtained "some" or "a lot" of their health-related information online. Forty-one percent reported being "very interested" in using MHV to help track their home blood glucose readings, a third of whom did not have home Internet access. Factors associated with being "very interested" were as follows: having access to the Internet at home (p < 0.001), "a lot/some" trust in the Internet as a source of health information (p = 0.002), lower age (p = 0.03), and some college (p = 0.04). Neither race (p = 0.44) nor income (p = 0.25) was significantly associated with interest in MHV. CONCLUSIONS: This study found that a diverse sample of older VA patients with sub-optimally controlled diabetes had a level of familiarity with and access to the Internet comparable to an age-matched national sample. In addition, there was a high degree of interest in using the Internet to help manage their diabetes.Item Open Access Does improved access to diagnostic imaging results reduce hospital length of stay? A retrospective study.(BMC Health Serv Res, 2010-09-06) Hurlen, Petter; Østbye, Truls; Borthne, Arne S; Gulbrandsen, PålBACKGROUND: One year after the introduction of Information and Communication Technology (ICT) to support diagnostic imaging at our hospital, clinicians had faster and better access to radiology reports and images; direct access to Computed Tomography (CT) reports in the Electronic Medical Record (EMR) was particularly popular. The objective of this study was to determine whether improvements in radiology reporting and clinical access to diagnostic imaging information one year after the ICT introduction were associated with a reduction in the length of patients' hospital stays (LOS). METHODS: Data describing hospital stays and diagnostic imaging were collected retrospectively from the EMR during periods of equal duration before and one year after the introduction of ICT. The post-ICT period was chosen because of the documented improvement in clinical access to radiology results during that period. The data set was randomly split into an exploratory part used to establish the hypotheses, and a confirmatory part. The data was used to compare the pre-ICT and post-ICT status, but also to compare differences between groups. RESULTS: There was no general reduction in LOS one year after ICT introduction. However, there was a 25% reduction for one group - patients with CT scans. This group was heterogeneous, covering 445 different primary discharge diagnoses. Analyses of subgroups were performed to reduce the impact of this divergence. CONCLUSION: Our results did not indicate that improved access to radiology results reduced the patients' LOS. There was, however, a significant reduction in LOS for patients undergoing CT scans. Given the clinicians' interest in CT reports and the results of the subgroup analyses, it is likely that improved access to CT reports contributed to this reduction.Item Open Access HIPAA and the Leak of "Deidentified" EHR Data.(The New England journal of medicine, 2021-06-05) Mandl, Kenneth D; Perakslis, Eric DItem Open Access How can innovative uses of technology be harnessed to improve medication adherence?(Expert review of pharmacoeconomics & outcomes research, 2012-04) Bosworth, Hayden BItem Open Access Impact of Dementia on Incidence and Severity of Postoperative Pulmonary Complications Following Hip Fracture Surgery Among Older Patients.(Clinical nursing research, 2023-11) Tsumura, Hideyo; McConnell, Eleanor S; Xue, Tingzhong Michelle; Wei, Sijia; Lee, Chiyoung; Pan, WeiPostoperative pulmonary complications (PPCs) are the leading cause of death following hip fracture surgery. Dementia has been identified as a PPC risk factor that complicates the clinical course. By leveraging electronic health records, this retrospective observational study evaluated the impact of dementia on the incidence and severity of PPCs, hospital length of stay, and postoperative 30-day mortality among 875 older patients (≥65 years) who underwent hip fracture surgery between October 1, 2015 and December 31, 2018 at a health system in the southeastern United States. Inverse probability of treatment weighting using propensity scores was utilized to balance confounders between patients with and without dementia to isolate the impact of dementia on PPCs. Regression analyses revealed that dementia did not have a statistically significant impact on the incidence and severity of PPCs or postoperative 30-day mortality. However, dementia significantly extended the hospital length of stay by an average of 1.37 days.Item Open Access Implementation of Changes to Medical Student Documentation at Duke University Health System: Balancing Education With Service.(Academic medicine : journal of the Association of American Medical Colleges, 2021-06) Gagliardi, Jane P; Bonanno, Brian; McPeek Hinz, Eugenia R; Musser, R Clayton; Knudsen, Nancy W; Palko, Michael; McNair, Felice; Lee, Hui-Jie; Clay, Alison SPurpose
When the Centers for Medicare and Medicaid Services (CMS) changed policies about medical student documentation, students with proper supervision may now document their history, physical exam, and medical decision making in the electronic health record (EHR) for billable encounters. Since documentation is a core entrustable professional activity for medical students, the authors sought to evaluate student opportunities for documentation and feedback across and between clerkships.Method
In February 2018, a multidisciplinary workgroup was formed to implement student documentation at Duke University Health System, including educating trainees and supervisors, tracking EHR usage, and enforcing CMS compliance. From August 2018 to August 2019, locations and types of student-involved services (student-faculty or student-resident-faculty) were tracked using billing data from attestation statements. Student end-of-clerkship evaluations included opportunity for documentation and receipt of feedback. Since documentation was not allowed before August 2018, it was not possible to compare with prior student experiences.Results
In the first half of the academic year, 6,972 patient encounters were billed as student-involved services, 52% (n = 3,612) in the inpatient setting and 47% (n = 3,257) in the outpatient setting. Most (74%) of the inpatient encounters also involved residents, and most (92%) of outpatient encounters were student-teaching physician only.Approximately 90% of students indicated having had opportunity to document in the EHR across clerkships, except for procedure-based clerkships such as surgery and obstetrics. Receipt of feedback was present along with opportunity for documentation more than 85% of the time on services using evaluation and management coding. Most students (> 90%) viewed their documentation as having a moderate or high impact on patient care.Conclusions
Changes to student documentation were successfully implemented and adopted; changes met both compliance and education needs within the health system without resulting in potential abuses of student work for service.Item Open Access Improved tuberculosis outcomes with daily vs. intermittent rifabutin in HIV-TB coinfected patients in India.(The international journal of tuberculosis and lung disease : the official journal of the International Union against Tuberculosis and Lung Disease, 2016-09) Jenks, JD; Kumarasamy, N; Ezhilarasi, C; Poongulali, S; Ambrose, P; Yepthomi, T; Devaraj, C; Benson, CASetting
Y R Gaitonde Centre for AIDS Research and Education, Chennai, India.Objective
To compare anti-tuberculosis treatment outcomes in individuals with human immunodeficiency virus (HIV) and tuberculosis (TB) co-infection on atazanavir/ritonavir (ATV/r) antiretroviral therapy (ART) plus daily rifabutin (RBT) 150 mg with those on ATV/r plus thrice-weekly RBT 150 mg.Design
A retrospective study was conducted of two HIV-TB co-infected cohorts between 2003 and 2014. Basic demographic and TB outcome data were obtained from an electronic database and patient records. The χ(2) and Fisher's exact test were used to compare daily and intermittent RBT treatment groups.Results
Of 292 individuals on an ATV/r-based ART regimen plus RBT, 118 (40.4%) received thrice-weekly RBT and 174 (59.6%) daily RBT. Patients in the two RBT treatment groups were similar in sex, age, previous history of TB, site of TB and acid-fast bacilli smear status. More individuals in the daily vs. the intermittent RBT group achieved clinical cure (73.0% vs. 44.1%, P < 0.001), with no significant differences in relapse/recurrence or all-cause mortality between groups.Conclusion
There were higher rates of clinical TB cure in individuals on a boosted protease inhibitor-based ART regimen with daily RBT compared to intermittently dosed RBT. Optimal RBT dosing in this setting requires further investigation.Item Open Access In-hospital outcomes of premature infants with severe bronchopulmonary dysplasia.(Journal of perinatology : official journal of the California Perinatal Association, 2017-07) Jackson, W; Hornik, CP; Messina, JA; Guglielmo, K; Watwe, A; Delancy, G; Valdez, A; MacArthur, T; Peter-Wohl, S; Smith, PB; Tolia, VN; Laughon, MMOBJECTIVE:To characterize in-hospital outcomes of premature infants diagnosed with severe bronchopulmonary dysplasia (BPD). STUDY DESIGN:Retrospective cohort study including premature infants with severe BPD discharged from 348 Pediatrix Medical Group neonatal intensive care units from 1997 to 2015. RESULTS:There were 10 752 infants with severe BPD, and 549/10 752 (5%) died before discharge. Infants who died were more likely to be male, small for gestational age, have received more medical interventions and more frequently diagnosed with surgical necrotizing enterocolitis, culture-proven sepsis and pulmonary hypertension following 36 weeks of postmenstrual age compared with survivors. Approximately 70% of infants with severe BPD were discharged by 44 weeks of postmenstrual age, and 86% were discharged by 48 weeks of postmenstrual age. CONCLUSIONS:A majority of infants diagnosed with severe BPD were discharged home by 44 weeks of postmenstrual age. These results may inform discussions with families regarding the expected hospital course of infants diagnosed with severe BPD.Item Open Access Incorporating informatively collected laboratory data from EHR in clinical prediction models.(BMC medical informatics and decision making, 2024-07) Sun, Minghui; Engelhard, Matthew M; Bedoya, Armando D; Goldstein, Benjamin ABackground
Electronic Health Records (EHR) are widely used to develop clinical prediction models (CPMs). However, one of the challenges is that there is often a degree of informative missing data. For example, laboratory measures are typically taken when a clinician is concerned that there is a need. When data are the so-called Not Missing at Random (NMAR), analytic strategies based on other missingness mechanisms are inappropriate. In this work, we seek to compare the impact of different strategies for handling missing data on CPMs performance.Methods
We considered a predictive model for rapid inpatient deterioration as an exemplar implementation. This model incorporated twelve laboratory measures with varying levels of missingness. Five labs had missingness rate levels around 50%, and the other seven had missingness levels around 90%. We included them based on the belief that their missingness status can be highly informational for the prediction. In our study, we explicitly compared the various missing data strategies: mean imputation, normal-value imputation, conditional imputation, categorical encoding, and missingness embeddings. Some of these were also combined with the last observation carried forward (LOCF). We implemented logistic LASSO regression, multilayer perceptron (MLP), and long short-term memory (LSTM) models as the downstream classifiers. We compared the AUROC of testing data and used bootstrapping to construct 95% confidence intervals.Results
We had 105,198 inpatient encounters, with 4.7% having experienced the deterioration outcome of interest. LSTM models generally outperformed other cross-sectional models, where embedding approaches and categorical encoding yielded the best results. For the cross-sectional models, normal-value imputation with LOCF generated the best results.Conclusion
Strategies that accounted for the possibility of NMAR missing data yielded better model performance than those did not. The embedding method had an advantage as it did not require prior clinical knowledge. Using LOCF could enhance the performance of cross-sectional models but have countereffects in LSTM models.
- «
- 1 (current)
- 2
- 3
- »