Search

Objectives Chronic hepatitis C virus (HCV) infection is one of the main causes of hepatocellular carcinoma. Before initiating a multilevel HCV screening intervention, we sought to (1) describe concordance between the electronic health record (EHR) data warehouse and manual medical record review in recording aspects of HCV testing and treatment and (2) estimate the percentage of patients with chronic HCV infection who initiated and completed HCV treatment using manual medical record review. Methods We examined the medical records for 177 patients (100 randomly selected patients born during 1945-1965 without evidence of HCV testing and 77 adult patients of any birth cohort who had completed HCV testing) with a primary care or relevant specialist visit at an academic health care system in Tampa, Florida, from 2015 through 2018. We used the Cohen κ coefficient to examine the degree of concordance between the searchable data warehouse and the medical record review abstractions. Descriptive statistics characterized referral to and receipt of treatment among patients with chronic HCV infection from medical record review. Results We found generally good concordance between the data warehouse abstraction and medical record review for HCV testing data (κ ranged from 0.66 to 0.87). However, the data warehouse failed to capture data on HCV treatment variables. According to medical record review, 28 patients had chronic HCV infection; 16 patients were prescribed treatment, 14 initiated treatment, and 9 achieved and had a reported posttreatment undetected HCV viral load. Conclusions Using data warehouse data provides generally reliable HCV testing information. However, without the use of natural language processing and purposeful EHR design, manual medical record reviews will likely be required to characterize treatment initiation and completion.

PURPOSE Oral chemotherapy challenges providers' abilities to safely monitor patients' symptoms, adherence, and financial toxicity. COVID-19 has increased the urgency of caring for patients remotely. Collection of electronic patient-reported outcomes (ePROs) has demonstrated efficacy for patients on intravenous chemotherapy, but limited data support their use in oral chemotherapy. We undertook a pilot project to assess the feasibility of implementing an ePRO system for patients starting oral chemotherapy at our cancer center, which includes both an academic site and a community site. METHODS Patients initiating oral chemotherapy were asked to participate. A five-question tool was built in REDCap. Concerning responses triggered outreach within one business day. The primary outcome was time to first symptom assessment. For comparison, we used a historical cohort of patients who had been prescribed oral chemotherapies by providers in the same disease groups at the cancer center. RESULTS Twenty-five of 62 (40%) patients completed ePRO assessments. Fifty historical charts were reviewed. Time to first symptom assessment was 7 days (IQR, 4-14 days) in the historical group compared with 3 days (IQR, 2-4 days) in the ePRO group. Time to clinical action was 14 days (7-35 days) in the historical group compared with 8 days (4-19 days) in the ePRO group. No statistically significant differences were detected in 30-day emergency department visit or hospitalization (12% for both groups) or 90-day emergency department visit or hospitalization rates (historical 28% and ePRO 20%). CONCLUSION An ePRO tool monitoring patient concerns about adherence, cost, and toxicities for patients with new oral chemotherapy regimens is feasible and improves time to symptom assessment. Further investigation is needed to improve patient engagement with ePROs and evaluate the long-term impacts for patients on oral chemotherapy.

PURPOSE In an effort to promote cost-conscious, high-quality, and patient-centered care in the palliative radiation of painful bone metastases, the National Quality Forum (NQF) formed measure 1822 in 2012, which recommends the use of one of the four dose-fractionation schemes (30 Gy in 10 fractions, 24 Gy in 6 fractions, 20 Gy in 5 fractions, or 8 Gy in a single fraction). We investigated whether a custom electronic health record (EHR) alert system improved quality measure compliance among 88 physicians at a large academic center and institutional network. METHODS In March 2018, a multiphase alert system was embedded in a custom web-based EHR. Prior to a course of palliative bone radiation, the alert system notified the user of NQF 1822 recommendations and, once prescription was completed, either affirmed compliance or advised a change in treatment schedule. Rates of compliance were evaluated before and after implementation of alert system. RESULTS Of 2,399 treatment courses, 86.5% were compliant with NQF 1822 recommendations. There was no difference in rates of NQF 1822 compliance before or after implementation of the custom EHR alert (86.0% before March 2018 v 86.9% during and after March 2018, P = .551). CONCLUSION There was no change in rates of compliance following implementation of a custom EHR alert system designed to make treatment recommendations based on national quality measure guidelines. To be of most benefit, future palliative bone metastasis decision aids should leverage peer review, target a clear practice deficiency, center upon high-quality practice guidelines, and allow flexibility to reflect the diversity of clinical scenarios.

PURPOSE Key oncology end points are not routinely encoded into electronic medical records (EMRs). We assessed whether natural language processing (NLP) can abstract treatment discontinuation rationale from unstructured EMR notes to estimate toxicity incidence and progression-free survival (PFS). METHODS We constructed a retrospective cohort of 6,115 patients with early-stage and 701 patients with metastatic breast cancer initiating care at Memorial Sloan Kettering Cancer Center from 2008 to 2019. Each cohort was divided into training (70%), validation (15%), and test (15%) subsets. Human abstractors identified the clinical rationale associated with treatment discontinuation events. Concatenated EMR notes were used to train high-dimensional logistic regression and convolutional neural network models. Kaplan-Meier analyses were used to compare toxicity incidence and PFS estimated by our NLP models to estimates generated by manual labeling and time-to-treatment discontinuation (TTD). RESULTS Our best high-dimensional logistic regression models identified toxicity events in early-stage patients with an area under the curve of the receiver-operator characteristic of 0.857 ± 0.014 (standard deviation) and progression events in metastatic patients with an area under the curve of 0.752 ± 0.027 (standard deviation). NLP-extracted toxicity incidence and PFS curves were not significantly different from manually extracted curves ( P = .95 and P = .67, respectively). By contrast, TTD overestimated toxicity in early-stage patients ( P < .001) and underestimated PFS in metastatic patients ( P < .001). Additionally, we tested an extrapolation approach in which 20% of the metastatic cohort were labeled manually, and NLP algorithms were used to abstract the remaining 80%. This extrapolated outcomes approach resolved PFS differences between receptor subtypes ( P < .001 for hormone receptor+/human epidermal growth factor receptor 2− v human epidermal growth factor receptor 2+ v triple-negative) that could not be resolved with TTD. CONCLUSION NLP models are capable of abstracting treatment discontinuation rationale with minimal manual labeling.

PURPOSE Knowing the treatments administered to patients with cancer is important for treatment planning and correlating treatment patterns with outcomes for personalized medicine study. However, existing methods to identify treatments are often lacking. We develop a natural language processing approach with structured electronic medical records and unstructured clinical notes to identify the initial treatment administered to patients with cancer. METHODS We used a total number of 4,412 patients with 483,782 clinical notes from the Stanford Cancer Institute Research Database containing patients with nonmetastatic prostate, oropharynx, and esophagus cancer. We trained treatment identification models for each cancer type separately and compared performance of using only structured, only unstructured ( bag-of-words, doc2vec, fasttext), and combinations of both ( structured + bow, structured + doc2vec, structured + fasttext). We optimized the identification model among five machine learning methods (logistic regression, multilayer perceptrons, random forest, support vector machines, and stochastic gradient boosting). The treatment information recorded in the cancer registry is the gold standard and compares our methods to an identification baseline with billing codes. RESULTS For prostate cancer, we achieved an f1-score of 0.99 (95% CI, 0.97 to 1.00) for radiation and 1.00 (95% CI, 0.99 to 1.00) for surgery using structured + doc2vec. For oropharynx cancer, we achieved an f1-score of 0.78 (95% CI, 0.58 to 0.93) for chemoradiation and 0.83 (95% CI, 0.69 to 0.95) for surgery using doc2vec. For esophagus cancer, we achieved an f1-score of 1.0 (95% CI, 1.0 to 1.0) for both chemoradiation and surgery using all combinations of structured and unstructured data. We found that employing the free-text clinical notes outperforms using the billing codes or only structured data for all three cancer types. CONCLUSION Our results show that treatment identification using free-text clinical notes greatly improves upon the performance using billing codes and simple structured data. The approach can be used for treatment cohort identification and adapted for longitudinal cancer treatment identification.

PURPOSE To inform precision oncology, methods are needed to use electronic health records (EHRs) to identify patients with cancer who are experiencing clinical inflection points, consistent with worsening prognosis or a high propensity to change treatment, at specific time points. Such patients might benefit from real-time screening for clinical trials. METHODS Using serial unstructured imaging reports for patients with solid tumors or lymphoma participating in a single-institution precision medicine study, we trained a deep neural network natural language processing (NLP) model to dynamically predict patients' prognoses and propensity to start new palliative-intent systemic therapy within 30 days. Model performance was evaluated using Harrell's c-index (for prognosis) and the area under the receiver operating characteristic curve (AUC; for new treatment and new clinical trial enrollment). Associations between model outputs and manual annotations of cancer progression were also evaluated using the AUC. RESULTS A deep NLP model was trained and evaluated using 302,688 imaging reports for 16,780 patients. In a held-out test set of 34,770 reports for 1,952 additional patients, the model predicted survival with a c-index of 0.76 and initiation of new treatment with an AUC of 0.77. Model-generated prognostic scores were associated with annotation of cancer progression on the basis of manual EHR review (n = 1,488 reports for 110 patients with lung or colorectal cancer) with an AUC of 0.78, and predictions of new treatment were associated with annotation of cancer progression on the basis of manual EHR review with an AUC of 0.84. CONCLUSION Training a deep NLP model to identify clinical inflection points among patients with cancer is feasible. This approach could identify patients who may benefit from real-time targeted clinical trial screening interventions at health system scale.

PURPOSE To facilitate identification of clinical trial participation candidates, we developed a machine learning tool that automates the determination of a patient's metastatic status, on the basis of unstructured electronic health record (EHR) data. METHODS This tool scans EHR documents, extracting text snippet features surrounding key words (such as metastatic, progression, and local). A regularized logistic regression model was trained and used to classify patients across five metastatic categories: highly likely and likely positive, highly likely and likely negative, and unknown. Using a real-world oncology database of patients with solid tumors with manually abstracted information as reference, we calculated sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). We validated the performance in a real-world data set, evaluating accuracy gains upon additional user review of tool's outputs after integration into clinic workflows. RESULTS In the training data set (N = 66,532), the model sensitivity and specificity (% [95% CI]) were 82.4 [81.9 to 83.0] and 95.5 [95.3 to 96.7], respectively; the PPV was 89.3 [88.8 to 90.0], and the NPV was 94.0 [93.8 to 94.2]. In the validation sample (n = 200 from five distinct care sites), after user review of model outputs, values increased to 97.1 [85.1 to 99.9] for sensitivity, 98.2 [94.8 to 99.6] for specificity, 91.9 [78.1 to 98.3] for PPV, and 99.4 [96.6 to 100.0] for NPV. The model assigned 163 of 200 patients to the highly likely categories. The error prevalence was 4% before and 2% after user review. CONCLUSION This tool infers metastatic status from unstructured EHR data with high accuracy and high confidence in more than 75% of cases, without requiring additional manual review. By enabling efficient characterization of metastatic status, this tool could mitigate a key barrier for patient ascertainment and clinical trial participation in community clinics.

The College of American Pathologists Cancer Protocols have offered guidance to pathologists for standard cancer pathology reporting for more than 35 years. The adoption of computer readable versions of these protocols by electronic health record and laboratory information system (LIS) vendors has provided a mechanism for pathologists to report within their LIS workflow, in addition to enabling standardized structured data capture and reporting to downstream consumers of these data such as the cancer surveillance community. This paper reviews the history of the Cancer Protocols and electronic Cancer Checklists, outlines the current use of these critically important cancer case reporting tools, and examines future directions, including plans to help improve the integration of the Cancer Protocols into clinical, public health, research, and other workflows.

BackgroundApproaches are needed to better delineate the continuum of opioid misuse that occurs in hospitalized patients. A prognostic enrichment strategy with latent class analysis (LCA) may facilitate treatment strategies in subtypes of opioid misuse. We aim to identify subtypes of patients with opioid misuse and examine the distinctions between the subtypes by examining patient characteristics, topic models from clinical notes, and clinical outcomes.MethodsThis was an observational study of inpatient hospitalizations at a tertiary care center between 2007 and 2017. Patients with opioid misuse were identified using an operational definition applied to all inpatient encounters. LCA with eight class-defining variables from the electronic health record (EHR) was applied to identify subtypes in the cohort of patients with opioid misuse. Comparisons between subtypes were made using the following approaches: (1) descriptive statistics on patient characteristics and healthcare utilization using EHR data and census-level data; (2) topic models with natural language processing (NLP) from clinical notes; (3) association with hospital outcomes.FindingsThe analysis cohort was 6,224 (2.7% of all hospitalizations) patient encounters with opioid misuse with a data corpus of 422,147 clinical notes. LCA identified four subtypes with differing patient characteristics, topics from the clinical notes, and hospital outcomes. Class 1 was categorized by high hospital utilization with known opioid-related conditions (36.5%); Class 2 included patients with illicit use, low socioeconomic status, and psychoses (12.8%); Class 3 contained patients with alcohol use disorders with complications (39.2%); and class 4 consisted of those with low hospital utilization and incidental opioid misuse (11.5%). The following hospital outcomes were the highest for each subtype when compared against the other subtypes: readmission for class 1 (13.9% vs. 10.5%, pConclusionsA 4-class latent model was the most parsimonious model that defined clinically interpretable and relevant subtypes for opioid misuse. Distinct subtypes were delineated after examining multiple domains of EHR data and applying methods in artificial intelligence. The approach with LCA and readily available class-defining substance use variables from the EHR may be applied as a prognostic enrichment strategy for targeted interventions.

Background While evidence-based psychotherapy (EBP) for posttraumatic stress disorder (PTSD) is a first-line treatment, its real-world effectiveness is unknown. We compared cognitive processing therapy (CPT) and prolonged exposure (PE) each to an individual psychotherapy comparator group, and CPT to PE in a large national healthcare system. Methods We utilized effectiveness and comparative effectiveness emulated trials using retrospective cohort data from electronic medical records. Participants were veterans with PTSD initiating mental healthcare (N = 265 566). The primary outcome was PTSD symptoms measured by the PTSD Checklist (PCL) at baseline and 24-week follow-up. Emulated trials were comprised of ‘person-trials,’ representing 112 discrete 24-week periods of care (10/07–6/17) for each patient. Treatment group comparisons were made with generalized linear models, utilizing propensity score matching and inverse probability weights to account for confounding, selection, and non-adherence bias. Results There were 636 CPT person-trials matched to 636 non-EBP person-trials. Completing ⩾8 CPT sessions was associated with a 6.4-point greater improvement on the PCL (95% CI 3.1–10.0). There were 272 PE person-trials matched to 272 non-EBP person-trials. Completing ⩾8 PE sessions was associated with a 9.7-point greater improvement on the PCL (95% CI 5.4–13.8). There were 232 PE person-trials matched to 232 CPT person-trials. Those completing ⩾8 PE sessions had slightly greater, but not statistically significant, improvement on the PCL (8.3-points; 95% CI 5.9–10.6) than those completing ⩾8 CPT sessions (7.0-points; 95% CI 5.5–8.5). Conclusions PTSD symptom improvement was similar and modest for both EBPs. Although EBPs are helpful, research to further improve PTSD care is critical.

ObjectiveAlcohol misuse is present in over a quarter of trauma patients. Information in the clinical notes of the electronic health record of trauma patients may be used for phenotyping tasks with natural language processing (NLP) and supervised machine learning. The objective of this study is to train and validate an NLP classifier for identifying patients with alcohol misuse.Materials and MethodsAn observational cohort of 1422 adult patients admitted to a trauma center between April 2013 and November 2016. Linguistic processing of clinical notes was performed using the clinical Text Analysis and Knowledge Extraction System. The primary analysis was the binary classification of alcohol misuse. The Alcohol Use Disorders Identification Test served as the reference standard.ResultsThe data corpus comprised 91 045 electronic health record notes and 16 091 features. In the final machine learning classifier, 16 features were selected from the first 24 hours of notes for identifying alcohol misuse. The classifier’s performance in the validation cohort had an area under the receiver-operating characteristic curve of 0.78 (95% confidence interval [CI], 0.72 to 0.85). Sensitivity and specificity were at 56.0% (95% CI, 44.1% to 68.0%) and 88.9% (95% CI, 84.4% to 92.8%). The Hosmer-Lemeshow goodness-of-fit test demonstrates the classifier fits the data well (P = .17). A simpler rule-based keyword approach had a decrease in sensitivity when compared with the NLP classifier from 56.0% to 18.2%.ConclusionsThe NLP classifier has adequate predictive validity for identifying alcohol misuse in trauma centers. External validation is needed before its application to augment screening.

Advance care planning (ACP) informs future medical decision-making, especially for patients with advanced age or serious illness. For clinicians to act on these preferences, or continue the ACP conversation as illness progresses, documentation of ACP discussions must be readily accessible within the electronic health record (EHR).Develop an intervention to improve accessible ACP documentation for hospitalized patients and assess its impact on viewing and documentation of ACP conversations within a specific EHR location.Adult patients age 75 or older or with serious illness discharged during a two-year period were included. The EHR's ACP Navigator was targeted as the intended location for documenting ACP-related activities. We implemented a hospital-wide, multipronged intervention that included increased ACP Navigator visibility and a process for workflow-congruent ACP documentation. Accessible ACP documentation was measured by documentation within the ACP Navigator and was analyzed by interrupted time-series analysis. ACP Navigator access was measured by user audit logs.After the intervention, 6703 of 16,117 (41.6%) patient encounters had accessible ACP documentation, compared to 3689 of 13,143 (28.1%) preintervention (P.001). In the intervention's first month, accessible ACP documentation increased 5.3% (P.001, CI 2.9%-7.6%), followed by a 1.3% monthly increase relative to the preintervention period (P.001, CI 1.0%-1.6%). ACP Navigator access for patients with ACP documentation increased in the intervention period (52.2% vs. 39.8%, P.001).An institution-wide intervention significantly increased accessible ACP documentation within a centralized location of the EHR. EHR usability changes improved rates of accessible ACP documentation and subsequent views of this documentation.

BACKGROUND: There are few studies examining the usage and utility of patient portals among seriously ill and end-of-life populations and their caregivers. OBJECTIVE: The aim of this study was to describe portal user characteristics among patients and their caregivers (proxy login) at two time points: (1) the 12 months following an electronic medical record flag for serious illness and (2) during the last 12 months of life. METHODS: A retrospective cohort analysis of Kaiser Permanente Colorado (KPCO) patients with serious illness, as defined by Kaiser Permanente's prognostic algorithm, and their proxy caregivers was performed for the two time periods. Use was characterized as (1) the discrete number of days the portal was used and (2) the number of days that portal features were accessed. Differences in use by user characteristics were assessed. RESULTS: Patients flagged for serious illness (N = 6129) were 70.4 ± 14.2 years of age, and used the portal on average 50.4 days. Patients (N = 6517) in the last year of life were 76.7 ± 13.7 years of age and used the portal on average 43 days. Caregiver proxy use of the portal was low in both cohorts. Patients who were older, female, non-White, and healthier were less likely to use the portal. CONCLUSIONS: In comparison with overall KPCO portal use and recent patient portal studies examining use patterns, patient portal use was high among patients flagged with serious illness and nearing the end of life. However, because use was associated with age, gender, and race, addressing barriers to portal adoption among underserved populations and caregiver proxies is key to better leveraging patient portal systems for palliative and end-of-life care.

Traumatic injuries remain one of the leading causes of death in the United States. Patients who survive traumatic injuries but return to the emergency department with repeat injuries are said to suffer from injury recidivism. Numerous studies have described trends in injury recidivism using trauma registry and survey data. To our knowledge, no prior study has leveraged electronic medical record (EMR) data to characterize injury recidivism. The EMR is potentially more comprehensive as it contains details of patients who visited the emergency department after injury but did not meet the criteria for inclusion in the trauma registry. Such injuries could be predictive of future recidivism. We therefore aimed to describe patterns of injury recidivism seen at a Level 1 trauma center using the EMR.A retrospective review was conducted of all injury-related encounters between January 2016 and December 2019. Manual review was conducted of all recidivistic encounters with11 months between encounters to ensure the recidivistic encounter was not a sequela of the index visit. A general estimating equation logistic regression adjusted for age, race, sex, and insurance payor, estimated odds ratios (ORs) and 95% confidence intervals (CIs) for the association between injury mechanism and odds of recidivistic encounter.A total of 20,566 index encounters was included during the study period. Of the 20,566 encounters, 7.6% (n = 1570) had a recidivistic encounter during the study period, half of which (n = 781) occurred within the first year of the index encounter. An over two-fold increased odds of recidivism was observed for blunt assault encounters (OR 2.53, 95% CI 2.03-3.15) and unintentional falls (OR 2.10, 95% CI 1.76-2.52). For both mechanisms, this increase was observed across the three years following the index encounter.Our study found that patients with assault injuries have the highest odds of injury recidivism and assault-related recidivistic encounters. These results demonstrate the feasibility and utility of incorporating EMR data, and suggest that the development of targeted interventions focused on mitigating assault injuries, such as hospital-based violence intervention programs, should be considered in our region.

Objective: To establish a structured and integrated platform of clinical data and biobank data, and a client to retrieve these data. Study Design: Initially, the hospital information system (HIS) a...

OBJECTIVES We investigated changes in prescriptions for antimicrobial agents to treat children with acute otitis media (AOM). METHODS A descriptive study using an electronic medical record database. Of 199,896 patients enrolled between 2001 and 2019, a total of 10,797 were aged

Aim: An augmented intelligence tool to predict short-term mortality risk among patients with cancer could help identify those in need of actionable interventions or palliative care services. Patients & methods: An algorithm to predict 30-day mortality risk was developed using socioeconomic and clinical data from patients in a large community hematology/oncology practice. Patients were scored weekly; algorithm performance was assessed using dates of death in patients’ electronic health records. Results: For patients scored as highest risk for 30-day mortality, the event rate was 4.9% (vs 0.7% in patients scored as low risk; a 7.4-times greater risk). Conclusion: The development and validation of a decision tool to accurately identify patients with cancer who are at risk for short-term mortality is feasible.

Relative value units (RVUs) are utilized to evaluate physician productivity in many fields, including radiology. The goal of this paper is to use the electronic medical record (EMR) to evaluate whether the current RVU system allows for fair comparison between radiologists' time effort.The study was approved by the local Institutional Review Board (IRB). Over 600,000 radiology studies with unique current procedural terminology (CPT) codes were evaluated, and after exclusion of studies interpreted in conjunction with trainees or interpreted using other software systems, a total of 241,627 studies remained. The median 25%ile, 50%ile, 65%ile, 75%ile and modal study ascribable times (SATs) for each CPT code was calculated across all radiologists. To evaluate the potential bias incurred using the current RVU system, the number of days required to achieve the Association of Administrators in Academic Radiology AAARAD 65%ile were calculated.RVU values were positively correlated with SATs (r = 0.69-0.71, p 0.001). The variability in the radiologists' time to achieve the AAARAD 65%ile benchmark was highest for musculoskeletal imaging, and lowest for thoracic imaging. The discrepancy in the number of days of work required to achieve the AAARAD 65%ile benchmark was 141.1% (197.7 days) for musculoskeletal imaging, 107.5% (161.9 days) for neuroimaging, 89.6% (185.9 days) for body imaging, and 72.2% (84.0 days) for thoracic imaging.The current RVU system is not strongly correlated with radiologist effort measured by radiologists' time. A time-based metric is more representative of radiologist work. However, there is no perfect method to measure radiologists' work.

BACKGROUND: The transport of the inpatients to and from locations inside the hospital can vary in complexity depending on the patient location, status, and logistical needs. Most transport systems have not developed at the same speed as other medically related technologies. We conducted a pilot study of a new automated transport system for patients within the hospital. METHODS: Our innovative system was introduced in January 2020. We present a retrospective case review of all in-patient transport request during April 15, 2020 through May 30, 2020 at the Shaare Zedek Medical Center, Jerusalem, Israel. The system is fully automated and works via smartphone and electronic medical record integration. Transfer requests are processed on the basis of priority, proximity, and availably. RESULTS: During the study period there were 15, 581 transfer requests. Mean times to hospital destinations ranged from 9:25 to 28:02 minutes. Overall, mean times were quicker for emergency and surgical services. Trip times by priority code were likely influence by unmeasured confounders. There were no reported patient identification adverse events. Peak requests occurred during 0900-1500, and at the beginning of the week. CONCLUSION: Our automated in-patient transfer system appears to be efficient, safe, well received, and capable of servicing our large tertiary care medical center. Future controlled studies are needed to assess efficacy, adverse events, and clinical outcomes.

Objective The aim of the current study was to evaluate primary care pediatrician (PCP) adoption of an electronic health record (EHR) documentation tool and their delivery of a behavioral activation (BA) intervention within their routine practice with adolescents who screened positive for depression. Methods We used the RE-AIM framework to describe PCP adoption and implementation of EHR documentation tools and brief evidence-based protocols. Utilization was assessed using a customized toolbar (ie, actions toolbar) via retrospective chart review. A pre-post design was used to measure changes in PCP-reported knowledge, comfort, and feasibility managing depression before and after they were trained. A mixed-effects logistic regression model was used to analyze associations of resource utilization with depression severity. Results PCPs used the actions toolbar to document responses to elevated Patient Health Questionnaire (PHQ-9) scores for 29.80% of encounters. The PCPs utilized the BA protocol for 10.5% of encounters with elevated PHQ-9 scores. Higher depression severity categories were associated with significantly higher odds of utilization relative to mild severity. The training was rated highly acceptable and PCPs reported significant post-training increases in comfort and feasibility. Conclusion This is the first study of its kind to implement and evaluate PCP utilization of an EHR documentation tool aimed to improve delivery of an evidence-based intervention for adolescent depression. Teaching PCPs to implement brief interventions has potential to increase access to evidence-based care; however, large-scale practice change requires an effective implementation strategy that does not increase provider burden and is fully integrated into physician documentation and workflow.

Introduction The use of an electronic health record may create unanticipated consequences for emergency care delivery. We sought to describe emergency department nursing task distribution and the use of the electronic health record. Methods This was a prospective observational study of nurses in the emergency department using a time-and-motion methodology. Three trained research assistants conducted 1:1 observations between March and September 2019. Nurse tasks were classified into 6 established categories: electronic health record, direct/indirect patient care, communication, personal time, and other. Nurses’ perceived workload was assessed using the National Aeronautics and Space Administration (NASA) Task Load Index. Results Twenty-three observations were conducted over 46 hours. Overall, nurses spent 27% of their time on electronic health record tasks, 25% on direct patient care, 17% on personal time, 15% on indirect patient care, and 6% on communication. During morning (7 am -12 pm ) and afternoon shifts (12 pm -3 pm ), the use of the health record was the most commonly performed task, whereas indirect patient care was the task most performed during evening shifts (3 pm -12 pm ). Using the National Aeronautics and Space Administration (NASA) Task Load Index, nurses reported an increase in mental demand and effort during afternoon shifts compared with morning shifts. Discussion We observed that emergency nurses spent more time using the electronic health record as compared to other tasks. Increased usability of the electronic health record, particularly during high occupancy periods, may be a target for improvement.

Background It is important to document the domains surrounding end-of-life (EOL) care in the electronic health record (EHR). No pediatric navigator exists for these purposes. Measures Medical charts were reviewed for documentation surrounding code status and care at the time of death from January 2017 to June 2019. Intervention Creation of a navigator in the EHR to consolidate advance care planning documents, code status orders and notes and EOL flowsheets. Outcomes After implementing the navigator, 96% code status changes had supporting documentation, an increase of 35%. The percentage of deaths supported by a psychosocial team (social worker, chaplain and certified child life specialist) increased by 25%. Post-mortem documentation became electronic. Patient level metrics began to be electronically collected. Conclusions/Lessons Learned Little has been published regarding use of the EHR to consolidate EOL documentation in pediatrics. Development of a systematic approach to documentation is critical to providing EOL care and standardizing care delivered.

Several recent trials have evaluated invasive versus medical therapy for stable ischemic heart disease. Importantly, patients with significant left main coronary stenosis (LMCS) were excluded from these trials. In the ISCHEMIA trial, these patients were identified by a coronary CT angiogram (CCTA), which adds time, expense, and contrast exposure. We tested whether a coronary artery calcium scan (CACS), a simpler, less expensive test, could replace CCTA to exclude significant LMCS.We hypothesized that patients with ≥50% LMCS would have a LM CACS score0. As a corollary, we postulated that a LM CACS = 0 would exclude patients with LMCS. To test this, we searched Intermountain Healthcare's electronic medical records database for all adult patients who had undergone non-contrast cardiac CT for quantitative CACS scoring prior to invasive coronary angiography (ICA). Patients aged50 and those with a heart transplant were excluded. Cases with incomplete (qualitative) angiographic reports for LMCS and those with incomplete or discrepant LM CACS results were reviewed and reassessed blinded to CACS or ICA findings, respectively.Among 669 candidate patients with CACS followed by ICA, 36 qualifying patients were identified who had a quantitative CACS score and LMCS ≥ 50%. Their age averaged 71.8 years, and 81% were men. Angiographic LMCS averaged 72% (range 50%-99%). Median time between CACS and ICA was 6 days. Total CACS score averaged 2,383 Agatston Units (AU), range 571-6,636. LM CACS score averaged 197 AU, range 31-610. Importantly, no LMCS patient had a LM CACS score of 0 vs 57% (362/633) of non-LMCS controls (P.00001).Our results support the hypothesis that an easily administered, inexpensive, low radiation CACS can identify a large subset of patients with a very low risk of LMCS who would not have the need for routine CCTA. Using CACS to exclude LMCS may efficiently allow for safe implementation of an initial medical therapy strategy of patients with stable ischemic heart disease in clinical practice. These promising results deserve validation in larger data sets.

PURPOSE: Chemotherapy-induced nausea and vomiting (CINV) is a common cause of unplanned healthcare utilization. The University of Michigan Rogel Cancer Center initiated the chemotherapy remote care monitoring program (CRCMP) to proactively identify patients experiencing CINV and intervene before the need for urgent evaluation. METHODS: High-risk patients for CINV are identified by neurokinin-1 (NK-1) antagonist administration, enrolled in the CRCMP, and received a daily text message survey for 7 days after chemotherapy administration to report symptoms. Responses above a set threshold trigger a message to the team pharmacist for intervention. The primary outcome of 14-day unplanned healthcare use was evaluated before and after CRCMP implementation. RESULTS: In 8 months, 652 patients received an NK-1 antagonist (2,244 cycles) and 387 patients were enrolled in the CRCMP (59%). Text message response rate was 94%. Clinical pharmacists provided 248 interventions in 121 patient episodes meeting threshold criteria. Fourteen-day unplanned healthcare use was decreased in the CRCMP-enrolled NK-1 episodes (6.68% v 4.53%, P = .02). Admissions were numerically lower for those enrolling in CRCMP when only admissions for nausea were considered (0.63% v 0.35%, P = .33). CONCLUSION: The CRCMP allowed for real-time management of patient-reported CINV symptom burden based on patient-reported outcomes (PROs) and an electronic medical record-integrated SMS text questionnaire. Clinical pharmacists were key team members to manage patient symptoms. Enrollment in CRCMP significantly reduced overall unplanned healthcare utilization. Although these changes were numerically small, any reduction in unnecessary care utilizing PROs can contribute to high-value care for patients with cancer.

Introduction As use of electronic medical records (EMRs) increases in healthcare, multiple accreditation organizations recommend training future clinicians on EMRs as part of the academic curriculum. Therefore, some pharmacy programs now utilize an academic EMR. Our objective was to examine pharmacy instructors' early experiences with a commercially available teaching EMR (tEMR) in order to identify current and future priorities along with benefits and barriers to academic EMR use in pharmacy education. Methods We conducted semi-structured interviews and a web-based survey with pharmacy instructors. Interview and survey data analyses consisted of hybrid inductive/deductive approaches to coding data and descriptive statistics, respectively. We mapped themes to the Pharmacists' Patient Care Process (PPCP). Results Seven instructors participated in interviews, and 10 completed the survey. We identified 25 interview themes that were generally congruent with survey responses. A noted benefit of the tEMR was providing a large repository of real-life patient data. Instructors' current priorities for tEMR activities primarily aligned with “collect,” “assess,” and “plan” steps of the PPCP. One reported barrier was insufficient training regarding the tEMR capabilities. Instructors offered innovative ideas for future academic EMR use, including pharmacokinetics, incident reporting, and longitudinal cases. Conclusions Study findings are most applicable to pharmacy programs in the early stages of academic EMR adoption. We identified opportunities for pharmacy programs to facilitate innovative, future use of academic EMRs, including curriculum planning, so pharmacy students gain experience using EMR functions for each PPCP step and develop more advanced EMR skills.

BACKGROUND: A significant proportion of surgical inpatients is often admitted longer than necessary. Early identification of patients who do not need care that is strictly provided within hospitals would allow timely discharge of patients to a postoperative nursing home for further recovery. We aimed to develop a model to predict whether a patient needs hospital-specific interventional care beyond the second postoperative day.METHODS: This study included all adult patients discharged from surgical care in the surgical oncology department from June 2017 to February 2020. The primary outcome was to predict whether a patient still needs hospital-specific interventional care beyond the second postoperative day. Hospital-specific care was defined as unplanned reoperations, radiological interventions, and intravenous antibiotics administration. Different analytical methods were compared with respect to the area under the receiver-operating characteristics curve, sensitivity, specificity, positive predictive value, and negative predictive value.RESULTS: Each model was trained on 1,174 episodes. In total, 847 (50.5%) patients required an intervention during postoperative admission. A random forest model performed best with an area under the receiver-operating characteristics curve of 0.88 (95% confidence interval 0.83-0.93), sensitivity of 79.1% (95% confidence interval 0.67-0.92), specificity of 80.0% (0.73-0.87), positive predictive value of 57.6% (0.45-0.70) and negative predictive value of 91.7% (0.87-0.97).CONCLUSION: This proof-of-concept study found that a random forest model could successfully predict whether a patient could be safely discharged to a nursing home and does not need hospital care anymore. Such a model could aid hospitals in addressing capacity challenges and improve patient flow, allowing for timely surgical care.

Background The intent of this review is to discover the types of inquiry and range of objectives and outcomes addressed in studies of the impacts of Electronic Medical Record (EMR) implementations in limited resource settings in sub-Saharan Africa. Methods A state-of-the-art review characterized relevant publications from bibliographic databases and grey literature repositories through systematic searching, concept-mapping, relevance and quality filter optimization, methods and outcomes categorization and key article analysis. Results From an initial population of 749 domain articles published before February 2015, 32 passed context and methods filters to merit full-text analysis. Relevant literature was classified by type (e.g., secondary, primary), design (e.g., case series, intervention), focus (e.g., processes, outcomes) and context (e.g., location, organization). A conceptual framework of EMR implementation determinants (systems, people, processes, products) was developed to represent current knowledge about the effects of EMRs in resource-constrained settings and to facilitate comparisons with studies in other contexts. Discussion This review provides an overall impression of the types and content of health informatics articles about EMR implementations in sub-Saharan Africa. Little is known about the unique effects of EMR efforts in slum settings. The available reports emphasize the complexity and impact of social considerations, outweighing product and system limitations. Summative guides and implementation toolkits were not found but could help EMR implementers. Conclusion The future of EMR implementation in sub-Saharan Africa is promising. This review reveals various examples and gaps in understanding how EMR implementations unfold in resource-constrained settings; and opportunities for new inquiry about how to improve deployments in those contexts.

Objective Biomedical text summarization helps biomedical information seekers avoid information overload by reducing the length of a document while preserving the contents’ essence. Our systematic review investigates the most recent biomedical text summarization researches on biomedical literature and electronic health records by analyzing their techniques, areas of application, and evaluation methods. We identify gaps and propose potential directions for future research. Materials and Methods This review followed the PRISMA methodology and replicated the approaches adopted by the previous systematic review published on the same topic. We searched 4 databases (PubMed, ACM Digital Library, Scopus, and Web of Science) from January 1, 2013 to April 8, 2021. Two reviewers independently screened title, abstract, and full-text for all retrieved articles. The conflicts were resolved by the third reviewer. The data extraction of the included articles was in 5 dimensions: input, purpose, output, method, and evaluation. Results Fifty-eight out of 7235 retrieved articles met the inclusion criteria. Thirty-nine systems used single-document biomedical research literature as their input, 17 systems were explicitly designed for clinical support, 47 systems generated extractive summaries, and 53 systems adopted hybrid methods combining computational linguistics, machine learning, and statistical approaches. As for the assessment, 51 studies conducted an intrinsic evaluation using predefined metrics. Discussion and Conclusion This study found that current biomedical text summarization systems have achieved good performance using hybrid methods. Studies on electronic health records summarization have been increasing compared to a previous survey. However, the majority of the works still focus on summarizing literature.

Objective: To develop a predictive model for identifying patients at high risk of all-cause unplanned readmission within 30 days after discharge, using administrative data available before discharge. Materials and methods: Hospital administrative data of all adult admissions in three tertiary metropolitan hospitals in Australia between July 01, 2015, and July 31, 2016, were extracted. Predictive performance of four mixed-effect multivariable logistic regression models was compared and validated using a split-sample design. Diagnostic details (Charlson Comorbidity Index CCI, components of CCI, and primary diagnosis categorised into International Classification of Diseases chapters) were added gradually in the clinically simplified model with socio-demographic, index admission, and prior hospital utilisation variables. Results: Of the total 99 470 patients admitted, 5796 (5.8%) were re-admitted through the emergency department of three hospitals within 30 days after discharge. The clinically simplified model was as discriminative (C-statistic 0.694, 95% CI [0.681-0.706]) as other models and showed excellent calibration. Models with diagnostic details did not exhibit any substantial improvement in predicting 30-days unplanned readmission. Conclusion: We propose a 10-item predictive model to flag high-risk patients in a diverse population before discharge using readily available hospital administrative data which can easily be integrated into the hospital information system.

PURPOSE: Management of neovascular age-related macular degeneration (nAMD) has evolved over the last decade with several treatment regimens and different medications. This study describes the treatment patterns and, importantly, visual outcomes over ten years in a large cohort of patients. DESIGN: Retrospective analysis of electronic health records from 27 National Health Service (NHS) secondary care healthcare providers in the UK. PARTICIPANTS: Treatment-naïve patients receiving at least three intravitreal anti-vascular endothelial growth factor (VEGF) injections for nAMD in their first six months of follow-up were included. Patients with missing data for age or gender and those aged less than 55 were excluded. METHODS: Eyes with at least three years of follow-up were grouped by years of treatment initiation, and three-year outcomes were compared between the groups. Data were generated during routine clinical care between 09/2008 and 12/2018. MAIN OUTCOME MEASURES: Visual acuity, number of injections, number of visits. RESULTS: A total of 15,810 eyes of 13,705 patients receiving 194,904 injections were included. Visual acuity (VA) improved from baseline during the first year, but dropped thereafter, resulting in loss of visual gains. This trend remained consistent throughout the past decade. Although an increasing proportion of eyes remained in the driving standard, this was driven by better presenting visual acuities over the decade. The number of injections dropped substantially between the first and subsequent years, from a mean of 6.25 in year 1 to 3 in year 2 and 2.5 in year 3, without improvement over the decade. In a multivariable regression analysis, final VA improved by 0.24 letters for each year since 2008, and younger age and baseline VA were significantly associated with VA at three years. CONCLUSION: Our findings show that despite improvement in functional VA over the years, primarily driven by improving baseline VA, patients continue to lose vision after the first year of treatment, with only marginal change over the past decade. The data suggest that these results may be related to suboptimal treatment patterns, which have not improved over the years. Rethinking treatment strategies may be warranted, possibly on a national level or through the introduction of longer-acting therapies.

Objective Little is known regarding variation among electronic health record (EHR) vendors in quality performance. This issue is compounded by selection effects in which high-quality hospitals coalesce to a subset of market leading vendors. We measured hospital performance, stratified by EHR vendor, across 4 quality metrics. Materials and Methods We used data on 1272 hospitals in 2018 across 4 quality measures: Leapfrog Computerized Provider Order Entry/EHR Evaluation, Centers for Medicare and Medicaid Services Hospital Compare Star Ratings, Hospital-Acquired Condition (HAC) score, and Hospital Readmission Reduction Program (HRRP) ratio. We examined score distributions and used multivariable regression to evaluate the association between vendor and score, recovering partial R2 to assess the proportion of quality variation explained by vendor. Results We found significant variation across and within EHR vendors. The largest vendor, vendor A, had the highest mean score on the Leapfrog Computerized Provider Order Entry/EHR Evaluation and HRRP ratio, vendor G had the highest Hospital Compare score, and vendor F had the highest HAC score. In adjusted models, no vendor was significantly associated with higher performance on more than 2 measures. EHR vendor explained between 1.2% (HAC) and 7.6 (HRRP) of the variation in quality performance. Discussion No EHR vendor was associated with higher quality across all measures, and the 2 largest vendors were not associated with the highest scores. Only a small fraction of quality variation was explained by EHR vendor choice. Conclusions Top performance on quality measures can be achieved with any EHR vendor; much of quality performance is driven by the hospital and how it uses the EHR.

Background: Electronic health records (EHRs) may help enable reliable, rapid data management for many uses, such as facilitating communication of advance care planning (ACP). However, issues with validity and accuracy of EHRs hinder the use of ACP information for practical applications. Design: We present a cross-sectional pilot study of 433 older adults with cancer from three large health care systems, participating in an ongoing multisite pragmatic trial (4UH3AG060626-02). We compared data extracted from dedicated structured EHR fields for ACP to a chart review of corresponding ACP documentation contained in the medical chart. Results: Structured ACP data existed for 43.2% of patients and varied by site (25.7% −48.9%). Of the identified structured ACP data elements, 59.2% of recorded elements were correct, 23.7% were incorrect, and 17.1% were duplicates with heterogeneity across sites. Conclusion: Structured ACP data in EHRs were frequently incorrect. This represents a problem for patients and their families, as well as quality improvement and research efforts. Clinical Trials Registration: NCT03609177.

Objective Substance use screening in adolescence is unstandardized and often documented in clinical notes, rather than in structured electronic health records (EHRs). The objective of this study was to integrate logic rules with state-of-the-art natural language processing (NLP) and machine learning technologies to detect substance use information from both structured and unstructured EHR data. Materials and Methods Pediatric patients (10-20 years of age) with any encounter between July 1, 2012, and October 31, 2017, were included (n = 3890 patients; 19 478 encounters). EHR data were extracted at each encounter, manually reviewed for substance use (alcohol, tobacco, marijuana, opiate, any use), and coded as lifetime use, current use, or family use. Logic rules mapped structured EHR indicators to screening results. A knowledge-based NLP system and a deep learning model detected substance use information from unstructured clinical narratives. System performance was evaluated using positive predictive value, sensitivity, negative predictive value, specificity, and area under the receiver-operating characteristic curve (AUC). Results The dataset included 17 235 structured indicators and 27 141 clinical narratives. Manual review of clinical narratives captured 94.0% of positive screening results, while structured EHR data captured 22.0%. Logic rules detected screening results from structured data with 1.0 and 0.99 for sensitivity and specificity, respectively. The knowledge-based system detected substance use information from clinical narratives with 0.86, 0.79, and 0.88 for AUC, sensitivity, and specificity, respectively. The deep learning model further improved detection capacity, achieving 0.88, 0.81, and 0.85 for AUC, sensitivity, and specificity, respectively. Finally, integrating predictions from structured and unstructured data achieved high detection capacity across all cases (0.96, 0.85, and 0.87 for AUC, sensitivity, and specificity, respectively). Conclusions It is feasible to detect substance use screening and results among pediatric patients using logic rules, NLP, and machine learning technologies.

Background: The use of electronic health record (EHR) systems encourages and facilitates the use of data for the development and surveillance of quality indicators, including pain management. Aim: to conduct an integrative review on pain management research using data extracted from EHR in order to synthesize and analyze the following elements: pain management (assessments, interventions, and outcomes) and study results with potential clinical implications, data source, clinical sample characteristics, and method description. Design: An integrative review of the literature was undertaken to identify exemplars of scientific research studies that explore pain management using data from EHR, using Cooper’s framework. Results: Our search of 1,061 records from PubMed, Scopus, and Cinahl was narrowed down to 28 eligible articles to be analyzed. Conclusion: Results of this integrative review will make a critical contribution, assisting others in developing research proposals and sound research methods, as well as providing an overview of such studies over the past 10 years. Through this review it is therefore possible to guide new research on clinical pain management using EHR.

Objective To determine eligibility for discontinuation of cervical cancer screening. Methods Women aged 64 with employer-sponsored insurance enrolled in a national database between 2016 and 2018, and those aged 64–66 receiving primary care at a safety net health center in 2019 were included. Patients were evaluated for screening exit eligibility by current guidelines: no evidence of cervical cancer or HIV-positive status and no evidence of cervical precancer in the past 25 years, and had evidence of either hysterectomy with removal of the cervix or evidence of fulfilling screening exit criteria, defined as two HPV screening tests or HPV plus Pap co-tests or three Pap tests within the past 10 years without evidence of an abnormal result. Results Of the 590,901 women in the national claims database, 131,059 (22.2%) were eligible to exit due to hysterectomy (1.6%) or negative screening (20.6%). Of the 1544 women from the safety net health center, 528 (34.2%) were eligible to exit due to hysterectomy (9.3%) or negative screening (24.9%). Most women did not have sufficient data available to fulfill exit criteria: 382,509 (64.7%) in the national database and 875 (56.7%) in the safety net hospital system. Even among women with 10 years of insurance claims data, only 41.5% qualified to discontinue screening. Conclusions Examining insurance claims in a national database and electronic medical records at a safety net institution led to remarkably similar findings: two thirds of women fail to qualify for screening exit. Additional steps to ensure eligibility prior to screening exit may be necessary to decrease preventable cervical cancers among women aged >65. CLINICAL TRIAL REGISTRATION: N/A

In this observational study, we compared the prognostic ability of an electronic health record (EHR)-derived risk score, the Rothman Index (RI), automatically derived on admission, to the first 24-hour Sequential Organ Failure Assessment (SOFA) score for outcome prediction in the modern cardiac intensive care unit (CICU). We found that while the 24-hour SOFA score provided modestly superior discrimination for both in-hospital and CICU mortality, the RI upon CICU admission had better calibration for both outcomes. Given the ubiquitous nature of EHR utilization in the United States, the RI may become an important tool to rapidly risk stratify CICU patients within the ICU and improve resource allocation.

Introduction Case reports of immune checkpoint inhibitor (ICI) overlap syndrome of myasthenia gravis, myositis and myocarditis, are increasing in the published literature. This is a potentially fatal adverse event of ICIs and emergency physicians need to be familiar with this triad when patients present to the emergency department (ED). Methods We performed a retrospective chart review of the electronic medical record between September 1, 2016 to March 9, 2020. We identified patients with the overlap syndrome who presented to our ED. Results Seven patients were identified. Most were male and treated with a programmed cell death-1 inhibitor. Most patients presented with abnormal vital signs and the most common symptoms were ptosis, diplopia, dyspnea and fatigue. Most required supplemental oxygen and had a prolonged length of stay. All received steroids in addition to other immunomodulators. Only two patients died. Discussion Presence of one of the diagnosis should lead to evaluation for the others. Suspicion should be raised by patients presenting with ptosis, muscular weakness, fatigue and dyspnea. Early recognition of this triad can allow for early administration of high-dose glucocorticoids (1–2 mg/kg of prednisone or equivalent), which is the mainstay of treatment. However, it is likely that patients will need further immunomodulators and therefore, will need hospitalization. Conclusion Emergency physicians should be aware of this potentially lethal triad in cancer patients receiving ICIs. The life-saving interventions in the ED include recognizing the triad, airway support, administration of high-dose glucocorticoids, and early involvement of a multidisciplinary team.

Objectives Little is known regarding the differences in microbiology associated with cellulitis or abscess with or without lymphangitic streaking. The objective of our study is to assess whether there are differences in the pathogens identified from wound cultures of patients with paronychia with and without associated lymphangitis. Methods Retrospective cross-sectional study at a tertiary pediatric emergency department over 25 years. We opted to assess patients with paronychia of the finger, assuming that these cases will have a greater variety of causative pathogens compared to other cases of cellulitis and soft tissue abscess that are associated with nail biting. Case identification was conducted using a computerized text-screening search that was refined by manual chart review. We included patients from 1 month to 20 years of age who underwent an incision and drainage (I&D) of a paronychia and had a culture obtained. The presence or absence of lymphangitis was determined from the clinical narrative in the medical record. We excluded patients treated with antibiotics prior to I&D as well as immune-compromised patients. We used descriptive statistics for prevalence and χ2 tests for categorical variables. Results Two hundred sixty-six patients met inclusion criteria. The median age was 9.7 years [IQR 4.7, 15.4] and 45.1% were female. Twenty-two patients (8.3%) had lymphangitic streaking associated with their paronychia. Patients with lymphangitis streaking were similar to those without lymphangitis in terms of age and sex (p = 0.52 and p = 0.82, respectively). Overall, the predominant bacteria was MSSA (40%) followed by MRSA (26%). No significant differences were found between the pathogens in the 22 patients with associated lymphangitis compared to the 244 patients without. Conclusion Staphylococcus aureus represent the majority of pathogens in paronychia, although streptococcal species and gram-negative bacteria were also common. Among patients with paronychia of the finger, there seems to be no association between pathogen type and presence of lymphangitic streaking.

BACKGROUND: Probabilistic linkage can link patients from different clinical databases without the need for personal information. If accurate linkage can be achieved, it would accelerate the use of linked datasets to address important clinical and public health questions. OBJECTIVE: We developed a step-by-step process for probabilistic linkage of national clinical and administrative datasets without personal information, and validated it against deterministic linkage using patient identifiers. STUDY DESIGN AND SETTING: We used electronic health records from the National Bowel Cancer Audit and Hospital Episode Statistics databases for 10,566 bowel cancer patients undergoing emergency surgery in the English National Health Service. RESULTS: Probabilistic linkage linked 81.4% of National Bowel Cancer Audit records to Hospital Episode Statistics, vs. 82.8% using deterministic linkage. No systematic differences were seen between patients that were and were not linked, and regression models for mortality and length of hospital stay according to patient and tumour characteristics were not sensitive to the linkage approach. CONCLUSION: Probabilistic linkage was successful in linking national clinical and administrative datasets for patients undergoing a major surgical procedure. It allows analysts outside highly secure data environments to undertake linkage while minimizing costs and delays, protecting data security, and maintaining linkage quality.

Introduction: Computable phenotypes allow identification of well-defined patient cohorts from electronic health record data. Little is known about the accuracy of diagnostic codes for important clinical concepts in pediatric epilepsy, such as (1) risk factors like neonatal hypoxic-ischemic encephalopathy; (2) clinical concepts like treatment resistance; (3) and syndromes like juvenile myoclonic epilepsy. We developed and evaluated the performance of computable phenotypes for these examples using electronic health record data at one center. Methods: We identified gold standard cohorts for neonatal hypoxic-ischemic encephalopathy, pediatric treatment-resistant epilepsy, and juvenile myoclonic epilepsy via existing registries and review of clinical notes. From the electronic health record, we extracted diagnostic and procedure codes for all children with a diagnosis of epilepsy and seizures. We used these codes to develop computable phenotypes and evaluated by sensitivity, positive predictive value, and the F-measure. Results: For neonatal hypoxic-ischemic encephalopathy, the best-performing computable phenotype (HIE ICD-9 /10 and [brain magnetic resonance imaging (MRI) or electroencephalography (EEG) within 120 days of life] and absence of commonly miscoded conditions) had high sensitivity (95.7%, 95% confidence interval [CI] 85-99), positive predictive value (100%, 95% CI 95-100), and F measure (0.98). For treatment-resistant epilepsy, the best-performing computable phenotype (3 or more antiseizure medicines in the last 2 years or treatment-resistant ICD-10) had a sensitivity of 86.9% (95% CI 79-93), positive predictive value of 69.6% (95% CI 60-79), and F-measure of 0.77. For juvenile myoclonic epilepsy, the best performing computable phenotype (JME ICD-10) had poor sensitivity (52%, 95% CI 43-60) but high positive predictive value (90.4%, 95% CI 81-96); the F measure was 0.66. Conclusion: The variable accuracy of our computable phenotypes (hypoxic-ischemic encephalopathy high, treatment resistance medium, and juvenile myoclonic epilepsy low) demonstrates the heterogeneity of success using administrative data to identify cohorts important for pediatric epilepsy research.

Electronic health records (EHRs) are a rich source of data for researchers, but extracting meaningful information out of this highly complex data source is challenging. Phecodes represent one strategy for defining phenotypes for research using EHR data. They are a high-throughput phenotyping tool based on ICD (International Classification of Diseases) codes that can be used to rapidly define the case/control status of thousands of clinically meaningful diseases and conditions. Phecodes were originally developed to conduct phenome-wide association studies to scan for phenotypic associations with common genetic variants. Since then, phecodes have been used to support a wide range of EHR-based phenotyping methods, including the phenotype risk score. This review aims to comprehensively describe the development, validation, and applications of phecodes and suggest some future directions for phecodes and high-throughput phenotyping.

ObjectivesTo compare the incidence of psoriatic arthritis (PsA) in patients with psoriasis (PsO) according to different treatments for their skin: topics/no treatment, conventional disease-modifying antirheumatic drugs (DMARDs) (cDMARDs) or biological DMARDs (bDMARDs).MethodsPatients with PsO without PsA followed at a university hospital were included in this retrospective cohort study. Patients were classified according to their treatment in topics (topics, phototherapy or no treatment), cDMARDs (methotrexate and cyclosporine) and bDMARDs (tumour necrosis factor inhibitors (TNFi), interleukin 17 inhibitors (IL-17i) and IL-12-23i ((interleukin (IL) 12/IL-23 inhibitor))) groups. Incident cases of PsA were attributed to one treatment if developed during the administration of that treatment. A Cox proportional hazards model was used to evaluate the adjusted risk of PsA development by treatment group.Results1719 patients with PsO contributed a total of 14 721 patient/years (py). 1387 (81%) patients were in the topics, 229 (13%) in cDMARDs and 103 (6%) in the bDMARDs group. During follow-up, 239 patients (14%) developed PsA (231 under topics, six under cDMARDs and two under bDMARDs). Global incidence was 1.6 per 100 py. The risk of developing PsA in patients with PsO treated with bDMARDs was significantly lower (incidence rate ratio (IRR)=0.26; 95% CI 0.03 to 0.94; p=0.0111), compared with topics, but not compared with cDMARDs (IRR=0.35; 95% CI 0.035 to 1.96; p=0.1007). Adjusted Cox proportional hazards regression analysis showed that male sex, nail involvement and higher body max index were associated with increased risk of developing PsA, while biologics use was protective (HR: 0.19; 95% CI 0.05 to 0.81).ConclusionTreatment with biologics in patients with PsO reduced the risk of PsA development.

Objective Diagnostic errors are major contributors to preventable patient harm. We validated the use of an electronic health record (EHR)-based trigger (e-trigger) to measure missed opportunities in stroke diagnosis in emergency departments (EDs). Methods Using two frameworks, the Safer Dx Trigger Tools Framework and the Symptom-disease Pair Analysis of Diagnostic Error Framework, we applied a symptom–disease pair-based e-trigger to identify patients hospitalized for stroke who, in the preceding 30 days, were discharged from the ED with benign headache or dizziness diagnoses. The algorithm was applied to Veteran Affairs National Corporate Data Warehouse on patients seen between 1/1/2016 and 12/31/2017. Trained reviewers evaluated medical records for presence/absence of missed opportunities in stroke diagnosis and stroke-related red-flags, risk factors, neurological examination, and clinical interventions. Reviewers also estimated quality of clinical documentation at the index ED visit. Results We applied the e-trigger to 7,752,326 unique patients and identified 46,931 stroke-related admissions, of which 398 records were flagged as trigger-positive and reviewed. Of these, 124 had missed opportunities (positive predictive value for “missed” = 31.2%), 93 (23.4%) had no missed opportunity (non-missed), 162 (40.7%) were miscoded, and 19 (4.7%) were inconclusive. Reviewer agreement was high (87.3%, Cohen’s kappa = 0.81). Compared to the non-missed group, the missed group had more stroke risk factors (mean 3.2 vs 2.6), red flags (mean 0.5 vs 0.2), and a higher rate of inadequate documentation (66.9% vs 28.0%). Conclusion In a large national EHR repository, a symptom–disease pair-based e-trigger identified missed diagnoses of stroke with a modest positive predictive value, underscoring the need for chart review validation procedures to identify diagnostic errors in large data sets.

Objective To propose an algorithm that utilizes only timestamps of longitudinal electronic health record data to classify clinical deterioration events. Materials and methods This retrospective study explores the efficacy of machine learning algorithms in classifying clinical deterioration events among patients in intensive care units using sequences of timestamps of vital sign measurements, flowsheets comments, order entries, and nursing notes. We design a data pipeline to partition events into discrete, regular time bins that we refer to as timesteps. Logistic regressions, random forest classifiers, and recurrent neural networks are trained on datasets of different length of timesteps, respectively, against a composite outcome of death, cardiac arrest, and Rapid Response Team calls. Then these models are validated on a holdout dataset. Results A total of 6720 intensive care unit encounters meet the criteria and the final dataset includes 830 578 timestamps. The gated recurrent unit model utilizes timestamps of vital signs, order entries, flowsheet comments, and nursing notes to achieve the best performance on the time-to-outcome dataset, with an area under the precision-recall curve of 0.101 (0.06, 0.137), a sensitivity of 0.443, and a positive predictive value of 0. 092 at the threshold of 0.6. Discussion and Conclusion This study demonstrates that our recurrent neural network models using only timestamps of longitudinal electronic health record data that reflect healthcare processes achieve well-performing discriminative power.

ObjectiveWomen with heart failure with reduced ejection fraction (HFrEF) may reach optimal treatment effect at half of the guideline-recommended medication dose. This study investigates prescription practice and its relation with survival of patients with HF in daily care.MethodsElectronic health record data from 13 Dutch outpatient cardiology clinics were extracted for HF receiving at least one guideline-recommended HF medication. Dose changes over consecutive prescriptions were modelled using natural cubic splines. Inverse probability-weighted Cox regression was used to assess the relationship between dose (reference≥50% target dose) and all-cause mortality.ResultsThe study population comprised 561 women (29% HFrEF (ejection fraction (EF)ConclusionPatients with HF seen in outpatient cardiology clinics receive half of the guideline-recommended medication dose. Lower ACEI/ARB dose was associated with improved survival in women with HFrEF. These results underscore the importance of (re)defining optimal medical therapy for women with HFrEF.

Patients with nonvalvular atrial fibrillation (NVAF) and type 2 diabetes are at risk of kidney, limb, and ophthalmic complications. We evaluated the rate of these complications and death in patients with NVAF and type 2 diabetes prescribed rivaroxaban or warfarin.We analyzed Optum de-Identified electronic health record (EHR) data from 11/2010-12/2019. We included adults with NVAF and T2D newly initiated on rivaroxaban or warfarin with ≥12 months of prior EHR activity. Patients with another indication for anticoagulation, valve disease, history of end-stage renal disease, major adverse limb events (MALE), diabetic retinopathy or pregnancy were excluded. We evaluated the incidence rate of developing a composite outcome of40% decrease in estimated glomerular filtration incidence rate (eGFR) from baseline, eGFR15 mL/minute/1.73 m2, need for dialysis or kidney transplant, MALE, diabetic retinopathy or death. Overlap weighting was used to balance baseline characteristics between cohorts while preserving sample size. Hazard ratios with 95% confidence intervals were calculated using propensity score-overlap weighted Cox regression.We included 24,912 rivaroxaban and 58,270 warfarin users. The mean ± standard deviation (SD) CHA2DS2VASc score was 4.3 ± 1.5 and modified HASBLED score was 1.5 ± 0.8. Thirty percent of rivaroxaban patients were started on 15 mg once daily, with the rest prescribed 20 mg once daily. Warfarin patients had a mean time in therapeutic range of 47 ± 28%. Patients were followed for a mean of 2.89 ± 1.95 years. Rivaroxaban was associated with a reduced hazard of the composite outcome (HR = 0.93, 95%CI = 0.91-0.95; absolute risk reduction = 1.97 events per 1000 patient-years; number needed-to-treat = 51) versus warfarin. Rivaroxaban was also associated with significant reductions in the relative hazard of 40% decrease in eGFR from baseline (HR = 0.96), need for dialysis or renal transplant (HR = 0.81), and limb revascularization or major amputation (HR = 0.85). Death occurred at a lower incidence rate with rivaroxaban (HR = 0.92, 95%CI = 0.89-0.95).Rivaroxaban was associated with reduced incidence rates of kidney and limb complications, and death in NVAF patients with type 2 diabetes compared to warfarin.

To determine the effectiveness of 2 different continuous quality improvement interventions in an integrated community urology practice. We specifically assessed the impact of audited physician feedback on improving physicians' adoption of active surveillance for low-risk prostate cancer (CaP) and adherence to a prostate biopsy time-out intervention.The electronic medical records of Genesis Healthcare Partners were analyzed between August 24, 2011 and September 30, 2020 to evaluate the performance of 2 quality interventions: audited physician feedback to improve active surveillance adoption in low-risk CaP patients, and audited physician feedback to promote adherence to an electronic medical records embedded prostate biopsy time-out template. Physician and Genesis Healthcare Partners group adherence to each quality initiative was compared before and after each intervention type using ANOVA testing.For active surveillance, we consistently saw an increase in active surveillance adoption for low risk CaP patients in association with continuous audited feedback (P.001). Adherence to the prostate biopsy time-out template improved when audited feedback was provided (P.001).The implementation of clinical guidelines into routine clinical practice remains challenging and poses an obstacle to the improvement of United States healthcare quality. Continuous quality improvement should be a dynamic process, and in our experience, audited feedback coupled with education is most effective.

Background Within-consultation recruitment to primary care trials is challenging. Ensuring procedures are efficient and self-explanatory is the key to optimising recruitment. Trial recruitment software that integrates with the electronic health record to support and partially automate procedures is becoming more common. If it works well, such software can support greater participation and more efficient trial designs. An innovative electronic trial recruitment and outcomes software was designed to support recruitment to the Runny Ear randomised controlled trial, comparing topical, oral and delayed antibiotic treatment for acute otitis media with discharge in children. A qualitative evaluation investigated the views and experiences of primary care staff using this trial software. Methods Staff were purposively sampled in relation to site, role and whether the practice successfully recruited patients. In-depth interviews were conducted using a flexible topic guide, audio recorded and transcribed. Data were analysed thematically. Results Sixteen staff were interviewed, including GPs, practice managers, information technology (IT) leads and research staff. GPs wanted trial software that automatically captures patient data. However, the experience of getting the software to work within the limited and complex IT infrastructure of primary care was frustrating and time consuming. Installation was reliant on practice level IT expertise, which varied between practices. Although most had external IT support, this rarely included supported for research IT. Arrangements for approving new software varied across practices and often, but not always, required authorisation from Clinical Commissioning Groups. Conclusions Primary care IT systems are not solely under the control of individual practices or CCGs or the National Health Service. Rather they are part of a complex system that spans all three and is influenced by semi-autonomous stakeholders operating at different levels. This led to time consuming and sometimes insurmountable barriers to installation at the practice level. These need to be addressed if software supporting efficient research in primary care is to become a reality.

Background Integrating medical scribes with clinicians has been suggested to improve access, quality of care, enhance patient/clinician satisfaction, and increase productivity revenue. Objective Conduct a systematic review to evaluate the effects of medical scribes in emergency departments. Methods Electronic databases from 2010 through December 2019. Two individuals independently reviewed study eligibility, rated risk of bias, and determined overall certainty of evidence. Data abstracted included study and population characteristics, outcomes (efficiency, patient or clinician satisfaction, financial productivity, documentation quality, cost, and training time), and the effect of compensation structure, qualifications, duties, and setting on outcomes. Results Twenty studies (18 observational) were included; 12 from two institutions. All utilized in-person rather than virtual scribes. Fifteen were rated as serious or critical risk of bias; five were rated moderate. Findings indicate that scribes may increase patients seen per day and decrease length of stay; however, effects were small and may vary by setting and outcome measured (low certainty). Scribes may increase financial productivity; however, costs associated with developing, implementing, and maintaining scribe programs were not adequately reported. Results were mixed for door-to-room or door-to-provider time, patients left without being seen, and patient/clinician satisfaction. No studies examined the effects of scribes based on compensation structure, qualifications or duties. Conclusions Although information quality, quantity, and applicability are limited, in-person medical scribes may improve emergency department efficiency and financial productivity. There was no information on virtual scribes. There was little information on patient or clinician satisfaction, scribe documentation quality, or whether results vary by in-house vs. contracted hiring and training.

Human fatty acid synthase (FASN) is the sole cytosolic enzyme responsible for de novo lipid synthesis. FASN is essential for cancer cell survival and contributes to drug and radiation resistance by up-regulating DNA damage repair but not required for most non-lipogenic tissues. Thus, FASN is an attractive target for drug discovery. However, despite decades of effort in targeting FASN, no FASN inhibitors have been approved due to poor pharmacokinetics or toxicities. Here, we show that the FDA-approved proton pump inhibitors (PPIs) effectively inhibit FASN and suppress breast cancer cell survival. PPI inhibition of FASN leads to suppression of non-homologous end joining repair of DNA damages by reducing FASN-mediated PARP1 expression, resulting in apoptosis from oxidative DNA damages and sensitization of cellular resistance to doxorubicin and ionizing radiation. Mining electronic medical records of 6754 breast cancer patients showed that PPI usage significantly increased overall survival and reduced disease recurrence of these patients. Hence, PPIs may be repurposed as anticancer drugs for breast cancer treatments by targeting FASN to overcome drug and radiation resistance.