A Mixed Methods Analysis of Standardized Documentation of Serious Illness Conversations Within an Electronic Health Record Module During Hospitalization.

BACKGROUND: Analysis of documented Serious Illness Conversations (SICs) in the inpatient setting can help clinicians align management to address patient and caregiver needs. METHODS: We conducted a mixed methods analysis of the first instance of standardized documentation of a SIC within a structured module among hospitalized general medicine patients from 2018 to 2019. Percentage of documentations that included a description of patient or family understanding of the patient's medical condition and use of radio buttons to answer the "prognostic information shared," "hopes," and "worries" modules are reported. Using grounded theory approach, physicians analyzed free text entries to: "What is important to the patient/family?" and "Recommendations or next steps planned." RESULTS: Out of 5142 patients, 59 patients had a documented SIC. Patient or family understanding of the medical condition(s) was reported in 56 (95%). For "prognostic information shared," the most frequently selected radio buttons were: 49 (83%) incurable disease and 28 (48%) prognosis of weeks to months while those for "hopes" were: 52 (88%) be comfortable and 27 (46%) be at home and for "worries" were: 49 (83%) other physical suffering and 36 (61%) pain. Themes generated from entries to "What's important to patient/family?" included being with loved ones; comfort; mentally and physically present; and reliable care while those for "Recommendations" were coordinating support services; symptom management; and support and communication. CONCLUSIONS: SIC content indicated concern about pain and reliable care suggesting the complex, intensive nature of caring for seriously ill patients and the need to consider SICs earlier in the life course of patients.

Investigating racial and ethnic disparities in interhospital transfer within an academic integrated healthcare system: A matched cohort study.

The presence of racial and ethnic disparities in interhospital transfer (IHT) within integrated healthcare systems has not been fully explored. We matched Black and Latinx patients admitted to community hospitals in our integrated healthcare system between June 2015 and December 2019 to White patients by origin hospital, age, time of year, and disease severity. We performed conditional logistic regression models to determine if race or ethnicity was associated with IHT in one of the tertiary academic medical centers in the system, adjusting for covariates. The sample contained 107,895 admissions (82.6% White, 7.8% Black, and 9.6% Latinx). Transfer rates were 2.2% versus 2.2% after the Black/White match and 1.8% versus 1.8% after the Latinx/White match. After adjusting for covariates, there was no association between race or ethnicity and IHT (Black vs. White odds ratio [OR]: 0.87, 95% confidence interval [CI]: 0.72-1.07; Latinx vs. White OR: 1.05, 95% CI: 0.79-1.40). This may be due to reduced barriers to transfer with an integrated healthcare system.

Implementation of a standardised accept note to improve communication during inter-hospital transfer: a prospective cohort study.

IMPORTANCE: The transfer of patients between hospitals (interhospital transfer, IHT), exposes patients to communication errors and gaps in information exchange. OBJECTIVE: To design and implement a standardised accept note to improve communication during medical service transfers, and evaluate its impact on patient outcomes. DESIGN: Prospective interventional cohort study. SETTING: A 792-bed tertiary care hospital. PARTICIPANTS: All patient transfers from any acute care hospital to the general medicine, cardiology, oncology and intensive care unit (ICU) services between August 2020 and June 2022. INTERVENTIONS: A standardised accept note template was developed over a 9-month period with key stakeholder input and embedded in the electronic health record, completed by nurses within the hospital's Access Centre. MAIN OUTCOMES AND MEASURES: Primary outcome was clinician-reported medical errors collected via surveys of admitting clinicians within 72 hours after IHT patient admission. Secondary outcomes included clinician-reported failures in communication; presence and 'timeliness' of accept note documentation; patient length of stay (LOS) after transfer; rapid response or ICU transfer within 24 hours and in-hospital mortality. All outcomes were analysed postintervention versus preintervention, adjusting for patient demographics, diagnosis, comorbidity, illness severity, admitting service, time of year, hospital COVID census and census of admitting service and admitting team on date of admission. RESULTS: Of the 1004 and 654 IHT patients during preintervention and postintervention periods, surveys were collected on 735 (73.2%) and 462 (70.6%), respectively. Baseline characteristics were similar among patients in each time period and between survey responders and non-responders. Adjusted analyses demonstrated a 27% reduction in clinician-reported medical error rates postimplementation versus preimplementation (11.5 vs 15.8, adjusted OR (aOR) 0.73, 95% CI 0.53 to 0.99). Secondary outcomes demonstrated lower adjusted odds of clinician-reported failures in communication (aOR 0.88; 0.78 to 0.98) and rapid response/ICU transfer (aOR 0.57; 0.34 to 0.97), and improved presence (aOR 2.30; 1.75 to 3.02) and timeliness (-21.4 hours vs -8.7 hours, p<0.001) of accept note documentation. There were no significant differences in LOS or mortality. CONCLUSIONS AND RELEVANCE: Among 1658 medical patient transfers, implementing a standardised accept note was associated with improved presence and timeliness of accept note documentation, clinician-reported medical errors, failures in communication and clinical decline following transfer, suggesting that improving communication during IHT can improve patient outcomes.

Towards Medical Billing Automation: NLP for Outpatient Clinician Note Classification.

Objectives: Our primary objective was to develop a natural language processing approach that accurately predicts outpatient Evaluation and Management (E/M) level of service (LoS) codes using clinicians' notes from a health system electronic health record. A secondary objective was to investigate the impact of clinic note de-identification on document classification performance. Methods: We used retrospective outpatient office clinic notes from four medical and surgical specialties. Classification models were fine-tuned on the clinic notes datasets and stratified by subspecialty. The success criteria for the classification tasks were the classification accuracy and F1-scores on internal test data. For the secondary objective, the dataset was de-identified using Named Entity Recognition (NER) to remove protected health information (PHI), and models were retrained. Results: The models demonstrated similar predictive performance across different specialties, except for internal medicine, which had the lowest classification accuracy across all model architectures. The models trained on the entire note corpus achieved an E/M LoS CPT code classification accuracy of 74.8% (CI 95: 74.1-75.6). However, the de-identified note corpus showed a markedly lower classification accuracy of 48.2% (CI 95: 47.7-48.6) compared to the model trained on the identified notes. Conclusion: The study demonstrates the potential of NLP-based document classifiers to accurately predict E/M LoS CPT codes using clinical notes from various medical and procedural specialties. The models' performance suggests that the classification task's complexity merits further investigation. The de-identification experiment demonstrated that de-identification may negatively impact classifier performance. Further research is needed to validate the performance of our NLP classifiers in different healthcare settings and patient populations and to investigate the potential implications of de-identification on model performance.

The Association of Standardized Documentation of Serious Illness Conversations With Healthcare Utilization in Hospitalized Patients: A Propensity Score Matched Cohort Analysis.

Background: Serious Illness Conversations (SICs) conducted during hospitalization can lead to meaningful patient participation in the decision-making process affecting medical management. The aim of this study is to determine if standardized documentation of a SIC within an institutionally approved EHR module during hospitalization is associated with palliative care consultation, change in code status, hospice enrollment prior to discharge, and 90-day readmissions. Methods: We conducted retrospective analyses of hospital encounters of general medicine patients at a community teaching hospital affiliated with an academic medical center from October 2018 to August 2019. Encounters with standardized documentation of a SIC were identified and matched by propensity score to control encounters without a SIC in a ratio of 1:3. We used multivariable, paired logistic regression and Cox proportional-hazards modeling to assess key outcomes. Results: Of 6853 encounters (5143 patients), 59 (.86%) encounters (59 patients) had standardized documentation of a SIC, and 58 (.85%) were matched to 167 control encounters (167 patients). Encounters with standardized documentation of a SIC had greater odds of palliative care consultation (odds ratio [OR] 60.10, 95% confidence interval [CI] 12.45-290.08, P < .01), a documented code status change (OR 8.04, 95% CI 1.54-42.05, P = .01), and discharge with hospice services (OR 35.07, 95% CI 5.80-212.08, P < .01) compared to matched controls. There was no significant association with 90-day readmissions (adjusted hazard ratio [HR] .88, standard error [SE] .37, P = .73). Conclusions: Standardized documentation of a SIC during hospitalization is associated with palliative care consultation, change in code status, and hospice enrollment.

Ambulatory Medication Safety: Vaccination and Laboratory Screening for Patients Receiving Immunomodulatory Therapies.

OBJECTIVE: Immunomodulatory therapies improve the management of chronic diseases but can be associated with infectious risk. The present study was undertaken to examine the laboratory screening practices for hepatitis B virus (HBV), hepatitis C virus (HCV), and tuberculosis (TB) and rates of vaccination for pneumococcal and influenza in patients prescribed select immunosuppressive agents at our institution. METHODS: A retrospective analysis was conducted to review patients who were prescribed a select immunosuppressive over 3 years. Data were extracted from electronic health records to identify rates of screening and vaccination prior to initiation or at any time. Logistic regression models were developed to identify predictors of screening and vaccination. RESULTS: We identified 2,396 patients prescribed immunosuppressive medications by rheumatology (52.6%) and non-rheumatology specialties. Rates of screening at any time point were 84.5% (2,025 of 2,396) for HBV, 76.7% (1,838 of 2,396) for HCV, and 71.8% (1,720 of 2,396) for TB. Patients who had either in-system primary care providers (PCPs) or rheumatologists were more likely to receive pneumococcal vaccinations (odds ratio [OR] 1.98 [95% confidence interval (95% CI) 1.55-2.54] and OR 4.08 [95% CI 2.76-6.02], respectively). Patients with dermatologic (OR 1.67 [95% CI 1.14-2.45]) or rheumatologic providers (OR 2.5 [95% CI 1.86-3.36]) were more likely to be vaccinated for influenza. CONCLUSION: More than 70% of patients were screened for either HBV, HCV, or TB at some point. Rates of pneumococcal vaccination were better than rates of influenza vaccination. Patients with in-system PCPs were more likely to be screened and vaccinated. Establishing and executing consistent processes for screening and vaccination prior to immunosuppressive treatment remains a priority in ambulatory settings.

Who Gets (and Who Should Get) a Serious Illness Conversation in the Hospital? An Analysis of Readmission Risk Score in an Electronic Health Record.

Serious Illness Conversations (SICs) explore patients' prognostic awareness, hopes, and worries, and can help establish priorities for their care during and after hospitalization. While identifying patients who benefit from an SIC remains a challenge, this task may be facilitated by use of validated prediction scores available in most commercial electronic health records (EHRs), such as Epic's Readmission Risk Score (RRS). We identified the RRS on admission for all hospital encounters from October 2018 to August 2019 and measured the area under the receiver operating characteristic (AUROC) curve to determine whether RRS could accurately discriminate post discharge 6-month mortality. For encounters with standardized SIC documentation matched in a 1:3 ratio to controls by sex and age (±5 years), we constructed a multivariable, paired logistic regression model and measured the odds of SIC documentation per every 10% absolute increase in RRS. RRS was predictive of 6-month mortality with acceptable discrimination (AUROC .71) and was significantly associated with SIC documentation (adjusted OR 1.42, 95% CI 1.24-1.63). An RRS >28% used to identify patients with post discharge 6-month mortality had a high specificity (89.0%) and negative predictive value (NPV) (97.0%), but low sensitivity (25.2%) and positive predictive value (PPV) (7.9%). RRS may serve as a practical EHR-based screen to exclude patients not requiring an SIC, thereby leaving a smaller cohort to be further evaluated for SIC needs using other validated tools and clinical assessment.

Engaging Patients in the Use of Real-Time Electronic Clinical Data to Improve the Safety and Reliability of Their Own Care.

OBJECTIVES: There is considerable evidence that providing patients with access to their health information is beneficial, but there is limited evidence regarding the effect of providing real-time patient safety-related information on health outcomes. The aim of this study was to evaluate the association between use of an electronic patient safety dashboard (Safety Advisor) and health outcomes. METHODS: The Safety Advisor was implemented in 6 adult medicine units at one hospital in the United States. Study participants were asked to use the Safety Advisor, which provides real-time patient safety-related information through a Web-based portal. The primary outcome was the association between the application usage and health outcomes (readmission rate and mortality rate) per 3 different usage groups, and the secondary outcome was the association of Patient Activation Measure (PAM) scores with use. RESULTS: One hundred eighty-one participants were included for the data analysis. Approximately 90% of users accessed the application during the first 4 days of enrollment: 51.6% of users only accessed it on 1 day, whereas 5.8% used it more than 3 days. Patients who used the application more had lower 30-day readmission rates (P = 0.01) compared with the lower-usage group. The PAM scores for users of Safety Advisor (71.8) were higher than the nonpatient portal users (60.8, P < 0.0001). CONCLUSIONS: We found an association between the use of Safety Advisor and health outcomes. Differences in PAM scores between groups were statistically significant. A larger-scale randomized control trial is warranted to evaluate the impact on patient outcomes among a high-risk patient population.

Personal Formularies of Primary Care Physicians Across 4 Health Care Systems.

Importance: More conservative prescribing has the potential to reduce adverse drug events and patient harm and cost; however, no method exists defining the extent to which individual clinicians prescribe conservatively. One potential domain is prescribing a more limited number of drugs. Personal formularies-defined as the number and mix of unique, newly initiated drugs prescribed by a physician-may enable comparisons among clinicians, practices, and institutions. Objectives: To develop a method of defining primary care physicians' personal formularies and examine how they differ among primary care physicians at 4 institutions; evaluate associations between personal formularies and patient, physician, and practice site characteristics; and empirically derive and examine the variability of the top 200 core drugs prescribed at the 4 sites. Design, Setting, and Participants: This retrospective cohort study was conducted at 4 US health care systems among 4655 internal and family medicine physicians and 4 930 707 patients who had at least 1 visit to these physicians between January 1, 2017, and December 31, 2018. Exposures: Personal formulary size was defined as the number of unique, newly initiated drugs. Main Outcomes and Measures: Personal formulary size and drugs used, physician and patient characteristics, core drugs, and analysis of selected drug classes. Results: The study population included 4655 primary care physicians (2274 women [48.9%]; mean [SD] age, 48.5 [4.4] years) and 4 930 707 patients (16.5% women; mean [SD] age, 51.9 [8.3] years). There were 41 378 903 outpatient prescriptions written, of which 9 496 766 (23.0%) were new starts. Institution median personal formulary size ranged from 150 (interquartile range, 82.0-212.0) to 296 (interquartile range, 230.0-347.0) drugs. In multivariable modeling, personal formulary size was significantly associated with panel size (total number of unique patients with face-to-face encounters during the study period; 1.2 medications per 100 patients), physician's total number of encounters (5.7 drugs per 10% increase), and physician's sex (-6.2 drugs per 100 patients for female physicians). There were 1527 unique, newly prescribed drugs across the 4 sites. Fewer than half the drugs (626 [41.0%]) were used at every site. Physicians' prescribing of drugs from a pooled core list varied from 0% to 100% of their prescriptions. Conclusions and Relevance: Personal formularies, measured at the level of individual physicians and institutions, reveal variability in size and mix of drugs. Similarly, defining a list of commonly prescribed core drugs in primary care revealed interphysician and interinstitutional differences. Personal formularies and core medication lists enable comparisons and may identify outliers and opportunities for safer and more appropriate prescribing.

Comparison of Medication Alerts from Two Commercial Applications in the USA.

INTRODUCTION: Medication organizations across the USA have adopted electronic health records, and one of the most anticipated benefits of these was improved medication safety, but alert fatigue has been a major issue. OBJECTIVE: We compared the appropriateness of medication-related clinical decision support alerts triggered by two commercial applications: EPIC and Seegnal's platform. METHODS: This was a retrospective comparison of two commercial applications. We provided Seegnal with deidentified inpatient, outpatient, and inpatient genetic electronic medical record (EMR)-extracted datasets for 657, 2731, and 413 patients, respectively. Seegnal then provided the alerts that would have triggered, which we compared with those triggered by EPIC in clinical care. A random sample of the alerts triggered were reviewed for appropriateness, and the positive predictive value (PPV) and negative predictive value (NPV) were calculated. We also reviewed all the inpatient and outpatient charts for patients within our cohort who were receiving ten or more concomitant medications with alerts we found to be appropriate to assess whether any adverse events had occurred and whether Seegnal's platform could have prevented them. RESULTS: Results from EPIC and the Seegnal platform were compared based on alert load, PPV, NPV, and potential adverse events. Overall, compared with EPIC, the Seegnal platform triggered fewer alerts in the inpatient (1697 vs. 27,540), outpatient (2341 vs. 35,134), and inpatient genetic (1493 vs. 20,975) cohorts. The Seegnal platform had higher specificity in the inpatient (99 vs. 0.3%; p < 0.0001), outpatient (99 vs. 0.3%; p < 0.0001), and inpatient genetic (97.9 vs. 1.2%; p < 0.0001) groups and higher sensitivity in the inpatient (100 vs. 68.8%; p < 0.0001) and outpatient (88.6 vs.78.3%; p < 0.0001) groups but not in the inpatient genetic cohort (81 vs. 78.5%; p = 0.11). We identified 16 adverse events that occurred in the inpatient setting, 11 (69%) of which potentially could have been prevented with the Seegnal platform. CONCLUSIONS: Overall, the Seegnal platform triggered 94% fewer alerts than EPIC in the inpatient setting and 93% fewer in the outpatient setting, with much higher sensitivity and specificity. This application could substantially reduce alert fatigue and improve medication safety at the same time.

The Burden of Opioid-Related Adverse Drug Events on Hospitalized Previously Opioid-Free Surgical Patients.

OBJECTIVE: Opioid analgesics are a mainstay for acute pain management, but postoperative opioid administration has risks. We examined the prevalence, risk factors, and consequences of opioid-related adverse drug events (ORADEs) in a previously opioid-free surgical population. METHODS: A retrospective, observational, cohort study using administrative, billing, clinical, and medication administration data from two hospitals. Data were collected for all adult patients who were opioid-free at admission, underwent surgery between October 1, 2015, and September 30, 2016, and received postoperative opioids. Potential ORADEs were determined based on inpatient billing codes or postoperative administration of naloxone. We determined independent predictors of ORADE development using multivariable logistic regression. We measured adjusted inpatient mortality, hospital costs, length of hospital stay, discharge destination, and readmission within 30 days for patients with and without ORADEs. RESULTS: Among 13,389 hospitalizations where opioid-free patients had a single qualifying surgery, 12,218 (91%) received postoperative opioids and comprised the study cohort. Of these, we identified 1111 (9.1%) with a potential ORADE. Independent predictors of ORADEs included older age, several markers of disease severity, longer surgeries, and concurrent benzodiazepine use. Opioid-related adverse drug events were strongly associated with the route and duration of opioids administered postoperatively: 18% increased odds per day on intravenous opioids. In analyses adjusted for several covariates, presence of an ORADE was associated with 32% higher costs of hospitalization, 45% longer postoperative length of stay, 36% lower odds of discharge home, and 2.2 times the odds of death. CONCLUSIONS: We demonstrate a high rate and severe consequences of potential ORADEs in previously opioid-free patients receiving postoperative opioids. Knowledge of risk factors and predictors of ORADEs can help develop targeted interventions to minimize the development of these potentially dangerous and costly events.

Clinical Characteristics and Severity of COVID-19 Disease in Patients from Boston Area Hospitals.

We summarize key demographic, clinical, and medical characteristics of patients with respect to the severity of COVID-19 disease using Electronic Health Records Data of 4,140 SARS-CoV-2 positive subjects from several large Boston Area Hospitals. We found that prior use of antihypertensive medications as well as lipid lowering and other cardiovascular drugs (such as direct oral anticoagulants and antiplatelets) all track with increased severity of COVID-19 and should be further investigated with appropriate adjustment for confounders such as age and frailty. The three most common prior comorbidities are hyperlipidemia, hypertension, and prior pneumonia, all associated with increased severity.

Lab monitoring and acute care utilization during initiation of renin angiotensin aldosterone inhibitors or diuretics in chronic kidney disease.

Renin angiotensin aldosterone system inhibitors (RAASi) and diuretics are among the most frequently prescribed anti-hypertensives. Individuals with chronic kidney disease (CKD) are particularly at risk for electrolyte disturbances and kidney injury but the appropriate use of lab monitoring following RAASi or diuretic initiation is uncertain in CKD.We describe the frequency and time interval of lab monitoring during initiation of RAASi and diuretics in CKD and assess whether close lab monitoring associates with one-year risk of emergency department (ED) visit or hospitalization.We evaluated an observational cohort of 8,217 individuals with stage 3-5 non-dialysis CKD newly prescribed a RAASi (52.3%) or diuretic (47.7%) from thirty-six primary care offices affiliated with Brigham and Women's Hospital and Massachusetts General Hospital between 2009 and 2011.Overall, 3306 (40.2%) individuals did not have pre-prescription labs done within 2 weeks, and 5957 (72.5%) did not have post-prescription labs done within 2 weeks which includes 524 (6.4%) individuals without post-prescription within 1 year. Close monitoring occurred in only 1547 (20.1%) and was more likely in individuals prescribed diuretics compared to RAASi (adjusted OR 1.39; 95%CI 1.20-1.62), with CKD stage 4,5 compared with stage 3 (adjusted OR 1.47; 95%CI 1.16-1.86) and with cardiovascular disease (adjusted OR 1.42; 95%CI 1.21-1.66). Close monitoring was not associated with decreased risk of ED visit or hospitalization.Close lab monitoring during initiation of RAASi or diuretics was more common in participants with cardiovascular disease and advanced CKD suggesting physicians selected high-risk individuals for close monitoring. As nearly 80% of individuals did not receive close lab monitoring there may be value in future research on electronic physician decision tools targeted at lab monitoring.

Interhospital Transfer: Transfer Processes and Patient Outcomes.

Though often undertaken to provide patients with specialized care, interhospital transfer (IHT) is associated with worse outcomes for select patients. Certain aspects of the transfer process have been suggested as contributors to these outcomes. We performed a retrospective cohort study including patients ≥ 18 years who underwent IHT to a tertiary care hospital between January 2005 and September 2013. We examined the association between "weekend" transfer, "nighttime" transfer, "time delay" between transfer acceptance and arrival, and admission team "busyness" on the day of transfer, and patient outcomes, including transfer to the intensive care unit (ICU) within 48 hours and 30-day mortality. We utilized multivariable logistic regression models, adjusting for patient characteristics. Secondary analyses examined detailed timing of transfer and evaluated 30-day mortality stratified by service of transfer. Among the 24,352 patients who underwent IHT, the nighttime transfer was associated with increased adjusted odds of ICU transfer (odds ratio [OR] 1.54; 95% CI 1.38, 1.72) and 30-day mortality (OR 1.16; 95% CI 1.01, 1.35). Secondary analyses confirmed the association between nighttime transfer and ICU transfer throughout the week and demonstrated that Sunday (and trend towards Friday) night transfers had increased 30-day mortality, as compared with Monday daytime transfer. Stratified analyses demonstrated a significant association between transfer characteristics and adjusted odds of 30-day mortality among cardiothoracic and gastrointestinal surgical service transfers. Our findings suggest high acuity patients have worse outcomes during off-peak times of transfer and during times of high care team workload. Further study is needed to identify underlying reasons to explain these associations and devise potential solutions.

Evaluation of 'Definite' Anaphylaxis Drug Allergy Alert Overrides in Inpatient and Outpatient Settings.

INTRODUCTION: Drug-allergy interaction (DAI) alerts are generated when a known adverse sensitivity-inducing substance is prescribed. A recent study at our institution showed that providers overrode most DAI alerts, including those that warned against potentially life-threatening 'anaphylaxis'. OBJECTIVE: The aim of this study was to determine the rate of anaphylaxis overrides, the reasons for these overrides, whether the overrides were appropriate, and if harm occurred from overrides. METHODS: All DAI alerts, with a reaction of 'anaphylaxis', were analysed for inpatients and outpatients within our health system between January 2009 and December 2011. Only alerts that were triggered by 'definite' alerts (i.e. same ordered medication as documented allergen) were included. Patient charts were reviewed to assess the appropriateness of overrides and potential harm, according to a predetermined set of criteria. RESULTS: A total of 202 inpatient and 16 outpatient alerts met the inclusion criteria. The rate of overrides for 'definite' anaphylaxis DAI alerts was high (inpatient: n = 93, 46.0%; outpatient: n = 11, 68.8%) but appropriate for most overrides in the inpatient (n = 78, 83.9%) and outpatient settings (n = 11, 100%). The most common override reasons in the inpatient and outpatient settings were 'administer per desensitization protocol' (n = 64, 31.7%) and 'patient does not have this allergy' (n = 7, 63.6%), respectively. No harm was associated with overrides in either setting, particularly because many medications were not administered. CONCLUSIONS: Overrides of 'definite' anaphylaxis DAI alerts were common and often appropriate. Most overrides were due to desensitizations. Allergy reconciliation for patients could further improve critical decision support.

Medication-related clinical decision support alert overrides in inpatients.

Objective: To define the types and numbers of inpatient clinical decision support alerts, measure the frequency with which they are overridden, and describe providers' reasons for overriding them and the appropriateness of those reasons. Materials and Methods: We conducted a cross-sectional study of medication-related clinical decision support alerts over a 3-year period at a 793-bed tertiary-care teaching institution. We measured the rate of alert overrides, the rate of overrides by alert type, the reasons cited for overrides, and the appropriateness of those reasons. Results: Overall, 73.3% of patient allergy, drug-drug interaction, and duplicate drug alerts were overridden, though the rate of overrides varied by alert type (P < .0001). About 60% of overrides were appropriate, and that proportion also varied by alert type (P < .0001). Few overrides of renal- (2.2%) or age-based (26.4%) medication substitutions were appropriate, while most duplicate drug (98%), patient allergy (96.5%), and formulary substitution (82.5%) alerts were appropriate. Discussion: Despite warnings of potential significant harm, certain categories of alert overrides were inappropriate >75% of the time. The vast majority of duplicate drug, patient allergy, and formulary substitution alerts were appropriate, suggesting that these categories of alerts might be good targets for refinement to reduce alert fatigue. Conclusion: Almost three-quarters of alerts were overridden, and 40% of the overrides were not appropriate. Future research should optimize alert types and frequencies to increase their clinical relevance, reducing alert fatigue so that important alerts are not inappropriately overridden.

Association of Changes in Creatinine and Potassium Levels After Initiation of Renin Angiotensin Aldosterone System Inhibitors With Emergency Department Visits, Hospitalizations, and Mortality in Individuals With Chronic Kidney Disease.

Importance: Renin angiotensin aldosterone system inhibitors (RAASIs) benefit individuals with chronic kidney disease (CKD). Elevations in serum creatinine and potassium levels are common reasons for discontinuation of this therapy, but their incidence and risks are not well characterized in community practice. Objective: To evaluate associations of increased creatinine levels, hyperkalemia, and therapy continuation with the risk of emergency department (ED) visits, hospitalizations, and mortality within 1 year after RAASI therapy initiation in individuals with CKD. Design, Setting, and Participants: This prospective cohort study included 4661 individuals with nondialysis CKD newly prescribed a RAASI or a diuretic who were treated at 36 outpatient primary care offices affiliated with Brigham & Women's Hospital and Massachusetts General Hospital, Boston, from January 1, 2009, through December 31, 2011. Individuals receiving a new prescription for a diuretic were used to provide context. All participants had a baseline measure of renal function and at least 1 follow-up measurement of creatinine and potassium levels within 90 days of the prescription. Data were analyzed from January 1, 2009, through December 31, 2012. Exposures: Changes in creatinine and potassium levels within 90 days after the prescription date and therapy discontinuation. Main Outcomes and Measures: Emergency department visits, hospitalizations, and mortality within 1 year. Results: A total of 4661 individuals were included in the analysis (2506 [53.8%] women; mean [SD] age, 71 [14]; 3931 [84.3%] white; and 4198 [90.1%] with CKD stage 3). Of these, 2354 individuals (50.5%) received RAASIs and 2307 (49.5%) received diuretics. Creatinine level increase of at least 30% after RAASI therapy initiation was found in 158 of 2354 individuals (6.7%); hyperkalemia of greater than 5.0 mEq/L, in 251 of 2354 (10.7%). Increases in creatinine level of at least 30% (unadjusted odds ratio [OR], 1.40; 95% CI, 0.89-2.21), hyperkalemia (unadjusted OR, 1.15; 95% CI, 0.64-2.06), and therapy discontinuation (unadjusted OR, 1.01; 95% CI, 0.71-1.46) were not associated with ED visits or hospitalizations, which was consistent with results from competing risk analyses. Initial increases in creatinine level of at least 30% were associated with mortality in the total cohort (adjusted OR [aOR], 2.17; 95% CI, 1.45-3.25). However, the effect was only independent for diuretics (aOR, 2.27; 95% CI, 1.41-3.66) and not for RAASIs (aOR, 1.82; 95% CI, 0.83-3.99). Conclusions and Relevance: Acute creatinine and potassium level disturbances after initiation of RAASI therapy in individuals with CKD appear to be sustained often often not sustained and not associated with ED visits or hospitalizations, despite therapy continuation. Findings from this study suggest that increases in creatinine level were independently associated with mortality among individuals prescribed diuretics but not RAASIs. Structured laboratory monitoring during RAASI therapy initiation may guide appropriate continuation of therapy in the outpatient setting.

Comparison of Overridden Medication-related Clinical Decision Support in the Intensive Care Unit between a Commercial System and a Legacy System.

BACKGROUND: Electronic health records (EHRs) with clinical decision support (CDS) have shown to be effective at improving patient safety. Despite this, alerts delivered as part of CDS are overridden frequently, which is of concern in the critical care population as this group may have an increased risk of harm. Our organization recently transitioned from an internally-developed EHR to a commercial system. Data comparing various EHR systems, especially after transitions between EHRs, are needed to identify areas for improvement. OBJECTIVES: To compare the two systems and identify areas for potential improvement with the new commercial system at a single institution. METHODS: Overridden medication-related CDS alerts were included from October to December of the systems' respective years (legacy, 2011; commercial, 2015), restricted to three intensive care units. The two systems were compared with regards to CDS presentation and override rates for four types of CDS: drug-allergy, drug-drug interaction (DDI), geriatric and renal alerts. A post hoc analysis to evaluate for adverse drug events (ADEs) potentially resulting from overridden alerts was performed for 'contraindicated' DDIs via chart review. RESULTS: There was a significant increase in provider exposure to alerts and alert overrides in the commercial system (commercial: n=5,535; legacy: n=1,030). Rates of overrides were higher for the allergy and DDI alerts (p<0.001) in the commercial system. Geriatric and renal alerts were significantly different in incidence and presentation between the two systems. No ADEs were identified in an analysis of 43 overridden contraindicated DDI alerts. CONCLUSIONS: The vendor system had much higher rates of both alerts and overrides, although we did not find evidence of harm in a review of DDIs which were overridden. We propose recommendations for improving our current system which may be helpful to other similar institutions; improving both alert presentation and the underlying knowledge base appear important.

Review of Nonformulary Medication Approvals in an Academic Medical Center.

BACKGROUND: The Joint Commission requires hospitals to formally review formulary medications at least annually based on new clinical information. Although review of nonformulary medication (NFM) use is not required, frequent and inappropriate use of NFMs has the potential to increase hospital costs, negatively affect quality of care, and increase medication errors. Limited resources may restrict an institution's ability to review NFM use in addition to the required annual formulary review. NFM use at Brigham and Women's Hospital (BWH) was reviewed to provide insight on how to best direct an NFM review that is both effective and efficient. How an NFM review may negatively affect cost, quality of care, and medication errors is also inferred. METHODS: All approved NFM requests between 2009 and 2012 from Brigham and Women's Hospital's computerized provider order entry system were extracted and categorized according to the American Hospital Formulary Service (AHFS) Pharmacologic-Therapeutic Classification System. RESULTS: Of the 15,356,016 new medication orders, there were 223,266 NFM approvals for 433 unique NFMs. NFMs were categorized into 91 AHFS, 14 combination, and 4 "Others" classes. Twenty-five AHFS Classes accounted for approximately the top 90% of all NFM approvals, and the top 2 NFMs in each class accounted for a majority of the NFM approvals. CONCLUSION: Only a few classes of medications and a few medications within each class accounted for most of the NFM use at BWH. Targeting review of the most frequently used NFMs in each class may be a feasible strategy to reviewing NFMs annually that is both effective and efficient in optimizing formulary benefits.

Evaluation of medication-related clinical decision support alert overrides in the intensive care unit.

PURPOSE: Medication-related clinical decision support (CDS) has been identified as a method to improve patient outcomes but is historically frequently overridden and may be inappropriately so. Patients in the intensive care unit (ICU) are at a higher risk of harm from adverse drug events (ADEs) and these overrides may increase patient harm. The objective of this study is to determine appropriateness of overridden medication-related CDS overrides in the ICU. MATERIALS AND METHODS: We evaluated overridden medication-related alerts of four alert categories from January 2009 to December 2011. The primary outcome was the appropriateness of a random sample of overrides based on predetermined criteria. Secondary outcomes included the incidence of adverse drug events (ADEs) that resulted from the overridden alert. RESULTS: A total of 47,449 overridden alerts were included for evaluation. The appropriateness rate for overridden alerts varied by alert category (allergy: 94%, drug-drug interaction: 84%, geriatric: 57%, renal: 27%). A total of seven actual ADEs were identified in the random sample and where the medication(s) was administered (n=366), with an increased risk of ADEs associated with inappropriately overridden alerts (p=0.0078). CONCLUSIONS: The appropriateness of medication-related clinical decision support overrides in the ICU varied substantially by the type of alert. Inappropriately overridden alerts were associated with an increased risk of ADEs compared to appropriately overridden alerts.