Veena V. Goel

Systematic Review of Physiologic Monitor Alarm Characteristics and Pragmatic Interventions to Reduce Alarm Frequency

BACKGROUND Alarm fatigue from frequent nonactionable physiologic monitor alarms is frequently named as a threat to patient safety. PURPOSE To critically examine the available literature relevant to alarm fatigue. DATA SOURCES Articles published in English, Spanish, or French between January 1980 and April 2015 indexed in PubMed, Cumulative Index to Nursing and Allied Health Literature, Scopus, Cochrane Library, Google Scholar, and ClinicalTrials.gov. STUDY SELECTION Articles focused on hospital physiologic monitor alarms addressing any of the following: (1) the proportion of alarms that are actionable, (2) the relationship between alarm exposure and nurse response time, and (3) the effectiveness of interventions in reducing alarm frequency. DATA EXTRACTION We extracted data on setting, collection methods, proportion of alarms determined to be actionable, nurse response time, and associations between interventions and alarm rates. DATA SYNTHESIS Our search produced 24 observational studies focused on alarm characteristics and response time and 8 studies evaluating interventions. Actionable alarm proportion ranged from <1% to 36% across a range of hospital settings. Two studies showed relationships between high alarm exposure and longer nurse response time. Most intervention studies included multiple components implemented simultaneously. Although studies varied widely, and many had high risk of bias, promising but still unproven interventions include widening alarm parameters, instituting alarm delays, and using disposable electrocardiographic wires or frequently changed electrocardiographic electrodes. CONCLUSIONS Physiologic monitor alarms are commonly nonactionable, and evidence supporting the concept of alarm fatigue is emerging. Several interventions have the potential to reduce alarms safely, but more rigorously designed studies with attention to possible unintended consequences are needed.

Learning statistical models of phenotypes using noisy labeled training data

OBJECTIVE Traditionally, patient groups with a phenotype are selected through rule-based definitions whose creation and validation are time-consuming. Machine learning approaches to electronic phenotyping are limited by the paucity of labeled training datasets. We demonstrate the feasibility of utilizing semi-automatically labeled training sets to create phenotype models via machine learning, using a comprehensive representation of the patient medical record. METHODS We use a list of keywords specific to the phenotype of interest to generate noisy labeled training data. We train L1 penalized logistic regression models for a chronic and an acute disease and evaluate the performance of the models against a gold standard. RESULTS Our models for Type 2 diabetes mellitus and myocardial infarction achieve precision and accuracy of 0.90, 0.89, and 0.86, 0.89, respectively. Local implementations of the previously validated rule-based definitions for Type 2 diabetes mellitus and myocardial infarction achieve precision and accuracy of 0.96, 0.92 and 0.84, 0.87, respectively.We have demonstrated feasibility of learning phenotype models using imperfectly labeled data for a chronic and acute phenotype. Further research in feature engineering and in specification of the keyword list can improve the performance of the models and the scalability of the approach. CONCLUSIONS Our method provides an alternative to manual labeling for creating training sets for statistical models of phenotypes. Such an approach can accelerate research with large observational healthcare datasets and may also be used to create local phenotype models.

Safety analysis of proposed data‐driven physiologic alarm parameters for hospitalized children

INTRODUCTION Modification of alarm limits is one approach to mitigating alarm fatigue. We aimed to create and validate heart rate (HR) and respiratory rate (RR) percentiles for hospitalized children, and analyze the safety of replacing current vital sign reference ranges with proposed data-driven, age-stratified 5th and 95th percentile values. METHODS In this retrospective cross-sectional study, nurse-charted HR and RR data from a training set of 7202 hospitalized children were used to develop percentile tables. We compared 5th and 95th percentile values with currently accepted reference ranges in a validation set of 2287 patients. We analyzed 148 rapid response team (RRT) and cardiorespiratory arrest (CRA) events over a 12-month period, using HR and RR values in the 12 hours prior to the event, to determine the proportion of patients with out-of-range vitals based upon reference versus data-driven limits. RESULTS There were 24,045 (55.6%) fewer out-of-range measurements using data-driven vital sign limits. Overall, 144/148 RRT and CRA patients had out-of-range HR or RR values preceding the event using current limits, and 138/148 were abnormal using data-driven limits. Chart review of RRT and CRA patients with abnormal HR and RR per current limits considered normal by data-driven limits revealed that clinical status change was identified by other vital sign abnormalities or clinical context. CONCLUSIONS A large proportion of vital signs in hospitalized children are outside presently used norms. Safety evaluation of data-driven limits suggests they are as safe as those currently used. Implementation of these parameters in physiologic monitors may mitigate alarm fatigue. Journal of Hospital Medicine 2015;11:817-823.

Changing Management of the Patent Ductus Arteriosus: Effect on Neonatal Outcomes and Resource Utilization

Objective This historical cohort study investigated how a shift toward a more conservative approach of awaiting spontaneous closure of the patent ductus arteriosus (PDA) in preterm infants has affected neonatal outcomes and resource utilization. Methods We retrospectively studied very low birth weight infants diagnosed with a PDA by echocardiogram (ECHO) in 2006‐2008 (era 1), when medical or surgical PDA management was emphasized, to those born in 2010‐2012 (era 2) when conservative PDA management was encouraged. Multiple regression analyses adjusted for gestational age were performed to assess differences in clinical outcomes and resource utilization between eras. Results More infants in era 2 (35/89, 39%) compared with era 1 (22/120, 18%) had conservative PDA management (p < 0.01). Despite no difference in surgical ligation rate, infants in era 2 had ligation later (median 24 vs. 8 days, p < 0.0001). There was no difference in clinical outcomes between eras, while number of ECHOs per patient was the only resource measure that increased in era 2 (median 3 vs. 2 ECHOs, p = 0.003). Conclusion In an era of more conservative PDA management, no increase in adverse clinical outcomes or significant change in resource utilization was found. Conservative PDA management may be a safe alternative for preterm infants.

Electronic Health Record–Enabled Research in Children Using the Electronic Health Record for Clinical Discovery

Initially described more than 50 years ago, electronic health records (EHRs) are now becoming ubiquitous throughout pediatric health care settings. The confluence of increased EHR implementation and the exponential growth of digital data within them, the development of clinical informatics tools and techniques, and the growing workforce of experienced EHR users presents new opportunities to use EHRs to augment clinical discovery and improve pediatric patient care. This article reviews the basic concepts surrounding EHR-enabled research and clinical discovery, including the types and fidelity of EHR data elements, EHR data validation/corroboration, and the steps involved in analytical interrogation.

Physiologic Monitor Alarm Rates at 5 Children’s Hospitals

Alarm fatigue has been linked to patient morbidity and mortality in hospitals due to delayed or absent responses to monitor alarms. We sought to describe alarm rates at 5 freestanding children’s hospitals during a single day and the types of alarms and proportions of patients monitored by using a point-prevalence, cross-sectional study design. We collected audible alarms on all inpatient units and calculated overall alarm rates and rates by alarm type per monitored patient per day. We found a total of 147,213 alarms during the study period, with 3-fold variation in alarm rates across hospitals among similar unit types. Across hospitals, one-quarter of monitored beds were responsible for 71%, 61%, and 63% of alarms in medical-surgical, neonatal intensive care, and pediatric intensive care units, respectively. Future work focused on addressing nonactionable alarms in patients with the highest alarm counts may decrease alarm rates.

Inpatient-Derived Vital Sign Parameters Implementation: An Initiative to Decrease Alarm Burden

Implementation of data-driven, age-based HR parameters and iteratively derived RR parameters decreased cardiorespiratory alarms by 21% in a pediatric acute care unit. OBJECTIVES: To implement data-driven vital sign parameters to reduce bedside monitor alarm burden. METHODS: Single-center, quality-improvement initiative with historical controls assessing the impact of age-based, inpatient-derived heart rate (HR) and respiratory rate (RR) parameters on a 20-bed acute care ward that serves primarily pediatric cardiology patients. The primary outcome was the number of alarms per monitored bed day (MBD) with the aim to decrease the alarms per MBD. Balancing measures included the frequency of missed rapid response team activations, acute respiratory code events, and cardiorespiratory arrest events in the unit with the new vital sign parameters. RESULTS: The median number of all cardiorespiratory monitor alarms per MBD decreased by 21% from 52 (baseline period) to 41 (postintervention period) (P < .001). This included a 17% decrease in the median HR alarms (9–7.5 per MBD) and a 53% drop in RR alarms (16.8–8.0 per MBD). There were 57 rapid response team activations, 8 acute respiratory code events, and no cardiorespiratory arrest events after the implementation of the new parameters. An evaluation of HRs and RRs recorded at the time of the event revealed that all patients with HRs and/or RRs out of range per former default parameters would also be out of range with the new parameters. CONCLUSIONS: Implementation of data-driven HR and iteratively derived RR parameters safely decreased the total alarm frequency by 21% in a pediatric acute care unit.

Integration of Single-Center Data-Driven Vital Sign Parameters into a Modified Pediatric Early Warning System

Objectives: Pediatric early warning systems using expert-derived vital sign parameters demonstrate limited sensitivity and specificity in identifying deterioration. We hypothesized that modified tools using data-driven vital sign parameters would improve the performance of a validated tool. Design: Retrospective case control. Setting: Quaternary-care children’s hospital. Patients: Hospitalized, noncritically ill patients less than 18 years old. Cases were defined as patients who experienced an emergent transfer to an ICU or out-of-ICU cardiac arrest. Controls were patients who never required intensive care. Cases and controls were split into training and testing groups. Interventions: The Bedside Pediatric Early Warning System was modified by integrating data-driven heart rate and respiratory rate parameters (modified Bedside Pediatric Early Warning System 1 and 2). Modified Bedside Pediatric Early Warning System 1 used the 10th and 90th percentiles as normal parameters, whereas modified Bedside Pediatric Early Warning System 2 used fifth and 95th percentiles. Measurements and Main Results: The training set consisted of 358 case events and 1,830 controls; the testing set had 331 case events and 1,215 controls. In the sensitivity analysis, 207 of the 331 testing set cases (62.5%) were predicted by the original tool versus 206 (62.2%; p = 0.54) with modified Bedside Pediatric Early Warning System 1 and 191 (57.7%; p < 0.001) with modified Bedside Pediatric Early Warning System 2. For specificity, 1,005 of the 1,215 testing set control patients (82.7%) were identified by original Bedside Pediatric Early Warning System versus 1,013 (83.1%; p = 0.54) with modified Bedside Pediatric Early Warning System 1 and 1,055 (86.8%; p < 0.001) with modified Bedside Pediatric Early Warning System 2. There was no net gain in sensitivity and specificity using either of the modified Bedside Pediatric Early Warning System tools. Conclusions: Integration of data-driven vital sign parameters into a validated pediatric early warning system did not significantly impact sensitivity or specificity, and all the tools showed lower than desired sensitivity and specificity at a single cutoff point. Future work is needed to develop an objective tool that can more accurately predict pediatric decompensation.

Expanding the phenotype of hawkinsinuria: new insights from response to N-acetyl-L-cysteine.

Hawkinsinuria is a rare disorder of tyrosine metabolism that can manifest with metabolic acidosis and growth arrest around the time of weaning off breast milk, typically followed by spontaneous resolution of symptoms around 1 year of age. The urinary metabolites hawkinsin, quinolacetic acid, and pyroglutamic acid can aid in identifying this condition, although their relationship to the clinical manifestations is not known. Herein we describe clinical and laboratory findings in two fraternal twins with hawkinsinuria who presented with failure to thrive and metabolic acidosis. Close clinical follow-up and laboratory testing revealed previously unrecognized hypoglycemia, hypophosphatemia, combined hyperlipidemia, and anemia, along with the characteristic urinary metabolites, including massive pyroglutamic aciduria. Treatment with N-acetyl-L-cysteine (NAC) restored normal growth and normalized or improved most biochemical parameters. The dramatic response to NAC therapy supports the idea that glutathione depletion plays a key role in the pathogenesis of hawkinsinuria.