A Validation Study of Maternal Early Warning Systems: A Retrospective Cohort Study

Blumenthal, Elizabeth; Hooshvar, Nina; McQuade, Miriam; McNulty, Jennifer

doi:10.1055/s-0039-1681097

Cited by 19 publications

(13 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…19,20 We reported that maternal early warning triggers (MEWT) would be most specific for morbidity in our population. 22 Implementation of a paper-based MEWT was previously shown to reduce centers for disease control (CDC) defined maternal morbidity in one large U.S. hospital system. 20 The MEWT (►Fig.…”

mentioning

confidence: 99%

“…[10][11][12][13] These systems have been devised, tested, and recommended for nonobstetric populations for the past two decades, 14,15 and have been shown to both predict and reduce severe morbidity and mortality. 14,16,17 The National Partnership for Maternal Safety recommends use of early warning systems 5 ; however, the validation studies of these various proposed systems are limited, [18][19][20][21][22] as is evidence that they decrease severe maternal morbidity (SMM). 19,20 We reported that maternal early warning triggers (MEWT) would be most specific for morbidity in our population.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Implementation and Evaluation of an Electronic Maternal Early Warning Trigger Tool to Reduce Maternal Morbidity

et al. 2020

Self Cite

View full text Add to dashboard Cite

Objective We compare maternal morbidity and clinical care metrics before and after the electronic implementation of a maternal early warning trigger (MEWT) tool. Study Design This is a study of maternal morbidity and clinical care within three linked hospitals comparing 1 year before and after electronic MEWT implementation. We compare severe maternal morbidity overall as well as within the subcategories of hemorrhage, hypertension, cardiopulmonary, and sepsis in addition to relevant process metrics in each category. We describe the MEWT trigger rate in addition to MEWT sensitivity and specificity for morbidity overall and by morbidity type. Results The morbidity rate ratio increased from 1.6 per 100 deliveries in the pre-MEWT period to 2.06 per 100 deliveries in the post-MEWT period (incidence rate ratio = 1.28, p = 0.018); however, in cases of septic morbidity, time to appropriate antibiotics decreased (pre-MEWT: 1.87 hours [1.11–2.63] vs. post-MEWT: 0.75 hours [0.31–1.19], p = 0.036) and in cases of hypertensive morbidity, the proportion of cases treated with appropriate antihypertensive medication within 60 minutes improved (pre-MEWT: 62% vs. post-MEWT: 83%, p = 0.040). The MEWT trigger rate was 2.3%, ranging from 0.8% in the less acute centers to 2.9% in our tertiary center. The MEWT sensitivity for morbidity overall was 50%; detection of hemorrhage morbidity was lowest (30%); however, it ranged between 69% for septic morbidity, 74% for cardiopulmonary morbidity, and 82% for cases of hypertensive morbidity. Conclusion Overall, maternal morbidity did not decrease after implementation of the MEWT system; however, important clinical metrics such as time to antibiotics and antihypertensive care improved. We suspect increased morbidity was related to annual variation and unexpected lower morbidity in the pre-MEWT comparison year. Because MEWT sensitivity for hemorrhage was low, and because hemorrhage dominates administrative metrics of morbidity, process metrics around sepsis, hypertension, and cardiopulmonary morbidity are important to track as markers of MEWT efficacy. Key Points

show abstract

mentioning

confidence: 99%

mentioning

confidence: 99%

Implementation and Evaluation of an Electronic Maternal Early Warning Trigger Tool to Reduce Maternal Morbidity

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…As technology was developed to automate vital sign monitoring for patients in hospital settings, frequent—or even continuous—vital sign assessment was adopted widely. The unanticipated consequence of this practice, however, is that many abnormal vital signs do not actually signify acute pathophysiology—in other words, they are not “actionable.” 1 If an alarm sounds to alert a clinician to abnormal data that do not actually trigger a response, that clinician will be inadvertently trained to ignore the alarms. Increasing burden of alarms that are not linked to adverse patient outcomes further dulls clinician responsiveness.…”

Section: Alarm Fatigue On Labor and Deliverymentioning

confidence: 99%

Reducing Alarm Fatigue in Maternal Monitoring on Labor and Delivery: A Commentary on Deimplementation in Obstetrics

et al. 2022

View full text Add to dashboard Cite

Hospital labor and delivery floors frequently operate like intensive care units (ICUs) – with continuous data feeds pouring into central monitoring stations against a background of blaring alarms. Yet the majority of obstetric patients are healthy and do not require ICU-level care. Despite limited organizational recommendations guiding the frequency of vital sign measurement, continuous pulse oximetry is used widely for laboring patients. There is also no evidence that morbidity prevention is linked to specific frequencies of vital sign monitoring in low-risk patients. In fact, studies examining the performance of maternal early warnings systems based on vital signs suggest that these may not reliably provide actionable information regarding maternal physiologic status. Furthermore, it is very possible that intrapartum maternal overmonitoring can impact care negatively by generating alarm fatigue, causing providers to miss actual abnormal vital signs that may precede morbidity.

show abstract

“…JAMA Network Open. 2022;5(10):e2234924. doi:10.1001/jamanetworkopen.2022.34924 (Reprinted) October 5, 2022 12/12 Downloaded From: https://jamanetwork.com/ on 10/10/2022…”

mentioning

confidence: 99%

Comparison of Natural Language Processing of Clinical Notes With a Validated Risk-Stratification Tool to Predict Severe Maternal Morbidity

et al. 2022

View full text Add to dashboard Cite

ImportanceRisk-stratification tools are routinely used in obstetrics to assist care teams in assessing and communicating risk associated with delivery. Electronic health record data and machine learning methods may offer a novel opportunity to improve and automate risk assessment.ObjectiveTo compare the predictive performance of natural language processing (NLP) of clinician documentation with that of a previously validated tool to identify individuals at high risk for maternal morbidity.Design, Setting, and ParticipantsThis retrospective diagnostic study was conducted at Brigham and Women’s Hospital and Massachusetts General Hospital, Boston, Massachusetts, and included individuals admitted for delivery at the former institution from July 1, 2016, to February 29, 2020. A subset of these encounters (admissions from February to December 2018) was part of a previous prospective validation study of the Obstetric Comorbidity Index (OB-CMI), a comorbidity-weighted score to stratify risk of severe maternal morbidity (SMM).ExposuresNatural language processing of clinician documentation and OB-CMI scores.Main Outcomes and MeasuresNatural language processing of clinician-authored admission notes was used to predict SMM in individuals delivering at the same institution but not included in the prospective OB-CMI study. The NLP model was then compared with the OB-CMI in the subset with a known OB-CMI score. Model discrimination between the 2 approaches was compared using the DeLong test. Sensitivity and positive predictive value for the identification of individuals at highest risk were prioritized as the characteristics of interest.ResultsThis study included 19 794 individuals; 4034 (20.4%) were included in the original prospective validation study of the OB-CMI (testing set), and the remaining 15 760 (79.6%) composed the training set. Mean (SD) age was 32.3 (5.2) years in the testing cohort and 32.2 (5.2) years in the training cohort. A total of 115 individuals in the testing cohort (2.9%) and 468 in the training cohort (3.0%) experienced SMM. The NLP model was built from a pruned vocabulary of 2783 unique words that occurred within the 15 760 admission notes from individuals in the training set. The area under the receiver operating characteristic curve of the NLP-based model for the prediction of SMM was 0.76 (95% CI, 0.72-0.81) and was comparable with that of the OB-CMI model (0.74; 95% CI, 0.70-0.79) in the testing set (P = .53). Sensitivity (NLP, 28.7%; OB-CMI, 24.4%) and positive predictive value (NLP, 19.4%; OB-CMI, 17.6%) were comparable between the NLP and OB-CMI high-risk designations for the prediction of SMM.Conclusions and RelevanceIn this study, the NLP method and a validated risk-stratification tool had a similar ability to identify patients at high risk of SMM. Future prospective research is needed to validate the NLP approach in clinical practice and determine whether it could augment or replace tools requiring manual user input.

show abstract

A Validation Study of Maternal Early Warning Systems: A Retrospective Cohort Study

Cited by 19 publications

References 19 publications

Implementation and Evaluation of an Electronic Maternal Early Warning Trigger Tool to Reduce Maternal Morbidity

Implementation and Evaluation of an Electronic Maternal Early Warning Trigger Tool to Reduce Maternal Morbidity

Reducing Alarm Fatigue in Maternal Monitoring on Labor and Delivery: A Commentary on Deimplementation in Obstetrics

Comparison of Natural Language Processing of Clinical Notes With a Validated Risk-Stratification Tool to Predict Severe Maternal Morbidity

Contact Info

Product

Resources

About