Derivation and validation of a machine learning record linkage algorithm between emergency medical services and the emergency department

Abstract

Objective Linking emergency medical services (EMS) electronic patient care reports (ePCRs) to emergency department (ED) records can provide clinicians access to vital information that can alter management. It can also create rich databases for research and quality improvement. Unfortunately, previous attempts at ePCR and ED record linkage have had limited success. In this study, we use supervised machine learning to derive and validate an automated record linkage algorithm between EMS ePCRs and ED records. Materials and Methods All consecutive ePCRs from a single EMS provider between June 2013 and June 2015 were included. A primary reviewer matched ePCRs to a list of ED patients to create a gold standard. Age, gender, last name, first name, social security number, and date of birth were extracted. Data were randomly split into 80% training and 20% test datasets. We derived missing indicators, identical indicators, edit distances, and percent differences. A multivariate logistic regression model was trained using 5-fold cross-validation, using label k-fold, L2 regularization, and class reweighting. Results A total of 14 032 ePCRs were included in the study. Interrater reliability between the primary and secondary reviewer had a kappa of 0.9. The algorithm had a sensitivity of 99.4%, a positive predictive value of 99.9%, and an area under the receiver-operating characteristic curve of 0.99 in both the training and test datasets. Date-of-birth match had the highest odds ratio of 16.9, followed by last name match (10.6). Social security number match had an odds ratio of 3.8. Conclusions We were able to successfully derive and validate a record linkage algorithm from a single EMS ePCR provider to our hospital EMR.

Linking Emergency Medical Services and Emergency Department Data to Improve Overdose Surveillance in North Carolina

“…Data linkage using patient identifiers, such as names or social security numbers, improves performance compared with data sets that lack patient identifiers. 12 , 13 …”

“…Data linkage using patient identifiers, such as names or social security numbers, improves performance compared with data sets that lack patient identifiers. 12,13 By implementing an iterative deterministic linkage approach, we substantially improved the performance of EMS-ED linkage in North Carolina. The iterative approach achieved additional linkages while limiting the number of records passing through less stringent linkage requirements, thereby reducing the risk of false linkages.…”

Artificial intelligence in emergency medicine: A scoping review

“…[39][40][41][42] Another intervention used supervised ML to automatically link EMS electronic patient care reports to ED records. 43 The out-of-hospital setting presents a unique setting where limited clinical variables are used to make prompt decisions (for example, whether or not to transport to hospital). This is well suited to be tackled by AI given its predictive power and ability to use various data points and predict outcomes.…”

“…Several ML algorithms used EMS data to predict outcomes for out‐of‐hospital cardiac arrest 39–42 . Another intervention used supervised ML to automatically link EMS electronic patient care reports to ED records 43 . The out‐of‐hospital setting presents a unique setting where limited clinical variables are used to make prompt decisions (for example, whether or not to transport to hospital).…”

Automatic Learning Framework for Pharmaceutical Record Matching

“…Regarding to the record linkage problem, there are multiple approaches based on machine learning. For example, some of them aim discover drugs [58] or relationships among medical records [59] [60]. But these kind of applications are domain-dependent [61], [62] and requires specific steps for concrete applications.…”