Temporal Cohort Identification for Alzheimer’s Disease with Sequences of Clinical Records

BACKGROUND Alzheimer’s Disease (AD) is a complex clinical phenotype with unprecedented social and economic tolls in an aging global population. Real World Data (RWD) from electronic health records (EHRs) offer opportunities to accelerate precision drug development and scale epidemiological research on AD. A precise characterization of AD cohorts is needed to address the noise abundant in RWD. METHODS We conducted a retrospective cohort study to develop and test computational models for AD cohort identification using clinical data from 8 Massachusetts healthcare systems. We mined temporal representations from EHR data using a novel transitive sequential pattern mining algorithm (tSPM) to train and validate our models. We then tested our models against a held-out test set from a review of medical records to adjudicate the presence of AD. We trained two classes of models using Gradient Boosting Machine (GBM) to compare the utility of AD diagnosis records versus the tSPM temporal representations (comprising sequences of diagnosis and medication observations) from electronic medical records for characterizing AD cohorts. RESULTS In a group of 4,985 patients, we identified 219 sequences of medication-diagnosis records for constructing the best classification models. The models with the sequential features improved AD classification by a magnitude of up to 16 percent (over the use of AD diagnosis codes). Six groups of sequences, which we refer to as temporal digital markers, were identified for characterizing the AD cohorts, including sequences that involved (1) a symptom or (2) a risk factor in the past, followed by an AD diagnosis, (3) AD medications, (4) indirect risk factors, symptom management, and potential side effects, (5) comorbidities with possible shared roots or side effects, and (6) plural encounters with of AD diagnosis codes. Discussions of how the identified sequential patterns can be interpreted are provided. CONCLUSIONS We present sequential patterns of diagnosis and medication codes from electronic medical records, as digital markers of Alzheimer’s Disease. Classification algorithms developed on the sequential patterns can replace standard features from EHRs to enrich phenotype modeling.