Synapse
⌘+K
Synapse
PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
July 25, 2016American Journal of Medical Quality

Going Beyond Administrative Data: Retrospective Evaluation of an Algorithm Using the Electronic Health Record to Help Identify Bleeding Events Among Hospitalized Medical Patients on Warfarin

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Does an electronic health record-based algorithm accurately identify bleeding events in hospitalized medical patients on warfarin compared to clinical review?

Population

Hospitalized medical patients on warfarin. 292 cases reviewed.

Comparison

Electronic health record-based algorithm for… vs Clinical review (manual review)

Design

Cohort

Key result

An electronic health record-based algorithm for identifying bleeding events demonstrated a sensitivity of 93.9% and a positive predictive value of 46.2% compared with clinical review.

Authors

JMJames P. MoriartyPDPaul R. DanielsDMDennis M. Manning

Discussion

Loading...

Member takes

Overview

Low PPV cautions against standalone clinical use without manual review; leaves open the need for refined algorithms and broader validation.

Study Design

Type

Observational (n=292)

Structured PICO

Does an electronic health record-based algorithm accurately identify bleeding events in hospitalized medical patients on warfarin compared to clinical review?

P
Population
292 cases from hospitalized medical patients on warfarin evaluated to validate an electronic health record-based algorithm for identifying bleeding events.
E
Exposure
Electronic health record-based algorithm for bleeding (based on ≥2 of 3 criteria: diagnosis indicating bleeding, lab value decrease suggesting bleeding, and blood product use)
C
Comparator
Clinical review (manual review)
O
Outcome
Validity and reliability of the algorithm (sensitivity, positive predictive value, interrater agreement)safety

An EHR-based algorithm using multiple criteria can identify bleeding events in hospitalized patients on warfarin with high sensitivity but low positive predictive value, requiring manual review for cases meeting only 2 criteria.

Limitations

  • Cases having exactly 2 criteria may require manual review for validation.
  • Cases having exactly 2 criteria may require manual review for validation

Cite This Study

Moriarty et al. (2016) conducted an observational in Bleeding events in hospitalized medical patients on warfarin (n=292). Electronic health record-based algorithm vs. Clinical review was evaluated on Algorithm sensitivity and positive predictive value. An electronic health record-based algorithm for identifying bleeding events demonstrated a sensitivity of 93.9% and a positive predictive value of 46.2% compared with clinical review.

synapsesocial.com/papers/6a9ac599dd93ec450a0e698bhttps://doi.org/10.1177/1062860616660757
View Full Paper
Ask AI
Bookmark
Share