Related Experiment Video
Updated: Jun 7, 2025
07:35
A Non-invasive Way to Isolate and Phenotype Cells from the Conjunctiva
Published on: July 5, 2017
10.3K
Development and Validation of Claims-Based Algorithms for Conjunctivitis and Keratitis.
Andrea K Chomistek1, Jessica M Franklin1, Rachel E Sobel2
1Optum Epidemiology, Boston, Massachusetts, USA.
Pharmacoepidemiology and Drug Safety
|November 13, 2024
Summary
Validated algorithms can now identify conjunctivitis and keratitis in atopic dermatitis (AD) patients using claims data. These tools will aid future research into ocular surface disorders in AD populations.
Area of Science:
- Ophthalmology
- Dermatology
- Health Informatics
Background:
- Ocular surface disorders are common in patients with atopic dermatitis (AD).
- Validated algorithms for identifying conjunctivitis and keratitis in AD patients using claims data are currently lacking.
- Accurate identification is crucial for understanding disease burden and treatment outcomes.
Purpose of the Study:
- To develop and validate algorithms for identifying conjunctivitis and keratitis in patients with atopic dermatitis (AD) using health insurance claims data.
- To establish reliable methods for detecting these specific ocular conditions within large patient populations.
Main Methods:
- Patients with AD were identified in a claims database (March 2017-November 2019).
- Candidate algorithms combining diagnosis codes and ophthalmic treatments were developed.
- Medical records of selected patients were reviewed for validation, with positive predictive values (PPVs) calculated.
Main Results:
- A conjunctivitis algorithm achieved a PPV of 81% (95% CI: 73%-87%).
- A keratitis algorithm, combining diagnosis codes with specific ophthalmic treatments, achieved an overall PPV of 80% (95% CI: 55%-93%).
- These algorithms met the pre-specified threshold of ≥70% PPV.
Conclusions:
- The study successfully developed and validated the first claims-based algorithms for conjunctivitis and keratitis in patients with atopic dermatitis (AD).
- These validated algorithms are now available for future research to better understand the prevalence and impact of these ocular conditions in AD patients.
