Low positive predictive value of International Classification of Diseases, 10th Revision codes in relation to rheumatic heart disease: a challenge for global surveillance.
Judith M KatzenellenbogenLee NedkoffJeffrey CannonDeborah KrugerFilippa PrettyJonathan R CarapetisKaren E DempseyJess De DasselMellise AndersonNicholas de KlerkJoseph HungPublished in: Internal medicine journal (2019)
International Classification of Diseases, 10th Revision codes for rheumatic heart disease (RHD) include valvular heart disease of unspecified origin, limiting their usefulness for estimating RHD burden. A cross-disciplinary national consultation developed an algorithm to improve RHD identification in hospital data. The algorithm has been operationalised and piloted. The algorithm developed categorised 32% of RHD-coded patients as probable/possible RHD. We outline a series of research initiatives to improve identification of RHD in administrative data thereby contributing to monitoring the RHD burden globally.
Keyphrases
- machine learning
- deep learning
- total knee arthroplasty
- pulmonary hypertension
- end stage renal disease
- rheumatoid arthritis
- big data
- electronic health record
- quality improvement
- ejection fraction
- chronic kidney disease
- newly diagnosed
- artificial intelligence
- palliative care
- total hip arthroplasty
- peritoneal dialysis
- emergency department
- risk factors
- aortic valve
- bioinformatics analysis
- acute care
- adverse drug