Validation of Algorithm to Identify Persons with Non-traumatic Spinal Cord Dysfunction in Canada Using Administrative...

Area of Science:

• Health Informatics
• Neurology
• Epidemiology

Background:

• Administrative health data, like the Discharge Abstract Database (DAD), offers potential for identifying patients with non-traumatic spinal cord dysfunction (NTSCD).
• Validation of algorithms using administrative data is crucial before clinical implementation to ensure accuracy and reliability.

Purpose of the Study:

• To validate a newly developed algorithm designed to accurately identify patients diagnosed with NTSCD using the Canadian Discharge Abstract Database (DAD).

Main Methods:

• The study utilized DAD records from Southern Alberta (2006-2016).
• An algorithm identified potential NTSCD cases, which were then validated through chart reviews of electronic and paper medical records.
• Key diagnostic accuracy metrics, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), were calculated with 95% confidence intervals (CI).

Main Results:

• The algorithm identified 280 potential NTSCD cases, with 252 included after exclusions.
• High diagnostic accuracy was observed: sensitivity 97% (95% CI, 94%-98%) and PPV 92% (95% CI, 88%-95%).
• Specificity was 60% (95% CI, 47%-73%) and NPV was 80% (95% CI, 65%-90%). Common causes included degenerative, infectious, malignant oncology, and vascular etiologies.

Conclusions:

• The developed algorithm demonstrates high sensitivity and PPV for identifying NTSCD patients via DAD.
• The algorithm also shows satisfactory specificity and NPV, making it a valuable tool for NTSCD surveillance.
• Awareness of the algorithm's limitations is essential for appropriate application in clinical and research settings.