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

Top Spinal Cord Inj Rehabil, 2017 · DOI: 10.1310/sci2304-333 · Published: January 1, 2017

Spinal Cord Injury Research Methodology & Design Public Health

Simple Explanation

The study validates an algorithm to identify patients with non-traumatic spinal cord dysfunction (NTSCD) using administrative health data. This is important because NTSCD is heterogeneous and difficult to identify. The researchers used hospital discharge records (DAD) from Alberta, Canada, and compared the algorithm's results with clinical chart reviews to check its accuracy. The algorithm links neurological impairment and NTSCD etiology codes to identify persons with NTSCD.

Study Duration

2006 to 2016

Participants

280 cases were identified

Evidence Level

Not specified

Key Findings

1
The algorithm has high sensitivity (97%) and positive predictive value (PPV) (92%) for identifying NTSCD cases.
2
The specificity of the algorithm was 60%, and the negative predictive value (NPV) was 80%.
3
The most common causes of NTSCD were degenerative (36.9%), infection (19.0%), oncology malignant (15.1%), and vascular (10.3%).

Research Summary

This study validated an algorithm for identifying persons with non-traumatic spinal cord dysfunction (NTSCD) using administrative health data in Alberta, Canada. The algorithm demonstrated high sensitivity and positive predictive value, making it a promising tool for research and clinical purposes. The most prevalent etiologies of NTSCD identified were degenerative, infection, oncology malignant, and vascular conditions.

Practical Implications

Improved Identification

The algorithm can help identify NTSCD patients who may not be connected to specialized services.

Resource Allocation

The algorithm can be used to calculate the prevalence of NTSCD and inform resource allocation.

Care Delivery

Using the algorithm with administrative data helps understand patient flow for NTSCD and improve care delivery.

Study Limitations

1
The algorithm requires further international consensus on the ICD-10 codes used.
2
The absence of a random sampling strategy may introduce bias.
3
The algorithm may not identify all inpatients with both a neurological impairment and NTSCD.

Your Feedback

Was this summary helpful?

Back to Spinal Cord Injury