Recommendations for optimal ICD codes to study neurologic conditions: a systematic review

Area of Science:

• Neurology
• Health Informatics
• Epidemiology

Background:

• Administrative health data are widely utilized for large-scale population studies.
• The reliability of these datasets for identifying specific neurologic conditions remains uncertain.

Purpose of the Study:

• To systematically review and assess the validity of administrative health data for identifying patients with various neurologic conditions.
• To evaluate the accuracy of International Classification of Diseases (ICD) code-based case definitions.

Main Methods:

• Systematic literature review of Medline and Embase databases.
• Independent assessment of study eligibility by two reviewers.
• Abstraction of study data including ICD code-based case definitions, sensitivity, specificity, positive predictive values (PPVs), and negative predictive values (NPVs).

Main Results:

• Thirty full-text articles met eligibility criteria, covering conditions like Alzheimer disease, brain tumors, epilepsy, motor neuron disease, multiple sclerosis, Parkinson disease, spinal cord injury, and traumatic brain injury.
• Performance metrics (sensitivity, specificity, PPVs, NPVs) varied widely across neurologic conditions.
• No studies met eligibility criteria for cerebral palsy, dystonia, Huntington disease, hydrocephalus, muscular dystrophy, spina bifida, or Tourette syndrome.

Conclusions:

• The accuracy of case definitions derived from administrative health data is a critical factor in interpreting population-based studies.
• Researchers must consider the varying validity of these data when analyzing neurologic conditions.