Article type
Year
Abstract
Background: Registers and databases of reports of clinical trials are valuable resources. However, their utility for retrieval is limited by variations in terminology used by different medical specialties and researchers. Coding them using standard international schemas may improve access.
Objectives: Our goal was to evaluate in a pilot study coding of health conditions using the International Classification of Disease (ICD-10) and interventions using the beta-test version of the International Classification of Health Interventions (ICHI) of systematic reviews and clinical trials in the Cochrane Eyes and Vision (CEVG) register of clinical trials within the Cochrane Central Register of Controlled Trials (CENTRAL).
Methods: Two raters coded all 43 CEVG systematic reviews published in The Cochrane Library by March 1, 2007. After discussing disagreements and agreeing on coding conventions, one rater coded a 10% random sample of approximately 10,000 entries in the CEVG register as of January 2007.
Results: Inter-rater agreement using the ICD-10 was 81% for the systematic reviews. Of 981 trial reports in the random sample, 716 (73%) could be coded using titles, and abstracts when available. ICHI was inadequate for coding interventions and was supplemented by ad hoc codes for outpatient procedures and drug classes.
Conclusions: We found ICD-10 to be a satisfactory schema for coding disorders of eyes and vision. However, we are seeking an alternative to the ICHI, designed primarily for hospital procedures; the majority of participants in trials in eyes and vision are outpatients. Also, we are conducting a pilot study to evaluate automated conversion of MeSH terms to ICD-10 codes.
Objectives: Our goal was to evaluate in a pilot study coding of health conditions using the International Classification of Disease (ICD-10) and interventions using the beta-test version of the International Classification of Health Interventions (ICHI) of systematic reviews and clinical trials in the Cochrane Eyes and Vision (CEVG) register of clinical trials within the Cochrane Central Register of Controlled Trials (CENTRAL).
Methods: Two raters coded all 43 CEVG systematic reviews published in The Cochrane Library by March 1, 2007. After discussing disagreements and agreeing on coding conventions, one rater coded a 10% random sample of approximately 10,000 entries in the CEVG register as of January 2007.
Results: Inter-rater agreement using the ICD-10 was 81% for the systematic reviews. Of 981 trial reports in the random sample, 716 (73%) could be coded using titles, and abstracts when available. ICHI was inadequate for coding interventions and was supplemented by ad hoc codes for outpatient procedures and drug classes.
Conclusions: We found ICD-10 to be a satisfactory schema for coding disorders of eyes and vision. However, we are seeking an alternative to the ICHI, designed primarily for hospital procedures; the majority of participants in trials in eyes and vision are outpatients. Also, we are conducting a pilot study to evaluate automated conversion of MeSH terms to ICD-10 codes.