The Canadian Assisted Reproductive Technologies Register (CARTR) Plus database: a validation study.

Area of Science:

• Reproductive medicine and health data management.
• Assisted reproductive technologies (ART) research and data quality assessment.

Background:

• Health databases are valuable research tools but require validation.
• The Canadian Assisted Reproductive Technologies Register (CARTR) Plus database lacks prior validation.
• Accurate documentation is crucial for research utilizing health registries.

Purpose of the Study:

• To evaluate the data accuracy and completeness of the CARTR Plus database.
• To assess the validity of key variables within the CARTR Plus database against patient charts.
• To inform potential improvements in the CARTR Plus data entry process.

Main Methods:

• Reabstracted 25 data elements from patient charts (reference standard) for 2015 IVF cycles.
• Included data from six Canadian clinics using purposive and fixed random sampling.
• Calculated validity measures including kappa coefficients and intraclass correlation coefficients (ICC).

Main Results:

• High agreement (ICC=0.99) for patient age and nearly perfect agreement (kappa > 0.90) for cycle type and embryo transfer.
• Moderate agreement for FSH levels (ICC=0.68) and oocyte origin (kappa=0.45).
• Lower agreement for elective single embryo transfer (kappa=0.55).

Conclusions:

• The CARTR Plus database demonstrates high data quality for most assessed variables.
• Identified specific variables (FSH, oocyte origin, elective single embryo transfer) requiring improved documentation.
• Further validation of additional CARTR Plus variables is recommended.