Measuring agreement of administrative data with chart data using prevalence unadjusted and adjusted kappa
2009

Agreement Between Administrative Data and Chart Data

Sample size: 4008 publication Evidence: moderate

Author Information

Author(s): Chen Guanmin, Faris Peter, Hemmelgarn Brenda, Walker Robin L, Quan Hude

Primary Institution: University of Calgary

Hypothesis

The study aims to assess the agreement between hospital discharge administrative data and chart review data using Kappa and PABAK statistics.

Conclusion

Kappa values varied more widely than PABAK values across the 32 conditions, indicating that PABAK should not be interpreted as measuring the same agreement as kappa, especially for conditions with low prevalence.

Supporting Evidence

  • The prevalence of conditions varied significantly depending on the sampling method used.
  • PABAK values were consistently higher than kappa values across all conditions.
  • The gap between kappa and PABAK values narrowed as the prevalence of conditions approached 50%.

Takeaway

This study looked at how well hospital records match with actual patient charts, finding that different methods of sampling can change the results.

Methodology

The study compared agreement using random sampling, restricted sampling by conditions, and case-control sampling from hospital discharge records.

Potential Biases

Bias may affect kappa values, as it reflects disagreement on the proportion of positive or negative cases.

Limitations

The study could not capture the 'true' value of kappa for the conditions in the administrative data, and the quality of chart data used as a 'gold standard' may vary.

Participant Demographics

Participants were hospital discharge records from four teaching hospitals in Alberta, Canada.

Digital Object Identifier (DOI)

10.1186/1471-2288-9-5

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication