Supplementing Tuberculosis Surveillance with Automated Data from Health Maintenance Organizations
1999

Using Pharmacy Data to Improve Tuberculosis Surveillance

Sample size: 350000 publication Evidence: moderate

Author Information

Author(s): Deborah S. Yokoe, Girish S. Subramanyan, Edward Nardell, Sharon Sharnprapai, Eugene McCray, Richard Platt

Primary Institution: Brigham and Women's Hospital, Boston, Massachusetts, USA

Hypothesis

Can automated data from managed care organizations enhance tuberculosis surveillance?

Conclusion

Pharmacy dispensing information from health maintenance organizations can effectively identify tuberculosis cases that may be missed by traditional public health methods.

Supporting Evidence

  • Automated pharmacy data identified 8 cases of tuberculosis that were unknown to public health.
  • The sensitivity of screening for two or more antituberculosis drugs was 89%.
  • Pharmacy dispensing information can complement traditional TB surveillance methods.
  • 18% of identified cases were not reported to the public health department.
  • Screening criteria based on pharmacy data had a positive predictive value of 30%.
  • Most cases were diagnosed at HMO centers with automated records.
  • Underreporting of TB cases without positive cultures may compromise surveillance.
  • Pharmacy data is available for most of the U.S. population, enhancing TB case identification.

Takeaway

This study shows that looking at pharmacy records can help find people with tuberculosis that doctors might miss.

Methodology

The study analyzed pharmacy dispensing data and medical records from managed care organizations to identify tuberculosis cases.

Potential Biases

There may be underreporting of tuberculosis cases that do not have positive cultures.

Limitations

The study may not account for all cases of tuberculosis, especially those treated outside the managed care system.

Participant Demographics

Participants included approximately 350,000 individuals with pharmacy coverage from various health care centers in Massachusetts.

Statistical Information

P-Value

p<0.05

Confidence Interval

95% CI = 76%, 96%

Statistical Significance

p<0.05

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication