Objective: To determine whether compliance with guidelines for cervical cancer screening, particularly use of the human papillomavirus (HPV) test in adult (aged ≥21 y) women, improves with the implementation of educational prompts in the electronic health record (EHR).
Materials and Methods: Two EHR-based interventions aimed at reducing unindicated HPV tests were implemented in the EHR in late June 2010. The Pap order form was revised with a descriptor next to the cotest (Pap plus HPV test) option advising that this is not for screening in women younger than 30 years, and a link to the American Society for Colposcopy and Cervical Pathology Web site was made available on the EHR home page. Charts of adult women with HPV results from January to December 2010 were reviewed. Appropriateness of HPV test ordering before (period A: from January to June) and after (period B: from July to December) the interventions were compared using the χ2 test of association.
Results: A total of 3,564 HPV tests were performed on adult women at Loyola University Medical Center in 2010. During period A, 1,709 tests were ordered compared with 1,855 tests ordered during period B (p = .014). The proportion of HPV tests without an appropriate indication decreased significantly after the EHR changes (20% for period A vs 13% for period B, p < .0005). This significant decrease was seen in both primary care (22% in period A vs 12% in period B, p < .0005) and obstetrics and gynecology (Ob/Gyn) (19% vs 13%, p < .0005).
Conclusions: Electronic health record–based tools improve compliance with cervical cancer screening guidelines.
Electronic health record&#x2013;based tools improve appropriate use of the human papillomavirus test in adult women.
Stritch School of Medicine, Loyola University Chicago, Maywood, IL
Correspondence to: Paula White, MD, Department of OB/GYN, Loyola University Medical Center, 2160 S 1st Ave, Maywood, IL 60153. E-mail: email@example.com
The authors have declared they have no conflicts of interest.
No funding was received for this work.