by Brian Jackson, MD and Brian Shirts, MD, PhD
We’ve heard a lot of talk lately about personalized medicine (i.e., using advanced diagnostics to guide customized therapy).
A great deal of research is going into creating new molecular and genetic tests. But whether the health care system is prepared to actually generate value from these advanced diagnostics remains an open question. Our group’s research, presented in a new study published in the Journal of Pathology Informatics, finds that one relatively new test is widely misused in ways that drive up costs without benefiting patients.
Cervical cancer screening with annual Pap smears has been a bedrock component of women’s health care for decades. It’s so engrained in the medical culture that many women continue to receive annual Paps even after total hysterectomy (see JAMA article). The personalized medicine philosophy would suggest that screening frequency (and, potentially, modality) should be customized to a patient’s individual risk of cancer rather than following a cookbook formula. And the most important component of that risk turns out to be persistent infection with certain HPV genotypes.
The American Society for Colposcopy and Cervical Pathology (ASCCP) has developed a detailed set of recommendations for personalized cervical cancer screening using HPV testing together with Pap smears. Our study, which analyzed more than 450,000 HPV tests performed at ARUP Laboratories between 2003 and 2009, was designed to assess whether doctors are following those recommendations.
Briefly, the ASCCP recommendations state:
- HPV testing is contraindicated in women under 21.
- In women 21 to 29, HPV testing should not be used as a first-line test but may be used for stratifying patients with certain Pap smear findings (e.g., patients with atypical squamous cells of undetermined significance).
- For women 30 and older, HPV testing may be used either for primary screening or for stratifying patients with those Pap smear findings.
- In evaluation of HPV-positive, colonoscopy-negative cervical lesions, repeat HPV testing is recommended after one year; shorter intervals are not indicated given the natural history of HPV infections.
- In women older than 30, negative-screening HPV and cytology allow the follow-up interval to be safely extended to three years, and annual screening is not necessary.
Although we did find HPV-ordering patterns are starting to trend toward the testing guidelines, overall, we found that nearly a quarter of all HPV tests ordered were unnecessary. Here are our key findings in the study:
- The proportion of HPV tests performed on women under 21 declined over the six-year study period from 20 to 5 percent. Teenage girls and young women under 21 generally do not benefit from HPV testing. In some cases, testing these patients may lead to unnecessary follow-up care, including colposcopy and cervical biopsy.
- One-third of tests on women between ages 21 and 29 arrived at our laboratory five or fewer days after collection, suggesting that these tests were ordered before the Pap smear result was known.
- For women 30 and older who were HPV positive, the test was often repeated three to six months later, which is too short a time interval to provide useful information. According to the guidelines, decisions about follow-up actions, such as colposcopy, do not require a repeat HPV test. Thus, a more rational interval before a subsequent HPV test (in conjunction with a Pap test) would be 12 months.
- For women older than 30 with a negative HPV test result, the most common time interval before the next test was 12 months, suggesting that annual screening is predominant in this group. For low-risk women with negative Pap and HPV results, however, screening every three years is safe and appropriate. Only 6 percent of follow-up tests after a negative HPV results had an interval of three years or more.
These findings aren’t too surprising given the large body of previous research showing that medical practice often doesn’t follow published recommendations. But regardless, the findings show that our health care system has not sufficiently figured out how to optimize and customize treatment based on test results. To achieve the promise of personalized medicine, we need more than new high-tech tests. We need to fundamentally change how we deliver care.
Brian Jackson is Medical Director of Informatics at ARUP Laboratories and Brian Shirts is a Molecular Genetic Pathology Fellow at the University of Utah School of Medicine and Assistant Medical Director of Informatics at ARUP Laboratories.
Submit a guest post and be heard on social media’s leading physician voice.