The use of diagnostic imaging has increased dramatically over the last 10 years. Concurrently, spectacular advances in imaging technology have allowed the detection of increasingly smaller anatomic lesions. The positive impact of these changes on the earlier and more precise diagnosis of a broad range of medical diseases cannot be overstated. However, one unavoidable consequence of the increased utilization and resolution of imaging is the diagnosis of increasing numbers of incidental findings (unexpected, asymptomatic findings) and indeterminate findings (findings related to the reason for imaging, but of unknown clinical significance). Although the risk of pathology associated with these lesions is considered low, uncertainty regarding their clinical significance and lack of guidelines for their management leads to extensive further testing and surveillance. While the identification and evaluation of these incidental/indeterminate findings may lead to the earlier identification and treatment of an occult malignancy, they also involve high costs due to additional physician visits and follow-up testing, as well as harms, such as anxiety, complications from further testing and unnecessary treatment. Although anecdotally radiologists have reported a rise in the how frequently imaging tests are obtained to further evaluate such lesions, the true frequency of these findings and the risk of cancer associated with specific incidental/indeterminate findings is unknown. Consequently, little evidence-based data are available to guide the management of these findings.
The aims of the proposed study are 1) to identify the indications that led to ultrasound imaging of the pelvis 2) to evaluate the prevalence of specific ultrasound morphological findings on these examinations and determine the risk of cancer and other accuracy statistics associated with a range of ultrasound findings (including benign appearing incidental findings, numerous indeterminate finding and suspicious findings). The risk of cancer associated with specific ultrasound findings will be calculated unadjusted and stratified for important patient factors (such as age, cancer risk factors including a family history of breast and ovarian cancer and Serum CA-125 levels) and then adjusted for important confounders. In order to complete this aim, we will first categorize the results of all ultrasound examinations and then calculate the risk of cancer by the type of lesion. Lastly for Aim 3) we will evaluate surveillance patterns following the identification of specific findings. The purpose of this aim is to help provide data on the costs of surveillance to balance the potential benefits of the identification of such lesions. This proposal focuses on ultrasound imaging, as after radiography (x-ray), ultrasound accounts for the largest proportion of imaging tests performed, and incidental/indeterminate findings are frequently identified on these examinations. The results of these analyses will clarify the magnitude of this problem by providing data on how frequently incidental findings are identified and further, will provide data on the risk of cancer associated with specific findings. To complete these aims, we will conduct a retrospective cohort analysis of patients enrolled in Group Health, a large not-for-profit healthcare system in Seattle Washington, whose 260,000 female members underwent over 110,000 ultrasound examinations over the years of the study (1997 - 2005). Group Health collects and stores automated and detailed clinical records of diagnostic testing, and these data are routinely linked with regional SEER tumor registry data. Further, detailed data on ovarian cancer risk factors are available for all enrollees. Using this large, population-based database, we will complete the aims, and the results of these analyses will inform the development of evidence-based guidelines for the management of findings identified on ultrasound imaging of the pelvis.
Over the last decade the utilization of medical imaging has increased dramatically. Concurrently, spectacular advances in imaging technology have allowed the detection of increasingly smaller anatomic lesions. The positive impact of these changes on the earlier and more precise diagnosis of a broad range of medical diseases cannot be overstated. However, one unavoidable consequence of the increased utilization and resolution of imaging is the diagnosis of increasing numbers of incidental and indeterminate findings. Although the risk of pathology associated with these lesions is considered very low, uncertainty regarding their clinical significance and lack of guidelines for their management can initiate extensive further testing and surveillance.
The aims of the proposed research are to identify the prevalence of incidental/indeterminate associated with ultrasound imaging, and to evaluate the risk of cancer associated with specific ultrasound finings identified in the pelvis. The goal is to develop management strategies for these incidental findings that will maximize the detection of occult cancer, while minimizing unnecessary imaging surveillance.
| Smith-Bindman, Rebecca; Lebda, Paulette; Feldstein, Vickie A et al. (2013) Risk of thyroid cancer based on thyroid ultrasound imaging characteristics: results of a population-based study. JAMA Intern Med 173:1788-96 |
| Levine, Deborah; Brown, Douglas L; Andreotti, Rochelle F et al. (2010) Management of asymptomatic ovarian and other adnexal cysts imaged at US: Society of Radiologists in Ultrasound Consensus Conference Statement. Radiology 256:943-54 |
| Levine, Deborah; Brown, Douglas L; Andreotti, Rochelle F et al. (2010) Management of asymptomatic ovarian and other adnexal cysts imaged at US Society of Radiologists in Ultrasound consensus conference statement. Ultrasound Q 26:121-31 |
