Challenges With Identifying Indication for Examination in Breast Imaging as a Key Clinical Attribute in Practice, Research, and Policy.

PURPOSE: To assess indication for examination for four breast imaging modalities and describe the complexity and heterogeneity of data sources and ascertainment methods. METHODS: Indication was evaluated among the Population-based Research Optimizing Screening through Personalized Regimens (PROSPR) breast cancer research centers (PRCs). Indication data were reported overall and separately for four breast imaging modalities: digital mammography (DM), digital breast tomosynthesis (DBT), ultrasound (US), and magnetic resonance imaging (MRI). RESULTS: The breast PRCs contributed 236,262 women with 607,735 breast imaging records from 31 radiology facilities. We found a high degree of heterogeneity for indication within and across six data sources. Structured codes within a data source were used most often to identify indication for mammography (59% DM, 85% DBT) and text analytics for US (45%) and MRI (44%). Indication could not be identified for 17% of US and 26% of MRI compared with 2% of mammography examinations (1% DM, 3% DBT). CONCLUSIONS: Multiple and diverse data sources, heterogeneity of ascertainment methods, and nonstandardization of codes within and across data systems for determining indication were found. Consideration of data sources and standardized methodology for determining indication is needed to assure accurate measurement of cancer screening rates and performance in clinical practice and research.

Effectiveness of Digital Breast Tomosynthesis Compared With Digital Mammography: Outcomes Analysis From 3 Years of Breast Cancer Screening.

IMPORTANCE: Breast cancer screening with digital breast tomosynthesis (DBT) combined with digital mammography (DM) decreases false-positive examinations and increases cancer detection compared with screening with DM alone. However, the longitudinal performance of DBT screening is unknown. OBJECTIVES: To determine whether the improved outcomes observed after initial implementation of DBT screening are sustainable over time at a population level and to evaluate the effect of more than 1 DBT screening at the individual level. DESIGN, SETTING, AND PARTICIPANTS: Retrospective analysis of screening mammography metrics was performed for all patients presenting for screening mammography in an urban, academic breast center during 4 consecutive years (DM, year 0; DBT, years, 1-3). The study was conducted from September 1, 2010, to September 30, 2014 (excluding September 2011, which was the transition period from DM to DBT), for a total of 44 468 screening events attributable to a total of 23 958 unique women. Differences in screening outcomes between each DBT year and the DM year, as well as between groups of women with only 1, 2, or 3 DBT screenings, were assessed, and the odds of recall adjusted for age, race/ethnicity, breast density, and prior mammograms were estimated. Data analysis was performed between February 16 and October 26, 2015. EXPOSURE: Digital mammography screening supplemented with DBT. MAIN OUTCOMES AND MEASURES: Recall rates, cancer cases per recalled patients, and biopsy and interval cancer rates were determined. RESULTS: Screening outcome metrics were evaluated for a total of 44 468 examinations attributable to 23 958 unique women (mean [SD] age, 56.8 [11.0] years) over a 4-year period: year 0 cohort (DM0), 10 728 women; year 1 cohort (DBT1), 11 007; year 2 cohort (DBT2), 11 157; and year 3 cohort (DBT3), 11 576. Recall rates rose slightly for years 1 to 3 of DBT (88, 90, and 92 per 1000 screened, respectively) but remained significantly reduced compared with the DM0 rate of 104 per 1000 screened. Reported as odds ratios (95% CIs), the findings were DM vs DBT1, 0.83 (0.76-0.91, P < .001); DM vs DBT2, 0.85 (0.78-0.93, P < .001); and DM vs DBT3, 0.87 (0.80-0.95, P = .003). The cancer cases per recalled patients continued to rise from DM0 rate of 4.4% to 6.2% (P = .06), 6.5% (P = .03), and 6.7% (P = .02) for years 1 to 3 of DBT, respectively. Outcomes assessed for the most recent screening for individual women undergoing only 1, 2, or 3 DBT screenings during the study period demonstrated decreasing recall rates of 130, 78, and 59 per 1000 screened, respectively (P < .001). Interval cancer rates, determined using available follow-up data, decreased from 0.7 per 1000 women screened with the use of DM to 0.5 per 1000 screened with the use of DBT1. CONCLUSIONS AND RELEVANCE: Digital breast tomosynthesis screening outcomes are sustainable, with significant recall reduction, increasing cancer cases per recalled patients, and a decline in interval cancers.

Baseline Screening Mammography: Performance of Full-Field Digital Mammography Versus Digital Breast Tomosynthesis.

OBJECTIVE: Baseline mammography studies have significantly higher recall rates than mammography studies with available comparison examinations. Digital breast tomosynthesis reduces recalls when compared with digital mammographic screening alone, but many sites operate in a hybrid environment. To maximize the effect of screening digital breast tomosynthesis with limited resources, choosing which patient populations will benefit most is critical. This study evaluates digital breast tomosynthesis in the baseline screening population. MATERIALS AND METHODS: Outcomes were compared for 10,728 women who underwent digital mammography screening, including 1204 (11.2%) baseline studies, and 15,571 women who underwent digital breast tomosynthesis screening, including 1859 (11.9%) baseline studies. Recall rates, cancer detection rates, and positive predictive values were calculated. Logistic regression estimated the odds ratios of recall for digital mammography versus digital breast tomosynthesis for patients undergoing baseline screening and previously screened patients, adjusted for age, race, and breast density. RESULTS: In the baseline subgroup, recall rates for digital mammography and digital breast tomosynthesis screening were 20.5% and 16.0%, respectively (p = 0.002); digital breast tomosynthesis screening in the baseline subgroup resulted in a 22% reduction in recall compared with digital mammography, or 45 fewer patients recalled per 1000 patients screened. Digital breast tomosynthesis screening in the previously screened patients resulted in recall reduction of 14.3% (p < 0.001; p for interaction = 0.21). The recall rate reduction for baseline screening was especially pronounced in women younger than 50 years (p = 0.005). DBT implementation resulted in an increase in cancer detection in the baseline subgroup of 40.5% versus an increase in the previously screened subgroup of 17.4%. DBT implementation resulted in an increase in PPV1 in the baseline subgroup of 85% versus 35.3% in the previously screened subgroup, although the p-interaction was not significant. CONCLUSION: If resources are limited, women younger than 50 years who are undergoing baseline screening or do not have prior available mammograms may benefit more from digital breast tomosynthesis than from digital mammography alone.

Clinical predictors of urgent findings on abdominopelvic CT in emergency department patients with Crohn's disease.

BACKGROUND: Patients with Crohn's disease (CD) are frequently exposed to diagnostic radiation in emergency departments (EDs). We aimed to examine clinical predictors of urgent abdominopelvic computed tomography (APCT) findings in this population. METHODS: A retrospective cross-sectional study was performed among adults with CD presenting to 2 emergency departments with a gastrointestinal chief complaint. The outcome, APON (abscess, perforation, obstruction, new or worsening non-CD-related findings), included APCTs with new or worsening CD-related or non-CD-related urgent findings. Variables with P < 0.05 in bivariate analyses were included in a multivariable logistic regression model, which was also used to develop a risk score for APON. RESULTS: A total of 481 APCTs were performed and 166 (34.5%) identified APON. Variables retained in the final model were history of intestinal obstruction (odds ratio [OR]: 3.78, 95% confidence interval [CI]: 2.27-6.28), history of intraabdominal abscess (OR: 2.64, 95% CI: 1.43 to 4.88), current hematochezia (OR: 0.38, 95% CI: 0.21 to 0.68), and white blood cell count >12,000/µL (OR: 2.49, 95% CI: 1.63 to 3.84). The c-statistic was 0.72. The risk score subtracts 1 point for hematochezia, and adds 1 point for each of the other variables. Among patients with a risk score of -1, the predicted and observed risk for APON was 9% and 6%, respectively. Any score greater than -1 had a predicted and observed risk of 19.8% and higher. CONCLUSIONS: An APON risk score of -1 is associated with a low risk of urgent APCT findings in patients with CD in the emergency department. Implementation of such a tool may support clinical decision-making in the ED setting.

Use of abdominopelvic computed tomography in emergency departments and rates of urgent diagnoses in Crohn's disease.

BACKGROUND & AIMS: In the United States, the use of abdominopelvic computed tomography (APCT) by emergency departments for patients with abdominal pain has increased, despite stable admission rates and diagnosis requiring urgent intervention. We proposed that trends would be similar for patients with Crohn's disease (CD). METHODS: We conducted a retrospective study of data from 648 adults with CD who presented at 2 emergency departments (2001-2009; 1572 visits). Trends in APCT use were assessed with Spearman correlation coefficient. We compared patient characteristics and APCT findings during 2001-2003 and 2007-2009. RESULTS: APCT use increased from 2001 (used for 47% of encounters) to 2009 (used for 78% of encounters; P = .005), whereas admission rates were relatively stable at 68% in 2001 and 71% in 2009 (P = .06). The overall proportion of APCTs with findings of intestinal perforation, obstruction, or abscess was 29.0%; 34.9% of APCTs were associated with urgent diagnoses, including those unrelated to CD. Between 2001-2003 and 2007-2009, the proportions of APCTs that detected intestinal perforation, obstruction, or abscess were similar (30% vs 29%, P = .92), as were the proportions used to detect any diagnosis requiring urgent intervention, including those unrelated to CD (36% vs 34%, P = .91). CONCLUSIONS: Despite the increased use of APCT by emergency departments for patients with CD, there were no significant changes in admission rates between the periods of 2001-2003 and 2007-2009. The proportion of APCTs that detected intestinal perforation, obstruction, abscess, or other urgent conditions not related to CD remained high.

Evaluation of the use of electronic health data to classify four-year mortality risk for older adults undergoing screening colonoscopies.

Current cancer screening recommendations often apply coarse age cutoffs for screening requirements without regard to predicted life expectancy. Using these cutoffs, healthier older patients may be under-screened, and sicker younger patients may be screened too often. Mortality risk classification using EHR data could be used to tailor screening reminders to physicians in ways that better align screening recommendations with patients who are more likely to live long enough to benefit from early detection. We have evaluated the performance of an existing prognostic index for 4-year mortality using data readily available in the electronic health record (EHR), and investigated the effect of the index in retrospective cohorts of adults age 65 and older undergoing screening colonoscopy. Risk scores in this adaptation of a four-year prognostic index were found to be associated with actual death rates and consistent with mortality rates from a national sample. Our results demonstrate that data extracted from electronic health records can be used to classify mortality risk. With improvements, including extension to a 5-year mortality model with inclusion of additional variables and extension of variable definitions, informatics methods to implement mortality models may prove to be clinically useful in tailoring screening guidelines.

Actual financial comparison of four strategies to evaluate patients with potential acute coronary syndromes.

OBJECTIVES: Small studies have shown that a negative computed tomography coronary angiogram (CTA) in low-risk chest pain patients predicts a low rate of 30-day adverse events. The authors hypothesized that an immediate CTA strategy would be as effective but less costly than alternative strategies for evaluation of patients with potential acute coronary syndrome (ACS). METHODS: The authors retrospectively compared four strategies for evaluation of patients after initial physician determination that the patient required admission and testing to rule out ACS. Patients were frequency-matched by age, race, gender, thrombolysis in myocardial infarction (TIMI) score, and initial electrocardiogram (ECG). The four groups were immediate CTA in the emergency department (ED) without serial markers (n = 98); clinical decision unit/observation unit (CDU) with biomarkers and CTA (n = 102); CDU evaluation with serial cardiac biomarkers and stress testing (n = 154); and usual care, defined as admission with serial biomarkers and hospitalist-directed evaluation (n = 289). The main outcomes were actual cost of care (facility direct and indirect fixed, facility variable direct labor and supply costs), length of stay (LOS), diagnosis of coronary artery disease (CAD), and safety (30-day death or myocardial infarction [MII). RESULTS: Patients in each group were of similar age (mean +/- standard deviation [SD] 46 +/- 9 years), race (62% African American), and gender (57% female) and had similar TIMI scores (100% between 0-2). Comparing immediate CTA versus CDU CTA versus CDU stress versus usual care, median costs were less ($1,240 vs. 2,318 vs. 4,024 vs. 2,913; p < 0.01), and LOS was shorter (8.1 hr vs. 20.9 hr vs. 26.2 hr vs. 30.2 hr; p < 0.01). Diagnosis of CAD was similar (5.1% vs. 5.9% vs. 5.8% vs. 6.6%; p = 0.95), but fewer patients had 30-day death/MI (0% vs. 0% vs. 0.7% vs. 3.1%; p = 0.04) or 30-day readmission (0% vs. 3.2% vs. 2.3% vs. 12.2%; p < 0.01). CONCLUSIONS: Compared to the other strategies, immediate CTA was as safe, identified as many patients with CAD, had the lowest cost, had the shortest LOS, and allowed discharge for the majority of patients. Larger prospective studies should confirm safety before immediate CTA replaces other strategies to rule out possible ACS.

Enriching knowledge domain visualizations: analysis of a record linkage and information fusion approach to citation data.

This article presents a study of the use of data preparation for data mining methodology to prepare biomedical citation data for visualization. Deterministic record linkage models were compared with probabilistic record linkage in a situation for which the truth is known through the use of gold standard or truth datasets. The linkages are evaluated on data from the Web of Science (WOS) and Medline citation databases. Sensitivity, specificity, and overall performance of record linkage models were empirically compared with ROC analysis. Data quality and visualization metrics are presented for datasets prepared with and without probabilistic record linkage and information fusion of Medline abstracts and MESH terms into WOS citation records. The major contributions of this work are to specifically develop a novel model of record linkage for biomedical citation databases, with the objective of improving and enriching biomedical knowledge domain visualizations.

CiteSpace II: visualization and knowledge discovery in bibliographic databases.

This article presents a description and case study of CiteSpace II, a Java application which supports visual exploration with knowledge discovery in bibliographic databases. Highly cited and pivotal documents, areas of specialization within a knowledge domain, and emergence of research topics are visually mapped through a progressive knowledge domain visualization approach to detecting and visualizing trends and patterns in scientific literature. The test case in this study is progressive knowledge domain visualization of the field of medical informatics. Datasets based on publications from twelve journals in the medical informatics field covering the time period from 1964-2004 were extracted from PubMed and Web of Science (WOS) and developed as testbeds for evaluation of the CiteSpace system. Two resulting document-term co-citation and MeSH term co-occurrence visualizations are qualitatively evaluated for identification of pivotal documents, areas of specialization, and research trends. Practical applications in bio-medical research settings are discussed.

Visual exploration of landmarks and trends in the medical informatics literature.

This study presents preliminary results from a visual study of a new dataset of forty years of citation data from publications of twelve journals in the medical informatics field covering the time period from 1964-2004. Highly cited and pivotal documents, areas of specialization within medical informatics, and emergence of research topics are visually mapped through a progressive knowledge domain visualization approach to detecting and visualizing trends and patterns in scientific literature.