Faculty Bibliography

PURPOSE/OBJECTIVE:The aim of this study was to assess national and state-specific changes in emergency department (ED) chest imaging utilization from 1994 to 2015. METHODS:Using aggregate 100% Medicare Physician/Supplier Procedure Summary Master Files for 1994 to 2015, the annual frequency of chest imaging in Medicare Part B beneficiaries in the ED setting was identified, and utilization was normalized to annual Medicare enrollment as well as annual ED visits. Using individual Medicare beneficiary 5% research-identifiable files, similar determinations were performed for each state. RESULTS:Between 1994 and 2015, per 1,000 beneficiaries, ED utilization of chest radiography and CT increased by 173% (compound annual growth rate [CAGR] 4.9%) and 5,941.8% (CAGR 21.6%). Per 1,000 ED visits, utilization increased by 81% (CAGR 2.9%) and 3,915.4% (CAGR 19.2%), respectively. Across states, utilization was highly variable, with 2015 radiography utilization per 1,000 ED visits ranging from 82 (Wyoming) to 731 (Hawaii) and CT utilization ranging from 18 (Wyoming) to 76 (Hawaii). Between 2004 and 2015, most states demonstrated increases in the utilization of both radiography (maximal increase of CAGR 11.0% in Vermont) and CT (maximal increase of CAGR 21.0% in Maine). Nonetheless, utilization of radiography declined in four states and utilization of CT in a single state. CONCLUSIONS:Over the past two decades, ED utilization of chest imaging has increased. This was related not only to an increasing frequency of ED visits but also to increasing utilization per ED visit. Across states, utilization is highly variable, but with radiography and CT both increasing, the use of CT seems additive to, rather than replacing, radiography.

The publisher regrets that this article has been temporarily removed. A replacement will appear as soon as possible in which the reason for the removal of the article will be specified, or the article will be reinstated. The full Elsevier Policy on Article Withdrawal can be found at https://www.elsevier.com/about/our-business/policies/article-withdrawal.

PURPOSE/OBJECTIVE:Radiologists have historically participated as individuals in CMS pay-for-performance programs, but little is known about how radiologists perform under increasingly available group participation. We aimed to assess radiologists' relative national performance on CMS quality metrics using group versus individual participation. METHODS:Radiologists' group- and individual-level 2016 performance on Physician Quality Reporting System (PQRS) and non-PQRS Qualified Clinical Data Registry (QCDR) measures were obtained from the CMS national Physician Compare database and compared. RESULTS:Radiology groups reported an average 4.6 ± 2.0 quality measures; individual radiologists reported 2.3 ± 1.2 (P < .001). At least six measures were reported by 31.5% of groups versus 1.0% of individuals. Only one measure was reported by 5.4% of groups versus 33.0% of individuals. Groups reported 21 unique measures (20 via registries and one via QCDR). For 8 of the 11 measures reported by 20 or more groups, the average group performance rate was 3% or better than the average performance rate among radiologists participating as individuals (maximum 14% improvement with group participation versus individual participation for any individual measure). Group and individual performance were similar for the remaining three such measures. For measures reported by 20 or more groups in which a higher score indicates better performance, average group performance rates ranged from 86.2% to 98.9%. CONCLUSION/CONCLUSIONS:Compared with individual participation in CMS quality performance programs, radiologists participating as a group reported larger numbers of quality measures and achieved higher performance rates on those measures. Radiology practices seeking success under Medicare's new Quality Payment Program should carefully explore group participation.

PURPOSE/OBJECTIVE:To assess characteristics of physicians and other providers frequently ordering intravenous pyelography (IVP). METHODS:The 2014 Medicare Referring Provider Utilization for Procedures data set was used to identify providers who ordered more than 10 IVP examinations ("high-ordering providers") in Medicare beneficiaries. The Medicare Provider and Other Supplier Public Use File and Physician Supplier Procedure Summary Master Files were used to obtain physician characteristics and total service counts, respectively. RESULTS:Of 18,344 IVPs performed in 2014 in Medicare fee-for-service beneficiaries, 6,321 (34.5%) were ordered by just 233 high-ordering providers. Of these, 220 (94.4%) were urologists. These urologists represented just 2.4% of all 8,981 Medicare-participating urologists and ordered an average of 27.1 IVPs (maximum 239). Urologists ordering IVPs (versus those not ordering IVPs) were more likely (P < .05) to practice in rural areas (6.4% versus 2.7%), be in practice more than 15 years (87.4% versus 71.2%), and be in practices with 100 members or fewer (71.3% versus 55.5%). They were also less likely (P < .05) to be female (3.2% versus 7.4%) and in academic practices (5.1% versus 10.7%). High-IVP-ordering urologists were more likely to practice in the South (54.1% versus 36.9%) or Midwest (30.0% versus 21.3%) and less likely to practice in the Northeast (5.0% versus 23.1%) or West (10.9% versus 18.6%). CONCLUSION/CONCLUSIONS:Although uncommonly performed, IVPs continue to be used in the Medicare population. Providers most likely to frequently order IVPs were later-career urologists in smaller and rural practices in the South. Targeting education and appropriate use criteria initiatives to high-ordering providers may help optimize utilization.

Radiologists play a critical role in helping the health care system achieve greater value. Unfortunately, today radiology is often judged by simple "checkbox" metrics, which neither directly reflect the value radiologists provide nor the outcomes they help drive. To change this system, first, we must attempt to better define the elusive term value and, then, quantify the value of imaging through more relevant and meaningful metrics that can be more directly correlated with outcomes. This framework can further improve radiology's value by enhancing radiologists' integration into the care team and their engagement with patients. With these improvements, we can maximize the value of imaging in the overall care of patients.

PURPOSE/OBJECTIVE:To determine whether the presence of an ultrasound hypoechoic region at the site of a MRI region of interest (ROI) results in improved prostate cancer (PCa) detection and predicts clinically significant PCa on MRI - ultrasound fusion targeted prostate biopsy (MRF-TB). MATERIALS AND METHODS/METHODS:Between July 2011 and June 2017, 1058 men who underwent MRF-TB and systematic biopsy by a single surgeon were prospectively entered into an IRB-approved database. MRI ROI were identified and scored for suspicion by a single radiologist. Each MRI ROI was prospectively evaluated for presence of a hypoechoic region at the site (ROI-HyR) by the surgeon and graded as 0,1,2 representing none, poorly demarcated, or well demarcated, respectively. Interaction of MRI suspicion score (mSS) and US grade (USG), and prediction of cancer detection rate (CDR) by USG, was evaluated by univariate and multivariate analysis. RESULTS:For 672 men, overall and Gleason Score (GS)≥7 CDR were 61.2% and 39.6%, respectively. CDR for USG 0,1,2 were 46.2%, 58.6%, 76.0% (p<0.001) for any cancer and 18.7%, 35.2%, 61.1% (p<0.001) for GS≥7, respectively. For MRF-TB only, GS≥7 CDR for USG 0,1,2 was 12.8%, 25.7%, 52.0% (p<0.001), respectively. On univariate analysis, among mSS 2-4, USG was predictive of GS≥7 CDR. Multivariable regression analysis revealed USG, PSAD, and mSS were predictive of GS≥7 PCa on MRF-TB. CONCLUSIONS:Ultrasound findings at the site of MRI ROI independently predict the likelihood of GS≥7 PCa, as men with a well demarcated ROI-HyR at the time of MRF-TB have a higher risk than men without.

PURPOSE/OBJECTIVE:To investigate the national coverage landscape for prostate MRI services, assessing the presence of updated and accurate coverage requirements by private payers. METHODS:The database Policy Reporter was used to evaluate private payer coverage related to prostate MRI for 81 plans covering 149 million people in the United States. Both the indications and requirements for prostate MRI coverage were recorded in a variety of clinical scenarios, including initial diagnosis, staging, active surveillance, and suspected recurrence. RESULTS:Overall, 11.1% of payers cover prostate MRI in biopsy-naïve patients with suspected prostate cancer, with the remaining 88.9% requiring a prior negative biopsy. Nearly all payers also require either a rising prostate-specific antigen (PSA) or abnormal digital rectal examination (DRE). Rarely, a planned future MRI-targeted biopsy serves as a basis for MRI coverage. Initial staging is covered by most payers, although typically with stringent indications (eg, PSA ≥ 20 ng/mL, Gleason score ≥7 or 8, stage T3 or T4, or ≥20% risk of nodal metastases). Only 10 payers discuss active surveillance, with 8 of these requiring a repeat biopsy before MRI. Coverage for detection of post-treatment recurrence often requires a rising PSA or abnormal DRE, and occasionally only if a CT is first performed; only 10 of 81 payers address coverage after androgen deprivation treatment. CONCLUSION/CONCLUSIONS:Prostate MRI coverage varies widely among private payers, fails to recognize major clinical scenarios, is overly restrictive, and is often not reflective of current clinical practice. This creates challenges for patients and referring physicians seeking to obtain ready access to prostate MRI services.

OBJECTIVE:To use Twitter to characterize public perspectives regarding artificial intelligence (AI) and radiology. METHODS AND MATERIALS/METHODS:Twitter was searched for all tweets containing the terms "artificial intelligence" and "radiology" from November 2016 to October 2017. Users posting the tweets, tweet content, and linked websites were categorized. RESULTS:Six hundred and five tweets were identified. These were from 407 unique users (most commonly industry-related individuals [22.6%]; radiologists only 9.3%) and linked to 216 unique websites. 42.5% of users were from the United States. The tweets mentioned machine/deep learning in 17.2%, industry in 14.0%, a medical society/conference in 13.4%, and a university in 9.8%. 6.3% mentioned a specific clinical application, most commonly oncology and lung/tuberculosis. 24.6% of tweets had a favorable stance regarding the impact of AI on radiology, 75.4% neutral, and none were unfavorable. 88.0% of linked websites leaned toward AI being positive for the field of radiology; none leaned toward AI being negative for the field. 51.9% of linked websites specifically mentioned improved efficiency for radiology with AI. 35.2% of websites described challenges for implementing AI in radiology. Of the 47.2% of websites that mentioned the issue of AI replacing radiologists, 77.5% leaned against AI replacing radiologists, 13.7% had a neutral view, and 8.8% leaned toward AI replacing radiologists. CONCLUSION/CONCLUSIONS:These observations provide an overview of the social media discussions regarding AI in radiology. While noting challenges, the discussions were overwhelmingly positive toward the transformative impact of AI on radiology and leaned against AI replacing radiologists. Greater radiologist engagement in this online social media dialog is encouraged.

RATIONALE AND OBJECTIVES/OBJECTIVE:To explore downstream costs associated with incidental pulmonary nodules detected on CT. MATERIALS AND METHODS/METHODS:The cohort comprised 200 patients with an incidental pulmonary nodule on chest CT. Downstream events (chest CT, PET/CT, office visits, percutaneous biopsy, and wedge resection) were identified from the electronic medical record. The 2017 Fleischner Society Guidelines were used to classify radiologists' recommendations and ordering physician management for the nodules. Downstream costs for nodule management were estimated from national Medicare rates, and average costs were determined. RESULTS:Average downstream cost per nodule was $393. Costs were greater when ordering physicians over-managed relative to radiologist recommendations ($940) vs. when adherent ($637) or under-managing ($166) relative to radiologists recommendations. Costs were also greater when ordering physicians over-managed relative to Fleischner Society guidelines ($860) vs. when under-managing ($208) or adherent ($292) to guidelines. Costs did not vary significantly based on whether or not radiologists recommended follow-up imaging ($167-$397), nor whether radiologists were adherent or under- or over-recommended relative to Fleischner Society guidelines ($313-$444). Costs were also higher in older patients, patients with a smoking history, and larger nodules. Five nodules underwent wedge resection and diagnosed as malignancies. No patient demonstrated recurrence or metastasis. Average cost per diagnosed malignancy was $3090. CONCLUSION/CONCLUSIONS:Downstream costs for incidental pulmonary nodules are highly variable and particularly high when ordering physicians over-manage relative to radiologist recommendations and Fleischner Society guidelines. To reduce unnecessary utilization and cost from over-management, radiologists may need to assume a greater role in partnering with ordering physicians to ensure appropriate, guideline-adherent, and follow-up testing.