Original article
Characterizing the Performance of the Nation’s Hospitals in the Hospital Outpatient Quality Reporting Program’s Imaging Efficiency Measures

https://doi.org/10.1016/j.jacr.2014.08.022Get rights and content

Abstract

Purpose

To describe the performance of the nation’s hospitals in terms of the Hospital Outpatient Quality Reporting Program’s imaging efficiency measures.

Methods

Data were obtained from the Hospital Compare website and reflect outpatient Medicare claims of 4,118 hospitals for 5 imaging efficiency metrics: (1) frequency of combination abdominal CT (performed with and without intravenous contrast); (2) combination chest CT (performed with and without intravenous contrast); (3) simultaneous brain/sinus CT; (4) mammography follow-up (diagnostic imaging after screening mammography); and (5) lumbar spine MRI for low back pain without prior conservative therapy. Metrics were summarized and compared with other hospital characteristics.

Results

Median frequency was 36.7% for lumbar spine MRI for low back pain and ranged from 1.6% to 7.8% for the remaining measures; however, extreme outliers were observed (maximal frequencies of 79.2%-95.2% for mammography follow-up and combination chest and abdominal CT). Essentially no correlation was found among measures, aside from combination abdominal and chest CT. For some measures, relatively poor performance was more commonly observed among critical access hospitals and physician-owned/proprietary hospitals, and less commonly observed among U.S. News & World Report “best” hospitals and primary residency teaching sites. Frequencies for combination abdominal and chest CT improved from 2013 to 2014 among hospitals with relatively poorer performance.

Conclusions

Although the imaging efficiency measures help identify individual hospitals and hospital categories with relatively inefficient imaging practices, they do not readily identify distinctly positively performing hospitals. Excess utilization was suggested for lumbar spine MRI. Frequency of combination abdominal and chest CT examinations improved over a short time interval.

Introduction

CMS implemented the Hospital Outpatient Quality Reporting (OQR) Program as a quality initiative aimed at improving hospital outpatient care in the United States through greater transparency to consumers and an emphasis on value-driven care [1]. Since taking effect in 2009, this program has required that hospitals collect and submit to CMS for public reporting a panel of standardized measures of care in order to receive the full annual update to their HOPPS payment rate. The initiative is intended to yield a uniform set of robust metrics that patients, payers, regulatory agencies, and hospitals themselves may use to compare performance among hospitals and conduct quality-improvement efforts.

The Hospital OQR Program incorporates 6 measures related to medical imaging for purposes of 2014 HOPPS payment determinations, all of which may be computed from standard Medicare fee-for-service claims data without any additional submission of data by hospitals [2]. Three of these measures pertain to “combination” CT scans: abdominal CT scans performed with and without intravenous contrast, chest CT scans performed with and without intravenous contrast, and simultaneously performed brain and sinus CT scans. One measure pertains to the performance of lumbar spine MRI for low back pain without documentation of previous conservative therapy; 1 measure pertains to the frequency of diagnostic breast imaging of any modality following screening mammography; and 1 measure pertains to the use of cardiac imaging for preoperative risk assessment for noncardiac low-risk surgery [2].

CMS indicates that these measures are intended to track potentially inappropriate medical imaging and that the reporting of the measures may lead to reduced cost and lower levels of exposure to radiation and intravenous contrast agents, in addition to improving adherence to evidence-based guidelines 3, 4, 5, 6, 7. CMS notes that lower percentages are generally more favorable, making exceptions for clearly indicated examinations (eg, combination abdominal CT for adrenal lesion evaluation [3]), as well as noting that for the diagnostic mammography measure, a percentage that is too low may also be inappropriate [6]. The extent of CMS’s concern regarding these measures is evidenced by its statements that such examinations constitute “indiscriminate use” that “represents a serious inefficiency of practice” with “enormous cost implications,” potentially relating to “a direct financial benefit to the service provider” 3, 4, 5, 6, 7. Subsequent to the initiation of tracking of these measures, concern regarding overutilization of combination chest CT studies was the basis of a front-page article published in The New York Times in 2011 [8].

Although these hospital-reported metrics are now publicly available and can be readily accessed via the Internet [9], data summarizing the performance of the nation’s hospitals and identifying potential trends are scarce. Such insights are important if the Hospital OQR Program is to achieve its intended purpose of catalyzing actual performance improvement. Likewise, any flaws in the metrics are important to uncover, given the resources involved in their collection and reporting and the potential of constructing alternative metrics. Therefore, the purpose of the current study is to describe the current level of performance of the nation’s hospitals in terms of the Hospital OQR Program’s imaging efficiency measures and to identify relevant associations and patterns of variation to help further characterize this data set.

Section snippets

Source of Data

As this study used solely aggregate data, institutional review board approval was not required. Data files were obtained from the publicly available Hospital Compare website [9], which is managed by CMS in conjunction with the Hospital OQR Program. CMS calculates the data based on claims for beneficiaries of traditional Medicare that are submitted by hospitals paid through HOPPS. Medicare patients treated in the inpatient setting, as well as non-Medicare populations treated in any setting, are

Results

The median values of the Hospital OQR Program imaging measures among all reporting hospitals within the United States were 1.6% for combination chest CT studies, 2.3% for simultaneous brain/sinus CT studies, 7.8% for combination abdominal CT studies, 8.3% for mammography follow-up rates, and 36.7% for lumbar spine MRI studies for low back pain (Table 1).

Figure 1 shows the distribution of the imaging measures among US hospitals. For all measures aside from lumbar spine MRI for low back pain, a

Discussion

The Hospital OQR Program aims to collect data that the public and other stakeholders may use to assess the quality of hospital care. The program’s metrics that are directly related to radiology assess “imaging efficiency” by determining the frequency with which hospital outpatient imaging departments perform examinations that are deemed to entail excessive, if not needless, resource utilization most of the time. The observed mean mammography follow-up rate of 9.1% satisfies a published target

Conclusions

The median frequency among the nation’s hospitals ranged from 2.3% to 7.8% for all of the Hospital OQR Program’s radiology-related imaging efficiency metrics, aside from frequency of lumbar spine MRI for low back pain (median 36.7%); in addition, the observed frequencies are comparable to published benchmarks for those 2 measures for which such benchmarks exist. Nonetheless, a considerable fraction of hospitals reported very high frequencies of the metrics that were outliers from the bulk of

Take-Home Points

  • Among the Hospital OQR Program’s 5 imaging efficiency measures, highest utilization among the nation’s hospitals was of lumbar spine MRI performed for low back pain (median frequency of 36.7%).

  • For the remaining 4 measures (combination abdominal CT, combination chest CT, simultaneous brain/sinus CT, mammography follow-up rate), the median frequency ranged from 1.6% to 7.8%.

  • A considerable minority of hospitals considered to be poorly performing were outliers in terms of exhibiting markedly

References (21)

  • CMS. Hospital outpatient quality reporting (OQR) specifications manual. V6.0b. Available at:...
  • QualityNet. Imaging efficiency measures. Available at:...
  • Agency for Healthcare Research and Quality. NQF-endorsed measure. Imaging efficiency: percentage of abdomen CT studies...
  • QualityNet. OP-14: simultaneous use of brain CT and sinus CT. Available at:...
  • Agency for Healthcare Research and Quality. NQF-endorsed measure. Imaging efficiency: ratio of thorax CT studies that...
  • Agency for Healthcare Research and Quality. NQF-endorsed measure. Imaging efficiency: percentage of patients with...
  • Agency for Healthcare Research and Quality. NQF-endorsed measure. Imaging efficiency: percentage of MRI of the lumbar...
  • Bogdanich W, Mcginty JC. Medicare claims show overuse for CT scanning. The New York Times, July 17, 2011. Available at:...
  • CMS. Official hospital compare data. Available at: https://data.medicare.gov/data/hospital-compare. Accessed April...
  • CMS. Medicare learning network. Rural health fact sheet series: critical access hospital. Available at:...
There are more references available in the full text version of this article.

Cited by (17)

  • Changes in Hospital Outpatient Quality Reporting Program Brain CT Efficiency Performance, 2013 to 2018

    2020, Journal of the American College of Radiology
    Citation Excerpt :

    Fifth, although hospitals acting as primary teaching sites for university-based radiology residency programs exhibited a trend toward superior brain CT efficiency in 2013, they performed less efficiently on average than other hospitals in 2018. This finding is somewhat counterintuitive, because some researchers have suggested that radiology trainees may exhibit a relatively high level of adherence to ordering and protocol guidelines compared with physicians in nonacademic settings [2,6]. However, the fact that hospitals acting as primary sites for radiology residency programs performed significantly worse than non-residency-affiliated hospitals on average in 2018 suggests this may not be the case.

  • National Quality Improvement Participation Among US Radiation Oncology Facilities: Compliance with Guideline-Concordant Palliative Radiation Therapy for Bone Metastases

    2020, International Journal of Radiation Oncology Biology Physics
    Citation Excerpt :

    In addition to these gross correlations, quality metrics for radiology have also been assessed using HCD. Rosenkrantz et al found that critical access hospitals, physician-owned/proprietary hospitals, and nonteaching hospitals were relatively low performing in a subset of imaging-specific quality metrics.16 Compared with the quality measures assessed in these other studies, guideline-concordant EBRT for bony metastases provides a more granular assessment of RO performance on a nationwide scale.

  • Hospital efficiency and utilization of high-technology medical equipment: A panel data analysis

    2018, Health Policy and Technology
    Citation Excerpt :

    Third, overuse of CT and MRI has been increasingly recognized a significant public health problem, raising serious concerns with wasteful costs and adverse health outcomes [14–16]. Given the complicated consequences for both health care costs and quality that are created by the widespread use of CT and MRI, some researchers have paid attention to the association between hospital characteristics and utilization of these types of equipment [17–19]. This study aims to take a step further by focusing on the relationship between hospital efficiency and utilization of CT and MRI.

  • Double Scan CT Rates: An Opportunity for Facility-Based Radiologist Measures in the Quality Payment Program

    2018, Journal of the American College of Radiology
    Citation Excerpt :

    Past studies have explored performance in double scan rates but used very distinct data sources and were thus unable to conduct the evaluation at the level of individual radiologists as we currently present. Of note, Rosenkrantz and Doshi used data from Hospital Compare to assess double scan rates solely at the level of hospital outpatient departments [4]. In addition, Flug et al [5] and Levin et al [23] used data from Physician Supplier Procedure Summary Master Files to assess double scan rates collectively across all radiologists nationally.

View all citing articles on Scopus
View full text