HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2353040473v1 235/3/934    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Beinfeld, M. T. Articles by Gazelle, G. S. Search for Related Content PubMed PubMed Citation Articles by Beinfeld, M. T. Articles by Gazelle, G. S.

Diagnostic Imaging Costs: Are They Driving Up the Costs of Hospital Care?¹

¹ From the Institute for Technology Assessment and Department of Radiology, Massachusetts General Hospital, Harvard Medical School, 101 Merrimac St, 10th Floor, Boston, MA 02114-4724 (M.T.B., G.S.G.); and Department of Health Policy and Management, Harvard University School of Public Health, Boston, Mass (G.S.G.). From the 2002 RSNA Annual Meeting. Received March 11, 2004; revision requested May 21; revision received July 19; accepted August 18. Address correspondence to G.S.G.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To retrospectively determine how changes in utilization of computed tomography (CT), magnetic resonance (MR) imaging, and other imaging technologies between 1996 and 2002 influenced costs of inpatient hospital care at one large academic medical center.

MATERIALS AND METHODS: Institutional review board did not require its approval or patient informed consent for studies with use of billing data. Patient anonymity was protected by removal of potentially identifying information. Data on hospital costs for 17 139 patients admitted to Massachusetts General Hospital, Boston, Mass, between 1996 and 2002 were downloaded from hospital cost-accounting system; sample was restricted to inpatients with diagnoses in diagnosis-related groups 014–015 (Stroke and TIA [transient ischemic attack]), 164–167 (Appendectomy), 082 (Lung Cancer), 182–183 (Upper Gastrointestinal Conditions), 148–149 (Colon Cancer), and 243 (Back Problems). For each patient, data on demographics, all products and services used, and costs associated with each product or service were obtained. By using institutional codes, we calculated costs of CT, MR imaging, and total imaging relative to total hospital costs. Statistical analyses were performed with Student t test and multiple linear regression analysis.

RESULTS: Between 1996 and 2002, number of inpatient CT and MR images obtained at the hospital more than doubled. In 2002, hospital costs were 155% those of 1996 levels; inpatient imaging costs were 151% those of 1996 levels. Total costs increased an average of 7.8% per year; imaging costs increased 8.3% per year. Although highly variable over the study period, as a percentage of total imaging costs, CT and MR imaging costs appeared to remain stable relative to costs of other imaging modalities.

CONCLUSION: Despite substantial increases in utilization of inpatient CT, MR imaging, and other imaging technologies, diagnostic imaging costs increased at approximately same rate as did total costs for inpatients with several diagnoses. CT and MR imaging do not appear to be driving the cost increases seen between 1996 and 2002.

© RSNA, 2005

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Computed tomography (CT) scanning is now the workhorse of most radiology departments and imaging centers in the United States. With many new indications and more efficient machines, CT scanning has forever changed our ability to diagnose and treat disease. Over the past 10 years, there have been substantial increases in the utilization of CT and magnetic resonance (MR) imaging (1–3). However, with medical care costs now composing 14% of the gross domestic product (4), there is concern that recent increases in the utilization of high-technology imaging studies are contributing heavily to the crisis in health care costs; some have even argued that diagnostic imaging has replaced prescription drugs as the new driver of health care costs. For instance, in a series of analyses of state-level data, "medical technology" (of which imaging is only a part) was one of the most important drivers of health care costs in a variety of settings. Medical technology accounted for 19% of the growth in hospital costs between 1998 and 2000 (5), 7% of the growth in outpatient costs between 1996 and 1999 (6), and 11% of the growth in physician services between 1990 and 2000 (7). Furthermore, in Massachusetts, diagnostic imaging costs increased faster than total patient costs between fiscal years 1999 and 2001. These trends have made diagnostic imaging a potential target for cost-cutting initiatives through such mechanisms as restrictions on physician-self referral, tiered copayment systems, or the development of evidence-based practice guidelines (8).

Though it is well established that the utilization of imaging technology has increased dramatically, few researchers have evaluated this trend in the context of the broader health care system. Findings in one study suggested a relationship between changes in the availability of CT, MR imaging, and other new technologies and health care spending by establishing an association between free-standing imaging centers and health care spending. The study, however, was limited by its short time span (9). Thus, the purpose of our study was to retrospectively determine how changes in the utilization of CT, MR imaging, and other imaging technologies between 1996 and 2002 have influenced costs of inpatient hospital care at a single large academic medical center.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Inpatients
Between 1996 and 2002, more than 250 000 patients were admitted to Massachusetts General Hospital, Boston, Mass, an 875-bed urban teaching hospital that also serves as a tertiary referral center and a major trauma center. We selected a sample of these patients on the basis of diagnoses by using diagnosis-related groups (DRGs). Six DRGs were included in our study because they are associated with a high utilization of imaging studies, are in clinical areas where there were substantial advances in imaging technology over the study period, and represent diseases that affect several different organ systems and therefore that involve different imaging tests. Thus, our study included 17 139 inpatients with diagnoses in the following DRGs: 4888, DRG 014–015 (Stroke and TIA [transient ischemic attack]); 1686, DRG 164–167 (Appendectomy); 1014, DRG 082 (Lung Cancer); 4332, DRG 182–183 (Upper Gastrointestinal Conditions such as esophagitis, gastroenteritis, and diverticulitis); 4068, DRG 148–149 (Colon Cancer); and 1151, DRG 243 (Back Problems) (10).

Data Collection
For each patient, one author (M.T.B.) downloaded information from our institution’s cost-accounting database (Transition Systems Inc; TSI, subsidiary of Eclipsys, Boca Raton, Fla) as follows: age (sex not downloaded), length of hospital stay, length of intensive care unit (ICU) stay, DRG, and Charlson Comorbidity Score (11,12), as well as all products and services used and the costs associated with each product and service. At the time our study was conducted, our institutional review board did not require its approval or patient informed consent for studies using billing data. Patient anonymity, however, was protected by the removal of any potentially identifying information.

In the database, relative value units are assigned to each product or service to reflect the time and/or complexity and are updated annually. Relative value units are then converted to costs by using a set conversion value. Unit costs of products and services consist of both direct and indirect components. The direct cost of each product or service could be variable or fixed, depending on whether or not costs correlate with fluctuations in volume. For example, variable direct costs include consumable supplies and personnel time costs, while fixed direct costs include some portion of the acquisition costs (for major equipment) amortized over the lifetime of the equipment. To calculate these costs, representative time-and-motion studies are used for labor costs, and actual acquisition costs are used for supplies. A portion of indirect or overhead costs is also allocated to each product or service. Overhead costs are allocated on the basis of several factors, such as square footage and utility use, among others.

The sum of unit costs for all products and services used represents the total cost for a patient’s hospitalization. Physician’s fees are excluded because they are maintained separately by our institution’s physician organization. This cost-accounting method has been described previously (13–15).

Utilization and cost of imaging were determined by using institution-specific product codes for CT, MR imaging, and all other imaging modalities. In the cost-accounting database, some CT scans and MR images that are normally ordered as bundled images (ie, images of abdomen and pelvis or of head and neck) are counted as two individual images. As a result, data on the number of images obtained per patient presented in this study may appear inflated. Furthermore, for the purposes of this study, we calculated all costs relative to 1996 levels. Costs were not adjusted for inflation because, theoretically, the force of inflation would drive up both imaging and total hospital costs equally and would not affect the relationship between the two.

Statistical Analysis
All analyses were performed by using computer software (Microsoft Access and Excel 2000 for Windows, Microsoft, Redmond, Wash; SAS, version 8 for Windows, SAS Institute, Cary NC). Comparisons of inpatient demographics and length of hospital stay were made by using the Student t test, where appropriate (two-tailed test, = .05). Multiple linear regression analysis was performed to determine the effect of covariates on total hospital costs and length of stay.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Inpatient Characteristics
In Table 1, demographic characteristics of the study sample according to DRG and of the overall sample are presented. In 2002, there were 752 more patients admitted with these diagnoses than there were in 1996. The mean age of the patients in the overall sample decreased significantly (from 59.6 to 57.9 years, P < .004) over the study period. There was also a significant decline in the mean Charlson Comorbidity Score, which declined from 2.05 in 1998 to 1.86 in 2002 (P < .001). Furthermore, the mean length of hospital stay decreased by 1.6 days (from 7.0 days in 1996 to 5.4 days in 2002, P < .001); however, the mean length of ICU stay remained stable at 0.4 day (P > .99).

We found that, in our sample, only stroke and transient ischemic attack and back problems (DRG 243) were associated with substantial utilization of MR imaging (ie, more than 50 images per year). Among patients with stroke and transient ischemic attack, the number of MR images obtained almost doubled between 1996 and 2002 because of a 47% increase in the number of MR images obtained per patient.

Costs
Between 1996 and 2002, hospital costs at our institution steadily rose at an average rate of 7.8% per year. By 2002, hospital costs were 55% higher than they were in 1996 (Fig 1). Although more variable, trends in inpatient imaging costs paralleled those of total costs. Over the study period, imaging costs increased an average of 8.3% per year. By 2002, imaging costs were 51% higher than they were in 1996.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Graph depicts total hospital and inpatient imaging costs at our institution according to fiscal year for years 1996 through 2002. All costs are presented relative to 1996 levels (ie, 1996 levels are set to 100%). In 2002, total hospital costs were 55% higher than they were in 1996, and inpatient imaging costs were 51% higher than they were in 1996.

View larger version (11K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Graph depicts inpatient imaging costs as a percentage of hospital costs at our institution according to fiscal year for years 1996 through 2002. Imaging costs, although variable between 1996 and 2002, represented approximately 10% of total hospital costs.

View larger version (15K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Graph depicts unit cost per MR image and CT scan at Massachusetts General Hospital, Boston, Mass, according to fiscal year for years 1996 through 2002. All costs are presented relative to 1996 levels (ie, 1996 levels are set to 100%). In 2002, the cost per MR image was 110% that of 1996 levels, whereas the cost per CT scan in 2002 was 59% that of 1996 levels.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Over the study period, total hospital costs increased at approximately the same rate as did inpatient imaging costs. As a percentage of total hospital costs, inpatient imaging costs remained relatively stable between 1996 and 2002, and they represented approximately 10% of the total costs. As a result, diagnostic imaging is unlikely to be a dominant driver of hospital costs. However, the rate of increase in imaging costs was tempered by a substantial reduction in the cost per examination, particularly per CT examination.

Imaging costs, broken down according to modality, were so highly variable over the study period that we were not able to determine whether there was any technology substitution of CT or MR imaging for other modalities. Although highly variable over the study period, costs of MR imaging appeared to represent a slightly greater percentage of imaging costs relative to CT, while the costs of other imaging modalities declined somewhat.

In multiple regression analyses, we found that imaging costs were not predictive of total hospital costs. We did find, however, that imaging costs were predictive of length of stay. According to the model, an increase of $385 in total imaging costs per patient was associated with a reduction of 1 day in total length of stay.

The principal limitations of this study relate to how costs were calculated and what costs were included in the study. In this study, we relied solely on the cost-accounting system at our institution. Although the costs derived from this system may accurately reflect costs at our institution, the generalizability of the findings to other institutions and health care settings may be limited. Specifically, how overhead costs are allocated and how relative value units are assigned to individual products or services may vary between institutions. For example, at our institution, imaging tests with higher relative value units, such as CT angiography and MR angiography, were introduced during the study period. This may have skewed the unit costs for CT and MR imaging in the later years of the study. Nevertheless, we believe that findings at our institution are likely comparable to those that would have been observed at other large urban teaching hospitals.

Another important limitation of our study relates to our analysis of only inpatient costs. Most imaging is now performed on an outpatient basis. If imaging studies that had been performed on an inpatient basis were shifted to the outpatient setting during the study period, our analysis would have underestimated the contribution of imaging costs to total costs.

Furthermore, costs presented in this study were not adjusted for inflation in medical care costs or changes in case mix. We did not think that adjustment for inflation was necessary because the objective of the study was to analyze how changes in diagnostic imaging costs have changed relative to total inpatient costs. Theoretically, the force of inflation would have the same effect on both diagnostic imaging costs and total inpatient costs and thus would not change the nature of the relationship between the two. In addition, we thought that adjustment for case mix was not warranted in this study. Case mix indexes are typically used to control for changes in distribution of diagnoses (ie, severity of disease) over time and are calculated by using DRGs. In this study, we held diagnoses constant by selecting patients with a fixed set of six DRGs. We further attempted to control for severity of disease by collecting data on patient characteristics, such as age and Charlson Comorbidity Score. Interestingly, we found that both age and Charlson Comorbidity Score declined significantly (P < .004 and < .001, respectively) over the study period, and this decline suggested that, among the diagnoses we chose to include in the study, patients were actually less ill over time. Thus, it is unlikely that severity of disease can explain increases in either diagnostic imaging costs or total inpatient costs in this study.

In an evaluation of the value of improved imaging technology, ideally we would like to show that the use of today’s high-technology imaging services has somehow resulted in improved patient outcomes and/or decreased costs. Although our study cannot address patient outcomes, we have shown, at the very least, that imaging may not increase hospital costs at the same rate that imaging utilization has increased. Our results would therefore suggest that across-the-board limits on imaging utilization would be ill-advised.

In conclusion, in our study, although inpatient diagnostic imaging costs increased dramatically over the 7-year study period, total hospital costs increased at approximately the same rate. Inpatient imaging costs remained stable at approximately 10% of total hospital costs. Thus, although increases in imaging costs contribute to increases in hospital costs, so too do many other factors, and imaging costs cannot fully explain the observed trends.

	FOOTNOTES

 
Abbreviations: DRG = diagnosis-related group, ICU = intensive care unit

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, M.T.B., G.S.G.; study concepts and design, M.T.B., G.S.G.; literature research, M.T.B.; data acquisition, M.T.B.; data analysis/interpretation, M.T.B., G.S.G.; statistical analysis, M.T.B.; manuscript preparation, definition of intellectual content, editing, revision/review, and final version approval, M.T.B., G.S.G.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Maitino AJ, Levin DC, Parker L, Rao VM, Sunshine JH. Nationwide trends in rates of utilization of noninvasive diagnostic imaging among the Medicare population between 1993 and 1999. Radiology 2003; 228:795-801.[Abstract/Free Full Text]
2. Khorasani R, Goel PK, Ma’luf NM, Fox LA, Seltzer SE, Bates DW. Trends in the use of radiology with inpatients: what has changed in a decade. AJR Am J Roentgenol 1998; 170:859-861.[Abstract/Free Full Text]
3. Livstone BJ, Parker L, Levin DC. Trends in the utilization of MR angiography and body MR imaging in the US Medicare population: 1993–1998. Radiology 2002; 222:615-618.[Abstract/Free Full Text]
4. Levit K, Smith C, Cowan C, Lazenby H, Sensenig A, Catlin A. Trends in U.S. health care spending, 2001. Health Aff (Millwood) 2003; 22:154-164.
5. Hay JW. Hospital cost drivers: an evaluation of 1998–2001 state-level data. Am J Manag Care 2003; 9(spec no. 1):SP13-SP24.
6. Hearle K, Koenig L, Rudowitz R, Siegel JM, Dobson A, Ho S. Drivers of expenditure growth in outpatient care services. Am J Manag Care 2003; 9(spec no. 1):SP25-SP33.
7. Koenig L, Siegel JM, Dobson A, Hearle K, Ho S, Rudowitz R. Drivers of healthcare expenditures associated with physician services. Am J Manag Care 2003; 9(spec no. 1):SP34-SP42.
8. Massachusetts Division of Health Care Finance and Policy. Diagnostic imaging: a new cost driver. Massachusetts Division of Health Care Finance and Policy Web site. Available at: www.mass.gov/dhcfp. Published April 2003. Accessed October 2003.
9. Baker L, Birnbaum H, Geppert J, Mishol D, Moyneur E. The relationship between technology availability and health care spending. Health Aff (Millwood) 2003; Nov:W3–537–W3–551. doi:10.1377/hlthaff.W3.537. Published November 5, 2003. Accessed December 17 2003.
10. DRG Guide 2003: a complete guide to confirming DRG assignment Reston, Va: St Anthony Publishing/Medicode, 2003.
11. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373-383.[CrossRef][Medline]
12. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992; 45:613-619.[CrossRef][Medline]
13. Brox AC, Filion KB, Zhang X, et al. In-hospital cost of abdominal aortic aneurysm repair in Canada and the United States. Arch Intern Med 2003; 163:2500-2504.[Abstract/Free Full Text]
14. Second American Thoracic Society Workshop. Understanding costs and cost effectiveness in critical care. Am J Respir Crit Care Med 2002; 165:540-550.[Abstract/Free Full Text]
15. Shwartz M, Young DW, Siegrist R. The ratio of costs to charges: how good a basis for estimating costs? Inquiry 1995; 32:476-481.

This article has been cited by other articles:

				G. W. L. Boland Enhancing CT Productivity: Strategies for Increasing Capacity Am. J. Roentgenol., July 1, 2008; 191(1): 3 - 10. [Abstract] [Full Text] [PDF]

				A. Kemmling, I. Noelte, L. Gerigk, S. Singer, C. Groden, and J. Scharf A Diagnostic Pitfall for Intracranial Aneurysms in Time-of-Flight MR Angiography: Small Intracranial Lipomas Am. J. Roentgenol., January 1, 2008; 190(1): W62 - W67. [Abstract] [Full Text] [PDF]

				S. I. Lee, A. Saokar, K. J. Dreyer, J. B. Weilburg, J. H. Thrall, and P. F. Hahn Does Radiologist Recommendation for Follow-up with the Same Imaging Modality Contribute Substantially to High-Cost Imaging Volume? Radiology, March 1, 2007; 242(3): 857 - 864. [Abstract] [Full Text] [PDF]

				H. H. Dib, S. Y. Hong, Z. Bin, and S. Hongliang Health Care Costs in China: Need for Intervention Journal of Health Management, January 1, 2007; 9(1): 85 - 103. [Abstract] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2353040473v1 235/3/934    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Beinfeld, M. T. Articles by Gazelle, G. S. Search for Related Content PubMed PubMed Citation Articles by Beinfeld, M. T. Articles by Gazelle, G. S.