| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Orthopaedic Surgery, Harborview Medical Center, University of
Washington, Box 359798, 325 9th Ave., Seattle, WA 98104-2499.
2
Department of Radiology, Harborview Medical Center, Harborview Injury
Prevention and Research Center, Seattle, WA 98104-2499.
Received May 23, 2000;
accepted after revision September 26, 2000.
Address correspondence to M. B. Henley.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. This was an observational study of imaging use patterns from 1993 to 1998. Data were segregated by patient type and imaging procedure. Annual hospital admissions were adjusted for severity of illness using the Health Care Financing Administration's case-mix index. Per-patient imaging trends for the emergency department and outpatients were assessed using a ratio of the total number of procedures to the number of patient visits. Linear regression models were used to assess the strength of associations between resource use, measured as relative value units (RVUs), and independent variables (calendar year, patient type, and examination type).
RESULTS. The RVUs for all imaging increased 53% for inpatients, 69% for outpatients, and 85% in the emergency department. No significant trend for use was found for the aggregate of inpatient imaging. There was a significant increase in the inpatient MR imaging RVUs (p = 0.04). No significant trend was found for the aggregated outpatient imaging RVUs. The trends were significant for angiography (p = 0.006), MR imaging (p = 0.002), and sonography (p = 0.04). The aggregated emergency room imaging RVUs showed a significant increase over time (p < 0.03).
CONCLUSION. The number of imaging procedures increased during the study period. There was no overall trend toward increasing use of imaging in inpatients once an adjustment for severity of illness was made. Increases in patient visit—adjusted emergency department use of CT, sonography, and nuclear medicine procedures resulted from changes in practice patterns. For emergency department and outpatient settings, adjusting for the number of patient visits explains a significant portion of the increase in utilization.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Khorasani et al. [1] showed that the number of imaging procedures per patient admission did not increase if utilization was adjusted for the severity of illness. Nonetheless, these researchers found accelerated use of CT and MR imaging; in addition, they suggested that technology substitution was occurring and might be a driver of future health care costs. We undertook this study to understand the utilization rates of radiology services at our hospital and to see whether the trends seen at our institution replicated the findings of previous studies assessing the use of diagnostic radiology services for hospitalized patients [1]. We broadened our study to include outpatient and emergency department radiologic procedures because these services also consume significant hospital resources. Analysis and understanding of utilization trends are important for strategic planning, budgeting (including scarce personnel resources), and assessing and controlling costs and quality of care especially when using CT, MR imaging, nuclear medicine, and other expensive technologies.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Study Design
We performed an observational study of diagnostic imaging utilization
patterns based on administrative data cumulated annually that were segregated
by patient type and imaging procedure. All radiologic studies were identified
and subclassified into imaging categories: nuclear medicine, imaging with
portable scanners, diagnostic (conventional radiography including mammography
and fluoroscopic examinations), sonography, MR imaging, CT, and angiography.
These studies were identified for three separate patient groups: inpatients,
outpatients, and emergency department patients. The inpatient category was
defined as patients undergoing radiologic procedures for all hospital
admissions including short stays. Outpatient examinations were defined as
procedures performed for patients attending the ambulatory clinics or the
emergency department. Emergency department procedures included those for
patients seen in a walk-in urgent clinic.
The total number of admissions was taken from the hospital financial database. To account for possible changes in the population's disease severity over time, we adjusted the number of admissions per year using the Health Care Financing Administration case-mix index [3]. An average case-mix index is based on the case-mix index of all diagnostic imaging—related groups. The "unadjusted" number of admissions, which excludes short-stay admissions, was multiplied by the average case-mix index to produce the number of complexity-adjusted case-mix admissions. The adjusted admission numbers were used for evaluating temporal trends in inpatient imaging use. Trends for emergency department and outpatient services radiologic utilization used patient visits for normalization.
Relative value units (RVUs) were used to evaluate trends in the radiology workload. RVUs have been assigned by the Health Care Financing Administration to each CPT (current procedural terminology) code [4] and are used by health care providers to report their services [3]. RVUs are a composite of physician work, malpractice expense, and practice expense [4]. The practice expense and malpractice components of each radiologic procedure can both be both partitioned into professional and technical components. Only the technical components (facility practice expense + malpractice expense) of the RVU were used for our study because these components represent hospital-based resources. The latest published RVUs (which, at the time of our study, were from 1999 [3]) were applied retrospectively to our data. None of the RVU components was adjusted by Health Care Financing Administration's geographic practice cost indexes. [3]
Statistical Methods
Linear regression, with year as the independent variable, was used to
evaluate whether admissions, both raw and adjusted average case-mix indexes,
and length of stay were significantly associated with time. We attempted to
replicate the method of Khorasani et al.
[1] for inpatients and expand
it to the outpatient and emergency department settings. We have extended the
method of Khorasani et al. using linear regression with RVUs as the dependent
variable. For inpatients, the independent variables were year of admission,
number of inpatient admissions, case-mix index, and complexity-adjusted
case-mix admissions. For outpatients and emergency department patients, the
independent variables included the number of emergency department visits,
outpatient visits, and year of visit. Univariate and bivariate analyses were
performed for descriptive statistics and to detect collinearity among study
variables.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
0.01). The case-mix index increased
significantly from 1.48 in 1993 to 1.57 in 1998 (p = 0.004). The
length of stay, however, decreased 5% over the same period, from 7.74 to 7.36
(p = 0.058). Outpatient visits increased 20%, from 225,099 to 269,261
(p = 0.04). Emergency department visits increased only 4%, from
61,282 to 63,682. The RVUs for all imaging types combined increased 53% for
inpatients, 69% for outpatients, and 85% in the emergency department
(Fig. 2). All these trends were
also significantly associated with time (p < 0.01). However, after
accounting for the increasing complexity-adjusted case-mix admissions, there
was no significant trend in inpatient imaging RVUs over time. Nor, after
adjusting for the number of outpatient visits, was there any association
between time and increasing outpatient RVUs. The emergency department imaging
RVUs, after accounting for the number of emergency department visits,
maintained a significant increase in RVUs over time (p <
0.03).
|
|
In the inpatient group, all types of imaging studies showed a significant increase over time by linear regression (Fig. 3). CT increased 32% (p = 0.004); angiography increased 67% (p = 0.01); diagnostic radiology increased 33% (p < 0.001); MR imaging increased more than twofold, 128% (p = 0.005); portable radiology increased 44% (p = 0.004); nuclear medicine, particularly cardiac studies, increased 54% (p = 0.01); and sonography increased 81% (p = 0.056). After the increasing complexity-adjusted case-mix admissions over time were accounted for, only the increase in MR imaging remained significant (p = 0.04).
|
In the outpatient group, only the diagnostic imaging group did not show a significant increasing trend over time (Fig. 4). CT increased 90% (p = 0.02); angiography increased 10-fold (p = 0.03); MR imaging increased 83% (p = 0.007); nuclear medicine, especially cardiac studies, increased almost twofold, 160% (p = 0.004); and sonography increased 49% (p = 0.001). After adjusting for the number of outpatient visits, the trends remained significant for angiography (p = 0.006), MR imaging (p = 0.002), and sonography (p = 0.04). The portable imaging group was not included in the analysis or graphic representation because of the low number of outpatients.
|
In the emergency department group, no trend toward increasing use was found for portable radiology, angiography, and MR imaging. However, angiography and MR imaging showed trends for increasing use during the early years of 1993-1996 (Fig. 5). CT increased more than twofold (111%, p = 0.03), and diagnostic radiology increased 27% (p = 0.002). Nuclear medicine (cardiac studies in particular) increased more than 100-fold, from 16 to 4560 RVUs (p = 0.03), and sonography increased almost three-fold, 166% (p = 0.003). Sonography, portable imaging, and nuclear medicine are not represented on the graph because of their small numbers of RVUs. In the emergency department, all these trends remained significant after accounting for the number of emergency department visits (CT, p = 0.03; diagnostic radiology, p = 0.01; nuclear medicine, p = 0.04; and sonography, p = 0.003). After accounting for visits, we found that portable imaging was also used more often (p = 0.01).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although our institution is an urban hospital designated as a level I trauma center and has a preponderance of emergency admissions (80% emergent, 20% elective), our general findings parallel those reported for a university-affiliated tertiary referral hospital with predominantly elective hospital admissions [1]. Khorasani et al. [1] found that overall imaging decreased in the inpatient setting, but MR imaging and CT increased. They concluded that newer imaging techniques were replacing older ones. To facilitate comparison, we reproduced the methodology of this prior study. In the inpatient setting, we did not have a decreasing trend, but after adjusting for complexity-adjusted case-mix admissions, only MR imaging showed a significant increasing trend. Our findings suggest that the increasing use of MR imaging in the inpatient setting is not a function of hospital, patient, or payer type.
We have expanded the observations of Khorasani et al. [1] to other hospital service areas and hospital-based outpatients. We found increases in imaging that suggest changing practice patterns in hospital-based outpatient and, in particular, emergency department settings toward the increased use of more sophisticated imaging procedures. We believe the introduction of new procedures (e.g., screening CT of the cervical spine for patients with blunt trauma at high-risk for cervical spine fracture; focused abdominal sonography for trauma patients) and clinical services (acute cardiac evaluation unit performing acute cardiac perfusion single-photon emission computed tomography) account for most of the increased use of CT and nuclear medicine studies in the emergency department.
One important conclusion of our study is that the reproducibility of aggregate assessments for resource use obtained with our method seems independent of many attributes of our health care organization (e.g., geographic location, public versus private, payer mix). This technique may be a powerful tool by which to characterize and track practice patterns for individual hospitals, clinics, and health care systems [5,6,7,8]. Such knowledge is important in the informed management of resource utilization, whether it is used for radiologic or clinical laboratory evaluation.
What drives these increases in use of radiology services? We believe there are several drivers. Perhaps the most important are the systematic advances in treatment and diagnosis and the increasing number of illnesses associated with an aging population.
Technologic advances have led to changes in practice patterns, with an increase in reliance on diagnostic procedures to diagnose, stage, and follow up the therapies of patients. In particular, improvements in imaging technology have led to rapid changes in practice patterns and radiology resource use for the diagnosis of chest pain (nuclear cardiology, helical CT pulmonary arteriography) and abdominal pain (helical CT of the abdomen and pelvis). For example, increases in radiology resource use follow the vital roles that helical CT plays in the diagnosis of many disorders that present emergently, including acute abdominal disease (e.g., appendicitis, diverticulitis, bowel obstruction, ruptured aneurysms of the abdominal aorta, and renal and ureteral calculi) and pulmonary embolus.
In a similar manner, an aging population, managed care, the transition to outpatient treatment for certain diagnoses, and improved transport and resuscitation of severely ill and injured patients seem to have led many facilities, including ours, to see "sicker" inpatients, as reflected by an increase in the case-mix index over the decade. Indeed, when inpatient imaging was adjusted for case-mix index, the apparent trend to increasing imaging over the period of the study was no longer significant. We could not find a similar validated severity adjustment in the outpatient and emergency room settings reported in the literature. Thus, we were unable to predict the potential impact of the changing patient population on use.
Furthermore, imaging is generally used earlier and more frequently in the processes of diagnosis and treatment monitoring than it has been historically [9]. This change may explain the significant increase we found in the use of imaging in all settings, especially in the emergency department. Secular changes, such as societal expectations for rapid and accurate diagnoses, and providers' fear of litigation have also increased the propensity for imaging. As a result, diagnostic tests tend to be complements rather than substitutes for imaging procedures, such as when abdominal CT is performed after positive findings on diagnostic peritoneal lavage in hemodynamically stable patients who have sustained blunt abdominal trauma.
Physicians experience an increase in the confidence of their diagnoses as the number of confirmatory findings from diagnostic tests increase. New clinical programs, often specific to particular institutions, may also be associated with increased use of new studies or existing technology in new ways, such as the establishment of acute cardiac evaluation units and brain attack centers. By shortening the time to definitive diagnosis, the greater use of CT and MR imaging may act to increase system capacity, patient throughput, and facility efficiency. System planning and budgeting must compensate for these trends and look for means of balancing increasing demand with flat or declining revenues.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  G. P. Krestin, J. C. Miller, S. J. Golding, G. G. Frija, G. M. Glazer, H. G. Ringertz, and J. H. Thrall Reinventing Radiology in a Digital and Molecular Age: Summary of Proceedings of the Sixth Biannual Symposium of the International Society for Strategic Studies in Radiology (IS3R), August 25 27, 2005 Radiology, September 1, 2007; 244(3): 633 - 638. [Full Text] [PDF]
|
|
|
|
 |
|
 V. C. Khasgiwala, M. D. Aronson, R. B. Davis, and M. P. Rosen Imaging utilization in the era of the hospitalist. Am. J. Roentgenol., July 1, 2006; 187(1): 2 - 7. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Matin, D. W. Bates, A. Sussman, P. Ros, R. Hanson, and R. Khorasani Inpatient Radiology Utilization: Trends over the Past Decade Am. J. Roentgenol., January 1, 2006; 186(1): 7 - 11. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= CT, 
= angiography,
= diagnostic radiology, 
= MR imaging, x = portable
radiography, 
= nuclear medicine, 
= sonography.
