| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Emergency Radiology |
1 From the Department of Diagnostic Radiology, Yale University School of Medicine, 330 Cedar St, TE-2, New Haven, CT 06510 and Department of Economics, Yale University School of Management, New Haven, Conn (D.D.S., H.P.F.); and Research Department, American College of Radiology, Reston, Va (M.B., J.H.S.). Received February 24, 2003; revision requested April 11; final revision received July 18; accepted August 22. Address correspondence to H.P.F. (e-mail: howard.forman@yale.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
MATERIALS AND METHODS: The authors contacted a random national sample of 114 hospitals by telephone and administered an "ED Radiology Coverage" questionnaire. The questionnaire included queries about daytime image interpretation duties, nighttime radiology coverage arrangements, and radiologist staffing needs. Results were stratified on the basis of ED patient volumes and trauma center designation and were analyzed statistically by using multivariate and logistic regression analyses.
RESULTS: Representatives of 97 EDs responded to the questionnaire. Community hospital radiologists performed daytime primary interpretation of radiographs at 39 (40%) of 97 EDs, computed tomographic (CT) scans at 91 (95%) of 96 EDs, and ultrasonographic images at 87.5 (93%) of 94 EDs. "ED-dedicated" radiologists performed this emergency radiology work in only two (2%) of 97 EDs. During the nighttime, eight (8%) of 97 EDs had no radiology coverage, 80 (82%) of 97 EDs used teleradiology services in some form, and nine (9%) of 97 EDs employed in-house, rotating "non–ED-dedicated" radiologists. Analysis of participant responses revealed that clinicians at 37 (38%) of 97 EDs were able to consult radiologists for nighttime radiography questions, and 87 (92%) of 95 EDs had nighttime CT scans read by radiologists in time for patient care decisions. Twenty-four (25%) of 97 EDs reported radiologist staffing shortages, but only one indicated that it was actively trying to recruit ED-dedicated radiologists. Results of logistic regression analysis indicated that higher ED patient volumes (P = .005) and the presence of a trauma center (P = .02) each significantly increases the probability of higher nighttime levels of radiologist coverage.
CONCLUSION: There is great variation in the current provision of emergency radiology services in private community hospitals.
© RSNA, 2004
Index terms: Emergency radiology • Images, interpretation • Radiology and radiologists • Radiology and radiologists, departmental management
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
In response, many academic centers and large hospitals have shifted away from a system in which emergency physicians or residents perform initial interpretation of images (which are later given a final interpretation [ie, are "over-read"] by general radiologists) toward one in which "ED-dedicated" radiologists perform and interpret the results of all imaging studies day and night. These ED-dedicated radiologists possess emergency radiology experience or fellowship training and center their practices on the radiologic care of injured and acutely ill patients. To facilitate this transition, some hospitals have placed radiologic facilities, and even entire radiology departments, within or adjacent to emergency centers (1). Academic programs have begun to emphasize emergency radiology as a unique discipline, and it is currently taught as a separate section with its own curriculum in approximately 40% of programs (2,3). In addition, several programs now offer fellowships in emergency radiology. The American Society of Emergency Radiology was established in 1988 to foster the growth of emergency radiology, and it now has more than 450 members.
Most studies to date regarding ED radiology coverage have focused solely on academic centers (2–5), and very little has been published about private community hospitals. Community hospitals often face different needs and constraints than do academic centers due to their generally smaller sizes, different patient populations, and financial limitations. The goal of this study was to investigate the methods employed by these private community hospitals for delivering ED image interpretation services.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
To achieve nationwide representation, all medical centers listed third down from the top of every fourth page were selected for inclusion in the sample to be used for the survey. Small hospitals were defined as those with fewer than 100 beds; medium hospitals, as those with between 100 and 250 beds; and large hospitals, as those with more than 250 beds. Excluded from the study were hospitals listed in the AHA Guide as public, academic, or "specialty" and medical centers that did not have EDs. Specialty institutions included children’s; psychiatric; rehabilitation-only; nursing-only; orthopedic; and eye, ear, nose, and throat hospitals. Whenever a chosen hospital was in one of the excluded categories, a substitute hospital was selected by choosing the hospital listed directly below it on the same page of the AHA Guide (ie, the fourth, fifth, or sixth, etc, hospital listed from the top of the given page).
Survey Administration
A telephone call was placed to the ED physician director at each of the 114 hospitals so that the "ED Radiology Coverage" survey illustrated in Figure 1 could be administered. A maximum of five separate attempts were made to speak with an ED physician at each hospital ED. If department staff denied access to the attending emergency physician on all five occasions, then no further effort was made to contact the ED, and that hospital was excluded from the study. For small hospitals whose EDs were staffed by ED physicians on an on-call basis only, the survey was administered to nurse managers instead. Nurse managers were contacted only after every attempt was made to reach an attending ED physician. Representatives of corresponding radiology groups or departments were not contacted for their responses to this questionnaire.
|
Data Collection
The interviewer categorized all participant responses into the corresponding answer choices listed in Figure 1. Although most responses clearly consisted of one of the listed answer choices, ambiguous responses occasionally necessitated that the interviewer ask clarifying follow-up questions not shown in Figure 1. For daytime coverage (assessed in questions 1–3 of the survey), "radiologists" was selected for the answer only when radiologists performed primary interpretation more than half of the time. Half credit (ie, a score of 0.5 instead of 1.0) was given for questions 1–3 if initial interpretation was split evenly between radiologists and emergency physicians.
For nighttime services (question 6), radiology coverage was sometimes reported as varying from night to night (eg, some EDs reported that they had nighttime radiology coverage from an in-house radiologist five nights a week but used teleradiology for CT scans only for the remaining two nights a week). In these cases, the assigned answer choice described the coverage arrangement occurring most often.
In contrast to the questions regarding nighttime services, no distinction was made between on-site or off-site (ie, teleradiology) review of images in questions regarding daytime services. For consistency, "ED-dedicated radiologist" was defined to all respondents as "either an ED fellowship–trained radiologist or a non–ED fellowship–trained radiologist who spends more than 50% of his/her time working exclusively in the ED." "Initial interpretation" was defined as an interpretation that leads to a treatment decision, regardless of documentation.
Data Analysis
All questionnaire results (regarding daytime image interpretation duties, radiologist over-reading of images, ED-dedicated radiologist assignments, nighttime coverage patterns, teleradiology capabilities, and radiologist staffing needs) were tabulated by D.D.S. and stratified on the basis of ED patient volume and trauma center level. Low-volume EDs were defined as those with fewer than 10,000 patient visits per year; medium-volume EDs, as those with between 10,000 and 25,000 visits per year; and high-volume EDs, as those with more than 25,000 visits per year. Trauma center designation was listed as that self-reported by interviewees.
To determine the statistical significance of the survey findings, we used multivariate and logistic regression analyses. All statistical analyses were performed by M.B. with the statistical analysis software SAS version 8.1 (SAS, Cary, NC). Simple logistic regression models were applied to examine the influence of patient volume and trauma level designation on the probability that an ED had a specific characteristic listed in the questionnaire. The models were used to predict the probability of occurrence of each of the ED characteristics according to ED patient volume and trauma level designation.
In addition, ordered logistic regression models (7) were used to analyze nighttime coverage patterns, from the "no nighttime radiologist" coverage level to the "in-house, non–ED-dedicated radiologist" coverage level. An ordered logistic model preserved the rankings across all levels of coverage without imposing any restrictions on how much higher one level was than the other. The model was used to predict the probability of each level of nighttime coverage according to patient volume and trauma level designation. Logistic regressions allowed us to examine the effect of each explanatory variable (patient volume and trauma center designation) while controlling for the effect of the other. For all logistic regression analyses, P values of less than .05 were considered to indicate a statistically significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Overall Findings
Tables 1–4 include the overall results from our ED radiology coverage survey. Of the 97 responding EDs, 26 were categorized as low-volume (with fewer than 10,000 patient visits per year), 30 as medium-volume (with 10,000–25,000 patient visits per year), and 41 as high-volume (with more than 25,000 visits per year) centers. Nineteen EDs were designated as level II trauma centers, and 78 EDs were non–trauma centers. There were no level I trauma centers in the sample.
|
|
|
|
Overall nighttime radiology coverage patterns are summarized in the first column of Table 2 and in Figure 2. Notably, 80 (82%) of 97 EDs reported using nighttime teleradiology in one form or another, and none reported employing an in-house, ED-dedicated radiologist for nighttime work. Thirty-seven (38%) of 97 EDs had the ability to consult a radiologist about radiographs during the nighttime, whereas 87 (92%) of 95 EDs had their nighttime CT scans read by radiologists in time for patient management decisions (Table 1). Two EDs reported that they did not have nighttime access to a CT scanner; hence, their responses were not included in the nighttime CT image interpretation data set (this includes the ED that reported that it did not have a CT scanner at all).
|
Findings Stratified by Patient Volume and Trauma Center Designation
Tables 1 and 2 display survey results and nighttime coverage patterns stratified by ED patient volume (low, medium, or high), while Tables 3 and 4 display these parameters on the basis of trauma center designation (non–trauma center vs level II trauma center). For the variables in Tables 1 and 3, simple logistic regression did not reveal any statistically significant relationships between the survey parameters and the independent variables of ED patient volume and trauma center designation. However, for nighttime CT consultation capability and radiologist over-reading of images first read by an ED physician, there was not enough variability in the dependent variable to determine if relationships existed. That is, almost all EDs had nighttime CT consultation capability, and all had radiologist over-reading of images initially read by the ED physician.
For nighttime radiologist coverage (in terms of categories described in Tables 2 and 4), as ED patient volumes increased from low to medium to high, the proportion of centers reporting "no nighttime radiology coverage" decreased from 23% (six of 26 EDs) to 7% (two of 30 EDs) to 0% (none of 41 EDs), respectively. The most complete coverage—the use of an "in-house, non–ED-dedicated radiologist"—was found most often in high-patient-volume centers (six [15%] of 41 EDs) and less often in medium-patient-volume (two [7%] of 30 EDs) and low-patient-volume (one [4%] of 26 EDs) centers.
The one low-volume ED that reported having "in-house, non–ED-dedicated radiologist" nighttime coverage actually employed a radiologist who took calls at home and came into the hospital as needed to interpret all CT scans and difficult radiographs. In addition, six (32%) of 19 level II trauma centers—versus only three (4%) of 78 non–trauma centers—reported employing in-house radiologists for nighttime coverage, while eight (10%) of 78 non–trauma centers—versus no (0%) level II trauma center—reported having "no nighttime radiology coverage." Results of ordered logistic regression analysis confirmed what is suggested by these results in revealing that higher patient volumes (P = .005) and the presence of a trauma center (P = .02) each significantly increased the probability of higher levels of nighttime coverage when the other variable was controlled for.
Also of note, 19 (73%) of 26 low-volume EDs, 26 (87%) of 30 medium-volume EDs, and 35 (85%) of 41 high-volume EDs reported utilizing teleradiology in some fashion. The percentage of EDs with radiograph transmission capabilities, as defined in survey question 7 and reported in Table 1, varied slightly between the groups, with the highest proportion occurring for high-volume centers (19 [46%] of 41 EDs). Of the 37 total EDs that had the ability to transmit radiographs off site for nighttime interpretation, only one high-volume center (2%) did so for all radiographs, whether there was a question or not.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
First, EDs tend to assign radiograph interpretation duties differently from the way they assign CT and US image interpretation duties, placing extra emphasis in having a radiologist read results of more advanced studies. In our study, radiologists performed primary daytime radiograph interpretation at only 40% of community hospitals, but almost always initially read CT scans (at 95% of hospitals) and US images (at 93% of hospitals). Furthermore, many small hospitals without full-time radiologists on site reported using teleradiology services to ensure that radiologists read advanced studies in time for patient care. Results of published studies of academic centers, in comparison, are varied. A 1992 report (4) indicated that radiologists performed initial daytime interpretation of radiographs at 77% of academic centers, while a more recent report (5) suggested that emergency physicians currently perform this function at 66% of centers. To our knowledge, no academic center data on initial CT and US image interpretation are available.
Second, we found that community hospitals currently utilize a high level of information technology, with more than 80% of all departments and 70% of low-volume departments benefiting from nighttime teleradiology services. Although 38% of hospitals in our study possessed the capability to transmit nighttime radiographs for off-site interpretation, only 34% did so for consultation and only 1% did so for all radiographs. This relatively low rate of teleradiology adoption for radiography, as compared with that for CT imaging, warrants further exploration, although it may be tied to the prevalence of digital radiography and the sometimes poor resolution resulting from secondarily digitizing and transmitting conventionally obtained images.
Third, irrespective of patient volume or trauma center designation, very few EDs in community hospitals (ie, only 2%) had ED-dedicated radiologists perform emergency radiology work. This finding is in stark contrast to findings in academic centers, where a 2000 survey by the American Society of Emergency Radiology revealed that emergency radiologists interpreted results of emergency imaging studies at 51% of responding radiology departments and that, on average, 2.3 full-time-equivalent emergency radiologists were assigned to each academic emergency center (2,3). There may be several reasons for this discrepancy. Many smaller community hospital EDs may not have large enough imaging volumes to justify an ED-dedicated radiologist, may lack the funding needed to pursue additional radiologists, or may not recognize any added value in having an ED-dedicated radiologist instead of a general one. At the same time, many ED fellowship–trained radiologists may prefer to remain in academic centers, which are better structured to make full use of their expertise and services.
Fourth, it appears that staffing needs become more pronounced as ED size increases. Trauma centers and high-patient-volume EDs reported the highest levels of perceived radiology staffing shortages, at around 32% (vs 15%–20% for low-volume and non–trauma EDs). This difference is not great, though. One explanation for this difference may be that smaller hospitals have lower imaging volumes, which can be more easily handled by the covering general radiologist. Larger hospitals can become overwhelmed on nights when multiple patients with complicated conditions arrive simultaneously and only one radiologist is on duty to assist. It is important to realize that this staffing question was not asked of members of radiology groups but rather of representatives of EDs, who are likely to feel shortages only after other on-site radiologists are unable to manage the volume. Therefore, the finding that staffing needs become more pronounced as ED size increases relates more to how the shortage of emergency radiologists (8–10)—as opposed to an increase in radiology group workloads—is affecting patient care.
Study Limitations
There were several limitations to our study that deserve mention. First, with only 97 hospitals responding, the statistical significance of even modest-sized differences is low. Limiting the sample size to 114 was done intentionally to achieve a study size that was manageable but at the same time sufficient for drawing general conclusions in this evaluation. Second, we recognize that a telephone survey, though a generally accepted method, is not the optimal method for obtaining the information required in this study. With this in mind, interviewees were not pressured to give information unwillingly, and only data from respondents who were forthcoming and confident in their answers were included in the results. On the other hand, administering the survey by telephone did allow the interviewer to clarify questions and probe for information in a manner that would not have been possible with a written questionnaire. Third, questions asked in the survey were very general in nature, especially as they pertained to daytime image interpretation, and thus possibly influenced the results obtained. Most affected would be our findings regarding initial interpretation of US images because US images were not differentiated by type (eg, pelvic US images vs abdominal US images vs cardiac US images), and this potentially could have led to some confusion.
Fourth, administration of the questionnaire to emergency physicians may have yielded answers different from those that would have been given if the questionnaire had been administered to radiologists. ED clinicians’ views would be expected to differ, in some circumstances, from the perspectives of their imaging colleagues—the persons responsible for managing imaging scheduling, operational data, and quality control. Radiologists might have been able to provide more accurate and detailed explanations of the services that they offer and of who provides them. Unfortunately, such a study was not possible, since many smaller hospitals do not possess in-house radiology departments. EDs were the units of interest in this study, so questions were asked of emergency service providers. We were more concerned with obtaining a candid assessment of the value of emergency radiology services as perceived from the user’s perspective than with outlining what radiologists themselves believe. Only by understanding clearly how radiologists do or do not fulfill the needs of others can we optimize the services that we provide. Also of note, this study did not address billing and compliance issues for the institutions surveyed, nor did it address the financial motivations for providing various services.
Implications
A discussion of who interprets emergency images is especially important because this concept lies directly at the heart of the issues of standards of care and self referral. Self referral—the practice of nonradiologists ordering, interpreting, and collecting reimbursement for imaging studies for their own patients—has repeatedly been shown to result in higher image misinterpretation rates, higher costs, higher imaging utilization rates, poorer technical quality, and lower patient safety (11,12). Numerous studies have also revealed that interpretations by emergency physicians in particular are less accurate, less sensitive, and less specific than those of radiologists (11,13–16). One report even suggested that emergency physicians’ performances fail to improve with increasing experience (17) and perhaps are more related to limits in imaging training. Therefore, our finding that radiologists primarily interpret radiographs in only 40% of private hospital EDs is somewhat disconcerting. Emergency physicians might point out that very few (between 0.2% and 1.1%) of their image misinterpretations lead to changes in patient care (15,18,19). Nevertheless, it is still vital for both quality patient care and cost control that radiologists, not emergency physicians, control and interpret emergency imaging studies whenever possible. As radiologists, it is our responsibility to inform and formulate the appropriate standard of care regarding imaging services, to provide imaging services in a cost-effective manner, and to be available and willing to perform our duties.
Although we found that radiologists rather than emergency physicians primarily perform emergency CT and US image interpretation, emergency radiograph interpretation is still at issue. This controversy has intensified in the past few years, as the Centers for Medicare and Medicaid Services, or CMS (formerly the Health Care Financing Administration, or HCFA), mandated in 1995 that only one professional fee for interpreting each image would be reimbursed, and priority would be given to the physician who supplied the initial report (20). Emergency physicians, able to cite long imaging turnaround times, lack of radiologist availability during nights and weekends, and minimal discordance in interpretations, can and do perform initial interpretations of radiographs themselves, as evidenced by the results of this study.
Thus, a central issue appears to be that of freeing radiologists to perform emergency services in a timely manner. Unfortunately, due to the current radiologist shortage, it will be very difficult to redirect practicing radiologists to work exclusively in the ED. As Mueller and Yu (1) have pointed out, creating an in-house, separate, dedicated emergency radiology section may require having four to six additional radiologists on staff. ED imaging volumes must be able to support these radiologists, so their salaries do not outweigh the quality and efficiency benefits provided by their presence.
All else being equal, a dedicated emergency radiology service is preferable to general radiologist coverage, which in turn is preferable to emergency physicians reading their own images. A recent report by Yoon et al (21) indicated that even among attending radiologists, the use of subspecialists for ED image interpretation resulted in a 30% discordance rate between the readings of specialty and those of general radiologists. In addition, the quality assurance program described by Yoon et al (21) modestly improved patient care while also providing for education at multiple levels. In the community setting, the key is to make such a preferred situation feasible within the constraints of the hospital. Radiology department reorganization and technologic advances should allow us to achieve these goals, but practicing radiologists need to continue taking an active role in implementing such changes.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, D.D.S., H.P.F., J.H.S.; study concepts, D.D.S., H.P.F.; study design, D.D.S., H.P.F., M.B.; literature research, D.D.S.; data acquisition, D.D.S.; data analysis/interpretation, D.D.S., M.B.; statistical analysis, J.H.S., D.D.S., M.B.; manuscript preparation, D.D.S.; manuscript definition of intellectual content and revision/review, D.D.S., H.P.F., J.H.S.; manuscript editing and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. H. Thrall Teleradiology * Part I. History and Clinical Applications Radiology, June 1, 2007; 243(3): 613 - 617. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Licurse, D. D. Saket, J. H. Sunshine, C. D. Maynard, and H. P. Forman Update on the diagnostic radiology employment market: findings through 2005. Am. J. Roentgenol., September 1, 2006; 187(3): W249 - W254. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. D. Saket, C. C. Nwanze, C. D. Maynard, J. H. Sunshine, and H. P. Forman Update on the Diagnostic Radiologist Employment Market: Findings Through 2004 Am. J. Roentgenol., December 1, 2005; 185(6): 1408 - 1415. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Larson, Y. S. Cypel, H. P. Forman, and J. H. Sunshine A Comprehensive Portrait of Teleradiology in Radiology Practices: Results from the American College of Radiology's 1999 Survey Am. J. Roentgenol., July 1, 2005; 185(1): 24 - 35. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
