| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University of Washington, Harborview Medical Center,
325 Ninth Ave., Box 359728, Seattle, WA 98104-2499.
2 Department of Radiology, Valley Medical Center, Renton, WA.
3 Department of Radiology, Virginia Mason Medical Center, Seattle, WA.
Received November 26, 2007;
accepted after revision May 6, 2008.
Address correspondence to R. O. Nathan
(ronathan{at}u.washington.edu).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. For each of 100 consecutive adult patients who presented to a community hospital emergency department from August 2006 to November 2006 and underwent CT of the appendix, the proposed treatment plan and the likelihood of appendicitis were recorded before CT and were compared with the actual treatment after CT. The primary outcome assessed was change in patient management after CT. The percentage likelihood of appendicitis, whether patient disposition changed after CT, and the presence or absence of appendicitis were examined. The accuracy of CT was also calculated.
RESULTS. The treatment plans of the emergency clinicians changed in 29 patients (29%). Appendicitis was ruled out on the basis of CT findings in 50% (9/18) of patients when appendicitis was considered probable and in 60% (3/5) when appendicitis was considered very likely. When appendicitis was considered unlikely, appendicitis was ruled out by CT in 100% (20/20) of patients. CT of the appendix was shown to have high sensitivity (94%), specificity (100%), positive predictive value (100%), negative predictive value (99%), and accuracy (99%).
CONCLUSION. CT of the appendix had an important therapeutic impact on patients presenting to a community hospital emergency department. The data suggest that CT can be withheld in patients in whom emergency clinicians rate the likelihood of appendicitis as unlikely but that CT findings are often of benefit even when appendicitis is judged to be very likely.
Keywords: abdominal imaging • appendicitis • CT • emergency radiology
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although a diagnostic imaging test must be accurate for it to be used effectively in a clinical situation, accuracy alone does not ensure that the test has value. CT of the appendix would be of little use despite its accuracy if the results rarely influenced physician diagnostic certainty or subsequent therapy [14]. An imaging test that has therapeutic impact influences the treatment choice after the results of the test are known [14]. Such a test is likely to improve patient outcome. Thus, high diagnostic accuracy, as seen with CT, is not sufficient to prove the value of an imaging study in patient care. Despite the large number of studies of diagnostic accuracy of CT of the appendix [4, 7, 11, 15–17], few studies have examined the therapeutic impact of CT of the appendix—that is, how the results of the test influence the subsequent treatment of patients.
The purpose of this study was to assess the impact of CT of the appendix on the management of patients in an emergency department. Further, to increase the generalizability of the results, the study was undertaken in a community hospital rather than an academic setting.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Between August 2006 and November 2006, we performed a prospective study of consecutive adult patients with clinical suspicion of appendicitis who presented to the emergency department of a 300-bed community hospital serving a suburban population. In this hospital, CT of the appendix is ordered by emergency clinicians routinely for all patients with suspected appendicitis. The study emergency department, with approximately 80,000 visits per year, is staffed by 13 board-certified emergency physicians and six certified physician assistants, all of whom participated in the study. The CT scans were interpreted by members of a private practice group of 26 board-certified radiologists, 23 of whom participated in the study.
Patients eligible for the study were adult men and women evaluated in the emergency department who underwent CT of the appendix because of signs and symptoms considered suggestive of acute appendicitis by emergency clinicians. Patients younger than 18 years old, pregnant women, and patients in whom IV contrast material was contraindicated were excluded from the study. None of the patients seen in the emergency department during the study period underwent an appendectomy without undergoing CT of the appendix first.
Data Collection
The emergency clinicians who requested CT completed a questionnaire before
CT was performed in which they were asked to choose the most likely treatment
plan of patients based on the clinical and laboratory findings. The choices
were surgical consultation for appendectomy, admission for observation, or
discharge from the hospital. They were also asked to rate the likelihood of
appendicitis on a continuous analog scale from 0% to 100%. Suggested ranges
were unlikely, 0–20%; possible, 21–40%; equivocal, 41–60%;
probable, 61–80%; and very likely, 81–100%. The category
"unlikely" was included because in our hospital surgical
consultations do not occur before CT scans are obtained. The treatments chosen
after CT were determined by reviewing medical records no sooner than 2 weeks
after the emergency department encounter.
Reference Standard
The presence or absence of acute appendicitis in patients who underwent
appendectomy was assessed by reviewing surgical pathology reports. In patients
who did not undergo appendectomy, the absence of appendicitis was assumed if
no interval appendectomy was performed at the study hospital at least 2 weeks
after the emergency department encounter. A CT finding was determined to be an
alternative diagnosis if the emergency clinician considered the finding to be
a plausible explanation of the patient's pain as indicated in the medical
record or if surgical pathology results confirmed an alternative
diagnosis.
Performance and Interpretation of CT Examinations
All patients underwent focused helical CT on a 16-MDCT scanner (LightSpeed
Pro 16, GE Healthcare) from the iliac crests to the pubic symphysis using a
1.25-mm collimation with a reconstruction slice thickness of 2.5 mm, 120 kVp,
and automatic milliamperage (mA) settings. Coronal images were constructed
from the axial data set using a 2.5-mm slice thickness. IV contrast material
(100–130 mL of iopamidol [Isovue 350, Bristol-Myers Squibb]) was
administered at a rate of 3 mL/s using the automated bolus-timing feature.
Oral contrast material was not administered.
The radiologists were informed of the study. They interpreted the CT scans independently and did not have access to the emergency clinician questionnaires before interpretation. They were given no specific criteria for diagnosis or exclusion of appendicitis. The CT results were coded by one investigator from the radiology reports in the medical record. The coder did not perform CT interpretations and was blinded to the emergency clinician questionnaires. Reports were coded as appendicitis, a normal appendix without an alternative diagnosis, equivocal findings for appendicitis, or a normal appendix with an alternative diagnosis. A report was coded as equivocal if a definitive statement about the presence or absence of appendicitis was not given. All studies were interpreted on a commercially available PACS (DR PACS version 6, DR Systems).
Data Analysis
The therapeutic impact of CT of the appendix was analyzed by comparing the
treatment plans of the emergency clinicians proposed before CT, as determined
from reviewing the questionnaires, with the treatment that patients underwent
after CT, as determined from reviewing the medical record. The percentage of
patients in whom the proposed treatment plan changed was calculated. The
percentage likelihood of appendicitis was assessed in relation to the change
in treatment plan and the presence or absence of appendicitis. The accuracy,
sensitivity, and specificity of CT were also tabulated using standard 2
x 2 tables.
Equivocal CT results were included in the test results in accordance with the Standards for Reporting of Diagnostic Accuracy (STARD) Initiative [18]. We decided a priori to code equivocal CT results as a normal appendix. From our prior experience, we felt that nonvisualization of the appendix would be the cause of most equivocal results. The results of a study by Nikolaidis et al. [19] suggest that the incidence of appendicitis is very low in the absence of appen diceal visualization and second ary inflammatory changes.
The data obtained were analyzed using statistics software (STATA/MP 10.0, Stata Corporation). CIs for sensitivity, specificity, and accuracy were obtained using CI for proportion calculations.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Clinical Disease State
Of the 19 patients who underwent appendectomy, appendicitis was confirmed
by surgical pathology results in 17. The remaining two appendectomy patients
had no evidence of appendicitis on surgical pathology. Eighty-one patients did
not undergo appendectomy at the study hospital at least 2 weeks after the
emergency department encounter. These patients were considered to be negative
for appendicitis: Four were discharged after observation in the hospital and
77 were discharged from the emergency department.
CT Results
The CT results are shown in Table
2. Of the 17 patients in whom CT showed appendicitis, appendectomy
was performed and surgical pathology confirmed appendicitis in all 17. There
were two perforations.
|
CT scans were reported to show a normal appendix without an alternative diagnosis in 64 patients. In 63 patients, appendectomy was not performed at our institution for at least 2 weeks after the emergency department encounter. One patient, in whom appendectomy was performed after a normal CT report, had a normal appendix on surgical pathology. CT scans were reported to show a normal appendix with an alternative diagnosis in 15 patients: ovarian cyst (n = 8), inflammatory or infectious bowel disease (n = 3), perforated sigmoid diverticulitis (n = 1), right ureteral stone (n = 1), mesenteric adenitis (n = 1), and epiploic appendagitis (n = 1).
Four patients, three women and one man, had CT findings coded as equivocal for appendicitis. In three patients, visualization of the appendix was poor or the appendix could not be visualized. In one patient the appendix was mildly enlarged with no periappendiceal inflammation. Appendectomy was performed in two patients with equivocal findings on CT. One patient had appendicitis based on surgical pathology and one had a normal appendix. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CT of the appendix were 94%, 100%, 100%, 99%, and 99%, respectively.
Therapeutic Impact
Surgical consultation for appendectomy was the treatment plan before CT in
38 patients, observation in five patients, and discharge in 57 patients
(Table 3). The initial
treatment plans changed in 29 patients (29%) after CT. The changes occurred in
21 of 75 women (28%) and eight of 25 men (32%).
|
Of the 38 patients with the initial designation of surgical consultation for appendectomy, the treatment plan after CT changed in 22 patients. There were 18 changes from surgical consultation for appendectomy to discharge and four changes from surgical consultation to observation. Of the five patients with the initial treatment plan of observation, the treatment plan after CT changed in all five. There were four changes from observation to discharge and one change from observation to surgical consultation for appendectomy. The patient whose treatment was changed to surgical consultation had acute appendicitis confirmed on surgical pathology. Of the 57 patients with the initial disposition of discharge, the final treatment plan changed in two patients. Both of the changes were from discharge to surgical consultation and both patients had acute appendicitis confirmed on surgical pathology. CT ruled out appendicitis in 24 of 38 patients initially designated for surgical consultation for appendicitis and four of five patients initially designated for observation. CT correctly identified appendicitis in two patients initially designated for discharge and one patient initially designated for observation.
All five patients in whom appendicitis was rated very likely (81–100%) had an initial treatment plan of surgical consult for appendectomy (Table 4). Appendicitis was confirm ed surgically in only two of those patients (40%), with both cases correctly identified by CT. Of the 75 patients in whom the likelihood of appendicitis was rated possible, equivocal, or probable (21–80%), surgical consultation for appendectomy was the initial treatment plan in 33, observation in five, and discharge in 37. Appendicitis was confirmed surgically in 15 patients (20%): CT correctly identified 14 patients and yielded equivocal findings in the remaining patient. In the 80 patients in whom the likelihood of appendicitis was 21–100%, there were 15 alternative diagnoses on CT in addition to the 17 cases of acute appendicitis. The initial treatment plan was discharge in all 20 patients in whom appendicitis was rated unlikely (0–20%). No patient in this group had appendicitis on clinical follow-up and all had negative findings on CT. Of the patients with a rating of unlikely for appendicitis, CT suggested an alternative diagnosis (a 6-cm ovarian cyst) in one patient.
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The results of our study suggest that CT of the appendix has an important therapeutic impact in the evaluation of suspected appendicitis in a community hospital emergency department. The initial treatment plans of the emergency clinicians changed in 29 patients (29%) after the results of the CT were known. CT had the greatest therapeutic impact in patients initially designated for surgical consultation for appendectomy or for observation. In those two categories combined, the final disposition changed in 27 of 43 patients (63%). These changes resulted in 22 patients being discharged from the hospital, none of whom returned to our hospital for interval appendectomy. On the other hand, of the 57 patients in our study with the initial designation of discharge, the final disposition changed in only two patients after the results of the CT were known.
Two appendectomies yielded negative surgical pathology findings in our series. One patient underwent appendectomy despite normal CT findings. In the second patient, with equivocal CT findings, the surgeon elected to operate on clinical grounds. As with previous reports [1], the sensitivity (94%), specificity (100%), positive predictive value (100%), negative predictive value (99%), and accuracy (99%) of CT of the appendix were high in our study.
In several previous studies, investigators have assessed the therapeutic impact of CT of the appendix. In a prospective study by Rao et al. [4] performed at an academic medical center, CT of the appendix with rectal contrast material was performed on 100 consecutive adult and pediatric patients with a clinical suspicion of appendicitis. CT resulted in 63 treatment changes. Thirteen patients avoided unnecessary appendectomy. Fifty patients avoided observation in the hospital: 21 before necessary appendectomy, 11 before treatment for alternative conditions diagnosed on CT, and 18 before discharge from the hospital.
Additional studies of the therapeutic impact of CT of appendicitis have shown that CT altered the initial treatment plans of surgeons in 25–35% of patients [7, 11, 15–17]. One study showed that CT altered the initial treatment plans of emergency physicians in 35% of patients [15]. In one study in which diagnostic as well at therapeutic impact was assessed, diagnostic certainty was increased by 22% as a result of CT [16]. CT was particularly useful in lowering the potential negative appendectomy rate in females in two studies [7, 11].
If CT of the appendix does significantly influence patient treatment, it should be used either routinely or selectively in patients with clinical suspicion of appendicitis. Several researchers have concluded that CT should be performed routinely in all patients suspected of having appendicitis [2–4]. However, CT entails cost; radiation exposure to the patient; and, if IV contrast material is used, risk of contrast reaction or contrast-induced nephropathy. For these reasons, selective scanning is desirable if clinically appropriate. Various selective strategies based on patient sex, signs and symptoms, and clinician-assessed likelihood of appendicitis have been advocated [5–11, 22].
Some researchers have advocated that CT of the appendix not be ordered before surgical consultation [8, 9] or that CT is not needed if the surgeon thinks that appendectomy is indicated [11, 19]. These studies were performed in either academic medical centers or community-based teaching hospitals in which surgical house staff members were readily available. Surgeons are less available in most community hospitals after hours. A study by Sarkaria et al. [22] in which more than 500 general surgeons were surveyed showed that most CT scans are ordered by emergency clinicians without prior surgical consultation.
Our study was designed to reflect community practice. In contrast to other studies of therapeutic impact in which the settings were academic medical centers [4, 7, 11], community-based teaching hospitals [16, 17], or a military medical center [15], our study was performed in a community hospital without emergency medicine or surgery residency programs. We evaluated how the CT results altered the initial treatment plans of emergency clinicians rather than surgeons because emergency clinicians are more likely to initiate CT studies in community hospital settings. Both emergency physicians and physician's assistants were included in the study. No criteria were given to the emergency clinicians to rate the likelihood of appendicitis, and radiologists were not given criteria for the CT diagnosis of appendicitis.
A possible limitation of our study is that the likelihood of appendicitis in some patients may have changed as their clinical courses evolved after the emergency clinicians completed the questionnaires. The study is also limited because we tracked patient follow-up only in the study hospital. It is possible that study patients could have undergone appendectomies at other institutions.
In summary, we found that CT of the appendix had not only a high degree of accuracy, but also an important therapeutic impact on patients presenting to a community hospital emergency department with suspicion of appendicitis. The impact was greatest in patients designated initially for surgical consultation for appendectomy or for observation. Appendicitis was ruled out by CT in 12 of 23 patients in whom the clinical likelihood was considered probable or very likely. CT had little impact on those patients initially designated for discharge from the hospital. In patients in whom appendicitis was unlikely, appendicitis was ruled out in 20 of 20 patients.
The data suggest that CT should be performed even in patients in whom emergency clinicians consider appendicitis to be probable or very likely but that CT could be withheld in patients in whom emergency clinicians rate appendicitis as unlikely. If that strategy had been followed in our patient population, 20% of the CT examinations would have been eliminated with no false-negatives. Larger studies are needed to confirm these findings.
Some researchers have maintained that negative appendectomy rates have not improved despite the widespread use of CT of the appendix and that this lack of improvement may be explained by the lower diagnostic accuracy of CT in general radiology practice than that reported from research centers [9, 12]. The diagnostic accuracy in our study, as well as those reported in other studies performed at community hospitals [23, 24], compares favorably with the diagnostic accuracy reported in studies performed at academic medical centers. The 99% negative predictive value achieved in this study suggests that emergency clinicians can have a high degree of confidence in a negative CT result, which should result in lower rates of negative appendectomy results.
Future prospective studies could evaluate whether these conclusions apply in other similar community hospital settings as well as in settings in which surgeons play a larger role in the decision to order CT. Development of clinical prediction rules to identify patients who are unlikely to have appendicitis would also be of value. Although we found CT of the appendix to be effective, we did not evaluate the cost of the examination, including the potential effects of radiation exposure.
Prior studies have shown that performance of appendiceal CT results in cost savings by reducing negative appendectomy rates and preventing unnecessary hospital admission for observation [4, 25]. Health effect metrics, however, were not included in these analyses. A full cost-effectiveness analysis of CT of the appendix, the sixth level in the hierarchical model of Fryback and Thornbury [20], would be an important addition to the health services literature.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  T. K. Chaudhuri and S. Fink Therapeutic Impact of CT of the Appendix in a Community Hospital Am. J. Roentgenol., September 1, 2009; 193(3): W255 - W255. [Full Text] [PDF]
|
|
|
|
|
|
R. O. Nathan, C. C. Blackmore, and J. G. Jarvik Reply Am. J. Roentgenol., September 1, 2009; 193(3): W256 - W256. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
