Effect of computed tomography of the appendix on treatment of patients and use of hospital resources 

Authors Rao P, Rhea J, Novelline R, Mostafavi A, McCabe C. 
Source New England Journal of Medicine. 338:141-6. January 15, 1998. 
Institution Massachusetts General Hospital, Boston. 
Support General Electric Medical Systems 


The diagnosis of acute appendicitis is difficult. Many cases of undiagnosed appendicitis and of unnecessary appendectomies occur each year. CT scanning, in particular helical CT after Gastrografin instillation into the colon, has been shown to be a very useful tool for the diagnosis of appendicitis. This study was designed to examine the effect of appendiceal CT scanning on overall hospital costs. 



Patients were eligible for the study if they presented directly to the emergency room (or were referred there) with suspected appendicitis and if a surgeon decided to hospitalize them, based on history, physical examination and labwork, for this diagnosis. 


Prior to appendiceal CT, the admitting surgeon estimated the likelihood of appendicitis as definite, probable, equivocal or possible. 

The patients then underwent focused helical appendiceal CT after instillation of Gastrografin saline solution into the colon. 

The scans were interpreted by one of three ER physicians as having a likelihood for appendicitis of: definite, probable, equivocal, probably not or definitely not. They were further classified as: appendicitis or an alternative diagnosis or normal appendix without an alternative diagnosis. 


Effect of CT on patient care

Final outcomes were determined by the results of surgery or by clinical follow-up at least two months later for those patients who did not undergo surgery. 

Treatment plan prior to CT (admission for observation or urgent surgery) was compared to actual treatment (discharge from the ER, admission for observation, treatment for another condition, urgent appendectomy and other surgery). 

Effect of CT on use of hospital resources

A number of determinations and calculations were made to determine the effect of CT on costs: 

  • The total hospital cost of an unnecessary appendectomy (removal of a normal appendix) was determined by retrospective analysis of the hospital's cost database. The savings to be gained by avoiding an unnecessary appendectomy was calculated as the total hospital cost of an unnecessary appendectomy minus the emergency room costs prior to CT scanning. 
  • It was assumed that each patient who would have been admitted for observation but was able to be discharged based on CT scan results would have been hospitalized for just one day of observation. Thus, the savings achieved by CT scanning was calculated as the cost of a day of observation for suspected appendicitis (based on the cost of one day of nursing care and hospital room at the lowest level of illness severity). 
  • The savings achieved by CT was calculated as the number of unnecessary appendectomies avoided times the savings per appendectomy avoided, plus the number of hospital admissions for observation avoided times the savings per admission avoided. From this sum, the cost of 100 appendiceal CT scans was subtracted to determine the overall cost savings from the use of CT scanning. The cost of a CT scan of the pelvis without contrast was used to estimate the cost of an appendiceal CT. 





    100 consecutive patients were included in this study. No patients declined participation. 

    Effect of CT on patient care

    There were 53 surgically confirmed cases of appendicitis. The other 47 patients did not have appendicitis, based on clinical follow-up (41 patients), appendectomy with normal appendix (3 patients) and other surgery (3 patients). 

    The appendiceal CT had a 98% sensitivity, specificity, positive and negative predictive value and overall accuracy. There was one false positive and one false negative CT scan. 

    CT scan was far more sensitive and specific than the pre-CT clinical diagnosis. For example, among 23 patients in whom the clinical diagnosis of appendicitis was rated "definite", there were 5 patients in whom it was ruled out. 

    Management changes based on CT findings

    Clinical management was changed (pre-CT results vs. post-CT results) in 59 patients. There were 63 management changes: 

    • Prevention of an unnecessary appendectomy in 13 patients. 
    • Prevention of an unnecessary admission for observation, prior to a needed appendectomy in 21 patients. 
    • Prevention of an unnecessary admission to the hospital for observation in 18 patients. 
    • Discovery of another condition on CT scan in 11 patients who would have been admitted for observation. 
    Cost analysis

    The cost savings for avoiding one unnecessary appendendectomy was calculated to be $3,637. Since 13 unnecessary appendectomies were avoided, this yields a savings of $47,281. 

    The cost savings for avoiding one day of hospitalization for observation was calculated to be $405. Since a total of 50 days of hospitalization were avoided, this yields a savings of $20,250. 

    The cost of 100 CT scans was judged to be $22,800. 

    The net cost savings was thus $47,281 + $20,250 - $22,800 = $44,731 saved per 100 patients. 

    Author's discussion

    According to the authors, had management been based on clinical evaluation only, 13 patients would have had unnecessary appendectomies and 21 would have had a needed appendectomy after a delay. Integrating the CT scan results into management decisions, three patients had an unnecessary appendectomy and none had a delay. 

    The authors feel that the cost savings they calculated were understated for several reasons. First, they feel that some of the patients who would have been hospitalized for observation would eventually have undergone unneeded appendectomies. Second, they estimated the cost of observation very conservatively. And third, they did not include other benefits from improved diagnosis and treatment such as reduced disability. 

    They note, however, that expanded use of CT in patients in whom the suspicion of appendicitis is lower than those studied here would decrease the savings. 



    In an accompanying editorial, Dr. Ian McColl (Guy's Hospital, London) notes that the diagnosis of appendicitis remains difficult, with frequent false positives and false negatives. Although ultrasound was thought to be a promising technique, appendiceal CT is more sensitive and specific. Ultrasound may still be more useful in children, since it is quicker, simpler and does not require any exposure to radiation. 

    He notes that although CT is a very useful technique, overreliance on technology should not be allowed to diminish clinical skills, which would lead to more and more indiscriminate scanning. The costs related to CT scanning will rise as more and more scans are performed. 



    In this study, the effect of appendiceal CT scanning on overall hospital costs was studied. In patients who were felt to be at high risk of appendicitis, CT scanning not only greatly improved diagnostic precision but also reduced overall hospital resource use, since unnecessary appendectomies and unnecessary delays in surgery were avoided. 

    I have one methodologic criticism of this study. The effect of CT scanning on patient management was determined by comparing surgeons' plans before CT scanning with their plans after the results of the scan were known. However, knowledge that patients were to undergo CT scanning might have influenced the surgeons' pre-scan assessments. 

    For example, a surgeon might be more likely to record "immediate surgery" rather than "observation" as the plan if he or she knew that a CT scan was about to be performed and that the final decision would be made after that. This would increase the number of pre-scan assessments pointing to surgery, and would increase the number of unnecessary appendectomies prevented by CT scanning. We cannot automatically assume that surgeons who knew that their patients were about to undergo a very precise diagnostic test would necessarily make the same decisions for the purpose of the study as they would have in caring for the patient in the absence of CT scanning. 

    A second point, noted by both the authors and the author of the editorial, needs to be emphasized. The cost savings were determined in a population at high risk for appendicitis; only patients who were felt to require either urgent surgery or admission for observation were eligible for this study. If the technique becomes more widespread and available, it will most certainly be applied to a population at much lower risk. In a lower risk population, the cost savings would be substantially lower and could easily turn into a cost increase. If every patient who comes to the emergency room with unexplained nausea and fever or vague abdominal pain is given the benefit of an appendiceal CT scan, no money is likely to be saved. This doesn't mean that the CT scan is less useful in a lower-risk population, only that the cost-savings are likely to disappear. 

    This paper contributes more evidence to the notion that appendiceal CT scanning is a better way to diagnose acute appendicitis. The paper's focus is on cost reduction, however. I do not believe these results, in terms of dollars saved per patient scanned, will be borne out in practice. 

    February 15, 1998 


    References related to this article from the NLM's PubMed database. 

    Reader Comments

    Date: Thu, 05 Mar 1998
    From: sachin dave <sdave@musom.marshall.edu> 

    I feel there is a possibility of observer bias more than what we realize. Surgeons may not have performed a detailed clinical exam, knowing that patients will undergo helical ct with gastrograffin anyways. Surgeons/ER physicians should have been blinded to the fact that patients were undergoing the study investigation. 

    Long term follow-up is not available. We do not know of the cost incurred from a ruptured appendix that should have been removed, but was not because of a negative ct scan (false negative). 

    I think we need a more well designed study. Clinical exam is still next to gold standard (surgical specimen). Comparision should be made between clinical exam and helical ct in a prospective manner. 

      I agree that the fact that surgeons were aware that patients were to undergo CT scanning may have introduced bias into the decision-making process. 

      The problem of false negatives causing increased costs that were not accounted for is a potential one, but patients were followed up 2 months after discharge. This would probably be long enough to catch most late complications. -- mj 

    Date: Tue, 07 Apr 1998
    From: "Dr. Zakko" <szakko@emirates.net.ae> 

    Dear sir 

    Both the article and the comment mentioned the use of CT and US for the diagnosis of appendicitis.I would like to add further that labelled WBC has been used for such application. First TC-WBC has been used. It is highly specific, but it takes at least 90 min. to prepare and further one hour to image. More recently An anti-bodies (fragments) labelled with Tc-99m were used to label the white cells IN-VIVO and image in an hour. 

    It has already been approved by the FDA recently. I think now it would be fair to note that Nuclear Medicine is all the time ignored. It is unfair! or IS it lack of awareness? 

    Dr Saad Zakko
    Dubai Hospital

    July 12, 1998

    Letters to the editor about this article, from the June 18, 1998 New England Journal of Medicine. 

    January 31, 1999

    Date: Mon, 25 Jan 1999
    From: "Patrick J. Dowling M.D." <pdowling@medicine.bsd.uchicago.edu>

    I feel there is another advantage to helical scanning for appendicitis in addition to the possible cost savings put forth in this article.  From the stand point of an emergency medicine resident operating in an academic center, scanning can accelerate the diagnosis in cases in which appendicitis is high on the differential, but the exam and lab findings are equivocal (as they often are). The time involved in having these patients evaluated by multiple levels of surgeons, at various points in their training, not to mention OB-Gyne consults frequently performed at the behest of surgeons in female patients, can lead to hours upon hours in the Emergency Room before a disposition is achieved.  In my experience, surgery almost inevitably requests a CT rather than taking the patient straight to the OR anyway, so we often order a CT in equivocal cases and call surgery based on the results.

    I believe this will become more standard in the future, and perhaps in 5-10 years appendicitis will be, if not an emergency medicine disease, a diagnosis made or ruled out reliably by emergency physicians rather than surgeons.

    Patrick J. Dowling M.D.
    University of Chicago Hospitals
    Section of Emergency Medicine

    Date: April 8, 1999
    From: Cranmax@aol.com

    Don't forget, the first walkin centers were started in chicago, just across the street from the ER, by combining time/cost considerations, without compromising patient care. Sounds like the time has come for CT to be the dx method of choice for appendectomy.

    Paul Appleton, MD

    Submit a comment

    Journal Club homepage

    Site Meter