| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary |
Department of Radiology, Central Middlesex Hospital, Northwest London Hospitals NHS Trust, Acton Lane, London NW10 7NS, UK
My interest in the appendix was kindled early in my career, and the incident responsible probably put paid to any pretensions I might have had about being a surgeon. My consultant decided it was time for me, a young house surgeon, to perform my first operation. An otherwise healthy 18-year-old woman was admitted with classic symptoms and signs of appendicitis. At surgery I was confronted by totally unfamiliar anatomy. After poking around for a while, and becoming very worried about what I had done, my consultant, with a sigh, took over. About an hour later, with the original incision considerably enlarged, he had laid open a large infected retroperitoneal cyst that extended beyond our view, leaving the normal appendix alone. The patient became septicaemic and was later transferred to a specialist unit. We never found out the aetiology of this abnormality.
This incident occurred before the advent of CT and MRI, and when ultrasound was in its infancy. Nowadays both the presence and origin of the cyst would have been identified pre-operatively by elegant cross-sectional imaging. If appropriate, percutaneous drainage would have been performed and the patient spared a disfiguring scar. But would, and should, that happen? Imaging of suspected appendicitis evokes polarized views. On the one hand there is the experienced surgeon who not only believes imaging causes unnecessary delay but who also accepts a negative appendectomy rate. Conversely there are those who are more cautious and believe that avoiding unnecessary laparotomies or laparoscopies is paramount. This issue is not trivial as it has considerable resource implications; if imaging of suspected appendicitis becomes the norm, this will be an even more common demand on out-of-hours imaging.
A number of questions need to be asked; how accurate is imaging at detecting appendicitis, what technique should be used and, most importantly, what is the risk:benefit balance for imaging?
There is now a wealth of accuracy data in the literature [1–9]. In the USA there is a strong consensus that CT is the best approach, with most arguments revolving around whether a complete abdominal and pelvic examination or just a "focused scan" should be performed, and whether oral, rectal or intravenous (iv) contrast should be given alone or in combination. In the UK and Europe there is greater emphasis on ultrasound.
The fundamental issues concerning all imaging include whether there is a sufficiently low false negative rate to justify avoiding an operation, whether the inevitable delay caused by imaging adds to morbidity, and whether, for techniques involving radiation, the long-term radiation penalty is less than the potential risk of a missed diagnosis or unnecessary laparotomy or laparoscopy. There are special factors in paediatric patients with suspected appendicitis that need to be considered separately. A great deal of literature concerns patients outside the paediatric age group and caution needs to be taken in using these data for deciding management protocols in children.
CT using modern helical scanners with 5 mm collimation and workstation review has an overall accuracy of 93–99% in diagnosing appendicitis [2, 3, 6–9]. Identification of a normal appendix is important in positively ruling out appendicitis, and this is achieved in most patients using CT (95–100%). In this issue of BJR, Ege et al [10] report a diagnostic sensitivity of 96%, specificity of 98% and false negative rate of 2% in 296 patients with suspected appendicitis, using no oral, iv or rectal contrast media. Accuracy may be lower in children, presumably owing to their size and paucity of intra-abdominal fat in comparison with adults. False negative rates are generally between 1% and 3% for CT and probably higher for ultrasound. The success of ultrasound in detecting a normal appendix is normally reported at a much lower rate than CT [11–15]. If one assumes that all patients with possible appendicitis will go to theatre unless imaging rules out the diagnosis, (and this is a big assumption as there is no doubt that the threshold for investigation drops when that investigation becomes easily available), then false positive rates are less important than false negative rates. On the whole, false positives rates are low for both ultrasound and CT, and where incorrectly diagnosed often indicate other acute inflammatory conditions that are mimicking appendicitis.
Impressive though CT results are, they are not without cost. A timely edition of Pediatric Radiology reports on a conference on the "as low as reasonably achievable" (ALARA) concept in paediatric CT intelligent dose reduction [16]. There is much information on ways to minimize radiation dose when dealing with children, but the most effective method is not to scan the child in the first place. There is no doubt that predominant practices in the USA and the UK (and, I suspect, Europe as well) [17] have been at two ends of the spectrum. In the USA radiation issues to children from CT were not considered a high priority until recently, and particularly following articles in AJR that received national coverage [18], whereas in the UK there has always been a reluctance to expose children to CT, especially for benign conditions. The risk:benefit issues surrounding radiation are now at the top of the agenda. Children are considered more sensitive than adults to the effects of radiation, probably by a factor of 10–15, and girls more so than boys. In the past, cancer risks from radiation were always extrapolated from studies of populations exposed to large doses and studied over a length of time, or from animal studies. The chilling fact is that we now have population cancer risk data from non-medical exposure at dose levels comparable with spiral CT, i.e. we no longer have to extrapolate. So, not only are children who undergo CT being given radiation doses shown to increase cancer risk, but this is occurring more often. In the USA, approximately 4% of all CT procedures pre-1993 were on children. In 1993 this had risen to 6% and is now thought to be approximately 10% [16]. With the introduction of multislice CT there is a danger that the dose for each scan will also increase. The figure that created the stir in the USA was that 500 eventual deaths from cancer were likely to have been produced by CT in the paediatric population each year; this figure has even been put as high as 1000 eventual deaths produced per year [16]. The percentage of these CT procedures performed to rule out appendicitis is not known, but it is undoubtedly increasing. In the "AuntMinnie" general radiology discussion group on the Internet, there was a recent contribution from a radiologist who covered 14 hospitals by teleradiology, reporting approximately 100 CT scans each night; 20% of these CTs were for ruling out appendicitis. Another contributor stated a rate of up to 3–4 appendicitis scans each hour [19].
Whilst the risk of solid tumour induction takes many years before an effect can be seen, the risk for leukaemia may be apparent at only 2–5 years but it does persist. With widespread use of CT, this increased risk may become detectable, with the numbers of some induced cancers in young people no longer swamped by those occurring naturally. Studies of radiation-exposed children compared with control populations need to be undertaken.
The dilemma is always that any radiation risk is relatively small in the individual and the perceived risk is often outweighed by the benefit of a diagnosis in the context of an acutely unwell child and their anxious parents. It is difficult to persuade all parents that an operation with the possibility of a "negative" appendectomy may be the safest option. In the past appendicitis was anything but a benign condition, with a mortality of 26% in the early part of the last century. With modern management mortality is reduced to virtually zero. With mortality rates so low there must be serious doubt whether radiation techniques with a measurable risk of cancer induction are ever justified except when complications have set in. There is no evidence in surveys of populations that current use of CT has improved morbidity, and mortality is too low to be measurably affected. Prospective studies [20] have shown that CT can reduce the negative appendectomy rate, but retrospective surveys [21, 22] have shown no change in these rates despite increasing use of imaging.
So, whilst it may be argued that it is unjustified to use CT, why not perform ultrasound? The main problem is that even the best ultrasound examination will have a significant false negative rate, and the examination cannot be used to rule out appendicitis unless (unusually) a normal appendix is seen in its entirety. Although some early studies did put the accuracy of ultrasound at 90% and above [11–14], more recent comparisons [23, 24] of CT and ultrasound in patients undergoing both investigations have shown that ultrasound has a significantly lower sensitivity, particularly in children over 10 years of age. Ultrasound in children can identify an alternative aetiology and may therefore be useful in directing alternative therapy, a different operative incision or even a different surgeon if a gynaecological problem is discovered. However, if appendicitis is a serious consideration a negative scan should not deny the patient surgery. The rate of appendiceal perforation is directly related to the time it takes to get the patient to theatre, and ultrasound has been implicated in contributing to this. Similarly, a negative scan should not be used for procrastination.
The truly equivocal patient, in whom signs, symptoms and laboratory results are against the diagnosis of appendicitis and yet doubt remains, is often referred for imaging. This group needs to be studied separately, as the incidence of appendicitis is likely to be lower than in patients with classical presentation, yet the imaging may be more pivotal, rightly or wrongly, on the decision to operate. It is difficult to tease out from the literature in which studies these truly equivocal cases have been recruited, although in one small study [11] ultrasound was recorded as having a sensitivity of 90% and a specificity of 100%. Radiolabelled white cell imaging has been shown to be effective in this situation [25], virtually ruling out intra-abdominal inflammation and having a negative predictive value of 98%, but practicalities mitigate against this approach.
The argument in adults is very different. In those of middle age and older, investigation is generally undertaken in the context of acute abdominal pain for which the differential diagnosis is much wider and appendicitis is less likely. Even if appendicitis is the diagnosis, complications benefiting from pre-operative imaging are more likely in the elderly. The radiation effects of CT are less relevant and the diagnostic accuracy of CT is very high. For patients between the paediatric age group and middle age the rationale is less clear. Acute right iliac fossa pain in young females does require ultrasound in view of the high prevalence of gynaecological causes [26], and the use of CT should be considered on individual merit.
Where does this leave us? Acute appendicitis in children is a unique entity; it is common, often has typical clinical findings, is cured by surgery, has an extremely small mortality and needs to be treated quickly to avoid complications. It is my belief that in the paediatric patient, clinical findings typical of appendicitis need no imaging investigation and the patient should go straight to theatre. An experienced surgeon will be correct 80–90% of the time [27, 28], but patients and their relatives should accept that there is a small chance that no abnormality will be discovered and a normal appendix will be removed. If clinical findings are atypical, ultrasound is indicated on the understanding that this investigation cannot be unduly delayed. If complications such as perforation or abscess formation are suspected, then imaging is indicated, as this may direct non-operative treatment. In this situation CT may be helpful if ultrasound is inconclusive. In older age groups ultrasound should be used in cases of suspected gynaecological disease. In other situations common sense should direct which is the appropriate investigation; a thin patient with right iliac fossa pain should have ultrasound first whereas an obese patient, or elderly patient with risk of compromised bowel, should undergo CT.
In 2002, what about the patient I described at the beginning of this article? She should probably have undergone ultrasound, likely followed by CT, but I think it would have been difficult to criticise if she had gone straight to theatre. At least now a "green" house surgeon would never have been asked to operate.
Received for publication July 2, 2002. Accepted for publication July 3, 2002.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
