Martin Necas, specialist sonographer, Waikato Hospital, Hamilton, New Zealand.
The diagnosis of acute appendicitis continues to be illusive. Clinical presentation, bloodwork and scoring systems are often non-specific or overlap with other pathologies, especially in women. Ultrasound is commonly employed as the first-line imaging test, however, ultrasound is subject to numerous limitations that ultrasound practitioners and clinicians must be aware of in order to make balanced clinical decisions.
Firstly, the sensitivity of ultrasound for the detection of acute appendicitis in the general population of emergency patients is disappointingly low - in the region of 50-60% even when performed by experienced sonographers and radiologists.1 A recent unpublished re-audit at Waikato Hospital showed that our sensitivity has improved over the years, but only from 50% to 59%. Some papers report much higher sensitivities (upto 99%), but this is largely due to patient selection biases and statistical acrobatics such as selective inclusion of only those patients where the appendix was positively visualised.2 The problem is that the appendix is often not visualised or the imaging is indeterminate.3
Figure 1: High resolution image of a normal appendix surrounded by unrelated free fluid
Perhaps the saving grace is that the negative predictive value of ultrasound is relatively high. That is, if absolutely nothing is found, the likelihood of acute appendicitis is quite low.
What are some of the barriers in the ultrasound diagnosis of appendicitis and how could they be overcome? There are several considerations:
Case pre-selection
Ultrasound user
Machine
Patient
Technique
Let’s have a look at each of these individually.
Case pre-selection
In our experience, only 14%-16% of patients referred with a clinical suspicion of appendicitis actually had surgically and histological proven appendicitis in the end. This indicates to us that referring doctors use ultrasound more as a triage tool rather than a targeted tool for the diagnosis of appendicitis. It is well known that the sensitivity of ultrasound increases if the pre-test probability of the condition increases. Therefore, careful pre-selection of patients can be helpful. It should be acknowledged that ultrasound is a powerful triage tool in patients with RIF pain. We found that alternate diagnoses were 3 times more prevalent than appendicitis itself, and included predominantly gynaecological abnormalities (ovarian cyst asccident, hydrosalpinx, mesenteric adenitis, terminal ileitis endometriosis, adnomyosis and others).
Figure 2: Imaging appearances of a wide variety of alternate diagnoses to appendicitis
Ultrasound user
When it comes to ultrasound, the user is often one of the weakest links. This is because the user is responsible for simultaneously acquiring as well analysing the images. And this requires a complex interaction of visuomotor and cognitive skills. The level of experience with ultrasound (and specifically bowel imaging), number of positive cases identified, and variety of appearances encountered makes a major difference when it comes to one’s ability to correctly identify the next case of appendicitis. Training ultrasound users is complex and relies on hands-on training by experts and on access to large image libraries to hone in the recognition of imaging patterns including correct identification of inflammatory fat (Figure 3), bowel wall signature, appendix base, appendix tip, appendicolith, distinguishing faecally loaded from inflamed appendix, identifying free fluid, reactive nodes, unruptured (Figure 4) versus ruptured appendix (Figure 5), appendiceal abscess and surrounding inflammatory changes affecting nearby structures, tortuous appendix (Figure 6), pelvic appendix, and correctly distinguishing appendicitis from ileitis (infective or IBD). Finally, in order to progress, users should regularly audit their work to learn from their successes and failures.
Figure 3: Echogenic inflammatory fat with appendix target in the centre as imaged on initial survey with a low-frequency curvilinear transducer
Figure 4: High frequency transducer images demonstrate inflamed intact appendix in exquisite detail including all layers of the wall: mucosa, muscularis mucosa, submucosa, muscularis, serosa and surrounding inflammatory fat.
Figure 5: Ruptured appendix with a surrounding appendiceal abscess
Figure 6: Tortuous appendix with ischaemic changes in the superficial portion as evidenced by reduced wall perfusion raising concern of impending rupture
Machine
The quality of the ultrasound machine makes a major difference to the image quality and the resultant diagnostic confidence of the user. Cheap, small, portable hand-held devices are a poor choice as their image quality often resembles the output of high-end ultrasound machines some 20 years ago. The good news is that the prices of ultrasound machines have steadily reduced over the years. I suggest that purchasing a refurbished high-end system is far better than a new low-end system. A pocket-size gadget may look sexy, but is unlikely to be of good use for this type of work.
Figure 7: Improvements in image quality of high-end systems through the ages
Patient
Patient acoustic characteristics are a major determinant of success or failure of ultrasound. And it’s not about size alone. Some large patients with soft “squishy” fat image beautifully, whilst some slim patients with dense fat image terribly. You won’t know until you try. If the patient is acoustically challenging, there is often little you can do, apart from paying greater attention to system setting and imaging the patient from a variety of approaches (see the section technique). Sometimes, small adjustments of multiple system controls significantly improve image quality. Things to try include: lower frequency, lower dynamic range, lower line density, trapezoidal versus rectangular field of view, tissue harmonic imaging (on), spatial compounding (on), greyscale maps (flick through the options to see what works best) and other speckle reduction algorithms.
Technique
The search for the appendix should be purposeful, not accidental. Sometimes, one gets lucky, puts the transducer where the patient is sore and viola – there is a beautiful appendix. The vast majority of time, the user won’t be this lucky. I strongly discourage users from just mindlessly “mowing the lawn” and waving the transducer purposelessly through the RLQ. The correct technique is to positively identify the landmarks that will lead to the location of the appendix. Use a low frequency transducer first to get the “lay of the land” and look for large features. Then swap to a high-frequency transducer and repeat the process. Commence the examination in the groin in transverse section and identify the external iliac artery and vein. Follow the vessels superiorly. The first segment of bowel that crosses the vessels and the psoas muscle is the terminal ileum. Follow the terminal ileum laterally through the ileo-caecal valve and the caecum which is identified by its capacious size (Figures 7a and 7b). This is the territory of the appendix (Figure 8). Many appendices are found underneath the terminal ileum. All are, of course, connected to the caecum. Use generous amount of pressure to reduce the thickness of the tissues requiring investigation and have a good look around for the direct and indirect features of appendicitis. If the appendix is difficult to identify, roll the patient into a decubitus position. This sometimes shifts the caecum to the midline revealing a retrocaecal appendix.
Figure 7a: Overlay of RIF anatomy including terminal ileum, ileo-caecal valve, caecum and surrounding structures
Figure 7b: Anatomy of the RIF without overlay
Figure 8: Positive identification of the appendix
I am not one to concede defeat easily, but let’s be honest, the appendix will likely continue to elude us. Ultimately, it is down to your clinical judgement whether the patient need to be explored surgically, imaged with ultrasound or imaged with CT. We could have a long debate about the optimal strategy. As ultrasound users, we are likely going to resort to ultrasound. In my view, it is almost always the right choice. But we have to work hard and smart to get the answers.