Appendicitis in Three Courses

Fearless chief Ted has organized the first of our Tuesday Night Journal Clubs (with apologies to Sheryl Crow). The goal of these meetings is to informally discuss several relevant and provocative papers on a certain topic of interest in EM. Particularly, we wanted to give attendees a foundation in the literature, and some ammunition when dealing with other services. We picked a dinner theme, with an appetizer, main course, and dessert.

Our appetizer came from the classic JAMA Rational Clinical Exam series, “Does This Patient Have Appendicitis?” (Wagner et al. JAMA Nov 20 1996, Vol 276, No. 19, PMID 8918857). Some interesting and quotable tidbits from this comprehensive review:

• RLQ pain had a positive likelihood ratio of about 8, and negative LR of between 0 and 0.28
• Rigidity wasn’t very sensitive, but had a positive LR of 3.8
• Migration of pain had a LR of 3.2
• Psoas sign, rebound, guarding, nausea, vomiting, fever and anorexia had relatively unimpressive positive and negative LR’s. Having ‘no similar pain previously’ gave a negative LR of 0.32.

The main course and desert follows below!

Our main course featured two papers on dealing with contrast CT in diagnosing appendicitis:

 

 

“Advantages of Focused Helical CT Scanning with Rectal Contrast Only vs. Triple Contrast in the Diagnosis of Clinically Uncertain Acute Appendicitis.” Mittal, Goliath et al. in Arch Surg 2004 (139) p495-500 (PMID 15136349).

The authors found the sensitivity, specificity, PPV and NPV were comparable in the triple-contrast vs. rectal-contrast groups. The rectal contrast group tolerated the study better (1.5 L in, by the way – compared to 900 oral and 900 rectal in the triple groups), cost less ($305 vs. $620) and had no missed diagnoses with a better negative appy rate. They also noted a lower perforation rate compared to triple-contrast, but that was likely due to a failure of randomization – the triple-contrast group waited about 40 hours before presenting, compared to about 30 hours for the rectal group. ED-to-OR times were comparable in both groups, despite the time saved avoiding PO contrast.

Our group had a lot of confusion about the methodology. This is mostly because the authors note that 684 appendectomies were performed during the study period, but only 91 were enrolled in the study. Why? Because “only patients with uncertain diagnosis of acute appendicitis” were entered. We took this to mean cases in which the ED docs consulted surgery before the CT was done. Of course, we have no idea how many CT’s were done without consulting surgery, and how these were performed. And we were surprised by their determination that only 91 out of 684 of their appendicitis diagnoses were “uncertain” and thus needed a CT – the rest went straight to the OR! But this is a paper out of Southfield, Michigan, and the data was collected in 2000-2002, perhaps a period before CT gained the prominence it has in New York these past few years. (It’s worth noting in the discussion section, one of the authors floats the idea of a moratorium on ED docs ordering CT scans before the surgeon sees the patient – in my limited experience with surgeons, they’d prefer a moratorium on consults pre-CT).

The second paper of the main course was, “A systematic review of whether Oral Contrast is Necessary for the CT diagnosis of Appendicitis in Adults,” by Anderson et al, Am J Surg 190 (2005) 474-478 (PMID 16105539).

This was a less-than-rigorous review – using only MeSH terms on pre-2004 papers – of patients 16 or over, in which CT scans were used to diagnose appendicitis. The papers they picked had to provide enough data to calculate sensitivity, specificity, and NPV and PPV. Studies had to also be consistent in how contrast was given, too, and could not be just on women or children. That narrowed things down from 189 hits to 23 English-language papers.

 

Their analysis of those 23 papers showed that CT techniques that didn’t involve oral contrast were equivalent, if not better (!) than those with oral contrast (so IV +/- rectal was better than triple contrast, or oral + IV, or rectal alone). How is that even possible? IV contrast enhances the wall of the inflamed appendix, and oral/rectal contrast works by opacifying the cecum, allowing for visualization of appendiceal obstruction. But we kinda believed oral contrast wouldn’t hurt interpretation. Yet they found that the sensitivity and PPV of CT, sans PO contrast, was better than with it (specificity and NPV were about the same). They can’t explain it, and suggest maybe publication bias was at fault – maybe studies diminishing the sensitivity of noncontrast CT weren’t published, or residents were reading those films.Also, the introduction of helical CT’s during the study period may be confounding results.

 

This 2005 study was also notable because it references a 2000 statement from the Group Health Cooperative that states on 20% of patients undergoing appy have had a preoperative CT scan – that was a few years after Rao’s landmark studies and several years before we achieved the near-100% rate we have now.

For ‘desert’, we wanted to end on a light but interesting note, so we chose a paper from the EM literature (as opposed to the surgery papers from the main course). When it comes to appendicitis, the EM literature has the benefit of enrolling all comers (or at least, more comers) and so may be more pertinent to our practice.

The paper discussed was, “The use of the clinical scoring system by Alvarado in the Decision to Perform CT for acute appendicitis in the ED,” by Robert McKay et al in AJEM 2007 (25) 489-493 (PMID 17543650). These authors applied the previously-devised Alvarado clinical scoring system to decide which ED patients need a CT. The scoring is described below — each criteria can be remembered by the irrelevant mnemonic, MANTRELS:

Migration: — 1 point
Anorexia-acetone:– 1 point
Nausea-vomiting — 1 point

Tenderness in RLQ — 2 points
Rebound pain — 1 point
Elev. Temp over 37.8C — 1 point
Leukocytosis — 2 points
Shift to the left — 1 point

For a maximum score of ten. The authors retrospectively looked at 150 charts of patients aged 7 and older, finding Alvarado scores 3 or lower were 96.2% sensitive for not finding appendicitis with a specificity of 67%. Patients with Alvarado scores 7 or higher had an incidence of acute appendicitis of 77.7% (28/36).

For those equivocal cases of scores between 4-6, the sensitivity of scores in predicting acute appendicitis was 35.6%, and the specificity 94%. The sensitivity and specificity of CT scans in patients with equivocal Alvarado scores remained
high, at 90.4% and 95%, respectively.

From this, the authors concluded that, “In the equivocal clinical presentation of appendicitis as defined by Alvarado scores of 4 to 6, adjunctive CT is recommended to confirm the diagnosis in the ED setting. If clinical presentation suggests acute appendicitis by an Alvarado score of 7 or higher, surgical consultation is recommended. Computed tomography is not indicated in patients with Alvarado scores of 3 or lower to diagnose acute appendicitis.”

Our problems with this paper stem from our problems with the Alvarado system: any patient with a fever, white count, left shift, and nausea has four points and is therefore in the “equivocal” zone for appendicitis — even if their chief complaint was diarrhea or Crohn’s flare. Still, this is a fairly easy-to-learn tool to help justify CT scans and calls to surgery, if you’re on the fence or the patient has an ambiguous presentation.

So, there you have it — our first Tuesday Night Journal Club, and some useful facts and guidance for navigating the diagnosis and disposition of appendicitis patients.

 

 

---

Possibly related posts: (automatically generated)

• I Can’t Win
• How Dangerous Are CT Scans?