SGEM#115: Complicated – Non-Operative Treatment of Appendicitis (NOTA)

Subscribe: RSS

[display_podcast]

Date: April 11th, 2015

Guest Skeptics: Dr. Bret Batchelor. Bret is a general practitioner with Enhanced Surgical Skills currently working in Vanderhoof, BC. He is the host of the newly created podcast that can be found on iTunes called Really Rural Surgery.

Case: A 35-year-old man presents to the emergency room with right lower quadrant pain for approximately 18 hours. You assess the patient and find that his Alvarado Score is 7. You then ask for an ultrasound, as his body mass index is in the normal range. The ultrasound shows that he has an enlarged appendix >6mm that is not compressible and there is no intraperitoneal fluid present. You make a presumptive diagnosis of uncomplicated acute appendicitis. You relay this finding to the patient and he turns to you and asks, “Hey doc, I heard that you can treat appendicitis with antibiotics now. Is that true?”

Background: Claudius Amyand did the first appendectomy in 1735. The standard treatment for acute appendicitis ever since Charles McBurney described it in 1889 has been appendectomy.

Omar et al (2008) showed just how safe laparoscopic appendectomies have become. They found in a study of over 230,000 UK patients under the age of 49 there were no deaths. Kluiber et al (1996) demonstrated the incidence of post-operative intra-abdominal wound infection to be about 2-5%.

Being that there are doctors out there without scalpels, and that diverticulitis has often been treated successfully with antibiotics (this also being an infection based on the same theory), some have put two and two together and postulated that perhaps acute appendicitis could be treated successfully with antibiotics.

Two meta-analyses have been done recently and interestingly enough; they looked at nearly the same studies on “uncomplicated” acute appendicitis and came up with two opposite conclusions. This is an example of why things in evidence-based medicine can be “complicated”.

Clinical Question: Are antibiotics in non-complicated acute appendicitis an effective and safe alternative to appendectomy?

Reference #1: Varadhan et al. Safety and efficacy of antibiotics compared with appendectomy for treatment of uncomplicated acute appendicitis: meta-analysis of randomised controlled trials. BMJ 2012

Population: 4 RCTs of adult patients (n=900)
Intervention: Antibiotics
Comparison: Appendectomy
Outcome:
- Primary Outcome: Complications (wound infection, perforated appendicitis, peritonitis)
- Secondary Outcomes: Length of stay, readmissions, efficacy of treatment, perforation, pain and body temperature.

Authors’ Conclusion: “Antibiotics are both effective and safe as primary treatment for patients with uncomplicated acute appendicitis. Initial antibiotic treatment merits consideration as a primary treatment option for early uncomplicated appendicitis.”

Quality Checklist for Therapeutic Systematic Reviews:

The clinical question is sensible and answerable. Unsure. Sensible question but the definitive diagnosis is pathology. That will always limit the group assigned to antibiotics to know whether they truly had uncomplicated appendicitis.
The search for studies was detailed and exhaustive. Yes
The primary studies were of high methodological quality. No
The assessment of studies were reproducible. Yes
The outcomes were clinically relevant. Yes
There was low statistical heterogeneity for the primary outcomes. Yes
The treatment effect was large enough and precise enough to be clinically significant. Unsure

Key Results:

Primary Outcome: Complications (wound infection, perforated appendicitis or peritonitis)

Relative Risk Reduction 0.69 (CI 0.54 to 0.89 P=0.004) favouring antibiotics

Secondary Outcomes:
- Length of Stay – No difference
- Readmissions: 68/345 (20%) of patients treated with antibiotics were readmitted with recurrence of symptoms. If one were to include the studies with crossover (which I think should be considered failure) this number would be 158/438 (36%).
- Efficacy: Antibiotics 274/470 (58%) and surgery 398/430 (93%). Failure as defined by normal pathology, which to me is a failure of diagnosis, not therapy.
- Pain and temperature were not analyzed in a meta-analysis format

Reference #2: Kirby et al. Appendectomy for suspected uncomplicated appendicitis is associated with fewer complications than conservative antibiotic management: A meta-analysis of post-intervention complications. J of Infection 2015.

Population: 3 RCTs of adult patients (n=531)
Intervention: Antibiotics
Comparison: Appendectomy
Outcome:
- Primary Outcome: Major complications (peritonitis or abscess formation after intervention). They excluded wound infection in this analysis.

Authors’ Conclusion: “Suspected uncomplicated appendicitis has a lower rate of major post-intervention complications when managed with primary appendicectomy compared to antibiotic therapy.”

Quality Checklist for Therapeutic Systematic Reviews: These are the same except for question #7

The clinical question is sensible and answerable. Unsure. Sensible question but the definitive diagnosis is pathology. That will always limit the group assigned to antibiotics to know whether they truly had uncomplicated appendicitis.
The search for studies was detailed and exhaustive. Yes
The primary studies were of high methodological quality. No
The assessment of studies were reproducible. Yes
The outcomes were clinically relevant. Yes
There was low statistical heterogeneity for the primary outcomes. Yes
The treatment effect was large enough and precise enough to be clinically significant. YES

Key Results: The primary outcome was major post-intervention clinical complications (peritonitis or abscess formation after intervention).

10.1% (27/268) with antibiotics vs. 0.8% (2/263) with appendectomy

You can alternatively report this as a Risk Ratio that was 7.71 (CI 2.3-25.5 p=0.0008) or a NNH =11. There was no statistical difference in perforated appendicitis in either group.

Diagnosis of Acute Appendicitis:
- In all studies this was different. There was no defined definition of the positive diagnosis of acute appendicitis. Some studies included certain lab tests (CRP etc.) others included +/- imaging that may have included CT or ultrasound or both. None of the studies used a defined scoring system for acute appendicitis (i.e. Alvarado Score).
- We know that even in the best circumstances, the negative appendectomy rate still ranges from 6%-30%. Diagnostic certainty is still quoted between 70-97% depending on where you look. This is a huge range. And, if studies don’t use rigorously defined diagnostic criteria for uncomplicated appendicitis with a known specificity, then we don’t know how many patients in each study labeled as “acute appendicitis” actually have “acute appendicitis” or some other entity.
- The only way to diagnosis appendicitis is on pathology. Thus, we will never truly know in any studies that randomize patients based on pre-pathology diagnosis whether we are treating a number of patients with “acute appendicitis” whom actually have another diagnosis. Thus, it will always be that the antibiotic arm will have an unknown in this regard as compared to the surgical arm that will always have a pathological confirmation of diagnosis.
Heterogeneity in Studies:
- Populations were different and didn’t include children, and didn’t include women in one study.
- Some studies used laparoscopic appendectomy, some used both, and others didn’t define the type of surgery.
- Each study used a different antibiotic regimen (though in the 2012 study, it showed that this didn’t have an effect on efficacy).
Intention To Treat Analysis (ITT):
- This means that patients were studied in the groups they were randomly allocated to even if they crossed over in the study.
- A crossover from antibiotics to an appendectomy means a failure of antibiotic therapy, rather than other reasons patients would fall out of an ITT analysis. Neither of the metanalyses tried to do a per-protocol analysis (though they did exclude one study with significant crossover in the 2012 metanalysis). What it would look like if all the crossover patients were studied in the groups in which appendicitis was actually treated successfully?
Treatment Difference:
- The type of antibiotic used did not have an impact on the results for the patients assigned to the antibiotic group. However, only two of the four studies in the Varadhan review reported peri-operative antibiotics prophylaxis in the surgical group. If antibiotics were given the post-op wound infection rate was about 3%. In the patients’ without documentation of antibiotics the infection rate was 12%.
- It is now standard care to use antibiotics peri-operatively for appendectomies because it decreases infection rate from 15% to 5%. So if antibiotics were not used in the surgical cases it stacks the deck in favor of the non-operative group. This means their higher composite outcome of complications in the surgical group could have been driven by post-operative wound infections
Included Studies:
- The systematic review by Kirby did not include the Hansson et al 2009 British Journal of Surgery study. This was an RCT of 369 patients with history, signs and laboratory tests suggestive of acute uncomplicated appendicitis. Not all the patients had imaging (CT and/or ultrasound) to confirm the diagnosis prior to randomization.
- Kirby also excluded Hansson because they included patients irrespective of the risk of perforation. A perforated appendix is not an uncomplicated case of appendicitis anymore. Antibiotic treatment of a perforated appendicitis can delay the diagnosis of complicated appendicitis and result in increased morbidity.

Comment on Author’s Conclusion Compared to SGEM Conclusion: It all depends on what you think is clinically relevant. These two meta-analysis came to opposite conclusions. They did so by choosing different studies to include and the outcomes that they felt (subjective) were clinically relevant.

If you think wound infection isn’t clinically relevant, then yes antibiotics increase risk of serious complications (abscess, peritonitis) compared to appendectomy. In contrast, if you think post-op wound infection is clinically relevant, then as a pooled outcome (wound infection, peritonitis and perforated appendicitis) overall it would look like antibiotics are safer in the treatment.

As long as you are not giving antibiotics prophylactically peri-operatively to prevent post-operative wound infections that is standard of care. And including studies that make no attempt to exclude patients with complicated appendicitis.

SGEM Bottom Line: Because of the diagnostic uncertainty of appendicitis, non-operative treatment of appendicitis (NOTA) will always be a very difficult thing to study. You will never know for sure in the antibiotics arm whether you were actually treating appendicitis. You will only know this in the surgical arm in which there is a pathology diagnosis. As a general practitioner with enhanced surgical skills, I would be concerned about treating suspected appendicitis with antibiotics because should it fail, this could lead to increased morbidity.

Case Resolution: You look skeptically at the patient. Being a person who routinely does surgery for acute appendicitis, your pre-conceived bias takes over and you say that yes, there have been some studies recently that have shown promise in this regard, but there also have been studies that demonstrate potential harm. Therefore, you stick with what you feel comfortable with and offer the patient a laparoscopic appendectomy as it has the lowest complication rate in regard to surgical management of acute appendicitis. In addition, it also has been shown to significantly reduce the negative appendectomy rate.

Dr. Bret Batchelor

Clinically Application: This represents a potential opportunity for shared decision making (SDM) between the patient and the surgeon. Barry and Edgman-Levitan NEJM 2012 describes SDM when the patient and the doctor collaborate on reaching a decision about a management strategy for a given clinical problem. For SDM to take place there must be more than one reasonable option. It also requires the physician give the patient the information they need to choose between competing acceptable strategies.

What Do I Tell the Patient: There are some studies showing antibiotics can be effective in the treatment of acute un-complicated appendicitis. However, it’s complicated, because there are other studies that show the opposite to be true. There are risks no matter what treatment you choose. I think the data still supports doing surgery right away but ultimately the decision is yours.

Keener Kontest: Last weeks winner was Cornelia Hartel from Germany. She knew San Antonio Medical Centre that used to be called the Brooke Army Medical Centre was named after Roger Brooke who had the rank of Brigadier General. His claim to medical fame was implementing routine CXR in military medicine.

Listen to the podcast for this weeks keener question. If you know the answer send an email to TheSGEM@gmail.com with “keener” in the subject line. The first correct response will win a cool skeptical prize.

Other FOAMed Resources: