Non-operative treatment of uncomplicated appendicitis (NOTA) has been demonstrated through major clinical trials to be surgery- and complication-sparing, with earlier acute pain resolution and less disability than appendectomy, and with outpatient management appearing possible in as many as 90% of cases, further cutting costs.
Additionally, write medical experts David Talan and Salomone Di Saverio, risks of computed tomographic (CT) scan–missed cancer and advanced disease progression in adults appear negligible, with recurrence being uncommon after a year, and later surgery posing no additional risk.
Yet, they write in JAMA Surgery, despite the American College of Surgeons and the World Society of Emergency Surgery endorsing NOTA as a safe first-line alternative to appendectomy, NOTA remains uncommonly practised.
A 2024 US National Inpatient Sample report indicates 18% of adults had NOTA. It was used more often among older adults with comorbidities and at academic centre, suggesting an even lower rate of use among patients without additional peri-operative risk at community hospitals.
Futility is the most commonly cited reason to recommend appendectomy over NOTA due to a substantial chance NOTA will fail and appendicitis will recur – estimated to occur in as much as 50% of cases by the online decision aid appyornot.org – incurring avoidable pain, medical care, and recovery.
But what if NOTA’s failure rate is closer to 15% than 50%?
The major clinical trials were monumental in challenging more than 100 years of practice and comparing NOTA and appendectomy in a large number of adults and children. However, a blinded, sham operation model was impossible. While providing a pragmatic perspective on acceptance and outcomes associated with a new treatment among physicians and patients at the time, the trials were open, with asymmetrical biases that we will show exaggerated NOTA’s failure rate and underestimated its actual clinical efficacy.
When these trials were conducted, physicians had little NOTA experience, and broad equipoise did not exist. Clinical response was subjectively decided by surgeons concerned for allowing rupture, which prematurely limited antibiotic trials and increased appendectomy rates.
The CODA (Comparison of Outcomes of Antibiotic Drugs and Appendectomy) trial had a management guideline recommending appendectomy if there were diffuse peritonitis, septic shock, or worsening after 48 or more hours. But this guideline was not enforced – appendectomy was performed before 48 hours in 22% of cases with and 8% without appendicolith, including approximately 9% and 6% within 24 hours, respectively.
Yet, guideline indications for surgery existed in less than 1% of cases.
The known association of appendicolith and rupture may have caused a tendency toward early surgery. Early surgeries also imbalanced counting complications, particularly in the appendicolith subgroup.
A National Surgical Quality Improvement Programme finding was counted if the observed abnormality were found any time after randomisation in participants assigned to NOTA, including those getting early surgery, but not when seen at the initial operation performed sometime later in participants assigned to surgery.
CT-missed rupture was found at operation in 24% and 13% of cases with and without appendicolith, respectively.
The APPAC (Appendicitis Acuta) trial stipulated that surgery could be performed if progressive infection, perforation, or peritonitis were suspected at any time, and the Midwest Paediatric Surgery Consortium (MPSC) trial specified that appendectomy “could occur” if there were no improvement after 24 hours, potentially leading to prematurely discontinuing antibiotic trials for patients with only localised symptoms.
In these open trials, patients also knew their treatment assignment and could not ethically be prevented from later electing the alternative treatment.
Many patients believe the now-disproven narrative that surgical delay results in appendiceal rupture and death. The CODA and MPSC trials included in their failure rates patients randomised to NOTA who later consented for surgery without evidence of clinical worsening, which accounted for as many as 20% to 30% of early antibiotic “failures”.
In the CODA and APPAC trials, surgeons performed some interval appendectomies after successful medical treatment.
For example, in APPAC, a higher proportion of appendectomies with negative pathology occurred among participants assigned to NOTA with “recurrence” than those assigned to surgery (8% vs 1%, respectively).
Along these lines, in the CODA trial, appendectomy rates were higher among those originally doubting NOTA’s effectiveness than those expecting its success: at 30 days, rates were 26% vs 14%, respectively. The former group likely included patients insufficiently reassured who had limited antibiotic trials; their attitudes may well have correlated with those of their physicians.
About half of the CODA trial’s participants treated with NOTA had emergency department (ED) discharge and outpatient management, which was found safe, further supporting a longer minimum antibiotic trial (eg, 48 hours).
Surgeon tendency toward premature appendectomy among participants assigned to NOTA was suggested by a lower appendectomy rate among those sent home with oral antibiotics as opposed to those cared for on surgical wards (at 30 days, 13% vs 20%, respectively), including when severity matched and in the appendicolith subgroup, with the difference extending over two years.
Before participating in the CODA trial, one of the authors’ (D.A.T.) groups reported results of a pilot randomised trial completed in 2015 of 30 patients (median [range] age, 33 [9-73] years) with uncomplicated appendicitis (5 patients [17%] with appendicolith).
In this trial, only two of 15 participants assigned to NOTA (13%; 95% CI, 2%-41%) followed up over one year had appendectomy; major complications occurred in two patients assigned to surgery and one assigned to NOTA. More participants assigned to NOTA achieved days three to five pain resolution (63% vs 7%, respectively).
While the small number of participants clearly limited the certainty of these results, our experience with this pilot trial and NOTA in other settings in the context of these observations provides clues as to why this low appendectomy rate may more accurately represent NOTA’s actual efficacy than that reflected by larger trials.
First, although the pilot trial had similar NOTA appendectomy indications as the CODA trial, in contrast, criteria were consistently enforced, and no early surgeries occurred; most patients improved by 24 hours, and all improved by 48 hours.
Patience has not traditionally been the practice in managing appendicitis.
Second, we noted that early on, upon enrolment, participants assigned to NOTA often did not have pain as well controlled as those assigned to appendectomy. Surgeons expressed concern for opiates masking progression and non-steroid anti-inflammatory drug bleeding risk, should surgery become necessary. However, once physicians were reassured by research disproving these notions, consistent pain control was achieved, which was not dictated by larger trials’ methods.
Severe but localised pain was treated medically, not surgically, allowing time for response.
Third, no participants elected crossover from NOTA to surgery.
Fourth, 93% of participants assigned to NOTA had ED discharge, which reduced the opportunity for appendectomy and extended the antibiotic trial absent concern for prolonging hospitalisation. Fifth, although appendicitis recurred in an additional patient treated with NOTA, that participant elected antibiotic retreatment and responded; appendectomy was required in APPAC and MSPC trials, and antibiotic retreatment was only rarely used in the CODA trial.
Finally, no NOTA-recovered participant had later interval appendectomy.
By design, the APPAC and MSPC trials had surgeon-defined NOTA futility primary outcomes (ie, one-year appendectomy rates >25% and >30%, respectively).
NOTA failed, with appendectomy rates of 27% and 33%, respectively, far higher than the negative appendectomy rate, which is observed to be as high as 15% in non-enrolled patients and in general practice.
CODA’s appendectomy rates were even greater, particularly in the appendicolith subgroup, with the rate around 50%. Based on these goalposts, it is unsurprising that physicians have not embraced this option and that NOTA is infrequently used in practice.
Further, it remains rewarding to help patients, as we have, by providing treatment that is a first-time cure.
Most patient candidates for appendectomy, however, have a different perspective than most physicians who perform appendectomy.
Surveys of adults now indicate most would choose NOTA even if the chance of surgery were 50%, a threshold also found acceptable among families in planning the MPSC study and justified by the absence of significant risk.
In other words, patients would move the goalposts back. But first, they have to have a chance at the ball. With greater understanding of biases associated with the major trials, recognition that the true NOTA appendectomy rate is likely to be closer to 15% than 50%, and awareness of ways to optimise NOTA’s success, physicians can more confidently offer NOTA – especially knowing that many patients would prefer an opportunity to avoid surgery and accrue NOTA’s other benefits.
David Talan, MD1 – Department of Emergency Medicine, Ronald Reagan UCLA Medical Center, David Geffen School of Medicine, University of California, Los Angeles
Salomone Di Saverio, MD, PhD – Department of General Surgery, Madonna del Soccorso Hospital, AST Ascoli Piceno, San Benedetto del Tronto, Italy
See more from MedicalBrief archives:
Antibiotics vs surgery for appendicitis
Antibiotics effective in 70% of appendix cases with no added risk – Duke study
Royal College Ireland: Surgery best for acute uncomplicated appendicitis