Should ambulatory appendectomy become the standard treatment for acute appendicitis?

Acute appendicitis is the most common indication for emergency surgery worldwide, with a lifetime risk of 8.6% for men and 6.7% for women [1, 2]. Despite the promotion of antibiotic therapy as a non-invasive alternative, laparoscopic appendectomy remains the standard treatment for acute appendicitis [3‐5]. It is a common surgery, performed mostly in young healthy patients, with standardized management, operating time under 1 h, low morbidity, and low average length of hospital stay, all of which render it safe to perform in ambulatory settings [6]. Over the past decades, appendectomy has been increasingly performed as ‘short stay’ (< 24 h hospital stay) [7‐10] or ‘ambulatory’ (no overnight hospital stay) [6, 11, 12] procedures, with satisfactory outcomes and low complication rates [13‐16]. Yet, conventional hospitalization (with overnight stay) remains the routine procedure for acute appendicitis [6, 16].

There is no present consensus on the selection of patients with acute appendicitis for ambulatory surgery (AS). Recently, Lefrançois et al. [6] proposed a new score, which classifies patients into 6 risk levels, to select patients for ambulatory appendectomy without compromising outcomes. In their series, only 20% were discharged on the same day of surgery. The percentage of appendectomy procedures performed in ambulatory settings reported in the literature varies greatly (20–88%) [6, 8, 11‐13, 15‐18], likely due to inconsistent definitions of AS and to different criteria for enrollment of ambulatory patients.

For over 5 years, the authors have been performing ambulatory surgery for increasingly complex procedures, including colectomy [19], and have implemented a standardized care protocol for appendectomy following the guidelines of AS for common elective procedures [20]. The primary goal of the present study was to evaluate a consecutive series of appendectomies, complicated and uncomplicated, performed at our center and compare characteristics and outcomes of patients operated in ambulatory vs. conventional settings. The secondary goal was to investigate factors associated with unexpected re-consultations and re-hospitalizations, and thereby improve patient selection for ambulatory appendectomy.

The authors retrospectively reviewed a consecutive series of 185 patients who underwent emergency laparoscopic appendectomy for acute appendicitis by 5 surgeons, between January 2015 and June 2016. The series comprised 104 men and 81 women aged 34.1 ± 18.2 (median, 28; range 8–87) at the index operation. Acute appendicitis was suspected on the basis of clinical signs (pain, abdominal tenderness, and muscle guarding in the right lower quadrant), laboratory test results (abnormal white blood cell (WBC) count and serum C-reactive protein (CRP) level), and systematically confirmed by ultrasound (53%) and/or Computed Tomography (CT) (47%).

Following the definition of the International Association for Ambulatory Surgery (IAAS), we defined true AS as ‘a procedure where the patient is discharged on the same working day,’ that is, a procedure with no overnight stay. Ambulatory appendectomy was offered to all patients diagnosed with uncomplicated or complicated acute appendicitis, who presented none of the following pre-operative contraindications: (i) cases that could not be operated before 5 pm and could not be postponed to the next day (ii) home-hospital journey over 1 h, (iii) patient living alone, and (iv) severe comorbidities which required monitoring. Furthermore, based on the experience of the emergency team, surgeons reverted to conventional appendectomy in the event of one or more of the following peri- and post-operative contraindications: (v) generalized peritonitis or abscess requiring administration of intravenous antibiotics (vi) anaphylaxis, (vii) enduring pain that could not be treated by oral analgesic, (viii) fever, or (ix) vomiting.

From the initial cohort of 185 patients, 117 (63.2%) were eligible for AS, of which 8 had peri-operative or post-operative contraindications (3 severe infections, 1 anaphylaxis, 2 generalized peritonitis, 2 excessive pain), which left a total of 109 (58.9%) patients operated in ambulatory settings. Their median length of stay, starting from hospital admission, was 8.5 h (range, 3.3–20.5). Compared to conventional cases, ambulatory cases were significantly younger (27.8 ± 13 vs 43.2 ± 20; p < 0.001), with lower preoperative CRP blood levels (27.8 ± 37 vs 63.7 ± 102; p = 0.016), smaller appendix diameters (10.0 ± 3.2 vs 11.7 ± 4.2; p < 0.001), fewer appendix perforations (10.1 vs 25%; p = 0.008), better Saint-Antoine scores (3.8 ± 1.1 vs 3.3 ± 1.2; p = 0.009), and shorter operation times (30.8 ± 12 min vs 37.7 ± 16.5 min; p = 0.004) (Table 1). Moreover, there were no statistical differences between ambulatory and conventional cases in terms of rates of unexpected re-consultations (11.9 vs 22.4%, p = 0.069), re-hospitalizations (4.6 vs 9.2%, p = 0.236), and re-operations (0.9 vs 4.0%, p = 0.306).

Of the 185 patients operated, postoperative complications were noted in 32 (17.3%), and included excessive pain or fever in 13 (7.0%), wound infections in 8 (4.3%), abdominal abscess in 5 (2.7%), liver abscess in 1 (0.5%), ileus in 2 (1.1%), phlegmon in 1 (0.5%), hematoma in 1 (0.5%), and acute urine retention in 1 (0.5%). The 109 ambulatory cases had a lower incidence of complications (11.9%) than the 76 conventional cases (25.0%) (p = 0.029). According to the modified Clavien system, the postoperative complications for ambulatory cases were of grade I in 7 (6.4%) patients and grade II in 6 (5.5%) patients, while for conventional cases, they were of grade I in 11 (14.5%) patients, grade II in 6 (7.9%) patients, and grade IIIb in 2 (2.6%) patients. Among the 32 patients with complications, 12 were re-hospitalized for the following reasons: pain or fever without abscess (n = 6), abdominal abscess (n = 3), liver abscess (n = 1), ileus (n = 1), and phlegmon (n = 1).

Uni- and multi-variable regressions for the entire cohort (n = 185) revealed that unexpected re-consultations were not significantly associated with any of the pre- or peri-operative variables (Table 2) but that unexpected re-hospitalizations were 4 times more likely for patients with appendicolith (OR, 4.32; p = 0.04) (Table 3). It is worth noting that among the 12 patients who were re-hospitalized, 7 (58%) had appendicolith, with a median time to rehospitalization of 3 days (range, 2–8).

Studies on appendectomy refer to various labels, definitions, and criteria for ‘ambulatory’ surgery, making it difficult to compare the proportions of procedures performed in true ambulatory settings. Some defined ‘ambulatory’ surgery as ‘outpatient procedures’ with a length of hospital stay < 24 h (with or without overnight stay). Although it seems logical to calculate length of stay from the time of hospital admission [8, 13, 15, 18], many authors calculate it from the end of surgery [6, 12, 13, 15, 17, 18]. Over the last decade, definitions of AS changed from day-case surgery with possible overnight stay [17] to a procedure with no overnight stay [6, 11, 12, 16]. We followed the current guidelines of the IAAS and defined ambulatory surgery as ‘a procedure where the patient is discharged on the same working day.’

In this series, 59% of all appendectomies were performed as ambulatory procedures, which is higher than the rates reported in studies with a similar definition of AS (Table 4), whose IIT population comprised complicated appendicitis [6, 11, 12]. Sabbagh et al. [12], Lefrançois et al. [6], and Grelpois et al. [11] reported ambulatory surgery in 18, 20, and 32% of cases, respectively. Cash et al. [13] and Frazee et al. [15, 18], who discharged patients within 24 h, and whose cohorts with intention to treat (ITT) comprised only uncomplicated acute appendicitis, claimed remarkable outpatient rates ranging from 65 to 88%. Our rate of AS is all the more remarkable that 27% of patients presented appendicolith and 16% had perforation signs. In our series, 18% of appendectomies could not be performed in ambulatory settings because they were diagnosed or operated outside standard working hours. This was also observed by Scott et al. [16] who demonstrated that surgeries completed after 8pm had 70% greater odds of retaining the patient overnight.

While there are guidelines for deciding on surgical or non-surgical treatment of acute appendicitis [27], there is no present consensus on medical indications to prefer conventional over ambulatory procedures for acute appendicitis [15], although ambulatory procedures have been shown to decrease post-operative morbidities [20]. In France, only 1.3% of appendectomy procedures in 2015 were performed in ambulatory settings [20]. Most surgeons see conventional hospitalization as the standard and look for patients eligible for AS. By contrast, we consider all patients eligible for AS by default, unless they (i) have severe comorbidities which require monitoring, (ii) cannot be operated before 5 pm and cannot be postponed to the next day, (iii) do not meet ambulatory discharge criteria [25], and/or (iv) have severe infections or complications.

Lefrançois et al. [6] created a score to select patients for AS based on 5 preoperative criteria (BMI < 28 kg/m², preoperative CRP levels < 30 mg/dL, preoperative WBC counts < 15,000/mm³, diameters of the appendix ≤ 10 mm, and no radiological signs of perforation). In our series, we found that patients with greater CRP levels, greater appendix diameters, perforation signs as well as older patients were less likely to have AS, but found no differences regarding BMI or WBC counts. Frazee et al. [15] and Cash et al. [13] also demonstrated that, in patients eligible for same-day discharge, pre-existing comorbidities as well as older age were factors that prompted conventional surgery. Scott et al. [16] underlined that the decision to discharge younger, healthier patients from the recovery room is often based on the subjective appreciation of operating surgeons rather than objective medical criteria. If so, then the practice of routine conventional hospitalization after appendectomy should be questioned.

In our series, complications were reported in 25% of conventional cases (Clavien grades I to IIIb) and in 12% of ambulatory cases (Clavien grades I and II). The higher rate of complications among conventional cases could be biased because they included patients with severe comorbidities. Among recent studies that used the definition of the IAAS, only Lefrançois et al. [6] reported the complication rate of 8% in their cohort of uncomplicated acute appendicitis treated by AS. In our series, the higher complication rate of 12% could be because we performed AS in a much larger proportion of patients, of whom 27% had appendicolith and 10% had perforation signs. Importantly, AS did not increase the rate of unexpected re-consultation, re-hospitalization, or re-operation. Our re-consultation, re-hospitalization, and re-operation rates amounted to 12, 4.5, and 1%, respectively, which is analogous to rates reported in recent studies [6, 11, 12, 16], even if our series included a larger proportion of ambulatory cases. Overall, the low rate of complications and re-hospitalizations of AS should challenge the current standard of routine conventional hospitalization for uncomplicated appendicitis. If AS is increasingly recognized as safe for uncomplicated cases, our findings suggest it could also be considered in some complicated cases without severe comorbidities.

Our uni- and multi-variable regressions revealed that no variable, including AS, were significantly associated with unexpected re-consultation. The only variable significantly associated with unexpected re-hospitalization was appendicolith, also known as fecalith, corpolith, or stercolith. Appendicolith, which is composed of firm feces and some mineral deposits, is a well-established risk factor for perforation [28‐30], inflammation [31], and failure of antibiotics treatment [32] or post-operative adverse events [31, 33, 34]. In our series, BMI was not associated with unexpected re-consultations or re-hospitalizations. In our experience, the Saint-Antoine score is a good predictor of complicated appendicitis, but is too restrictive for selecting patients for AS, notably because of the criterion of BMI > 28. Our uni- and multi-variable regressions revealed that patients with appendicolith were 4 times more likely to be re-hospitalized. It is worth noting that Lefrançois et al. [6] did not include appendicolith in their multivariable analysis on which they based their Saint-Antoine score. Furthermore, as highlighted by Di Saverio et al. [35], each variable selected in the calculation of the total score was given an equal weight of 1 point instead of adjusting the propensity of each variable to its odds ratio. Interestingly, our series presented comparable proportions of appendicolith in ambulatory and conventional cases, which enabled detection of its associated risks, independently from AS or conventional hospitalization. Given the association of appendicolith with re-hospitalization, the authors recommend that appendicolith should therefore trigger closer monitoring.

Appendicolith is a known risk-factor for non-operative treatment of appendicitis [20, 36, 37], and its surgical treatment is recommended [38]. Before conducting this study, the authors did not expect that appendicolith was also a risk-factor in surgical treatment of appendicitis; therefore, they did not manage patients with appendicolith specifically. Recent studies have identified that retained appendicolith after appendectomy is associated with abscess formation and infection [39‐42] and suggest that risk of spillage of appendicolith may be reduced by ensuring that the distal end of the appendix is intact after division [41] and by performing a double ligature of appendicular base [43]. Our findings suggest that appendicitis with appendicolith should be considered as “complicated,” and therefore be treated with antibiotics post-operatively. Patients with appendicolith must be informed of increased risks of rehospitalization and be monitored regularly after appendectomy whether performed in conventional or ambulatory setting.

The limitations of this study are typical of retrospective investigation. Out of 185 patients operated, we could not re-contact 33 (17.8%) patients, which may have led us to underestimate the number of minor complications. The strengths of the study include a strict definition of AS rendering comparisons with the literature rigorous, a consecutive series, a multi-operators database, and a large inclusion of patients.

In the present consecutive series of complicated and uncomplicated appendectomies, 59% were performed as true ambulatory procedures, with patients discharged on the same working day. There were no significant differences in rates of unexpected re-consultations, re-hospitalizations, or re-operations between cases performed in ambulatory versus conventional settings. Therefore, ambulatory surgery seems safe and feasible for selected patients with both uncomplicated and complicated acute appendicitis. To better inform patients and manage their expectations, surgeons should be aware that appendicolith is an independent risk factor of unexpected re-hospitalization for appendectomy and should therefore trigger closer monitoring.