Study design and participants
This is a single center randomized controlled trial consisting of two treatment groups (NCT05401188). This study was designed in accordance with the Declaration of Helsinki guidelines and approved by the Ethics Committee of Pere Virgili Institute (CEIm 081/2019). Patients were recruited at University Hospital of Tarragona Joan XXIII in Spain from the June 1st 2019 to June 1st 2021.
All patients admitted in the emergency department with AA diagnosis were recruited for the study. Saint Antoine’s criteria were evaluated in all patients(17). These criteria were the following: Leukocytosis lower than 15.000; C-reactive protein lower than 30mg; Body mass index lower than 30kg/m2; appendicular diameter lower than 10 mm and no radiological sign of complication(17). The radiological diagnosis could be made by ultrasonography (US) or Computed Tomography (CT) scan.
Eligible patients were approached, and informed consent was taken. After LA had been performed and a non-complicated AA was confirmed, the patients were randomly assigned to the hospitalization group (HG) or outpatient group (OG).
The inclusion criteria were the following: patients over 18 years old with diagnosis of AA, who met at least 4 out of 5 Saint Antoine criteria, ASA less than or equal to 3, patients who did not live alone at a maximum distance of 30 minutes from the hospital with an adequate cognitive capacity.
The exclusion criteria were the following: pregnancy or breastfeeding, complicated AA, or a surgical time longer than 90 minutes.
If the criteria were not met or presenting any exclusion criteria, the patient was excluded from the study and followed the usual clinical management according to our center protocol.
Anesthetic and surgical protocol
To ensure consistency, an anesthetic and surgical protocol were established for both groups. A modified ERAS protocol(4) was followed in both groups.
During preadmission, the ERAS protocol is not available in the emergency setting. The patient was fasted before surgery. The abdomen was prepared using clippers to remove hair from the incision area. The patient was asked to void the bladder before surgery.
The ERAS protocol for anesthetic management was preoperative evaluation. Intraoperative monitoring included: non-invasive blood pressure, EKG, oxygen saturation, heart rate, Bispectral Index(BIS™) and train of four(TOF™). Pre-oxygenation was performed. Rapid sequence induction was achieved using propofol at a dose of 2mg/kg, remifentanil in continuous infusion starting at 0.2picograms/kg/min and rocuronium at a dose of 1 mg/kg/min. One minute after induction, intubation was carried out using direct Mackintosh laryngoscopy. Intravenous fluid with balanced crystalloids was started at a rate of 5 ml/kg/h as a replacement for potential losses. Anesthetic status was maintained with Remifentanil TIVA(Total Intravenous Anesthesia) between 0.1–0.2 picograms/kg/min continuous infusion, and propofol at a dose of 10mg/kg/min in the first 10 minutes and at 8mg/kg/min during the next 10 minutes, and finally between 5-7mg/kg/min for the rest of the surgery according to BIS™. Nausea prophylaxis of dexamethasone 4mg after induction, and Ondansetron 4mg 20 minutes prior to completion was completed. Multimodal analgesia preformed with infiltration at the surgical site prior to incision of 10ml of 2% lidocaine, with an additional subfascial infiltration before aponeurotic closure with 10ml of 0.2% of Ropivacaine. Intraoperative analgesia is administered with Dexketoprofen 50mg and Metamizole 2gr. Reversal of neuromuscular blockade with Suggamadex→ is achieved according to TOF at the end of the procedure. Finally, the patient is extubated and transferred to the recovery unit.
For the surgical protocol, the patient is positioned supine. Laparoscopic approach begins with pneumoperitoneum performed through a Hasson’s Trocar placed in an umbilical incision after infiltration with local anesthesia. A slow and progressive insufflation is performed. Low-pressure (8–9mmHg) pneumoperitoneum is established. A 5mm trocar is inserted in the hypogastric area and a 5/10mm trocar in the left iliac fossa, after infiltration with local anesthesia. The meso-appendix is dissected either using diathermy or by diving the vessels with endoscopic clips. The closure of the appendix stump is commonly carried out with Endoscopic loop. The specimen is retrieved through an Endoscopic bag. Right iliac fossa and the Douglas cul-de-sac are flushed with saline solution. The trocars are removed under direct vision, the residual pneumoperitoneum is eliminated through them. Finally, the closure of umbilical fascia is conducted with long-term monofilament resorbable material nd staples are placed in the skin.
Following the postoperative ERAS care protocol, postoperative pain intensity was measured at rest on the visual analog scale (VAS) immediately after the patients regained responsiveness and then after 2, 6, 12 and 24 h following the surgical intervention. Analgesia was provided on demand for patients with pain intensity > 5 cm (VAS). Those who developed postoperative nausea and vomiting (PONV) received antiemetics. No intravenous infusions were given postoperatively to any of the patients.
The patients were discharged home if they had no complications, and they accomplished a ALDRETE score
A telephone survey was conducted on days 2 and 30 after discharge from hospital. The patients were asked about the presence of pain, the episodes of fever and indigestion, any complications and readmissions.
Randomization and intervention
Randomization was performed using a computer-generated program. Surgeons on-call were responsible for enrollment and treatment allocation, according to sequentially numbered envelops. Enrollment was unblinded for patients and physicians, due to the type of intervention. To reduce biases, the investigators assessing the outcome did not participate in the follow-up or discharge of patients. All patients received detailed written information about their diagnosis and study treatment plan.
After obtaining the informed consent of the study, the patients were admitted and operated following the same anesthetic technique and the same surgical technique by all the members of the on-call team. Once the patient was operated and an uncomplicated appendicitis was confirmed, the patient was randomized to one of the two branches: the HG or the OG.
Patients in the HG, once the surgical intervention was finished, were transferred to the postoperative recovery unit, and later they were discharged to the usual hospital ward. Patients received adequate intravenous fluid resuscitation, based on their individual hemodynamic parameter and fluid balance, and they received analgesia according to personal requirement. In the hospital ward, the usual patient management protocols were followed until a complete recovery, and consequently discharged according to the conventional criteria.
Patients in the OG, once operated, were transferred to the surgery-day unit and were later discharged home if they met the ALDRETE criteria (18), in less than 23 hours after the intervention(19) (following the same day surgery criteria). If the patient was operated during the night shift, following the advice of major outpatient surgery (19, 20) where overnight stays are allowed, the patient was admitted to the post-anesthetic recovery unit and discharged the next day, always in less than 23 hours. In case of being discharged after 23 hours or not meeting ALDRETE criteria, it was considered a failure of the outpatient treatment.
Study Endpoints
We consider the length of stay (LOS) the best indicator to confirm the safety and efficacy of outpatient management of laparoscopic appendectomy.
The primary endpoint was the LOS. The LOS was calculated from the time of admission in the surgical area to the time of discharge. The secondary endpoints included the failure of outpatient management, complications, readmission, unplanned hospital appointments within 30 days and cost.
Data Collection
The anthropomorphic, radiologic characteristics and blood samples were collected in the emergency department. Quantitative data were obtained from the blood results and from the appendicular diameter. Categorical data were obtained from radiological features. The postoperative complications were evaluated according to the Clavien-Dindo(21) and Comprehensive Complication Index scores(22). The day and time of admission to the surgical unit and time of discharge, were collected as a time variable. Follow-up after discharge, consultations to the emergency department and its reason were assessed. Outpatient follow-up was carried out up to 6 months. The cost analysis study was also conducted.
Data on each patient were collected in a standard form by the research coordinator ensuring the anonymity of the patients. Data was monitored and included in a database using Access© Microsoft 2013 and the statistical study was performed with Stata® inc. 16 version.
Sample size and Statistical analysis
In the hospital reference population, the median of LOS was 2.75 days (SE 0.42), according to a previous review of cases of AA, admitted to the hospital between 2015 and 2018. The population sample was calculated using Stata®, based on a 5%-α and 10% β-risks. A dropout rate of approximately 30% was assumed. Based on this, to detect a reduction in hospital stay which meets the requirements for surgery without admission (contemplated up to a maximum of 23h, or what is the same as 0.95 days), a total of 60 patients per group was assigned (a total of 120 patients). An analysis was conducted, with the exclusion of patients in whom AA diagnosis was incorrect or in whom no exclusion criteria was met (it was decided before any analysis by the data monitoring committee whose members did not know the treatment assignments).
Differences between groups were analyzed using Fisher exact test for categorical variables and t-test or ANOVA test for quantitative variables. P-values < 0.05 were considered significant. If after performing the Shapiro-wilk test the variable did not follow a normal distribution, the p-values were obtained through a permutation test. Multiple comparison was conducted using Wilcoxon nonparametric test for the categorical variables. Fisher and McNemar test were used to analyze the results of contingency tables. Valid variables were considered if they presented a maximum of 10% of missing values, choosing a multiple imputation with a matrix of 20 values to solve it. All analyses were performed by a statistician.