Study Setting and Population
Overview
The study population will be all children presenting to participating ANZ EDs with abdominal pain. Study sites all have EDs that manage pediatric patients and have a pediatric on-site general surgical service with the capacity to provide input, admission, and surgical intervention if needed.
Perth Children’s Hospital (Western Australia, Australia)
Children's Hospital Westmead (New South Wales, Australia)
Gold Coast University Hospital (Queensland, Australia)
Monash Health (Victoria, Australia)
Queensland Children's Hospital (Queensland, Australia)
Royal Children’s Hospital (Victoria, Australia)
Royal Hobart Hospital (Tasmania, Australia)
Sunshine Coast University Hospital (Queensland, Australia)
Sunshine Hospital (Victoria, Australia)
Starship Hospital (Auckland, New Zealand)
Participant Inclusion Criteria
To be eligible to participate in the study, a participant must meet the following criteria:
Age 5 to <18 years,
Presentation with generalized or right-sided abdominal pain,
Duration of pain for ≤7 days (ie, ≤168 h),
Clinician’s concern for the possible diagnosis of appendicitis as defined by any one of the following:
Investigations performed (blood sample, imaging, or both, including external investigations before ED attendance),
Surgical consultation to assess the patient for possible appendicitis,
Senior ED clinician consultation to assess the patient for possible appendicitis,
Period of observation in ED to reassess patient for possible appendicitis.
For those hospitals with mixed ED and limited pediatric surgical capacity to operate on children under a certain age, the site-specific “pediatric” age range will be applied (eg, >10 years and <16 years). The age range of 5 to < 18 years was chosen due to the rarity and the atypical presentation of appendicitis in younger children [64], and many CPSs required variables on history or examination that might be difficult to determine accurately in younger children. Acute abdominal pain was defined as the duration of pain for ≤7 days in keeping with a previous study looking at abdominal pain and appendicitis [65], and a previous feasibility study found several patients with pain lasting for more than 5 days diagnosed with appendicitis [43].
Participant Exclusion Criteria
The following conditions exclude participation in the study:
Abdominal trauma that required medical review within the preceding 7 days,
Previous significant abdominal surgery (eg, appendectomy, short gut, ileostomy, and Hirschsprung disease),
Chronic illnesses that may affect the abdomen (including inflammatory bowel disease, chronic pancreatitis, cystic fibrosis, and sickle cell anemia),
Pregnancy,
Inability to obtain accurate history (eg, parent or guardian unavailable, language other than English and where no interpreter was available, or patients who are nonverbal due to pre-existing medical condition).
Outcome Measures
Overview
The primary outcome is the presence of appendicitis; this will be confirmed through the histopathology report when available, or the operation report if not available. Patients diagnosed with appendicitis but managed nonoperatively will be followed up to confirm the presence of appendicitis in cases where interval appendectomy was performed or excluded from analysis in cases treated conservatively with antibiotics only. A nonappendicitis diagnosis will be confirmed on follow-up via either recording of the primary diagnosis on hospital discharge if admitted to hospital, or contact at 30-60 days to ensure no progression in symptoms or representation to hospital for those discharged directly from the ED. The area under the receiver operating characteristic (ROC) curve (AUC) will be calculated for the various published CPS for the primary outcome of appendicitis.
Secondary outcomes include rates of uncomplicated and complicated appendicitis (defined as the presence of perforation, extraluminal fecalith, diffuse intraperitoneal pus, or a well-formed abscess) [62], negative appendectomy cases (based on histopathological findings), and missed appendicitis cases. These will be confirmed through the histopathology report when available, and the operation report if not available. To identify missed appendicitis cases, families of children who did not undergo appendectomy and were discharged directly from the ED will be contacted within 30-60 days of hospital discharge to assess for postdischarge status and whether their child had an appendectomy in the interim. For families unable to be contacted, medical records will be reviewed at 60 days to assess for representation. Other secondary outcome measures include length of hospitalization, rates of investigations, and rates of interventions.
Patient Recruitment and Data Collection
All suitable patients with abdominal pain will be identified on ED presentation and approached by a research nurse (if available) or the treating clinician for consent for (1) study participation, data access, and collection and (2) follow-up contact. Patients who are not captured on the index visit (ie, “missed” recruitment) will be identified by the research team through a review of the daily ED attendance record using the terms “abdominal pain,” “appendix” or “appendicitis” on triage text or discharge diagnosis text “appendicitis” or “abdominal pain.”
Basic demographic data of missed patients (including any interventions performed and management) will be recorded on a separate datasheet and collected at the earliest possible time. Missed cases, if assessed by the research team to be eligible, will have a retrospective case report form completed by the treating clinician, if possible, and the family will be contacted via phone or text message to obtain informed verbal consent for inclusion retrospectively for (1) study participation and data access and collection and (2) follow-up contact. Follow-up will not be conducted for patients who are admitted to the hospital under an inpatient medical or surgical team or those who experience severe morbidity, such as a new diagnosis of chronic illness or permanent disability, or are deceased. Participants who decline consent for data access and collection will be excluded from the study.
The treating clinician will complete, at the time of the ED visit, or retrospectively if missed initially, a case report form recording:
demographic data (sex and age),
eligibility (inclusion and exclusion criteria),
caregiver verbal consent,
the clinician's seniority and their perceived likelihood of appendicitis on a Visual Analog Scale from 0 being “Extremely unlikely” to 10 being “Extremely likely” and whether this perceived likelihood was formed before or after investigations (bloods, imaging, or both),
history of the presenting complaint (location of initial pain, duration of pain, highest pain score, highest reported temperature during this illness, right lower quadrant (RLQ) or right iliac fossa (RIF) pain on arrival to ED, pain migration to RLQ or RIF, progression of pain, gradual onset of pain, pain pattern, history of pain with walking, anorexia, nausea, vomiting, difficulty with micturition, pothole tenderness, family history of appendicitis, respiratory tract infection in last 2 weeks, and bowel habit), and
examination findings (RLQ tenderness, RIF tenderness, tenderness worst in RLQ, tenderness outside RLQ, presence of hop, cough, or percussion tenderness, degree of rebound tenderness, abdominal guarding, Rovsing sign, nature of bowel sounds, palpable abdominal mass, and abdominal rigidity).
Clinician management will proceed independently of study participation.
All data collected at sites will be transcribed to a REDCap (Research Electronic Data Capture; Vanderbilt University) database [66]. The following additional data will be collected from the patient's medical records: blood test results (white cell count, absolute neutrophil count, neutrophilia, and C-reactive protein), urinalysis (presence of leukocytes or nitrites), which will then allow the various CPSs to be automatically calculated in REDCap.
Between 30-60 days after the ED visit, the site research team will collect the following parameters in all eligible patients through a medical record review:
detailed demographics (postcode, Medicare status, race or nationality, and referral source),
prehospital management,
management and imaging were undertaken both pre-ED in the community and the study hospital,
time-related data (times of triage, clinician evaluation, ED, and hospital discharge),
duration of ED and hospital stay,
admission status and specialty unit consultations, including intensive care admission,
medication use in ED, including antibiotics and analgesia,
related surgical and nonsurgical interventions,
other significant pathology or adverse events, including mortality, and
disposition.
Follow-Up
All eligible patients will be contacted for a follow-up survey except when (1) the parent or guardian declined follow-up on the initial presentation, (2) the participants were admitted under an inpatient medical or surgical team with no further representations to ED within the follow-up period of 30 days, or (3) participant identified by the site research team to have experienced clinically significant adverse events. Any approved parental contact will occur via a telephone call, email, or text, with scripts for each contact method provided to participating sites. Follow-up emails will be automatically generated and sent out twice during the follow-up period with identifiable data accessible only to the local study site and the central coordinating study team, and will be removed from the REDCap database at 61 days postenrollment date by the local study research team. A maximum of 3 contact attempts will be made. If more than 60 days have elapsed from the time of presentation, or if there have been three failed contact attempts, the patient follow-up will be regarded as unsuccessful. The medical record for patients unable to be contacted will be reviewed and if sufficient information is documented in the medical record, this information may be used to substitute for the failed contact.
The follow-up survey will determine the following:
if the child had any problems with abdominal pain since being discharged from the ED,
if the child missed any childcare or school because of abdominal pain,
if the carers took any time off work because of the child’s ongoing pain, and
if the child has been seen by another doctor, health professional, or hospital since the visit to the ED.
It will also collect the following details: details of ongoing problems; the number of missed days of school or childcare; the number of days off work; types and numbers of medical reviews; additional medical diagnoses; need for and duration of ED stay and hospital admissions including investigations (with a review of the relevant medical record); and related surgical interventions. Any eligible participants whose parent or guardian consented for data access and collection but declined follow-up will have their data collected and analyzed, but will not be contacted for follow-up. If participants withdraw consent and discontinue participation in the study at any time, there will be no effect on their medical care or access to treatment. If a participant withdraws before completing the study follow-up period, any known reason for withdrawal will be documented in the database. All information already collected as part of the study will be retained for analysis, but no further efforts will be made to follow up or obtain additional information regarding the participant. As an observational study, there are no anticipated adverse events related to the research. During the follow-up contacts, if any medical issues are identified by research staff, they will be referred to the managing clinicians (for patients in hospital), or the site investigator (for questions during follow-up), who will refer, if required, the patient to their general practitioner or ED for review.
To minimize information bias, the calculation of the CPSs will be performed by the research team and not by the treating clinician. If the study participant is transferred from a nonstudy site to a study site, the data collection form will be completed by the initial treating physician at the study site; if the study participant is transferred from a study site to a second study site, this will ideally be completed by the initial treating physician at the first study site.
Sample Size and Power Calculation
Based on a feasibility study performed at PCH ED before the multicenter study [43], we estimate a true appendicitis rate of 30% in the multicenter study, and that all the CPSs will be calculable in at least half of the patient cohort. We aim to have a sample size of 2400 children with undifferentiated abdominal pain enrolled, which will result in all CPSs being calculable in a minimum of 360 enrolled cases of true appendicitis. This will provide a level of precision such that a 95% CI around the AUC will range from 0.016 for an AUC estimate of 0.95-0.032 for an AUC point estimate of 0.75 [67]. It will also provide more than 85% power for a 2-sided test, with α=.05, to find a difference of .05 between 2 AUCs if both values are ≥0.80, assuming the correlation of risk scores amongst appendicitis cases and that amongst noncases have an average of at least 0.5 [68]. Under this same assumption, if both AUC values are ≥0.9, the planned sample numbers would provide more than 90% power to detect a difference of 0.04.
Statistics and Interim Analysis
We will follow Standards for the Reporting of Diagnostic accuracy studies (STARD) 2015 guidelines for statistics [69] and the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [70], and validation of CPSs will be performed and reported in according to the 2014 updated methodologic standards for interpreting clinical decision rules in emergency medicine [71]. The AUC will be calculated for the various published CPSs and the clinician's gestalt for the outcome of appendicitis. The sensitivity, specificity, negative predictive value, positive predictive value, negative likelihood ratio, positive likelihood ratio, and missed appendicitis rate will be calculated based on the published cutoffs for each score [2-31]. For the pARC, subanalysis will be performed for the following pARC categories based on previous validation studies [4,56]: ≤5%, 6%-15%, 16%-25%, 26%-50%, 51%-75%, 76%-90%, and >90% for pediatric EDs and <5%, 5%-14%,15%-24%, 25%-49%, 50%-74%, 75%-84%, and ≥85% for mixed ED. For clinician’s gestalt, subanalysis will be performed for each point on the Visual Analog Scale. The AUC will also be calculated for complex appendicitis to assess the optimal cutoffs for each CPS to detect complex appendicitis.
The best-performing CPS will be identified by the highest AUC. Sensitivity and positive predictive value will be calculated to establish the CPS’s ability to identify high-risk patients, and the misclassification rate (proportion of patients identified as low-risk who had acute appendicitis) and negative likelihood ratio will be calculated to establish the CPS’s ability to identify low-risk patients. Accuracy will be reported as the proportion of patients not misclassified as either low-risk or high-risk as defined by published cutoffs. Baseline analyses of patient characteristics based on the risk groups as calculated by the best-performing CPS will be conducted by presenting simple counts and percentages with interquartile ranges.