Pediatric Appendicitis Pathway Study

Methods:

Overview:

This study will consent and enroll a prospective cohort as well as incorporate a retrospective before and after chart review based on a HIPAA waiver of informed consent.

The prospective portion will study pediatric patients ages 3-18 being evaluated for acute appendicitis in the OHSU/Doernbecher ED. Patients will be prospectively consented and enrolled by the CRISP program. The treating physician will document specific clinical predictors of acute appendicitis using a paper form. These criteria will be combined with routine laboratory measurements to identify patients at high and low risk for acute appendicitis based on two published and validated scoring systems.

High risk patients will be directly admitted to the pediatric surgery service. Low risk patients will be observed further in the ED or discharged home with close outpatient follow-up based on the ED attending physician's judgment. Low risk patients will not be routinely consulted by pediatric surgery unless by the discretion of the ED attending. Patients at intermediate risk based on the scoring systems will continue evaluation in the ED at the discretion of the attending physician. Patients who are not definitively diagnosed with acute appendicitis by operative findings will be contacted by telephone at two weeks by a trained research assistant to assess for missed outcomes, the same follow up method used by Schneider et al. in another validation study of prediction rules for pediatric appendicitis.

As this clinical protocol is being introduced in the OHSU/Doernbecher ED independently of the research study, patients not participating will likely also be treated according to a similar algorithm. The only difference will be that non-participating patients will have no data recorded except for that which is part of their regular clinical record, and will not be contacted for follow-up.

Outcomes of interest for the prospective portion of the study include the sensitivity and specificity of the scoring systems for identifying acute appendicitis, as well as the accuracy of various diagnostic elements including history, exam, laboratory studies, and imaging.

In order to assess the effect of the protocol on hospital-level outcomes, a retrospective before and after chart review study will also be conducted comparing the year before implementation to the year after. These patients will be retrospectively enrolled based on a HIPAA waiver of informed consent. Patients who consent to be included in the prospective portion of the study will also be included in this retrospective analysis. Outcomes of interest will include emergency department length of stay, time from arrival to inpatient bed or surgery, time to specialty pediatric surgery consultation, use of diagnostic imaging, and admission rate.

Study Subjects:

We will prospectively enroll a consecutive sample of all patients aged 3 through 18 presenting to the OHSU ED for suspicion of appendicitis during the hours in which study staff are available. The high risk (Alvarado) criteria have been validated in this age group. Clinical Research Investigative Study Program (CRISP) volunteers who are on duty in the emergency department between the hours of 7 am and 11 pm will identify possible patients based on suggestive chief complaints (e.g abdominal pain, vomiting, etc) and approach the attending physician about whether the patient is being evaluated for possible appendicitis. If so, they will approach the family, consent them for participation, and prospectively collect data on a paper form (see data collection section below).

Patients transferred from another facility to the ED will be included, including those with prior diagnostic imaging. Patients with symptoms greater than seven days will be excluded.

For the before and after portion of the study, patients will be enrolled retrospectively based on screening of eligible charts in EPIC. Screening will be based on the chief complaints and diagnoses in EPIC including abdominal pain, vomiting, appendicitis etc. Charts identified by initial screening will be reviewed by a trained physician and included if they judge that there was concern on the part of the attending physician for appendicitis. Their information will be entered in an electronic database (see below). Patients will be entered in the retrospective portion of the study regardless of whether they were enrolled in the prospective section.

Number of Study Subjects:

Based on a projected admission rate of 500 patients per year for suspected appendicitis, and an admission rate of 50%, we expect 1000 eligible patients to be seen in the OHSU emergency department annually. As the CRISP volunteers are present only 16 hours per day, with the expectation that 75% of the patient volume presents during those hours, and based on a 50% enrollment of eligible patients, we expect to prospectively enroll 375 patients in one year. Based on a prevalence of appendicitis of 30% and specificity of the high risk criteria of 78.8% in a previous study evaluating the high risk criteria, our projected sample size will yield of confidence interval for specificity of +/- 4.5%. For the low-risk criteria's previously demonstrated sensitivity of 98%, our sample size should yield a confidence interval for sensitivity of less than +/- 4%.

The before and after portion of the study will enroll patients for a full year before and after implementation to avoid the seasonal variation which is especially inherent in a teaching hospital. Expected enrollment will therefore be 2000 patients.

Prevalence rates in previous studies have ranged from 10 to 92% depending on inclusion criteria. Those studies with the highest prevalence rate included only patients who received who underwent surgery for suspected appendicitis. In studies with broad inclusion criteria similar to ours, (e.g. ED evaluation for acute abdominal pain, surgical evaluation for possible appendicitis), prevalence ranged from 10 to 61%.

Special Patient Populations:

Patients will be selected by the entry criteria described above. No special groups or vulnerable populations are specifically chosen for inclusion, but may be included within the general study population. There will be no specific exclusions on the basis of gender or ethnic groups. Consent forms will be translated into Spanish by ViaLanguage. Spanish speaking parents of limited English proficiency will be consented by professional in-person or telephone translator. Other limited English proficiency parents will be excluded.

Consent:

For the prospective portion of the study, eligible parents will be consented for enrollment by the CRISP volunteers using the attached consent forms. Children ages 7-18 will be asked to assent using the attached form.

For the retrospective portion, patients will be enrolled based on a HIPAA waiver of informed consent, with all data coming from the clinical record.

Study Procedure:

Patients enrolled in the prospective cohort portion of the study will be treated according to the clinical pathway (see attached). This protocol was developed in collaboration between the departments of emergency medicine and pediatric surgery. The pathway standardizes the routine treatment of pediatric patients being evaluated for appendicitis (e.g. lab ordering, pain medications, nursing assessments) as well as triages patients to further evaluation based on validated low and high risk scoring systems including historical, physical exam, and laboratory elements. Deviation from the pathway will be allowed at the discretion of the treating physician.

As this pathway is being implemented as a departmental protocol (similar to many other clinical pathways already in use in the emergency department) independent of the study, non-enrolled patients will also be treated according to it at the discretion of the treating physician. The only difference in treatment between enrolled and non-enrolled patients will be that data will be recorded independently from the patient's clinical record for those enrolled in the study, and enrolled patients will receive telephone follow-up two weeks after discharge if needed to determine outcomes.

The pathway and form for tabulating low and high risk clinical and laboratory criteria will be available for clinician reference for both enrolled and non-enrolled patients, but will not be retained after the visit for those not participating in the study.

The retrospective portion of the study will be based on EPIC chart review of patients being evaluated for appendicitis in the twelve month periods before and after implementation of the study.

Data Collection and Analysis:

Both the retrospective and prospective portions of the study will collect the following data: Patient demographics including age and gender. Duration of symptoms. Visit characteristics including referral from clinic, inter-hospital transfer, or presentation directly to the OHSU ED. Studies performed within the 72 hours before and after emergency department arrival: CT scan and abdominal/pelvic ultrasound, complete blood count, differential, urinalysis, urine pregnancy tests, metabolic panel, liver function tests, lipase. Time variables: time of emergency department room assignment, time of emergency department disposition, time of incision for patients undergoing surgery. Disposition will be classified as home, ED observation (defined by an ED stay >8 hours), surgery (including admission to inpatient with operation within 4 hours), inpatient observation (no operation within 4 hours of admission).

Patients who were seen at another facility for this same condition within 72 hours of coming to OHSU should have had the relevant part of that medical record sent with the patient, by fax, or by electronic medical record (Kaiser HealthConnect or PACS imaging system). If those records were not sent, the CRISP volunteer will obtain written permission on the HIPAA release of information (see attached). This will be immediately faxed to the referring facility so that this vital information is available to the treating physician for the patient's care in the emergency department as well as for the study purposes.

For the prospective portion of the study, consented patients will also have the specific signs and symptoms required for the calculation of the two scores documented on a form by the ED attending physician (see attached). This form will also ask the physician to rate their level of clinical suspicion for appendicitis prior to completing the scoring system. In the prospective study, sensitivities and specificities of the high and low risk criteria will be calculated for diagnosing appendicitis. Imaging study outcomes will be based on the interpretation of the attending radiologist in the clinical chart. These interpretations will then be classified as positive, negative, or indeterminate by a physician blinded to the final diagnosis. Data will be analyzed both including and excluding patients with imaging performed before transfer, as this may alter the use of the pathway.

Subjects in the prospective portion of the trial who do not have a surgically confirmed diagnosis of appendicitis made during their hospitalization will be contacted by phone by a research assistant (medical student or physician). The primary outcomes of interest will be whether the patient was subsequently diagnosed with appendicitis. Other data will include all subsequent medical visits and diagnostic studies, time until cessation of symptoms, and alternative diagnoses made.

Time outcomes and rates of imaging and laboratory utilization will be compared in the periods before and after implementation of the protocol in appropriate regression models incorporating variables considered to be of importance a-priori including age, gender, and pediatric ED census at the time of arrival by six hour block. Rates and location of pediatric surgery consultation (ED versus inpatient) will also be recorded. Key outcomes will also be analyzed in an interrupted-time-series fashion, looking at outcomes month by month in order to track the immediate versus sustained effect of the protocol implementation.

Paper data collection forms will be transcribed to a Microsoft Access database kept on a secured server at OHSU. Analysis will be conducted using STATA 11. Data analysis will be performed at the Oregon Health Sciences University Department of Emergency Medicine. No data will be shared outside of OHSU.

Ethical Considerations:

Once the records have been entered in a Microsoft Access database kept on a secure server with limited access at OHSU, records will be de-identified prior to analysis. The date of visit will be changed to include only the month and date of birth will be converted to age in order that they will not be identifiable. With these safeguards, we are confident that we can adequately protect patient confidentiality.

Risks:

The risk of participation in this study is that the clinical decision rules could result in treatment that was inferior to that which would have occurred without this treatment pathway. However, as the pathway is being implemented in the emergency department as a care standard, patients would likely be treated according to it whether or not they participated in the study. The other risk of participation is loss of privacy if confidential data were to be compromised.

As these risks are minimal, interim analysis or a data safety monitoring plan is unnecessary.

Data Safety Monitoring Plan:

Safety will be monitored by the investigators and the department of emergency medicine continuous quality assurance (CQI) process. The CQI process automatically flags all returns to the emergency department within 48 hours of discharge and forwards them for review monthly by the CQI committee. Cases of missed appendicitis, and other diagnoses require specific treatment or hospitalization identified by two-week telephone follow up will be reviewed by the CQI committee as well as the investigators and the division chief of pediatric emergency medicine (Dr. Spiro). Cases of patient admitted to emergency department observation who were subsequently admitted to the hospital will be identified by the research assistant and reviewed by the investigators.