Past Prospective StudiesAbdominal Pain Study
Logo of the MET Program

Logo of the Telfer School of Management

Abdominal Pain Study

MET system in action

Picture of a physician holding a mobile device

At CHEO, there are over 3000 ED patient visits annually seeking treatment for the presenting complaint of abdominal pain. This includes patients who are experiencing abdominal pain as a result of trauma. On average this results in 8-9 visits on a daily basis. Many more patients have abdominal pain as part of their presentation, though this is not captured as their presenting complaint. Approximately 240 children per year are admitted with acute appendicitis, with an additional 300 patients admitted with other causes. Of these patients, approximately 100 are discharged from hospital with non-specific abdominal pain following an average 2-day hospital stay.

Assessing children with abdominal pain is time and resource consuming. Patients wait an average of 60-90 minutes to be assessed, and then remain an additional 150-180 minutes for investigations and observation before a disposition is made. For patients found to have appendicitis, the total time in the ED is greater than 300 minutes, while those admitted with other causes of abdominal pain have an average length of stay greater than 380 minutes. During the ED visit, about 55% of all patients have laboratory investigations and 26% undergo imaging procedures.

Based on such numbers, we concluded that there is clear potential to reduce resources used to assess, investigate and observe patients with abdominal pain. A valid triage recommendation provided by a clinical DSS could lead to a correct disposition decision earlier and with fewer diagnostic tests.

A prospective trial was conducted from July 2003 through February 2004. During that time, 2255 patients with abdominal pain visited the ED. 1098 of these patients were eligible to be recruited for the trial, and 632 of them were approached and asked to participate with only 38 refused consent. Finally, 574 patients completed the follow-up period and were used for analysis – 457 were seen by the EPs and 339 by the residents.

The MET1 system with the abdominal pain module (MET1-AP) was used by over 150 members of the ED clinical staff – none of them were involved in development of the system. Due to ethical considerations, medical personnel participating in the trial were blinded to triage suggestions given by MET1-AP.

The overall accuracy of MET1-AP on patients who were seen by EPs was comparable to the accuracy of EPs themselves. The system's performance was similar to that of EPs for patients who could be discharged and who required specialist consult. However, its accuracy was lower for patient cases that required further investigations and/or observations. This result, consistent with what is being reported in the medical literature, might be attributed to the heterogeneous nature of this group of patients, as it includes patients ranging from those who suffer abdominal pain because of by gynecological problem to those who have abdominal pain resulting from acute illness in other body systems (i.e., pneumonia).

For the prospective study, the MET1 computing architecture was customized to integrate with information systems at CHEO (EPIC for storing patient records and DataGate for exchanging information between various systems). The system was also extended with a module for auditing patient records and completing follow-ups to facilitate the trial management process. The MET1 client was added a triage checklist to verify and record eligibility of patients participating in the study.

MET1 customized and extended for the prospective study
A diagram with customized and extended architecture of the MET1 system