Can Palm Pilots influence the decisions of emergency doctors?
This question is now on its way to being answered as Ottawa researchers prepare to enter a pivotal stage in a study involving handheld devices and child patients with abdominal pain.
So far, Dr. Ken Farion, an emergency
physician at the Children’s Hospital of Eastern Ontario (CHEO), has conducted a blind study where he is using a Palm to record the symptoms of abdominal patients. He plugs in 13 variables, including the point of pain and patient temperature, after which the handheld device uses a clinical algorithm to make three prioritized options: the child needs a surgical consult for a probable appendicitis, the child needs further observation for a related problem, or the child has a benign problem that requires no further investigation.
The intent is to see whether these options — all of which receive a score out of 100 and are presented to users in a bar graph format — have the power to influence a doctor’s decisions and ultimately affect the patient’s healthcare experience for the better, Farion said.
After the algorithm receives further tweaking and validation, conductors of the study plan to “”unmask”” the Palm’s findings this summer so those doctors who work with the group of 525 patient participants can see the handheld in action.
“”We want to see how the physician incorporates this information into the care of the patient,”” said Farion.
The algorithm is designed to replicate the reasoning of a doctor without telling him or her how to think, explained Dr. Wojtek Michalowski of the University of Ottawa’s school of management. For several years, Michalowski and CHEO’s Dr. Steven Rubin have worked on developing the algorithm and incorporating it into an operating system.
“”Our system is not a diagnostic one. It simply suggests all possible options,”” added Michalowski.
Farion emphasized that if this experimental use of PDAs ever goes mainstream, the devices aren’t expected to replace the decisions of medical experts. Rather, a PDA equipped with an algorithm is designed to assist doctors in making decisions, acting as a valuable bedside tool that can instantaneously access a database of past patients with similar symptoms, he said.
When the clinical algorithm kicks into action, it goes through several hundred steps, using a quantification scale to assign a score to each of the three options.
The algorithm was created by collecting retrospective patient data from roughly 700 patients, including patient charts and segments of physical exams, said Farion. The data has been mined to determine the patterns that help discriminate between the handheld device’s three decision categories, said Michalowski.
The prospective data from the existing 525 participants in the study is expected to further validate the algorithm, Farion said, adding the patient group is expected to grow to 600 by the end of February.
Using a Palm Pilot in this way could be especially handy in a rural setting, adds Dr. Farion, adding a second opinion from a PDA could mean avoiding an unnecessary trip to an urban emergency room. Eliminating unnecessary trips means shorter wait times for those who need medical attention, he said.
Using the device could also act as a safety net.
“”(The PDA) could say ‘You haven’t considered that this child might have an appendicitis when in fact his pattern matches quite closely to other children who have had it – you might want to get a surgeon involved.’””
So far, the reaction from patients and their parents has been positive, said. Farion.
“”A lot of the parents are fascinated by the potential of this. Because it doesn’t take long, and our ultimate goal is to ensure patients are on the correct path sooner, they see it as a true benefit if it all works out.””
This summer, the research team will be testing the system’s interoperability with other hospital systems. Possible applications to other patient groups are now being considered.
The work is funded by the Natural Sciences and Engineering Research Council of Canada (NSERC) and Physicians’ Services Incorporated (PSI).
Comment: [email protected]
