Computerized prescriptions can reduce medical errors
A pilot group of Vanderbilt doctors has laid aside pens and prescription pads and is turning instead to the computer to write prescriptions for their patients.
A prescription writer program created at Vanderbilt will significantly limit opportunities for medical error, said the developers and the doctors who are testing it. The program is integrated with the Medical Center’s electronic medical record system.
Besides producing prescriptions that are 100 percent legible, the program will automatically check each new prescription against a patient’s known problems and other medications, looking for contraindications such as drug allergies and potentially dangerous drug-drug interactions. The program will issue an alert if anything appears wrong with a prescription.
The prescription writer will also eventually use the patient’s weight and age (already available from StarPanel) to calculate and recommend a safe and effective dosage.
The program is designed to operate with StarPanel, an application that works off the StarChart electronic medical record application to support more efficient clinic workflow, easier communication of patient information, and quality improvement.
Prescriptions created with the program will automatically show up in the problems section of the patient’s electronic medical record, helping greatly to improve communication regarding which Vanderbilt patients are taking which medications. This information can be crucial when a patient lands in the emergency department.
The prescription writer will roll out across Vanderbilt Medical Group within a year, said project director Kevin B. Johnson, M.D., associate professor and vice chair of Biomedical Informatics.
Johnson said decision support capabilities of the prescription writer will grow as more of the Vanderbilt electronic record — the problem list, in particular — becomes coded information (field-bound information interpreted within a data base) rather than free-text information.
“Sixty to 70 percent of this project concerns understanding workflow in the clinics,” Johnson said.
As development of the program got underway last summer, Johnson conducted a series of focus groups with clinicians and made visits to teams of prospective users in various clinics. The advisory committee for development of the prescription writer includes physicians and clinic managers from across Vanderbilt Medical Group.
“If there is a net time savings for clinicians with this program, it will come in renewing patient prescriptions,” Johnson said. The nurse can send a previous electronic prescription to a doctor’s StarPanel work queue. Updating or refilling this prescription is accomplished by the doctor with a few clicks of the mouse. Around 60 percent of prescriptions are refills, Johnson said.
“The learning curve is fairly rapid, and most people would find that using the prescription writer is exceedingly fast,” said one tester of the program, cardiologist Waleed N. Irani, M.D., assistant professor of Medicine.
For practiced users, writing a new prescription will be a matter of typing in a string of terms to identify the drug, dosage, duration, and so on. Using a library of terms and clinical logic, the program takes the user’s more or less raw input and suggests a more complete and standard grammatical form for the prescription.
Less practiced users may instead choose to work through the program’s succession of pick lists to complete new prescriptions. The program knows, for example, what strengths a particular drug is sold in, and the common directions associated with the drug, and it offers these as canned options.
The program’s eventual links to insurance company formularies will help Vanderbilt doctors learn more quickly the important information about which drugs are covered by a patient’s insurance plan. Also, a database of local pharmacy fax numbers is in development; this will allow users to click from within the program to fax a prescription to the patient’s pharmacy.
Doctors at Vanderbilt have for several years written inpatient orders, including inpatient medication orders, exclusively with use of networked computers.