Patient Safety and Quality

Automated E-mail system helps identify adverse drug events

Sending automated E-mails to patients after they are prescribed new drugs may help detect and prevent adverse drug events, a new study finds. Saul N. Weingart, M.D., Ph.D., of Dana-Farber Cancer Institute, and colleagues examined actions 267 patients took after receiving 391 automated E-mail messages asking them about new medications they were prescribed from April 2001 to June 2002.

E-mails were sent 10 days after prescriptions were issued at three primary care practices, and patient responses were sent to the prescribing physician. Of the 267 patients, 128 responded to the initial E-mail. Seventy-seven percent of them opened the initial E-mail within a day of its being sent, and 13 percent sent responses. For example, they asked about the drug's effectiveness, drug-related side effects, and the dose. In return, 68 percent of physicians responded to the patients' E-mails, usually within a week.

Typical responses included asking questions, providing information, writing a new prescription, or changing the dose. During chart reviews, physicians identified 17 adverse drug events that were brought to light because of the E-mail exchanges. These ranged from drug-related constipation and diarrhea to nausea and dizziness.

The authors contend that the E-mail messages helped continue the dialogue between patient and clinician after a therapeutic intervention. They believe the system could be useful in following up with surgical patients and those with chronic diseases, as well as promoting healthy lifestyles. However, for this system to be effective, patients and providers must be timely in their responses to E-mail messages. This study was funded in part by the Agency for Healthcare Research and Quality (HS11644).

See "Medication safety messages for patients via the web portal: The MedCheck intervention," by Dr. Weingart, Hope E. Hamrick, Sc.M., B.A., Sharon Tutkus, R.N., B.S.N., and others in the March 2008 International Journal of Medical Informatics 77(3), pp. 161-168.

Return to Contents
Proceed to Next Article