Enhancing Doctor-Patient Communication Using Email: A Pilot Study

Shou Ling Leong, MD; Dennis Gingrich, MD; Peter R. Lewis, MD; David T. Mauger, PhD; John H. George, PhD


J Am Board Fam Med. 2005;18(3):180-188. 

In This Article


This Institutional Review Board-approved study was conducted from December 2001 to July 2002 in a faculty-based clinic affiliated with The Penn State Milton S. Hershey Medical Center. The majority of the practice's patients come from middle- to upper-income communities with high employment rates. Almost all patients have health insurance, including 24% Medicare and 3% Medical Assistance. Only 3% of the patient population have no insurance. Eight family physicians agreed to participate in the study. The investigator assigned the physicians into one of 2 groups: the intervention group, which offered e-mail communication to patients, and the control group, which did not. This assignment was used to achieve a comparable profile for the 2 groups in terms of physician age, sex, and attitude toward offering e-mail access to patients, as well as age of patient population ( Table 1 ).

The inclusion criteria for patients were regular use of e-mail and an existing doctor-patient relationship with one of the participating physicians. During the enrollment period, the office nurses screened all patients who came in for a scheduled appointment with 1 of the 8 participating physicians. The screening consisted of the nurses' asking whether the patient used e-mail regularly. Patients who said yes were then asked whether they were regular patients of the participating physicians. The nurses also review the patients' medical records to confirm that they have seen the physician at least 3 times in the past. Those patients identified as regular patients of the participating physicians and that use e-mail regularly were offered an opportunity to participate in the study. Consented patients of doctors in the intervention group were given their physician's e-mail address for communication. Patients of physicians in the control group were not given e-mail access.

A separate e-mail account was established for each of the 4 physicians in the e-mail group following pre-existing guidelines for doctor-patient e-mail at the Hershey Medical Center. These guidelines were adopted from a white paper by Kane on "Guidelines for the Clinical Use of Electronic Mail with Patients."[23] Patients were advised not to use e-mail for emergency or urgent matters that could not wait 3 to 5 days, or issues they felt were sensitive/confidential. An auto-reply feature informed patients of the receipt of their message or if the physician was away for an extended period of time. The e-mail system used for this study was set up by the information technology department specifically for this study and was established within a secure server. Patients in both the e-mail and control group were referred to the patient education web site, http://www.familydoctor.org , created by the American Academy of Family Physicians for health and drug information.

During office hours, clinic secretaries recorded all phone calls from patients to the e-mail and control group physicians on standard message slips. Using these message slips, the physicians recorded their responses to patients, which were later filed in the patients' records. For data analysis purposes, a copy of all patient messages was collected over a week at baseline and 3 selected weeks during the 6-month study period for data analysis. Phone messages received after normal hours were not tracked. E-mail messages were automatically recorded electronically, and copies of the initial message, physician reply, and subsequent communications (if any) were filed into the patients' charts. During the 3 selected weeks, all physicians completed a time log and recorded the number of messages received and time spent in answering patient messages each day. All participating patients and physicians completed satisfaction questionnaires at the onset of the study and at the 3-month and 6-month intervals.

On the patient questionnaire, participants were asked their age, sex, number of years in the practice, level of education, and comfort and frequency of e-mail usage. Patients were also asked whether they used the Internet for various purposes, including locating health information. In part 2 of the patient questionnaire, patients were asked to rank on a 5-point Likert scale their satisfaction regarding ease and convenience of communication with their physician, promptness of the reply, quality and amount of information in the reply, and the quality of home care instruction they received. They also were asked to rank the health information web site in terms of usefulness and frequency of use. On the physician questionnaire, the physicians were asked similar questions, except that in part 2, they were asked to rank their satisfaction with the practice's message system overall and the amount of time to reply to messages, their promptness in responding to messages, volume of messages, convenience of the system, quality of the messages, and the amount of patient education offered. Both patients and physicians were asked whether patients should be able to e-mail their physicians and whether e-mail should be used for prescription refills and scheduling appointments. Participants in both the e-mail and control groups completed their respective patient or physician questionnaires.

Patient and physician responses to questionnaires were summarized by frequency tables for categorical responses and descriptive statistics for continuous responses. Changes in the level of patient satisfaction across time were assessed within each group using the Friedman test for ordinal responses and repeated measures analysis of variance for continuous responses. Differences between the study and control groups with respect to patient satisfaction were assessed using analysis of variance controlling for physician effects. Given the few physicians in the study, changes in the level of physician satisfaction across time were not statistically significant. However, these data provided some general observations on physician response to e-mail communication with patients. Statistical analyses for other outcome measures were descriptive in nature. These data included the number of messages received, time physicians spent on answering messages, types (categories) of messages received, and response time (time lapse between when a message was received and when the physician replied). The response time and categories of messages were generated for both e-mail and phone messages. All analyses were conducted using SAS statistical software version 8.0 (SAS Institute, Cary, NC).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: