Current issues of ACP Journal Club are published in Annals of Internal Medicine

Quality Improvement

Computer-generated reminders increased the use of advance directives

ACP J Club. 1998 Jul-Aug;129:24. doi:10.7326/ACPJC-1998-129-1-024

Source Citation

Dexter PR, Wolinsky FD, Gramel-spacher GP, et al. Effectiveness of computer-generated reminders for increasing discussions about advance directives and completion of advance directive forms: a randomized controlled trial. Ann Intern Med. 1998 Jan 15;128:102-10.



For elderly persons and patients with chronic conditions, can computer-generated reminders given to physicians during clinic visits increase the frequency of discussions and establishment of advance directives?


Randomized controlled trial.


A primary care practice in a U.S. urban public teaching hospital.


1009 patients (mean age 65 y, 55% African American, 66% women) who were at high risk for deterioration because of advanced age (≥ 75 y) or were ≥ 50 years old and had chronic conditions (cardiac ischemia, heart failure, chronic lung disease, cancer other than nonmelanomatous skin cancer, cerebrovascular disease, renal insufficiency, or cirrhosis). Patients were excluded if they lived in a nursing home or a prison, had cognitive dysfunction, could not hear, did not speak English, or had completed advanced directives.


During scheduled clinic visits, all physicians received computer-generated reminders about giving preventive care, noting abnormal test results, and avoiding potential adverse drug interactions. During the 32 weekly half-day sessions, physicians were allocated to usual care (no additional reminders) (253 patients, 49 physicians, 603 visits), computer-generated reminders for instruction directives (219 patients, 30 physicians, 575 visits), reminders for proxy directives (260 patients, 31 physicians, 650 visits), or reminders for instruction and proxy directives (277 patients, 37 physicians, 663 visits).

Main outcome measures

Discussion of advance directives assessed at the end of the clinic visit and completion of an advance directive assessed at subsequent clinic visits or at the end of the study.

Main results

Odds ratios (ORs) were calculated for patients whose physicians received reminders compared with patients whose physicians received no reminders. Compared with the no-reminder group, the other 3 groups had more discussions about advance directives (4% for no reminders; 8% for proxy reminders [OR 2.5, 95% CI 1.1 to 5.5]; 14% for instruction reminders [OR 4.4, CI 2.1 to 9.4]; and 24% for both reminders [OR 7.7, CI 3.4 to 18]). Patients whose physicians received instruction or proxy reminders or both sets of reminders were more likely to have completed 1 or both advance directives (OR 3.0, CI 1.1 to 8.0 and OR 7.0, CI 2.9 to 17, respectively).


Computer-generated reminders given to physicians increased the frequency of discussion and establishment of advance directives in elderly persons and patients with chronic conditions.

Source of funding: Agency for Health Care Policy and Research.

For correspondence: Dr. W.M. Tierney, Regenstrief Institute for Health Care, 1001 West 10th Street, Indianapolis, IN 46202, USA. FAX 317-630-6962.


How can we ensure that outpatients have the opportunity to discuss their preferences for end-of-life care with their physicians? Building on their experience of using computer-generated prompts to improve clinical care, Dexter and colleagues show that prompts located on the front of the chart resulted in a higher frequency of outcomes of interest: Patients and their physicians discussed end-of-life preferences, and patients completed an advance directive or selected a proxy. This intervention was successful despite the disparate needs of a population that was ill and receiving care in the typical turmoil of a busy primary care practice. As the use of sophisticated information systems in mainstream medical practice grows, physicians can expect that these systems will not only enable easy access to patient information but will also generate attempts to influence physician behavior at many levels.

Although physicians received advance-directive prompts for every patient in the intervention groups, most patients (76% to 92%) completed the study without expressing a preference about end-of-life care. Other than age, the presence of lung disease, and a preference for aggressive care, the authors were unable to find factors that predicted which patients were most likely to complete this type of intervention. Also, little information exists on the physician characteristics that could predict whether these propts will result in discussions with patients. Finally, we do not know the most important outcome of all: whether the study patients who successfully completed an advance directive and discussed them with their physician will have their preferences honored.

Other studies have shown high levels of patient and physician interest in advance directives (1), but it is likely that physician reminders will need to be combined with other interventions to ensure that patients have every opportunity to successfully plan for end-of-life care before, during, and after their outpatient visits.

Jaan Sidorov, MD
Penn State Geisinger Health SystemDanville, Pennsylvania, USA


1. Miles SH, Koepp R, Weber EP. Advance end-of-life treatment planning. A research review. Arch Intern Med. 1996;156:1062-8.