Despite widespread support for the use of advance medical directives, including their encouragement by federal law in the form of the Patient Self-Determination Act of 1990, important questions remain regarding their effectiveness.

The ADVANCE (Advance Directives, Values Assessment, aNd Communication Enhancement) Project was initiated by Peter Ditto, Ph.D., Joseph Danks, Ph.D., and William Smucker, M. D. to address three key questions underlying the effective use of advance medical directives:

1. Do advance medical directives effectively communicate the life-sustaining treatment preferences of incapacitated patients to family members and physicians?
2. Are decisions about the use of life-sustaining treatments stable over time and across changes in health status such that preferences recorded in advance of medical problems will continue to reflect individuals' wishes at the time actual treatment decisions must be made?
3. Are the decisions that healthy individuals make with respect to hypothetical medical scenarios the same decisions that those individuals would make if they were to actually experience the medical scenario?

The ADVANCE Project was a three-phase study involving 400 elderly adults and their self-designated surrogate decision makers for life-support decisions (mostly spouses and children, mean length of relationship with patient -- over 45 years). The goals of each phase were:

• Phase 1: Used a randomized controlled trial to compare the ability of four different methods of collecting advance directive information to improve the accuracy of substituted judgment (measured as the agreement between patients' preferences and surrogates' predictions for life-sustaining treatment in several hypothetical medical scenarios).
• Phase 2: Followed patients and surrogates longitudinally for two years to measure the stability of life-sustaining treatment decisions over time and across changes in health state.
• Phase 3: Interviewed hospitalized study patients and their surrogates to compare treatment preferences and predictions made prior to hospitalization to those made during actual illness episodes.

Phase 1 of the ADVANCE project was completed in May 1997. Phase 2 and Phase 3 data collection were completed in June 1999.

The project members have made several presentations of this data at regional and national conferences. An overview of the project, findings and implications was presented in an invited address by Peter Ditto at the annual meeting of the American Psychological Association in Boston in August 1999.