Purpose: Wide geographic variations in spending for Medicare beneficiaries in the last 6 months of life don't appear to reflect differences in either access to or quality of end-of-life (EOL) care. These variations are warranted if they reflect genuine variations in patients' informed preferences for elective EOL interventions. The study purpose was to determine whether there are within- or across-region differences in the preferences for EOL care reported by Medicare beneficiaries residing in regions with known variations in Medicare spending.

Methods: Medicare beneficiaries in 3 regions (Minneapolis MN, Rochester NY, Miami FL) were randomly selected from the sampling frame of a larger, national telephone survey that had addressed different research purposes. Participants were interviewed in person by trained Community Interviewers using a standardized interview schedule. Respondents considered 5 scenarios involving a choice between a relatively lower- or higher-intensity elective intervention in EOL care: A. own home vs. nursing home vs. hospital as care site; B. no drugs vs. life-prolonging drugs that lower quality of life; C. no drugs vs. quality-of-life enhancing drugs that reduce length of life; D. no respirator vs. respirator to extend life by 1 month; and E. no respirator vs. respirator to extend life by 1 week. For each scenario, participants categorically indicated if they'd prefer the relatively lower- or higher-intensity intervention.

Results: 585 Medicare beneficiaries were selected, 366 agreed to be contacted, and 179 participated in the in-person interview (participation rate among those contacted = 49%). Within each geographic region, the majority reported that they would choose: their own home for EOL care (80%) ; no life-prolonging drugs that lower quality of life (91%); and no respirator to extend life by either 1 month (84%) or 1 week (88%). However, the majority reported that they would choose quality-of-life enhancing drugs that reduce length of life (77%). Chi-square tests detected no significant across-region differences in the distributions of these preferences.

Conclusions: The majority of Medicare beneficiaries may prefer lower-intensity EOL interventions, whether they reside in regions with lower (Minneapolis MN, Rochester NY) or higher (Miami FL) spending for EOL care. This, in turn, implies that – at least in regions with higher Medicare spending – beneficiaries may not actually be getting the care they want in their last 6 months of life.