End of Life: A Primary Care “Road Less Traveled”

In many instances, our patients receive invasive and aggressive subspecialty care at or near the end of life. Such care can become burdensome and painful without a favorable outcome. As advocates for our patients, we must ensure that their pain, shortness of breath, and other bothersome symptoms are alleviated, thus leading to a "good death." We also need to be sensitive to these patients' preferences for end of life care.

A recent study can teach us both about quality indicators in this specific cohort and about how we can do better for this growing and vulnerable group.1 Some of the measures studied were a result of planning by primary care providers long before patients reached the ICU.

ASSESSMENT OF END OF LIFE CARE

Sixteen quality indicators for end of life care were assessed in an appropriate target population; more than 50% of the group experienced end-stage disease and 28% were aged 75 years and older. Some of the indicators received very low scores for compliance. For example, if a "vulnerable elder" was admitted to the ICU and survived 48 hours, that person's preferences for further care should be considered or an attempt should be made to identify them. Unfortunately, this important indicator was performed only 46% of the time.

Another indicator, specifically for persons who were not expected to survive and who had had ventilator support withdrawn, stated that the patient should be assessed for dyspnea and, if present, symptomatically treated with sedation. Assessment for dyspnea after ventilator withdrawal occurred in only 29% of those studied. Furthermore, for patients who had an implantable defibrillator and who were terminally ill or enrolled in hospice, it should be documented in the medical record that the defibrillator had been inactivated or the reasons why this had not been done should be explained. This activity was performed 25% of the time in the study population.

Certain important indicators had high response rates. Appropriate documentation before placement of feeding tubes took place 100% of the time. Advance directives for "do not resuscitate" status were known and followed in 96% of the occurrences in the study.

PLANNING FOR THE FUTURE

This study had some weaknesses. Although it was of sufficient size (N = 496), it was performed at 1 center and was retrospective. However, it is a good "snapshot" of where we do well and not so well in end of life care in those patients who frequent the ICU.

Since many of our patients spend time in critical care units near the end of their lives, we as their primary caregivers—oftentimes also with the longest relationship— should be comfortable with the quality indicators used in this study. It would be worthwhile for each of us to determine how we score in their implementation. Our patients will come to expect oversight of end of life quality from us.