Skip to content

Today is National Healthcare Decisions Day. The day was created as “an initiative to encourage patients to express their wishes regarding healthcare and for providers and facilities to respect those wishes, whatever they may be.” Much of what will be going on today at various outreach programs will be focused on the first part of this initiative: educating the public on the importance of advance care planning.  I’d like to take a second though and pose the following question to our audience:

Should the failure to follow end-of-life preferences be considered a medical error?

This question is derived from a recent JAMA Internal Medicine editorial by GeriPal contributors Theresa Allison and Rebecca Sudore.  In it they make a persuasive argument that the disregard of patients’ preferences is indeed a medical error.  Here is an excerpt:

“Discussions about goals of care and code status constitute a medical procedure every bit as important to patient safety as a central line placement or a surgical procedure. Much as we have developed systems to improve patient safety in surgical procedures, we need to develop systematic approaches to discussing patient values and goals of care.”

The Importance of this Question on National Healthcare Decisions Day

To understand the importance of this question, one needs to only turn to the study that accompanies the editorial by Allison and Sudore.  Thestudy, conducted by Heyland and colleagues, involving 278 patients and 255 family members in 12 Canadian hospitals.  To be included, patients needed to be age 80 years or older and have either advanced pulmonary, cardiac, or liver disease, or metastatic cancer, or be to be someone whose death within the next 6 months would not be a surprise to any member of his or her care team.

Of these elderly patients who were at high risk of dying in the next 6 months, 76% reported that they thought about what kind of life-sustaining
treatments they would or wouldn’t want prior to hospitalization.  And what did they want?

  • Few (12%) wanted aggressive use of heroic measures and artificial life-sustaining treatments, including CPR with a goal of life prolongation
  • A little more (18%) wanted “Full medical care” but no CPR in the event their heart or breathing stopped
  • Most (31%) would want their care either focused on 1) comfort if their health deteriorated including avoiding CPR and other life-sustaining technologies, 2) trying to fix problems but if not getting better, switch to focusing comfort even if it hastens death

The good news is that most individuals discussed their preferences with someone (88%).  The bad news is that most patients just talked to their family members about these wishes.  Only 30% had talked to their family physician, and only about half (55%) had talked to any member of their health care team.

The other good news is that nearly half of patients (48%) reported having completed
a written advance care plan, and 73.3% had documented who they would like to be their surrogate
decision maker.  The other bad news though is that only one in four (25%) of these patients were asked about these prior discussions or written documents on admission to the
hospital.

The most unfortunate finding of this study was that more than two-thirds of patients’ preferences were either not documented at all or documented incorrectly in the medical record.  When the authors of the study looked at the 199 patients who expressed a preference for care and for whom a written goals-of-care order was present in the patient’s records, they found:

  • only a 30% agreement between patients’ expressed preferences and the documented goals-of-care order in the medical record
  • 28.1% of patients (56 of 199) preferred comfort
    measures only, but this preference was documented
    in only 4.5% (9 of 199) of stated goals (Figure 2A).

So, the question I have for you on this National Healthcare Decisions Day is whether you agree with Drs Allison and Sudore that this failure to document end-of-life preferences should be considered a medical error?

by: Eric Widera (@ewidera)

Back To Top
Search