skip to Main Content

by: Gretchen Schwarze, @GretchenSchwa10

My husband, also a surgeon, was recently discussing donor-nephrectomy with a
healthy 65 year-old woman in his transplant clinic.  He told her the 30-day survival for this
operation was excellent.  She was shocked
he would quote such metrics, “I don’t want to live for just 30 days!”  Yet for years, we in surgery have judged our
successes by the number of patients who don’t die within 30 days of surgery.

I know it is easy to find examples of surgeons behaving
badly over a postoperative patient whose death seems to threaten his “numbers,”
but as a surgeon who has been there I can tell you that it is viscerally
painful to watch a patient you operated on succumb to postoperative
complications.  Deep notions of error and
responsibility are ingrained in our culture and are reinforced in many ways
including M&M conferences and NSQIP (National Safety Quality Improvement
Program) reports. While I’m all for improving safety and quality, I shudder to
think about the consequences of 30-day mortality reporting for our older
patients when increasingly our successes and failures are measured (and
reported) by who is alive and who is dead.

Here’s an example of why I am worried.  One newer measure of safety in surgery is
“failure to rescue.” (FTR)  Studies show
complication rates for major operations are actually pretty constant across
hospitals with varying levels of quality however; mortality differences between
low and high performers come from the ability to “rescue” patients with
postoperative complications, i.e. mortality differences derive from whether the patient is able to survive the complication.  

This measure
morphed precariously in a (very well done) study that matched the health status
of surgical patients with DNR orders to those without and demonstrated a
substantially higher postoperative mortality for patients with DNR orders. The
authors called this “failure to pursue rescue.” Sheesh,
if you want to get a surgeon to do something, tell him that not doing it would
constitute failure – this is the language we speak! If anything, the mortality
difference demonstrates a success in honoring patient preferences but the
characterization of death as the worst possible outcome doesn’t recognize this
important difference.

We are told that these measures are risk-adjusted but risk adjustment
doesn’t adequately manage the needs of patients who would benefit from
palliative operations or patients who are high-risk and would authentically prefer
to take their chances in the operating room.
Risk adjustment is probably not the answer, stratification accounting for
palliative operations and robust measures of patient engagement would be more
aligned with the goals of patients and their families who likely don’t value
“just living to 30 days.”

The National Quality Forum (NQF) is proposing 30-day risk
adjusted mortality following CABG (heart bypass surgery) as a quality metric. This measure is up for appeal until December 12.   While
the Society of Thoracic Surgeons has used this metric for years for their
internal database, formalization of this metric for public reporting threatens
to exacerbate its unintended consequences.  

I recently ran a “town hall meeting” at the American College of Surgeons
annual Clinical Congress about this issue which involved some pretty
treacherous terrain: surgeons were clear that mortality measures are
increasingly influencing both their pre and postoperative surgical decisions.  

The 30-day metric is a game-able metric that harms patients and
families and the surgeon patient relationship.  It also fails to capture important
safety information such as the patient who has a straight-forward postoperative
course is scored equivalently to one who has a prolonged ICU stay and is
transferred to palliative care on postoperative day 31.  I hope the National Quality Forum will reconsider their decision
and focus on measures that are more aligned with patient-reported goals.

Back To Top