Abstract
The article by Chang et al in this issue (“Evaluating Metrics for Quality: Death on the Same Day of Elective Pediatric Surgery”) is a provocative work on many fronts. First, it demonstrates the growing use of large (>800 000 pediatric admissions), easily accessible administrative databases with the goal to identify and hopefully spark improved clinical quality. Second, the study provides an incidence of a clinical event that has not yet received a significant amount of national focus despite representing a parent’s nightmare: their child dying after being handed over to a care team for “elective” surgery. Finally, it brings up the provocative question of what to do with the data.
With the simultaneous surge of electronic medical records, large databases are becoming commonplace. These databases are powerful repositories of information ripe to be turned into knowledge. The authors must be complimented as they do not fall into the trap of assigning causality when the temptation will always be there and abused by others. This study should provoke curiosity and not judgment. Based on the report, we cannot tell the exact cause of death; location of the deaths; whether there was an accurate assignment of “elective”; whether a checklist could have made a difference; whether there were proper equipment, medications, communication, and handoffs of care among team members; whether the anesthesiologist and surgeon were the right choices for the case; or, given the medical complexity of the child, whether the risk of surgery was indeed high but has significantly improved. Addressing this last point, the data as presented do tend to indicate that children younger than 1 year of age have a rate of death greater than 6 times higher than average and that patients who undergo transplantation or congenital heart disease surgery also demonstrate the highest rates of mortality following elective surgery. Fortunately, specialty societies are beginning to capture the specialty-specific variables in additional databases that hopefully will be used to determine those variables and processes that can be employed to determine practice based on evidence rather than eminence. In the meantime, are our children who have elective surgery safe enough?
I disagree with the authors on 1 key assertion. In the discussion they state, “because of its infrequent occurrence, death on the day of elective surgery is of limited use as a measure of direct comparison across institutions or providers.” Although the overall incidence of death is 2.1/10 000 cases and seems low (0.06-17.4 deaths per 10 000 cases), if one puts this into Six Sigma terms, this range, taking into consideration age and diagnosis, ranges from 4.2 to 6.0 sigma. And this is only on the day of surgery! It will only get worse as one reports 30-day postoperative mortality. To put it in relative terms, this reported incidence range is slightly better than an error in baggage handling but far less than the reported mortality rate from airline crashes annually.
I do not want to advocate the comparison of apples to oranges, but this study does make me wonder whether we need to devote more study at a quicker pace in order to learn from and identify opportunities for system improvements every time a child dies after elective surgery. The airline industry, under the due diligence of the Federal Aviation Agency, thoroughly investigates each crash, and as a result, Six Sigma is now the annual benchmark for the airline industry. We often do not share the analysis and lessons learned, but now is the time for greater cross-institutional transparency. The alternative, in this case, may be lives needlessly lost.
Efforts are under way to foster the open investigation and shared learnings in the format of a “Patient Safety Organization.” Campaigns similar to those sponsored by organizations such as the Institute for Healthcare Improvement have helped champion efforts to reduce central line infections and other causes of mortality, and they should be applauded; however, they work after a best practice is discovered and then readily disseminated. There is another format that attempts to foster change but is more financially punitive than educational: classifying something as a “never event.” A never event is a clinical occurrence that represents a clinical situation that should never have occurred if all the correct practices had been adopted. As a result of a never event, a hospital’s reimbursement is reduced to a level comparable to if that event had not taken place. Whether a never event process actually drives or ever will drive system change is yet to be clearly evidenced.
I hope the Chang et al article provides a bit of a wake-up call regarding not a specific case and incident but the more global issue of pediatric mortality after elective surgery. Although not as directed a topic as the drive to lower central line infection rates and urinary tract infections, in terms of long-term impact, a reduction of this rate could result in many more years of valued life.