SURGERY, NSQIP, COMPLICATIONS & DEATH
As a practicing cancer surgeon, complications following surgery (including death, which can be considered the “ultimate” complication of surgery) are of great concern to me. As federal and state government officials continue to grapple with the tremendously complex and seemingly insoluble deficiencies in our current health care system here in the United States, one area that has been receiving increased attention, lately, has been patient outcomes following surgery.
Traditionally, surgeons have reviewed and analyzed their complications during regular morbidity and mortality (“M&M”) conferences. Each surgical complication is presented by the operating surgeon during such conferences, and the surgeon, and his or her peers, then analyze the patient’s clinical course. A consensus is then, hopefully, reached as to the proximate causes of the patient’s complications. Therefore, the goal of M&M conferences is to retrospectively identify patient risk factors for complications, as well as to examine the quality of care that patients have received, in an effort to identify areas where patient care can potentially be improved, and complications, perhaps, prevented.
Unfortunately, there are several obvious weaknesses associated with this approach to quality improvement in patient care. The first and perhaps greatest weakness of M&M conferences is their retrospective nature. A great deal of subjectivity is injected into the analysis of specific patient care factors when the operating surgeon reviews his or her own patient complications, in hindsight, and then presents selected patients to other surgeons participating in the M&M conference. Whether by accident or by intention, important lapses in diagnosis and/or treatment are commonly withheld during such conferences, which often leads to an incomplete picture of the events leading up to patient complications and deaths.
Another pitfall of M&M conferences is directly related to the interpersonal and professional dynamics between surgeons participating in such conferences. Some medical centers’ M&M conferences have a justly earned reputation for being ruthlessly aggressive in holding individual surgeons accountable for their complications, in an effort to improve the quality of patient care. At the same time, in many other medical centers, M&M conferences are relatively benign and quasi-social affairs among collegial groups of surgeons, and a critical evaluation of surgeons’ care of their patients is, instead, substituted with an affable and superficial review of patient complications and deaths that, too often, fails to drill down to specific potential patient care deficiencies. Both extremes in approaches to surgeon morbidity and mortality conferences tend to obscure the true causes and events associated with patient complications and unexpected deaths following surgery, due to the many biases that are injected into purely retrospective M&M conferences.
In an attempt to overcome the intrinsic biases and limitations associated with a purely retrospective assessment of surgical complications, most surgeons at morbidity and mortality conferences present clinical research studies, published in peer-reviewed medical journals, in an effort to inject some scientific objectivity into the discussion. However, once again, personal biases still often arise despite attempts to present published clinical data relevant to the complication being discussed, as it is almost always possible to find a couple of published papers that appear to support the decisions that were made by the operating surgeon.
In response to growing concerns regarding the quality of surgical care at Veterans Administration hospitals in the United States, a paradigm-shifting approach to the analysis of surgical complications was initiated by the Veterans Administration (VA) in 1991. Out of an abundance of concern over the high rate of postoperative complications and deaths at several VA medical centers, the National Surgical Quality Improvement Program (NSQIP) was born. Between 1991 and 2001, VA medical centers prospectively collected data encompassing multiple patient risk factors, as well as data related to 30-day postoperative morbidity (complications) rates, and 30-day postoperative mortality (death) rates. Using this prospectively collected data to improve surgical care at all of the VA’s 132 medical centers that perform surgery, the VA was able to subsequently demonstrate some striking improvements in patient outcomes. Specifically, between 1991 and 2001, 30-day death rates following surgery decreased by 27 percent, while postoperative complications occurring within 30 days of surgery were decreased by a whopping 45 percent. The average length of stay after surgery also declined by a rather incredible 50 percent (from an average of 9 days, to 4 days). Needless to say, if the NSQIP results from these 132 VA hospitals were to be universally replicated in all of the nearly 6,000 hospitals in the United States, the potential for improvement in patient morbidity and mortality, and consequently in the cost of delivering high quality surgical care in the United States, would be enormous (currently, only 243 civilian hospitals are listed as voluntary NSQIP participants by the program’s steward, the American College of Surgeons).
As surgeons finally begin to move from their long tradition of well-intentioned but seriously flawed retrospective analysis of postoperative complications, and into the dawning era of prospective data collection and analysis, a detailed evaluation of the data from programs such as NSQIP will continue to yield important new insights into the causes of preventable surgical complications and deaths. Indeed, a newly published study in the prestigious New England Journal of Medicine, from the University of Michigan, evaluated NSQIP data collected on 84,730 surgical patients between 2005 and 2007. Unlike most prior studies that have used NSQIP data to assess the potential causes of preventable surgical complications and mortality, the authors of this research study were particularly interested in the role of patient management in preventing postoperative deaths once complications had, in fact, already occurred.
The authors of this important clinical study ranked the NSQIP-participating hospitals according to their overall death rates for patients undergoing inpatient surgery, and divided them into five different groups, based upon their mortality rankings. The researchers then evaluated and compared the extensive NSQIP data in each of these five groups of hospitals. The results of this analysis were both intriguing and, seemingly, rather counterintuitive.
The first important finding of this study was that the actual incidence of complications following surgery did not significantly vary among the hospitals studied. The second illuminating finding of this study is that, unlike complication rates, death rates following surgery did significantly vary among these same hospitals (from 3.5 percent among the best performing hospitals, to 6.9 percent among the poorest performing hospitals).
A comprehensive review of the NSQIP data for these nearly 85,000 surgical patients confirmed that the two-fold difference in death rates that was observed between the best performing and worst performing hospitals appeared to be directly related to the way that patients with major complications were managed once the complications occurred, and not due to any underlying difference in the actual incidence of complications among the various hospitals. Thus, the authors concluded that our attention must not only continue to focus on preventing complications, but that we should also more aggressively concentrate on our actual management of complications following surgery, once they occur, in our ongoing efforts to reduce the incidence of preventable postoperative deaths.
In view of the landmark Veterans Administration NSQIP findings linking the prevention of complications with a decrease in postoperative death rates, it may seem counterintuitive that the two-fold difference in postoperative death rates observed in this new study appeared to be unrelated to the actual incidence of complications among the hospitals studied. However, it is important to remember that these same hospitals had already previously demonstrated their strong commitment to reducing preventable postoperative complications through their voluntary participation in the NSQIP program. Therefore, the finding of this study that complication rates did not vary considerably among these particular hospitals may not be applicable to the vast majority of hospitals that have not yet adopted NSQIP guidelines (or other comparable, prospective surgical quality improvement programs).
The results of this study clearly show that, while the implementation of the NSQIP program is helping participating hospitals to make significant progress in reducing the incidence of complications following surgery, we still have much work to do in devising and implementing evidence-based clinical pathways for managing those complications that, despite our best efforts, still continue to occur. Based upon the results of this important clinical research study, it appears that we can further and dramatically reduce the incidence of unnecessary deaths following surgery while, at the same time, significantly reducing the already excessive cost of delivering quality health care in the United States.
Disclaimer: As always, my advice to readers is to seek the advice of your physician before making any significant changes in medications, diet, or level of physical activity
is an oncologic surgeon, a professor
of surgery, a widely published author, and the
Physician-in-Chief for Surgical
Oncology at the Kaiser Permanente healthcare system in Orange County,
(Anticipated Publication Date: March 2010)
(Click above image for TV36 interview of Dr. Wascher)
Copyright 2007 - 2009
Robert A. Wascher, MD, FACS
All rights reserved