Surgery, NSQIP, Complications & Death

The information in this column is intended forstay after surgery also declined by a rather incredible
informational purposes only, and does not constitute50 percent (from an average of 9 days, to 4 days).
medical advice or recommendations by the author.Needless to say, if the NSQIP results from these 132
Please consult with your physician before making anyVA hospitals were to be universally replicated in all of
lifestyle or medication changes, or if you have anythe nearly 6,000 hospitals in the United States, the
other concerns regarding your health.potential for improvement in patient morbidity and
As a practicing cancer surgeon, complicationsmortality, and consequently in the cost of delivering
following surgery (including death, which can behigh quality surgical care in the United States, would
considered the "ultimate" complication of surgery) arebe enormous (currently, only 243 civilian hospitals are
of great concern to me. As federal and statelisted as voluntary NSQIP participants by the
government officials continue to grapple with theprogram's steward, the American College of
tremendously complex and seemingly insolubleSurgeons).
deficiencies in our current health care system here inAs surgeons finally begin to move from their long
the United States, one area that has been receivingtradition of well-intentioned but seriously flawed
increased attention, lately, has been patientretrospective analysis of postoperative complications,
outcomes following surgery.and into the dawning era of prospective data
Traditionally, surgeons have reviewed and analyzedcollection and analysis, a detailed evaluation of the
their complications during regular morbidity anddata from programs such as NSQIP will continue to
mortality ("M&M") conferences. Each surgicalyield important new insights into the causes of
complication is presented by the operating surgeonpreventable surgical complications and deaths. Indeed,
during such conferences, and the surgeon, and his ora newly published study in the prestigious New
her peers, then analyze the patient's clinical course. AEngland Journal of Medicine, from the University of
consensus is then, hopefully, reached as to theMichigan, evaluated NSQIP data collected on 84,730
proximate causes of the patient's complications.surgical patients between 2005 and 2007. Unlike most
Therefore, the goal of M&M conferences is toprior studies that have used NSQIP data to assess
retrospectively identify patient risk factors forthe potential causes of preventable surgical
complications, as well as to examine the quality ofcomplications and mortality, the authors of this
care that patients have received, in an effort toresearch study were particularly interested in the role
identify areas where patient care can potentially beof patient management in preventing postoperative
improved, and complications, perhaps, prevented.deaths once complications had, in fact, already
Unfortunately, there are several obvious weaknessesoccurred.
associated with this approach to quality improvementThe authors of this important clinical study ranked
in patient care. The first and perhaps greatestthe NSQIP-participating hospitals according to their
weakness of M&M conferences is theiroverall death rates for patients undergoing inpatient
retrospective nature. A great deal of subjectivity issurgery, and divided them into five different groups,
injected into the analysis of specific patient carebased upon their mortality rankings. The researchers
factors when the operating surgeon reviews his orthen evaluated and compared the extensive NSQIP
her own patient complications, in hindsight, and thendata in each of these five groups of hospitals. The
presents selected patients to other surgeonsresults of this analysis were both intriguing and,
participating in the M&M conference. Whether byseemingly, rather counterintuitive.
accident or by intention, important lapses in diagnosisThe first important finding of this study was that the
and/or treatment are commonly withheld during suchactual incidence of complications following surgery did
conferences, which often leads to an incompletenot significantly vary among the hospitals studied.
picture of the events leading up to patientThe second illuminating finding of this study is that,
complications and deaths.unlike complication rates, death rates following
Another pitfall of M&M conferences is directlysurgery did significantly vary among these same
related to the interpersonal and professional dynamicshospitals (from 3.5 percent among the best
between surgeons participating in such conferences.performing hospitals, to 6.9 percent among the
Some medical centers' M&M conferences have apoorest performing hospitals).
justly earned reputation for being ruthlesslyA comprehensive review of the NSQIP data for
aggressive in holding individual surgeons accountablethese nearly 85,000 surgical patients confirmed that
for their complications, in an effort to improve thethe two-fold difference in death rates that was
quality of patient care. At the same time, in manyobserved between the best performing and worst
other medical centers, M&M conferences areperforming hospitals appeared to be directly related
relatively benign and quasi-social affairs among collegialto the way that patients with major complications
groups of surgeons, and a critical evaluation ofwere managed once the complications occurred, and
surgeons' care of their patients is, instead, substitutednot due to any underlying difference in the actual
with an affable and superficial review of patientincidence of complications among the various
complications and deaths that, too often, fails to drillhospitals. Thus, the authors concluded that our
down to specific potential patient care deficiencies.attention must not only continue to focus on
Both extremes in approaches to surgeon morbiditypreventing complications, but that we should also
and mortality conferences tend to obscure the truemore aggressively concentrate on our actual
causes and events associated with patientmanagement of complications following surgery, once
complications and unexpected deaths followingthey occur, in our ongoing efforts to reduce the
surgery, due to the many biases that are injectedincidence of preventable postoperative deaths.
into purely retrospective M&M conferences.In view of the landmark Veterans Administration
In an attempt to overcome the intrinsic biases andNSQIP findings linking the prevention of complications
limitations associated with a purely retrospectivewith a decrease in postoperative death rates, it may
assessment of surgical complications, most surgeonsseem counterintuitive that the two-fold difference in
at morbidity and mortality conferences present clinicalpostoperative death rates observed in this new
research studies, published in peer-reviewed medicalstudy appeared to be unrelated to the actual
journals, in an effort to inject some scientificincidence of complications among the hospitals
objectivity into the discussion. However, once again,studied. However, it is important to remember that
personal biases still often arise despite attempts tothese same hospitals had already previously
present published clinical data relevant to thedemonstrated their strong commitment to reducing
complication being discussed, as it is almost alwayspreventable postoperative complications through their
possible to find a couple of published papers thatvoluntary participation in the NSQIP program.
appear to support the decisions that were made byTherefore, the finding of this study that complication
the operating surgeon.rates did not vary considerably among these
In response to growing concerns regarding the qualityparticular hospitals may not be applicable to the vast
of surgical care at Veterans Administration hospitals inmajority of hospitals that have not yet adopted
the United States, a paradigm-shifting approach toNSQIP guidelines (or other comparable, prospective
the analysis of surgical complications was initiated bysurgical quality improvement programs).
the Veterans Administration (VA) in 1991. Out of anThe results of this study clearly show that, while the
abundance of concern over the high rate ofimplementation of the NSQIP program is helping
postoperative complications and deaths at severalparticipating hospitals to make significant progress in
VA medical centers, the National Surgical Qualityreducing the incidence of complications following
Improvement Program (NSQIP) was born. Betweensurgery, we still have much work to do in devising
1991 and 2001, VA medical centers prospectivelyand implementing evidence-based clinical pathways for
collected data encompassing multiple patient riskmanaging those complications that, despite our best
factors, as well as data related to 30-dayefforts, still continue to occur. Based upon the results
postoperative morbidity (complications) rates, andof this important clinical research study, it appears
30-day postoperative mortality (death) rates. Usingthat we can further and dramatically reduce the
this prospectively collected data to improve surgicalincidence of unnecessary deaths following surgery
care at all of the VA's 132 medical centers thatwhile, at the same time, significantly reducing the
perform surgery, the VA was able to subsequentlyalready excessive cost of delivering quality health
demonstrate some striking improvements in patientcare in the United States.
outcomes. Specifically, between 1991 and 2001,Disclaimer: As always, my advice to readers is to
30-day death rates following surgery decreased byseek the advice of your physician before making any
27 percent, while postoperative complicationssignificant changes in medications, diet, or level of
occurring within 30 days of surgery were decreasedphysical activity.
by a whopping 45 percent. The average length of