herfindahl measures of surgical specialization correlate with improved mortality rates based on...
TRANSCRIPT
![Page 1: Herfindahl measures of surgical specialization correlate with improved mortality rates based on NSQIP data](https://reader031.vdocuments.net/reader031/viewer/2022020604/575073931a28abdd2e90238c/html5/thumbnails/1.jpg)
TlPBC
IAapv
MldcD(pOawf
R238wtwai
Cn
HcNEBW
Isctrvdbvec
MStHclwpc
Riuppl
Csiif
RiNU
Ilailt
Mqftp(rrv
Rcegoi(f
S75Vol. 207, No. 3S, September 2008 Surgical Forum Abstracts
he number of surgical procedures in an Americanifetime in 3 stateseter HU Lee MD, MPH, MS, Atul A Gawande MD, MPH, FACSrigham and Women’s Hosptial, Caritas St Elizabeth’s Medicalenter, Boston, MA
NTRODUCTION: The total number of operations the averagemerican has in a lifetime is not known. Such an estimate of the typend number of surgical procedures performed in the U.S. wouldrovide a snapshot of the state of current surgical practice and pro-ide a reference for following past and future trends.
ETHODS: The number of surgical procedures in an 85-yearifespan was estimated by summing the rates of surgical proce-ures for each year of age between 0 and 84. These rates werealculated for CO, FL, and NJ, based on the 2002 State Inpatientatabase (SID) and the State Ambulatory Surgery Database
SASD) of the Health Care Utilization Project (HCUP). Surgicalrocedures were categorized as inpatient, outpatient, or non-R. The ten most common surgical procedures for both men
nd women, as well as total rates for various surgical specialtiesere determined. Rates for 2002 and 1997 were also compared
or NJ.
ESULTS: Based on the state of medical and surgical practice in002, the average American has 3.41 inpatient, 2.56 outpatient, and.20 non-OR, for an overall total of 9.17 surgical procedures in an5-year lifespan. The most common OR procedures are PTCA,ound debridements, and groin hernias for men, and cesarean sec-
ions, cholecystectomies, and lens and cataract procedures foromen. Compared with 1997, there was a 4.5% decrease in OR but28.5% increase in non-OR procedures, for an overall 4.9% increase
n any procedures.
ONCLUSIONS: The average American has 5.97 OR and 3.20on-OR procedures in an 85-year lifetime.
erfindahl measures of surgical specializationorrelate with improved mortality rates based onSQIP datasther Lee BS, Barton H Hamilton PhD,ruce L Hall MD, PhD, MBAashington University in Saint Louis, Saint Louis, MO
NTRODUCTION: There is great interest in the degree to whichurgical specialization affects outcomes, particularly given theurrent drive to measure and reward quality in healthcare. Al-hough surgical specialization has been previously analyzed withespect to outcomes, most studies have treated it as a dichotomousariable based on academic credentials, which confounds proce-ural concentration with case volume and skill sets impartedy additional training. We treat specialization as a continuousariable defined quantitatively by procedural diversity, in anffort to isolate the procedural concentration component of spe-
ialization. sETHODS: We used 2002-2005 patient data from the Nationalurgical Quality Improvement Program for Barnes Jewish Hospi-al in Saint Louis, MO. To quantitate procedural specialization,erfindahl indices for each surgeon were calculated using billing
odes. These indices were calculated according to three differentevels of procedural aggregation. Using conditional logit models,e examined the relationships between these indices and 30-dayostoperative mortality rates, adjusting for case volume, emergen-ies, and other characteristics of surgeons and patients.
ESULTS: Surgeon specialization was inversely related to mortal-ty rates after adjusting for case volume, when indices were calculatedsing the intermediate (Odds Ratio 0.539; 95%ile CI: 0.315,0.921;�0.024) or low (OR 0.471; CI: 0.257,0.864; p�0.015) levels ofrocedural aggregation. No relationship was observed at the high
evel of aggregation.
ONCLUSIONS: The procedural concentration component ofurgical specialization is correlated with improved mortality ratesndependently of case volume. Thus, specialization is associated withmproved mortality. How broadly or narrowly ‘specialization’ is de-ined has an impact on this relationship.
egionalization in laparoscopic cholecystectomy: Ist necessary?icholas G Csikesz BSc, Shimul A Shah MDniversity of Massachusetts, Worcester, MA
NTRODUCTION: Numerous reports in the 1990s pointed to aearning curve of laparoscopic cholecystectomy (LC) critical inchieving acceptable outcomes defined by open conversion, bile ductnjury and mortality. As LC is now standard therapy for acute cho-ecystitis (AC), we aimed to determine if surgeon volume is still vitalo patient outcomes.
ETHODS: The Nationwide Inpatient Sample (NIS) was used touery 83,612 emergent/urgent cholecystectomies performed for ACrom 1999-2005 for 12 states with available surgeon/hospital iden-ifiers. Surgeon volume was divided into tertiles based on # of LCerformed per year for AC (lowest tertile (LV) 8/yr; 8 medium tertileMV) 17; highest tertile (HV) 17/yr). Primary outcomes were theate of open conversion, bile duct injury and mortality. Logisticegression models were created to further assess the effect of surgeonolume on primary endpoints.
ESULTS: The number of cases performed by HV surgeons in-reased from 22% to 45% from 1999-2005. There were no differ-nces in preoperative comorbidity/demographics among the threeroups. With increasing volume, there was a lower overall rate ofpen conversion (13.6% vs. 10.9% vs. 7.5%, p�.0001), bile ductnjury (0.27% vs. 0.11% vs. 0.08%; p�0.02) and lower mortality0.84% vs. 0.81% vs. 0.65%, p�0.03). Logistic regression con-irmed that volume was an independent predictor of open conver-
ion.