High-volume hospitals have low rates of adverse events for high-risk surgeries
Several surgical procedures, such as heart and gastric
bypass surgeries and abdominal aortic aneurysm
repair, have lower mortality and better outcomes if
they are performed at hospitals that conduct a high
volume of such procedures. Do these hospitals also
have lower rates of surgical adverse events for these
procedures as well? A new study suggests the answer
is "yes." It found that high-volume hospitals had
significantly lower rates of adverse events compared
to low-volume hospitals.
Researchers collected data from the Nationwide
Inpatient Sample discharge database from 2005
through 2008. Patients undergoing one of the three
procedures were categorized according to their type of
hospital: high-, mid-, or low-volume. A set of patient
safety indicators (PSIs), established by the Agency for
Healthcare Research and Quality (AHRQ), were used
to determine the occurrence of surgical adverse
events.
Patients undergoing surgery for abdominal aneurysm
at high-volume hospitals had lower rates for a number
of PSIs compared to patients at low-volume hospitals.
These included lower rates of sepsis, blood clots, and
bloodstream infections. In similar fashion, patients
undergoing heart bypass surgery at high-volume
hospitals experienced lower rates of in-hospital death,
bloodstream infections, postoperative hemorrhage,
postoperative respiratory failure, and other adverse
events. Patients undergoing gastric bypass surgery at
high-volume hospitals had lower rates on all PSIs
except for sepsis, for which mid-volume hospitals had
lower rates. Overall, low-volume hospitals had four
times more gastric bypass adverse events compared to
high-volume hospitals. The study was supported in
part by AHRQ (HS18558).
See "Relationship between patient safety and hospital
surgical volume," by Tina Hernandez-Boussard, Ph.D.,
M.P.H., John R. Downey, M.D., M.P.H., Kathryn
McDonald, M.S., and John M. Morton, M.D., M.P.H.,
in the April 2012 Health Services Research 47(2), pp.
756-769.
— KB
Return to Contents
Proceed to Next Article