HOW IS PERFORMANCE IMPROVED?
One significant barrier to improving performance is that preoperative
care is a particularly complex and dynamic process. As such, it seems unlikely that
one approach will be effective in all patient populations and all operative settings,
just as one measure cannot provide a complete picture of quality. Strategies that
combine different approaches are often more successful in changing performance compared
with any single approach.[36]
Two strategies that
have been successfully employed in the aviation industry to improve performance include
interventions to reduce complexity and to create redundancies in the system to ensure
that critical processes occur.[3]
These strategies
have not been fully evaluated in anesthesia.
Much of health care is delivered in complex processes. Because
each step in a process has an independent probability of failure, care processes
that require more steps are more likely to fail than processes that require fewer
steps. We found that there were 107 steps (from writing an order to giving the medication)
to administer medications in our ICU. Given this information, the high rate of
medication errors could be predicted. As Paul Bataldin and other quality improvement
leaders say, "Every system is perfectly designed to achieve exactly the results it
gets." To improve, we must change the systems, not individual providers. To decrease
complexity, providers need to understand the system and reduce complexity by decreasing
the number of steps in a process.[82]
One approach to decrease complexity is to bundle care processes.
Care bundles are a group or collection of evidence-based
interventions that are linked in time and space. For example, a ventilator bundle
for patients requiring mechanical ventilation may include elevating the head of the
bed, providing deep venous thrombosis and peptic ulcer disease prophylaxis, holding
sedation, and assessing for readiness to extubate. Measuring performance on care
bundles may also increase the likelihood of observing the anticipated effect on patient
outcomes if the evidence supports more than one care process. For example, we may
not see an improvement in mortality for ventilated patients if patients are always
placed in semirecumbent position, but providers fail to evaluate the ability to wean
daily or fail to titrate sedation such that the patient can follow commands at least
once each day. Additional areas that are ripe for measurement as a care bundle include
care for the sepsis patient or efforts to minimize transfusion requirements.
A key concept used to improve safety in the aviation industry
is independent redundancy. If something is a critical step in a process, we can
engage independent caregivers to ensure the process occurs. The use of checklists,
for example, is an independent redundancy that is believed to have significantly
improved safety in aviation and anesthesiology. Unfortunately, we have not fully
applied the concept of independent redundancy widely in perioperative care.
