One crucial metric for hospitals to minimize is the surgery complication proportion—a percentage indicating adverse outcomes following surgical procedures, varying from low for minimally invasive procedures to elevated for high-risk surgeries.
Not only does this number differ across types of surgeries, but it also varies among surgeons due to the specialized nature of the skill. Many other factors can affect the number as well, so it is important for hospitals to rule out as many outside factors as possible before analyzing individual performance.
After surgery, often a surgeon is asked whether the setup for the surgery was exactly how he or she expected it to be. The purpose of this question is twofold: first, a surgeon’s time is expensive and limited, so having a room that is set up correctly saves time, and secondly, not having the tools needed for the surgery in the places that the surgeon needs them could lead to delays, potentially causing adverse outcomes for the patient.
An example of this could be the surgical lighting. Proper lighting is crucial so the surgeon is clearly able to see what they are doing. If the lights are not positioned correctly or if there are issues with their intensity or focus, it can significantly impair the surgeon’s ability to perform the procedure safely and accurately. This could, in theory, lead to complications during or following the procedure.
In our scenario, we decided to collect data from two months of surgeries among three different surgeons. We recorded whether or not the surgeon affirmatively stated that everything was set up correctly or not. Then, we measured if there was at least one complication recorded with each patient in the subsequent 30 days post-operation.
Ultimately, we wanted to see if our null hypothesis was correct; we assumed that there would be an association between the room setup and patient complications, and we wanted to prove that there was a statistically significant difference between the two outcomes that could not be explained away by chance. Obviously, with the limited data we collected, we would be unable to prove causation, but proving an association could be a strong starting data point for continuous improvement.
We entered our data from 105 surgeries into Minitab Statistical Software and performed a Chi-Square Test for Association. We opted to use Chi-Square instead of ANOVA since the results were a binary “yes, there were complications” or “no, there were no complications”―the data were not continuous. Here are the results Minitab produced:
With a p-value of less than .001, it is safe to conclude that there was indeed an association between the complication proportion and the setup of the operating room. If you look at the percent profiles chart, you will see that on average 64% of procedures do not result in any type of complication. If our null hypothesis was incorrect, we would expect the “correct” and “incorrect” bar charts to follow the same pattern, but they clearly do not. Proportionally, surgeries where the room was set up correctly was associated with a much lower complication proportion.
Moreover, in the percent difference between observed and expected counts, you can see a long red bar on the incorrect setup portion of the chart. This demonstrates that complications happen way more than you would expect when the operating room is set up incorrectly.
Further, Minitab produced a diagnostic report and a report card for this analysis:
The report card confirms that this analysis was indeed valid and can be used confidently to demonstrate an association since all the samples were large enough to obtain sufficient expected counts―easy language to understand for statisticians and non-statisticians alike. We can confidently conclude that our p-value is accurate.
Now, we have the data needed to begin to make data-driven improvements. It’s safe to say that it’s likely in this scenario that improving the setup of the operating room may lead to improved patient outcomes in the form of lower complication proportions. Again, while this test does not demonstrate causation, the association with our data was very strong.
Armed with the data, hospitals can then use Minitab Engage to ideate, track, manage and implement improvement projects. Leaders can measure the effectiveness of certain teams of surgery technicians and brainstorm new ways to improve the setup of the room to not only protect bandwidth and time, but also generate better outcomes for patients and increase patient satisfaction.
Additionally, if you want to take your analysis a step further, Minitab’s Automated Machine Learning in Minitab’s Predictive Analytics module can be leveraged to gauge the impact of multiple different variables on the complication proportion. This can help to identify other areas where improvement efforts should be focused.
Ultimately, making these improvements can lead to a reduced complication proportion for the patients within your network, resulting in a positive outcome that is beneficial for all involved.