Many years ago, I was the Quality Director at a large automobile assembly plant in the Midwest. When I was first promoted to the job and moved into my new (and much nicer) office, I found on the wall next to the desk a plaque with these words on it: “In God We Trust, Everybody Else Bring Data.”
This was the early 90’s and Toyota was far ahead of the curve in the learning, teaching and application of the Toyota Production System and lean methodology. My company (which shall remain anonymous, but referred to in this article only as GM), had not yet embraced lean but was very active in their problem-solving efforts through the use of statistical process control tools or SPC. X-bar & R charts and control charts were everywhere and terms like data and metrics were part of our daily language.
But now many years later, as my focus on improvements has shifted to the arena of healthcare, at times I feel like I have stepped back into time when we talk about data and metrics.
It is not that the healthcare organizations do not have data; they often have reams and reams of data. It is more to the point of how that data is used—first to know where the problems are and second to understand what the data is really telling you when it comes to continuous improvement and countermeasure activities.
Recently I was working with a hospital that had issues (as many do) with unplanned readmissions. I was asked to work with the compliance office and facilitate a Kaizen event focused on decreasing unplanned readmissions. One of the first things I asked for as I prepared for the event was the data. “What does the data tell us?” “Why does the organizational leadership consider this a problem?” “What is the objective?”
The data I was given told us the hospital was last in the hospital’s organizational group of six hospitals in percentage of unplanned readmissions. That answered the question of why the hospital considered this a problem. So at least one of the metrics seemed fairly straightforward – reduce unplanned readmissions.
But as the team started to examine the data and understand the current condition, it quickly became apparent why the problem had existed with little change for quite some time. When it came to the data – that was basically the extent of the data – we are last in our group and this is our percentage of unplanned readmissions.
There was no Pareto chart of the top reasons why patients were returning to the hospital. There was no internal data being gathered to understand what was happening. There was no data…just a perceived problem.
Essentially, what we had was not uncommon in healthcare organizations. The data and metrics were at a very high level that caused concern and reactive actions. There was no follow-up or “drilled down” data to help us understand what and where the issues really were.
The focus of the Kaizen now had to shift. Before the team could designate where emphasis could be focused on specific issues and problems, we had to first obtain and then analyze the data.
What we knew at this point was that any patient that returned within thirty days of discharge was considered a readmission and reported as such. If that patient was not at the time of discharge scheduled for readmittance, then it was considered an unplanned readmission. But any further information on why the patient returned was not readily available. The compliance personnel would early in a month go through the records of the previous month and define these returning patients as planned or unplanned (as best they could) and report that information to corporate. But since this could be up to forty days after the incident, even with studying the medical records it was difficult to understand the circumstances; it was very time-consuming to gather any information.
The team developed a flow chart tool to analyze the information. It started with two pathways, clinical issues and patient issues. From there the tool analyzed options allowing the information to be drilled down and sorted to form a Pareto of the top issues affecting unplanned readmissions. Now the team was able to schedule further problem solving on the root causes of those issues.
The key was the process had now changed to be data driven. The initial data was now being reported to the compliance office on the day of readmittance. The analysis of why the patient had returned was now being examined in a standard manner through the flowchart parameters and entered into the data where it could be compared on a Pareto chart. For example, one immediate result was the confirmation that infection was a primary reason for return. When a team focused on that particular issue and gathered more data through direct observation, it was discovered that the instructions for wound care given to the patients at discharge were not consistently applied and contained little descriptive information (no visuals) to guide the patient. The packet and instructions were changed and the staff was retrained as initial countermeasures.
Data. It is defined as “Factual Information, often in the form of facts or figures obtained from experiments or surveys, used as a basis for making calculations or drawing conclusions.”
Some of you may remember the old Dragnet TV series from the 60’s. (Dun ta dunt dunt…Dun ta dunt dunt…daaa!). One of the two detectives, Joe Friday, when questioning the witnesses or the perps, every week would say, “Just the facts, ma’am, just the facts”. That is what data is all about….not speculation, not I think, not passing judgment, not jumping to conclusions or being reactive….just the facts.
Through the use of data and data driven metrics, anyone—including healthcare organizations, are able to prioritize problems, develop countermeasures to address the problems and use the data as the basis for the metrics to measure the change. In God we trust…everybody else bring data.