A case study in structured problem solving: How we cut $3MM in overtime costs at a hospital by applying structured problem solving.
One of the most rewarding engagements I’ve had as a management consultant was with a hospital that had been under severe budget pressure for years due to excessive overtime costs. I’d never worked for a hospital, so the expertise I was bringing wasn’t hospital management but a structured problem-solving method that I’d learned as a consultant. Here is an overview of each step I used. You can use these same steps to solve your challenging problems.
The Problem Statement. I started by clarifying with my client the problem I was going to solve. I did this with a simple one-page introduction with three parts.
– Situation: A definition of the client and their normal situation.
– Complication: A definition of what has changed in their normal situation that is causing a problem.
– Question: A definition of the problem the consultant is being hired to solve that clearly outlines what success would look like.
By clarifying the problem, we defined success as “substantially reducing overtime without negatively impacting patient care.”
The Logic Map. Next, I created a logical framework – a logic map – to identify all the potential root causes of overtime and categorize them into distinct categories. I chose “Supply-Driven” versus “Demand-Driven” at the top of my framework to organize the potential causes. Potential overtime root causes in the “Supply” category would be things related to the supply of labor – e.g., ‘not enough staff.’ Potential causes in the “Demand” category would be drivers of the demand for labor – e.g., ‘number of patients increased.’
The Interview Guide. I translated that logical framework into an interview guide that I could use in the meetings with the top executives of the hospital. As I walked through the list of all the possible root causes of the overtime problem with each executive, I got a variety of information and opinions on where the biggest, fixable root causes were. After those meetings, I adjusted and added to my logic tree of potential root causes.
Prioritizing Potential Hypotheses. Now that I had a good list of all the potential root causes, I prioritized each based on the probability of being the best answer to the problem. This is important because it takes work to get the data and do the analysis to test each hypothesis. I was able to put the whole ‘Demand’ part of my logic map at the bottom of the list because of the information I uncovered in the manager interviews. The problem had to be related to ‘Supply’ – staffing.
Defining First Hypothesis to Test. Within the ‘Supply” side, I decided to start with an “Excessive sick leave is driving overtime” hypothesis. My initial conversations with managers and staff led me to believe this was the best bet because it had not been looked at yet and it seemed like a logical driver of overtime.
The Best Guess Map and Assertion Tree. Now that I had a hypothesis, I had to figure out the things – the “assertions” – that would have to be true for the hypothesis to be true. I settled on three assertions – sick leave drives overtime, it is rampant, and it can be cut.
The Blank Deck / Storyboard. With a clear top hypothesis to test first, I focused my first data request from the hospital’s IT team clearly on that issue. I mapped out the charts I would want data to fill in so I could avoiding gathering unneeded or “nice to have” data.
The Analysis. The raw data came back quickly and I analyzed it using Tableau software. As soon as Tableau rendered the first charts, I knew we had cracked the case with our first hypothesis.
Sharing Answer with Client. Once I had all my charts filled in and inserted in a PowerPoint slide deck, I showed my findings to the hospital CEO and his top deputies. They quickly understood the opportunity.
Helping Client Implement. The CEO asked me to be a change agent to help him create the urgency in the line managers by presenting it to them at their next all-managers meeting. I presented the findings in a provocative way to energize the managers. By being the “bad cop,” I positioned the CEO to be the “good cop” who would help them implement.
The Result. The pay periods immediately after my presentation showed overtime reduced by 46%, or about $3 million a year, from their steady pattern the year before. When I looked back and added up the hours I spent on this project, I estimated it took me about 80 hours of my time. That’s a pretty good ROI for the client!
If you’d like to learn this structured problem solving approach, contact us to learn more about our training class Structured Thought: Problem Solving. We’ll be more than happy to help you and your team learn how to apply this method.
– Victor Prince is a Principal with thoughtLEADERS. He specializes in teaching Structured Thought: Problem Solving and Structured Thought and Communications. He is a former Bain & Company consultant, who was also a marketing executive with Capital One and the Chief Operating Officer of the US Consumer Financial Protection Bureau. He has an MBA in Finance from Wharton.
Did you enjoy this post? If so, I highly encourage you to take about 30 seconds to become a regular subscriber to this blog. It’s free, fun, practical, and only a few emails a week (I promise!). SIGN UP HERE to get the thoughtLEADERS blog conveniently delivered right to your inbox!