10/02/2014 10:52 am ET Updated Dec 02, 2014

Health Care Warranties: The Risks of Patient Risk Assessment

I came into process improvement many years ago from manufacturing industries. In the modern manufacturing facility, there is a lot of focus on making sure the quality and characteristics of the incoming materials is understood. For instance, the metallurgy variation might create a need for different machining tools, adjustments to processes such as heat treating time or temperature, or the work instructions at different points of the process. Once the metallurgy of the incoming batch of material is understood, recommended changes are dialed into the process to ensure that the outgoing product is not defective. Good parts have happier customers and minimal warranty costs.

What would happen to the warranty costs if the raw material metallurgy was analyzed, and recommendations made for process change, but the manufacturing facility was filled with a combination of old machines that didn't work, new machines which weren't well understood, workers on one shift that were new to the job, instructions that varied based on who was doing the work, etc.? You would expect to have lots of returned defective parts, unhappy customers, and significant warranty costs.

I began to think about this example recently during a webinar regarding hospital readmissions. Another participant asked if there were patient readmission risk assessment tools that could be used help meet the new goals that the Centers for Medicare & Medicaid Services CMS is creating. Within a few moments several people responded that they could share their tools and a decision was made to disseminate the tools to everyone on the webinar.

That exchange made me wonder: If the readmission risk assessment tools were so effective, why were the people who used them participating in a webinar to learn how to reduce readmissions? Why were the readmission rates about the same for those who didn't have the tools, and those who were currently already using them (from my personal observation)? How are these types of tools working for YOU in terms of identifying and adapting to specific needs of patients to reduce readmissions and other types of harmful care?

These are important questions, and here's why: The effective warranty cost of our current health care system is primarily realized in harmful care during acute care episodes (e.g., medication errors, hospital-acquired infections, and pressure ulcers) and avoidable readmissions where something occurred (or didn't occur) during the acute care episode that creates the need to readmit a patient a short time after discharge. CMS reports the 2012 readmission rate to have hovered around 18 percent.

The cost of this defective care has traditionally been borne by the patient in one way or the other: There has been limited accountability for the health care providers to bear the cost of this "warranty work." However, the employers and insurance companies are now beginning to reject this assumption of risk and are holding health care institutions more accountable for the quality of the patient outcomes.

Health care reform is also restructuring the CMS payments to eliminate payments for care that is not done right the first time.

In short, health care providers have to get better and more efficient at delivering safer patient care, or risk losing the game.

Specific to the warranty cost of readmissions, there has been a lot of focus on avoidable readmissions both from a community health and health care cost perspective. In reaction to this, many institutions are implementing patient assessments to identify those patients at high risk for readmission due to correlated factors such as medications, diagnosis, past history, etc.

From my perspective, there are two fundamental assumptions in relying on this type of tool: First, the condition of the patient is the primary "risk" for readmissions. Second, by identifying these risky patients at the point of acute care admission it is assumed that our health care processes will adjust to accommodate the patient's condition. Remember the metallurgy example? The parts left the factory without defects only if the downstream processes are stable and robust enough to adjust to changes in the material they are working with in the first place.

Perhaps our assumption should be changed: Is it possible that ANY patient is at risk for readmission if we don't have stable, predictable care processes spanning acute and post-acute care? In other words, should the care process be assessed as the primary "risk" for readmission? Recent research, such as Project BOOST, has done great work in identifying ways to engage the patient differently, new roles and responsibilities of care givers, and checklists of critical factors which influence the chances of readmission. The bad news? All of this great information and effective tools may not make much of a difference when you factor in inefficient and sloppy processes that are not capable of incorporating this thinking.

Even if the acute care setting had perfect internal processes, critical handoffs with post-acute care can render the best of acute-care intentions inadequate. It seems we have as much variation in our care processes as we see in the patients we care for. How about focusing on the specific issues which drive the specific readmissions in our community? Why not throw out our current assumptions (which have not served us well, quite frankly) and create a different assumption: The processes we create and manage are not effective in minimizing readmission. We don't need a PATIENT risk assessment form: We will need a PROCESS risk assessment form!

Fixing the Process

If those who deliver health care want to do so more efficiently, here are a few questions they need to be asking themselves:

• What do we specifically understand about our population of readmissions: Where do they come from (e.g., patient home, tertiary care) and what proportion comes from each locations?

• What are the presenting causes of readmissions and what are the root causes of these?

• Which handoffs within the acute setting aren't performing well, and how good do they need to be?

• Which handoffs between acute and post-acute settings aren't performing well and how good do they need to be?

• What information needs to be communicated better?

• What actions need to be coordinated better?

• And above all, how do we know when things go well or don't go well by using data inside our organizations as opposed to waiting for the payors to tell us about readmissions months after they occur?

So, if we want to fix health care and make it more cost effective, safe, and efficient, let's shine the spotlight where it needs to be shined, the one place that will lead to a better future: not on the patient, but on the process. There are things we can learn from other industries, including manufacturing, to improve healthcare. The first step on the road to success is WANTING to be accountable for the services we provide.