Better heath care with better IT, part 1: Outcomes registries

Health reform promises better care at a better price. But are we measuring outcomes to ensure that it's delivering on that promise? This is the first in a series of entries exploring health care outcomes, looking at the practical approach that Kaiser Permanente has adopted for tracking quality that is rooted in the functional use of health information technology. First we'll introduce the paradigm and Kaiser's solution, subsequent posts will cover implementation challenges, and keys to success.

600,000 total joint replacements are performed each year in the US, with demand over the next 20 years expected to increase over 600% for new knees, and 174% for hips. Hospital charges for replacements are forecasted to increase by over 300% in the next 5 years. Despite the savings intentioned by health reform, all those surgeries contribute significantly to health care costs.

At the same time, we have gross measures of process (the things that are done before, during and after a procedure), but few reliable measures of outcome. Theoretically, better process leads to better outcomes, but not always. And not all processes contribute equally to better outcomes. We certainly know how much we pay. But with high cost, high volume procedures such as joint replacement, a lack of outcomes measures makes it difficult to assess value: How are all those hips and knees holding up over time? Are costs well spent in terms of patient-reported outcomes like pain and mobility? What are the real-world revision and infection rates?

Clinical trials are often the gold standard for answering such questions, but they aren't always best for assessing joint replacement. Trials cherry pick ideal patients whose experience may not match your experience, with best-in-class surgeons who may not match your surgeon; they rarely follow patients as long as we'd ideally like; mortality rates are exceedingly low, and adverse events are relatively low, meaning that trials need to be of very large size (and therefore cumbersome and expensive to run) to measure these events; and randomization to comparative treatment protocols is unethical.

Enter the registry. In 2001 orthopedic surgeons at Kaiser Permanente created joint replacement registry that allows them to track every joint (and associated patient) that they've replaced. The registry was initially created with five general goals in mind:

1. Notification during recalls

2. Developing clinical best practice

3. Assessing risk factors among patients

4. Evaluating effectiveness of implants

5. Creating a database for research.

The registry has many advantages over a clinical trial. It follows real people in real time through the rest of their real lives. Each year 17,000 joints are replaced at Kaiser facilities, creating a potential research population that is big enough to allow even relatively uncommon adverse events or risk factors to emerge. And when products are recalled, surgeons can quickly and easily access a list of patients walking around on potentially defective devices.

There were challenges to implementation, and huge rewards. We'll cover those in upcoming postings.