PLP: PRHI is a collaborative working group of about 42 hospitals in the Pittsburgh region. We also include insurance plans, like Highmark, UPMC and Aetna, as well as the payors – both large and small companies. We have affiliations with a number of governmental agencies, such as the Allegheny County Health Department, the state Health Department, AHRQ, CMS and the CDC. We are engaged in a number of activities seeking to improve the quality of health care, its timely delivery and access to the population, as well as reducing the overall costs of health care. We facilitate forums where the best practices developed in our hospitals can be shared in safe environments where clinicians discuss things gone wrong, get ideas for improvements, and share ideas with other clinicians.

Probably the best example is our cardiac initiative, in which we have 12 of the 14 cardiac surgery units in this region tracking over a hundred variables involved in coronary artery bypass surgery. They can look at the performance of their team in terms of complying with best practices around these variables. You need very large numbers to do that: a typical cardiac surgeon might do 250 surgeries a year and only see one or two deaths that are associated with a particular condition. That cardiac surgeon doesn’t know whether those outcomes are significant or not, and could never explain or attempt improvement on his own or her own. When we have, say, 3,000 cases from the entire region and everyone is tracking the data consistently, we can actually draw conclusions and make hypotheses about improvements. We have clinician working groups within that initiative who are taking the data and developing improvements. We also have a relatively new initiative in the critical care and emergency medicine area.

We do one-on-one work with hospitals that are committed to trying techniques to achieve radical improvement in the way they deliver health care. We’ve hosted a number of forums around eliminating hospital-acquired infections and we continue to support a large group of hospitals in the region that are working on that goal, based on understanding the best practices in the region and setting up hospital units so that they have rapid feedback loops.

For example I’ve been working with the Veterans Hospital in the eradication of hospital-acquired infections that are caused by a strain of the staphylococcus pathogen. In that one-on-one engagement, we have been applying methods from the Toyota production system which are targeted at solving problems very rapidly as they are identified in the course of work, applying countermeasures that prevent those problems from recurring again, and taking the learnings and spreading them throughout the organization.

We have a pilot project, or "learning line," on depression specifically focused on the readmission of patients after they are discharged for treatment for depression. What we’ve seen there is that follow-up care is critically important for the eventual success of those treatment plans, and our work is to understand what is needed to ensure that people who are reporting depression are getting the appropriate follow-through so those therapy programs can be successful.

In the chronic disease area, we are primarily focused on diabetes care. While there is a great knowledge about how to treat diabetics and manage that disease, we still find low compliance rates on some of the best practices – getting appropriate follow-up to monitor blood, and foot and eye exams done as they should. Those are examples of preventive measures that can signal the early onset of adverse conditions in diabetics and prevent patient outcomes like amputations.

We also have a presence in policy discussions and participate on the regional, state and national level with different government agencies on regulations, trainings and demonstrations. We bring evidence from the front lines where care is given into those discussions, as an advocate for patient safety issues.

PLP: There have been improvements. The Pittsburgh region posted a 55 percent reduction in central line-associated bloodstream infections between 2001 and 2004. These devastating infections typically afflict patients who are already very sick, and result in death about half of the time. These results have attracted the attention of the Centers for Disease Control and Prevention. The infection rate from the antibiotic-resistant MRSA strain at the inpatient surgery unit at the VA Pittsburgh Healthcare System’s main hospital, after two years of implementing the Perfecting Patient Care System, with help from a grant from the CDC, has decreased by over 85 percent. This continued decline was accomplished despite an increase in patient acuity: the number of surgical enhanced care beds increased by 50 percent – 12 versus eight last year. The MRSA elimination effort is now being rolled out in other units and at the VA’s long-term care facility.

PLP: I would say the programs have not changed. The only thing that does differ is that we used to have a number of hospitals participating in an experiment on "real-time" problem-solving, which was an offshoot of the Toyota quality improvement approach. The experiment was to report everything gone wrong in a hospital and accumulate that information every day. Our experience was that such an approach to reporting did not result in increasing the capacity of the organization to make improvements. After having tried that experiment for two years, we decided at the end of 2004 to discontinue it. I might say, though, that it is consistent with our principles that we formulate approaches as experiments, declare hypotheses and go forward. In this case we decided that was not an effective way to drive change, and was certainly not an effective premise for engaging hospitals who wanted to work with us. We are in effect saying that we want to focus on what we do know produces results, and those are the clinical initiatives I mentioned earlier.

PLP: Since January of 2005. I’ve been with PRHI for 3 BD years now as implementation manager for VA Pittsburgh Healthcare Center’s University Drive hospital, facilitating that organization in learning how to think about their problems from a Toyota perspective. The staff on the surgery unit learning line has achieved a greater than 85 percent reduction in hospital-acquired infections from methicillin-resistant staphylococcus. Prior to PRHI, I was manager of process improvement for the e-commerce firm FreeMarkets, and before that I had spent 10 years at Alcoa in various engineering and managerial roles.

PLP: Paul O’Neill has been a great leader in the patient safety movement and was one of the three co-founders of PRHI seven years ago. I understand from Paul that he had become frustrated with the slow pace of change and had decided that he would like to experiment with a for-profit consulting firm to offer consulting services for hospital improvement. Seven individuals who have left PRHI were associated with the real-time reporting system experiment that we had attempted, and when we made the strategic decision to end that experiment it made sense for them to pursue other employment. Ken Segel was a special advisor to Paul O’Neill and I understand Ken has elected to join Paul O’Neill and his for-profit consulting venture. While we are talking about the departure of seven people, which was approximately a third of our staff, PRHI is actually a very large organization when you consider all the health care workers who come to our forums and working to advance the cause of the Initiative. There’s a lot of momentum in Pittsburgh that has not been diminished by those departures. We have since added senior staff that are super-charging the work that we have refocused on.

PLP: No. Hospitals have not scaled back their participation in PRHI initiatives. There have been reports in the press about differences between Paul O’Neill and UPMC Health System. Despite that "flap" in the press, we do many projects with UPMC staff and in UPMC settings. They are definitely a leader in pursuing improvements for patients and health care workers. We are engaged with them on the floor of the hospitals. UPMC managers, health care workers, nurses and doctors have gone through our training programs. So, it’s a productive relationship that we have with them.

PLP: Well, one of the things we’ve had to overcome is to find a way to create a safe forum. All these hospitals are correctly concerned about protecting patient privacy, so we have to have secure ways to collect and analyze data. We limit participation to certain doctors to discuss certain issues. We believe that there are some strong reasons for a regional approach. Addressing patient safety issues and eliminating medical errors is hard and requires some hard thinking about what the solutions need to be. That is greatly facilitated when health care workers and physicians are able to go to forums and have face-to-face interactions with peers, and are able to commiserate about the challenges.

PLP: Yes, you’d think that. One of the very interesting learnings from our experience is that there’s much attention given up front to the fact that these health care systems are very competitive, and they are. But when it actually comes down to the work, and you place these patient safety issues on the table, we just don’t see the competitive environment. I go to meetings where all the leading heart surgeons in the region come in: they hang their guns up at the door and they’re all about sharing ideas and working out best practices. Sometimes they share the challenges they face within their own systems in a fair way, for example, by saying, "You know, I can’t do that because of my system," and then they figure out ways to solve the issue together. Our experience has been that the initial anxiety that competition will rule out regional cooperation goes away as long as health care workers focus on the actual patient care issues.

PLP: It has not inhibited the flow of data. We are respectful of the fact that these are competing systems. We make sure we have a balanced approach in terms of staffing and who’s invited to participate at meetings. We make a special effort to make sure we’re inclusive and are not presenting ourselves in ways that make it awkward for these different systems to interact. At the end of day I think there’s a constructive attitude among all the systems.

PLP: The most constructive role that we have seen is for regulators to understand the problem on a floor level, because often in solving a problem regulations are a last resort. We have found that, if they can be a facilitator of finding solutions, a lot of the hospitals will come forward and track performance, and do it in a collaborative way, standardizing the way that it’s done with a much better result than if you came in and posed regulations. Let me give you an example. When PHC4 first started requiring the reporting of hospital acquired infections, there was a striking contrast between the number of infections reported and the number of infections that were treated and billed. Although it seems like a very easy problem, there may be subtle differences in the way things get defined, and in fact the data that was reported may be accurate according to the definitions that were given. And what was billed may be entirely accurate, too. There are enough loopholes in the definitions that what happened could very well have happened without hard work to try to go around regulations. Voluntary, collaborative reporting is the preferred approach because it really allows the people who are working on the problem to collect data in a way that helps them solve the problem. There was initial resistance about reporting infection data when the idea floated to the doctors was based on their assessment that the data gathered would not be useful to them. It really doesn’t empower them to find root causes of problems when we pile up a lot of mandated data.