KS: Value Capture is a new organization that seeks to work with the leaders of health care institutions who believe that, good as they might be, there is a vast opportunity to improve in safety, quality of clinical care, satisfaction and financial performance of the organization – and they are determined to close that gap. We help them understand the capability of everyone in their organization to identify and solve problems, identify opportunities for improvement and make them – in comparison to organizations that handle complexity very well: places like Alcoa, Toyota and select places in health care. The company was started a few months ago by a small group of us, including Paul O’Neill, who serves as the non-executive chairman; Paul O’Neill Jr., who is our managing director. Geoff Webster and several colleagues who met primarily through the Pittsburgh Regional Healthcare Initiative (PRHI). The consulting arm of the company provides direct support for health care institutions with significant clinical operations that are seeking to radically improve safety and quality of care. We also have a non-profit arm called the Value Capture Policy Institute that works to provide a practical, ground-level view of what pure safety and outstanding performance on quality take in the real world of health care, and promote the concept that key public policy areas that either impede or don’t adequately support high performance can be adjusted – everything from malpractice laws to the way we reimburse for health and clinical care.

KS: Mr. O’Neill and many of the rest of us have been actively engaged for a number of years in several critical areas of health care policy, including encouraging people to fundamentally re-look at the current, tiresome debate on malpractice, and asking people to focus on what the core objective of the system ought to be – which in our mind should be to aid safety. That requires making it safe, indeed necessary, to get information about every error that occurs out and known publicly so it could be learned from openly. The practical application of that idea is to socialize the cost of things gone wrong: we as a society ought to pay victims of medical mistakes a fair economic recompense for their injuries. We don’t think physicians and institutions should have to pay for medical malpractice insurance costs. In exchange for that, the professional community’s obligation should be that, as soon as an error is detected, it needs to be contributed to a national learning database immediately, as well as shared with the patient and family so that no one else has to recreate the same mistake. Folks who didn’t do that would subject themselves to significant damages and potential loss of license. We suggest that the medical malpractice system would be considerably more effective in reducing injury and providing just compensation if it were switched to this learning basis rather than the current punishment basis.

We’ve also been actively involved nationally in issues of the reimbursement system. For example, what are wise ways to begin rewarding improved outcomes for patients through improved reimbursement? How we can get health care away from keeping two sets of books – one price that is theoretical and is sent out to people including those who don’t have any health insurance, and another that is paid for at a much lower rate by HMOs and others. In Pennsylvania last year we had hospitals reimbursed 30 cents of every dollar that they billed, and that was the exact level they expected. We think that kind of two-book system has a distorting impact on managers and makes it harder to understand what economic inputs in the system are producing good outcomes for patients. It also has a corrupting effect on the environment because we’re asking staff at hospitals and other places to spend a good part of their day dealing with numbers that are complete fictions, and we find that has a corrosive impact on employees.

It is quite possible with good intentions to make changes in reimbursement systems that make things worse, not better. What we encourage is deep, first-hand understanding of what it looks like to provide care perfectly, understanding the costs associated with that, and basing reimbursement schedules and improvement opportunities on that first-hand knowledge. We have helped several hospitals to discover that in their intensive care units with patients for whom a nosocomial infection occurs, they are in fact losing a great deal of money on virtually every patient who gets an infection, despite the fact that they do get higher payments for those patients. The payments do not adequately compensate them for the extra care that they provide. That was largely unknown by the health care leaders we work with because of the tendency to over-aggregate measures in hospital accounting. That incentive, one might say though, is properly aligned once the detailed accounting work in done. In the same institutions we see that for significant complications of care – including avoidable complications in the ICUs that fall short of infection – hospitals are deriving significant revenue above and beyond the cost of care – in effect, a profit. One might say that that is a bit of a negative incentive to getting care right when there are hundreds of thousands of dollars following from care that is unnecessarily complicated. Not that anyone gets up in the morning trying to do a bad job, but certainly there’s not a burning fire to get it right.

We’re working with government officials and leaders of interested health care institutions around the country to try to fund a significant accounting study of three to five very fine hospitals – a mix of academic, urban, and rural – to get a ground-level look at all of their core processes and get a picture of the level at which they’re performing when held up against an ideal: if everything happened perfectly, what would it look like, in terms of clinical quality, safety and financial inputs associated with their care. The idea is to provide a practical managers’ map of where processes stand and how they could be improved that we hope would have a significant impact in the health care community and also help policy makers focus on what it’s really going to take to get to the next levels of performance in our health care system.

KS: We’ll let the clinical folks who are involved to define "perfect" from their perspective, while study design work will be done to create a baseline for outcomes and interventions. The thing that will be different about this study is that, in addition to looking at classic outcomes data sources – including CMS data sets – the real data points are going to be drawn from activity-based cost accountants who are in the institutions, married up with clinical folks and people from an industrial systems background to get a detailed picture of the actual processes of care and the resources they’re consuming, versus the outcomes they’re producing. So, the benchmark is going to be different here. It’s not just going to say, "When are we getting 100 percent of our diabetic foot exams day completed?" It’s going to say, "Here are the resources we’re investing in that process and here is what it might look like if everything we wanted to happen for the patient happened like clockwork in our offices and when they go home." There is a body of literature using this methodology, largely drawn from industry, and is the kind of detailed process mapping that many companies use to get a deep understanding of their operations. There have been a few experiments in health care over the years in activity-based costing, which is about as close as it gets to this.

KS: We start work with the executives of an institution such as a hospital by actually going out on the floor of the hospital and observe how work is being done together. We will have them watch what happens when, say, a nurse has a problem meeting a patient need or can’t do his or her work and, over the course of an hour make a series of observations and go back and reflect as an executive team on things that went wrong. Things go wrong all the time. At one recent observation of four staff members, 80 things went wrong inside an hour. The staff catch them heroically before any injury is done to the patient. We’ll then ask the question, "Were any of the things gone wrong investigated immediately to root cause?" The answer is usually none or perhaps one. anything done to put in place a solution that might keep the same problem from happening again to another staff person, or the same staff person the next day?" The answer is usually zero. "How many of these ‘near misses’ were actually called out and made it into the formal problem-solving system?" The answer is usually zero or almost zero. The next day, we’ll work with the same leadership team to take a reflective look at their formal problem-solving and risk management systems and ask them to look at the issues they do root cause analysis for, their incident reporting system, etc., and compare it to the nature of work that they saw out on the nursing floors. They will often see that there’s not as tight a fit in terms of the relevant utility of the formal problem-solving system to the problems their staff are really having as they’d like. We then help them reflect on some changes and experiments they’d like to launch to try to tighten those two things up.

KS: We do. We have The Wellmont Health System in the tri-cities area in eastern Tennessee. We have one client in this region, who I prefer not to name at this point, and others being considered in the pipeline. We think that, if we can contribute to the emergence of two or three hospitals or health systems that are performing so far above their peers on measures of safety, clinical quality and efficiency, that it will take away the excuse that "no one else is doing much better so we must be doing okay," and serve as places where the rest of the country can come and learn from what those health systems are accomplishing. We thought that would be a very positive contribution to make over the next few years, and that would require us to not be exclusive in a focus on southwestern Pa.

KS: Value Capture’s approach is complementary to a scientific approach such as the creation of a cardiac registry across our community here in Pittsburgh, which PHRI facilitated and is a tremendous step forward. But, on an institutional basis, one wants to build upon those scientific platforms, registries and tools to make sure that in the everyday practice of care and operations across a very complex institution, the systems for identifying and solving problems are running on all cylinders. That is a person-by-person, employee-by-employee issue of skill development and building a frame of mind from the leadership down to create those conditions. Registries are powerful tools for deciding what care produces the best outcome. But in an institutional context, knowing what care you want to provide doesn’t mean that you’re able to provide it 100 percent of the time to every patient. In fact, one of the things that health care institutions get quite frustrated by is that, on even elementary elements of care that are agreed upon, such as beta blockers for heart attacks, they find themselves having a hard time getting the numbers to where they want them to be. That is typically because their systems for identifying problems and making improvements are not robust at the local level. If you want to apply what you want to do rigorously, every time, you have to be very good at problem-solving and making improvements in the real world of clinical application. That’s where these advanced techniques of problem solving are tremendously helpful.

In that context, investigating problems when they occur, with the people whose work it was where the processes broke down, and designing solutions rapidly, as soon as a root cause is attained, is much more productive and powerful than trying to collect data for many months, arguing about it in off-line conference rooms, typically without the people who actually run the processes that are being discussed, and trying to come up with solutions on that basis. You’re much more likely to get an accurate read on what really happened if you’re investigating rapidly and you’re much more likely to get the appropriate solution designed to prevent it from happening again. The work of health care staff is so intense that going back to, say, a pharmacist three days after an incident occurs and try to get them to remember what were the circumstances around which a particular process breakdown process occurred - it dooms us to unproductivity in our problem-solving. As a consequence, we collect data for a long time and over-aggregate several different kinds of problems into one bucket and then are frustrated when our 20,000-foot solutions don’t solve the clinical problem.

KS: A registry-centered approach is enough to produce a certain level of improvement in the institutions that are eager to apply it. The next level of improvement comes from a determination to apply it, and that means getting into the nitty-gritty detail of the processes and getting very good at problem-solving improvement that I’ve described. Do I have confidence that a government mandate attached to the use of best practice information would have a further improvement effect, beyond the will and insight of health care leaders themselves? I think I’m doubtful on that count. The role of government in making levels of performance transparent in areas of medicine where the issue has been settled, in terms of the contribution of a particular action to quality – as CMS and the Joint Commission and others have begun to do by collaborating around a common data set – I think is very appropriate. Beyond that, any use of coercion to use and apply specific practices – I’m not sure it will produce that next level of increment toward perfection. I think marketplace financial incentives can work. They need to be applied through progressively bolder experiments, and they need to begin in areas where the issue is settled, clinically – where you don’t get people having significant broad-scale disagreements about whether the particular care process contributes the right outcome or not – and where the measurements are reasonably reliable. I think financial incentives for doing the right thing can be marginally effective, but I don’t think they’re going to produce the level of true excellence that can only come from the people whose place it is driving the processes.

KS: Mr. O’Neill believes, as do many of us, that the best possible use of his time is not to continue to try to convince specific health care institutions that there is a continuing opportunity to radically improve on safety, on quality, on efficiency, and that it must be seized – but instead to work with the leaders of health care institutions that don’t need to be convinced, that see an opportunity to improve their value creation by 30, 40 or 50 percent, and that would try and drive those changes in their institutions along the way. So ironically, even though we’ve created a small, for-profit consulting firm, part of the reason is to get him out of marketing. For Mr. O’Neill, in issues like the changes at UPMC that were reported widely in the papers, the issue isn’t personal. It’s that he recognizes the opportunity to build on the work of many good people and to kick it up three or four gears in terms of pace, scope, breath and vision, and believes that his role ought to be to help accelerate that process where it’s a willing process. My decision, and the decision of a few others to go and support that, are really just personal decisions in which we agree that that’s going to be a valuable contribution to try to make.

When Mr. O’Neill took over PRHI, pretty early on he began to push the hospital community to look for that next level of improvement. One of the things he suggested might be a worthy focus was to try to comprehensively take on the issue of illegible or incomplete physician orders. He knows from his experience that, if you let the same problems roll on day after day, it’s hard to claim that you’re particularly serious about safety or improvement. This isn’t an issue of doctors trying to be bad, it’s an ongoing issue that needs rapid investigation of solutions to help insure that when a doc hands a script over the party on the other end can understand it and carry out the doctor’s order as intended. That series of suggestions about taking on that issue fundamentally and comprehensively, beginning with telling ourselves honestly how often it’s happening every day, was met with an overall, decided lack of enthusiasm from a broad sample of the hospital community in Pittsburgh. I think it was partly a signal to him that, in his current capacity, he wasn’t having the kind of conversation where he thought he was going to be able to have the most value.