Wednesday, June 01, 2011

Getting transparency right

This is about transparency, when it is useful and when it is not. The term is now an established part of the health care lexicon, but there is little substantive discussion about how it is being used.

As I said in an article in Business Week over three years ago:

There are often misconceptions as people talk about "transparency" in the health-care field. They say the main societal value is to provide information so patients can make decisions about which hospital to visit for a given diagnosis or treatment. As for hospitals, people believe the main strategic value of transparency is to create a competitive advantage vis-à-vis other hospitals in the same city or region. Both these impressions are misguided.

Transparency's major societal and strategic imperative is to provide creative tension within hospitals so that they hold themselves accountable. This accountability is what will drive doctors, nurses, and administrators to seek constant improvements in the quality and safety of patient care.

Now, there rises an additional misconception. The perversion of the transparency concept that has evolved rides on the desire of CMS and private insurance companies to use publicly published outcome data to financially reward or penalize hospitals. As expected, this is raising hackles. The complaints often heard from hospitals are ones we have discussed before: "The data are wrong." "Our patients are sicker."

I am not going to accept those complaints, but I am going to suggest that the usual government mandates for transparency of data provide little basis for the kind of process improvement we need in hospitals. What's wrong with these mandates?

For one thing, the data are old. While you cannot manage what you do not measure, trying to manage with data that are a year or two or more older is like trying to drive viewing the road through a rearview mirror. The principles of Lean process improvement and other such systems suggest that real time "visual cues" of how the organization is doing are essential. Why? Because that kind of data is indicative of the state of the organization right now, not what existed months or years ago. Such data are collected in hospitals on a current basis. If their main purpose is to support process improvement, they do not need external validation or auditing to be made transparent in real time.

For another thing, the choice of data in the government's approach to transparency is externally imposed. Process improvement occurs when the people who do the work jointly decide what areas of change are important. We need to trust that the clinicians and administrators in hospitals, working with their patients and boards of trustees, are better able to decide on quality and safety priorities than the government or its agents. We want the hospitals to be transparent about the metrics they choose, knowing that their doctors, nurses, other staff will value the results highly and act on them.

Finally, the payers' approach to transparency creates attention on meeting certain outcomes, rather than stimulating a desire to design and implement a comprehensive structure to achieve better outcomes. A wise colleague said recently, "Obsession with outcome without obsession with structure will fail."

Captain Sullenberger talked about this in another respect: "A checklist alone is not sufficient. What makes it effective are the attitude, behavior, and teamwork that goes along with the use of it."

In summary, transparency of data alone is not sufficient. What makes it powerful in establishing creative tension in an organization are: The currency of the data; the fact that the metrics being made transparent have been chosen by those involved in the process improvement efforts; and the fact that the transparent outcomes are supported by a structure of ongoing process improvement.

As we have seen by examples on this blog, those hospitals that have been most effective in the challenge of process improvement have not done so because a government agency is making their clinical outcomes transparent. They have done so because the administrative and clinical leadership, strongly supported and encouraged by boards of trustees, have made it clear that this kind of effort is a top priority. More and more places each month have discovered the importance of transparency in supporting their efforts. How this takes place will be specific to each hospital, but it is clear that, to be effective and sustainable, change must come from within.


Anonymous said...

Beautifully written, Paul, but how are we to accomplish that this century, when the vast majority of leaders still don't 'get it'? It has already taken way too long.


Paul Levy said...

It has taken a long time to build up current patterns of care. It could take some time to change them.

Perhaps a "barely restrained crowd" of patients could help bring this about more quickly. Imagine if employers and individual patients became active in their communities, pressuring their hospitals to engage in quality and safety improvement along the lines I have set forth.

Those who use the hospitals and truly pay the bills are a strong, but latent, constituency for change.

Perhaps, too, insightful medical schools will take the lead in training MDs about the science of care delivery. Lucien Leape has made this point very forcefully.

I am sure of one thing, though: If the government persists in the current manner, we will have the impression of action, but it will prove to be illusory. Still more years will be wasted, leaving clinicians resentful and unengaged.

Anonymous said...

It will take a long time but there is visible progress. There will be major back slips and successes.

I went through this with the automotive industry in the 1970s and 1980s. The change was slow and hard. It takes decades for this kind of change to propagate through an industry. They continue to work at it in automotive several decades later. Healthcare will be harder and slower because there is no Japanese car maker as a motivating existential threat.

But from aviation, automotive, materials manufacture, and other industries there are examples of what can be accomplished and people who are teaching what is needed.

We still have a huge pool of people (including politicians and managers) who think that a gentler "Pour encourager les autres" is the way to achieve improvement. But that pool is shrinking. Words of encouragement and examples of better approaches do make a difference.

So continue to encourage the successes and have the patience needed for a multi-decade effort.

Barry Carol said...

To deal with the issue of whether or not some institutions have sicker patients than others, I wonder if risk scoring could be incorporated into the analysis. We already have such concepts as age adjusted death rates from heart attacks. On the insurance side, Germany’s Central Fund looks at 80 separate factors to determine how much in premiums will be paid to a sickness fund or insurer to cover a specific individual. Perhaps we could do this kind of scoring for four separate age groups – 0-1; 1-18; 18-64 and 65 and over. Hospitals could then track both risk adjusted and unadjusted infection rates, falls, and other appropriate patient safety related metrics.

Paul Levy said...

It is done all the time, Barry, although people always argue about the right risk adjustment factors to use. But the virtue of self-enacted transparency is that you don't have to worry about that. Your goal is to improve relative to yourself, not to compare with others.

Sure, you can use benchmarks or others as helpful explanatory comparisons, but the goal of transparency is to hold yourself accountable to the standard of care that you have decided is the right one.

When O'Neill started at Alcoa, he didn't design safety standards that were based on the relative riskiness of his aluminum company compared to others. He just said, we are going to eliminate industrial accidents. Period, end of story.

At my former hospital, likewise, our goals were to eliminate central line infections, achieve 100% compliance with the ventilator bundle, and so on. We didn't water those down and say, "but only where the patients aren't so sick." We didn't compare ourselves with others. We compared ourselves with our prior performance.

High performing and innovative organizations don't rank themselves against others on operational metrics. Do you think that Apple gives a squat about what Dell is doing?

Barry Carol said...

I hear what you’re saying, Paul. I know at my own employer, our CEO embraced workplace safety as a core priority and it’s made a positive difference. When I look back on my personal experience regarding goal setting and workplace challenges, I developed a concept I call Carol’s law of optimum pressure. It means that if the goals are too easy or I have more time than needed to accomplish them, my performance is likely to be sub-optimal. On the other hand, if I perceive the objective as impossible, I’ll be frustrated and probably won’t be able to accomplish it. The bottom line is that there needs to be pressure but the objectives need to be reasonable too.

As a financial guy, I’m probably super sensitive to the need for monetary incentives. Hospital patient safety goals, for example, could be incorporated into the bonus formula, at least for senior executive and clinical leadership while lower ranking staff could be eligible for a modest bonus as well if objectives are achieved. At the same time, if the hospital loses revenue due to flaws in the payment model while it improves patient safety, that’s a headwind that needs to be fixed by policy makers and insurers.

Anonymous said...


This is great. However, could you explain one point? Here: "Sure, you can use benchmarks or others as helpful explanatory comparisons, but the goal of transparency is to hold yourself accountable to the standard of care that you have decided is the right one."

You seem to be advocating that hospitals and other providers should be able to set the standard of care that they deem is appropriate. Isn't that what we have already, its just that the standard is exceptionally low? I think one of the reasons CMS has to force the issue is that because some hospitals, obviously there are many exceptions, but some hospitals do very little performance improvement work. Its not part of their business model and, in general, they have never had the incentive to do it. Payment isn't tied to outcomes and, in turn, providers manage very little performance risk (malpractice aside). In a perfect world, I would think CMS and other payers wouldn't have to develop programs like the Hospital VBP or the shared-savings model because providers would already be doing the work themselves. Again, I understand the work you've done at BIDMC, but you're an exception - not the norm...

Lori Nerbonne said...

'Anonymous' is right on the mark. For too long, (most, not all) hospitals have failed to engage in data-driven improvement. CMS is leading the way (who else is doing it?) toward using data and outcomes to: #1) provide patients/employers/all payers with what they need to begin to make informed decisions AND determine the value they're getting #2) force healthy competition #3)provide some counter-intelligence to hospital marketing ads #4) begin paying for performance instead of volume.
#5)Drive improvement from both the inside and outside-in.

If healthcare administrators/leaders had been proactive for all these years, then CMS wouldn't have to be called out as the bad guy now. It's taken run-away costs, an unacceptable degree of patient injuries and over-use to get there.

Doesn't CMS have a duty to provide this data and make it as user-friendly as possible for consumers, considering the revenue that hospitals take in from Medicare and Medicaid? We are footing this bill so dont' we deserve some accountability for how it's being spent and for knowing which hospitals are investing in patient safety & quality care?

We should be asking: Why aren't ALL insurance companies doing this? Aren't they supposed to be representing their paying customers? Instead too many of them are really bed-fellows with hospitals via the lucrative contracts they engage in. It's almost unbelievable that we haven't had this data for years; 'til now only hospitals, hospital associations, and state public health agencies have had access to it, but it appears they've not been doing much with it.

"Let the sun shine in."

Daniela said...

I would imagine that to some extent the ability of a hospital to be transparent has to do with it's ability to collect and manage data. Have phenomena like electronic health records had an impact on hospital transparency, or is it just another tool on top of the same old problem?