Thursday, November 30, 2006

Hi, I'm from WalMart and I'm here to help.....

Yesterday’s WSJ reported on a WalMart/Intel collaboration in digital health records. I don’t have a well-formed opinion yet on whether this is good news or bad news for the HIT adoption effort that many of us are engaged in, partly because the article didn’t provide a whole of detail on what this collaboration is actually doing. So let me proceed, but with caution. John McDonough asks whether this might be a “disruptive technology”. I don’t think so. He also asks whether this will complement the work of MAeHC and others involved in promoting HIT adoption. I do think so.

On the technology question, it’s not obvious what’s meant by “digital records for employees” and “portable electronic records.” Patients don’t document medical care, physicians do. And only 10-15% of physicians have EHRs, and most of the country’s 7600 hospitals don’t have accessible data either, so unless this is really a program giving digital records to physicians – and then giving patients access to those records – I don’t see how patients will benefit much.

Perhaps the WalMart model will be based on models that are already out there for the two types of data that are already electronic: claims and prescriptions. Health insurers are well down the road toward providing claims-based PHRs for patients, and AHIP has even brokered a deal for portability of the data across health plans. Revolution Health is going to build a portal that allows patient access to health financial information and health education information. KatrinaHealth is a patient-centric digital record of prescription information. None of these incorporate any hospital or physician information (ie, what we typically think of as our medical records) for the same reason noted earlier, namely, the data isn’t accessible electronically.

So, I don’t think this is a “disruptive technology” from a technical or innovation perspective – I personally don’t believe that there’s a technology magic bullet out there (though we all keep wishing for one!). The main obstacles, as always, are structural (our health care delivery and financing system is broken) and cultural (providers are notoriously independent and resistant to change, and patients think they get the best care in the world, even though there’s tons of evidence that they don’t).

I also don’t think it’s a “shift left” a la Andy Grove. You can’t get data out until someone puts it in, so I don’t think there are any good shortcuts here. It also has to be good data -- you can’t aggregate data that isn’t structured, so having physicians use word processors rather than real EHRs won’t facilitate data warehouses and will actually set them back 10-15 years. I agree that we don’t want to have complex technology be a barrier to adoption, but it needs to be sophisticated enough to deliver value.

That said, I do think this WalMart effort might exert “disruptive pressure” which could push the agenda forward and be very helpful to efforts such as MAeHC. The problem in HIT is that there’s no compelling reason for physicians to adopt EHRs or for providers to link up their systems once they have them. Most efforts to date have focused on the supply-side (ie, providers) because there’s been no real pressure from the demand-side (patients and employers, and their proxies, the insurers). Pay-for-performance may be an indirect means of forcing technological transformation, but it’s indirect. By contrast, when working with their other supply chains, WalMart, GM, Intel, and others insist that their vendors set up electronic data interchange systems that allow real-time inventory management, order management, delivery tracking, etc. If employers start thinking of their health care supply chain in the same way – and require that providers have EHRs and interoperability – they will fundamentally alter the pace of change by creating urgency, where none really exists today. Patients will be the main beneficiaries in the end.

I think it’s fair for all of us to be concerned about anything related to healthcare that WalMart is involved in, because their business success is based on cost-reduction, not on maintaining high quality products or service, and they apply this approach to their suppliers and to their employees alike.

I also worry that there could be an element of coercion in their model as described. Will they derive revenue from selling the de-identified data from the warehouse? Will they ask patient permission to sell this data (HIPAA doesn’t require it)? Will they share the revenues with their employees? My fear is that the answers to these questions aren’t on the side of their employees. WalMart of course would argue that the data is theirs since they’re holding it, paying for it, and de-identifying it (I guess possession is 9/10 of the law, or something like that). Yet another reason that we should move away from our system of employer-sponsored health benefits, but I’ll wait for John McDonough to open up that can of worms……

Tuesday, November 28, 2006

Promising data on clinical performance measurement

A lot of us hold the faith that pay-for-performance will be key to getting wider adoption of health IT systems in physician offices. This faith rests on the assumption that we can get doctors to use EHRs to record clinical data in ways that are meaningful, easy to gather, and comparable across physicians and over time. An article in the most recent issue of the Archives of Internal Medicine highlights the opportunities and challenges here and I think will be just the beginning of a much richer and more realistic discussion of these issues.

The authors of "Assessing the Validity of National Quality Measures for Coronary Artery Disease Using an Electronic Health Record" looked at quality measurement at a large internal medicine practice using a "commercial EHR" (they don't say which EHR they're using). They found that the actual performance of the physicians along several measures was better than their estimated performance, which was calculated from data automatically extracted from their EHRs. They conclude that:
"Profiling the quality of outpatient CAD care using data from an EHR has significant limitations. Changes in how data are routinely recorded in an EHR are needed to improve the accuracy of this type of quality measurement."

My reading of their study is that the authors raise legitimate concerns about the difficulty of using such data, but their conclusion overstates their case. They correctly point out that the real issues are not about technology, per se, but about process -- physicians don't routinely enter data in a way that makes it easy to do accurate calculations. For example, if you don't enter blood pressure readings as numeric data in the blood pressure fields of the EHR, you won't get "credit" for having taken the blood pressure.

More generally, they point to four sources of error:

  1. Wrong diagnosis (ie, person diagnosed as having CAD when they didn't);
  2. Data not entered as numeric or structured data (ie, they may have treated a person with aspirin, for example, but they buried that fact in a text note rather than entering it in the medication list)
  3. Exclusion criteria are not standardized (ie, there's no standard way to record the fact that a patient may not qualify for the treatment in question, which shouldn't count against the physician); and
  4. Measures don't account for patient non-adherence (ie, the patient doesn't comply with the physician's treatment decision, for example, doesn't take a lipid-lowering drug even though the physician recommended it and prescribed it).

Frankly, the only one of these problems that I find new and thus troubling is the third one regarding exclusion criteria, because the others are well known and they will become less severe over time as people get used to them. Entering consistent exclusion criteria is very complicated, however, because they are so measure-, condition- and patient-specific, there are no standards out there that I'm aware of, and the EHRs I'm familiar with don't have a good way to record this information systematically anyway. So, we need to figure out a way to address this issue.

That said, it's not clear how big these problems are in the scheme of things, however. It turns out that even with these problems, the automated measures performed pretty well -- the physicians were at 82% success on the measure they did worst on, which improved to 87% once corrections were made for the issues noted above. On the measure they did best on their scores went from 98% success to 99% after adjustment. While we'd obviously like these measures to be as accurate as possible -- particularly when quality and compensation are on the line -- this level of mis-measurement is surprisingly low given that we're really at the beginning of the beginning of clinical performance measurement.

Perhaps a bigger issue that the study doesn't address is what to do about so-called "ceiling effects." How are we going to tell the difference between physicians who are already high-performing? Is there really a difference between physicians performing at 98% vs 99%? And how do we tell the difference between physicians who are both performing at 100%? This suggests that we'll need increasingly granular measures to show variation in performance, if that's what we want to show. Or, do we just care that physicians get to an acceptably high level?

No one has any answers to this yet, of course, but the data from this study suggest that we may have to figure it out long before at least I thought we would. And doing this without having physicians feel that the goalposts are always being moved could be a much bigger issue than the technical arguments about measures that dominate the conversation today.

Tuesday, November 14, 2006

What would Say say?

The French economist Jean Baptiste Say is famous among economists for Say's Law, which essentially states that supply creates its own demand. When I was taught Say's law in school, it was positioned from the start as a discredited theory that was important to a fringe element, but otherwise largely dismissed by John Maynard Keynes and others in the mainstream of neoclassical economists. Today's article in the New York Times about gastroenterologists made me think that Say's Law, while maybe not generally applicable, might be alive and well in certain sectors of the economy. Health care delivery, for example.

The article describes how gastroenterologists might soon face a crisis as technological advances in imaging technology may make virtual colonoscopies preferable over actual colonscopies. Since colonoscopies are performed by gastroenterologists and virtual colonscopies by radiologists, this could represent a signficant shift of patient traffic away from gastroenterologists.

In a well-functioning market, this trend would, over time, foster decreases in the supply of gastroenterologists. The Times article suggests that this won't be the case, however, because gastroenterologists will find something else to do to make up for the lost income, such as performing other types of procedures. Thus, while the number of colonscopies performed may go down, they will be replaced by other types of gastroenterology procedures. As one physician noted in the article, "We have a lot of organs. The esophagus, the stomach, the small bowel, the liver, the pancreas. I think we've got a lot to do."

This phenomenon is not unique to gastroenterologists. Dr. Jack Wennberg and his colleagues at Dartmouth have made a cottage industry of showing how regional variations in medical practice are directly linked to the relative supply of specialists who perform the practice. Certain regions have more C-sections not because they have more high-risk pregnancies, but rather, because there are more ob/gyn's in the area. The ghost of Say lives on.

Of course, Say's Law doesn't apply everywhere. Purveyors of health IT have implicitly applied Say's Law in the past, but with disasterous consequences. About 20 years ago the CHINs (community health information networks) adopted this Kevin Costner view of economics ("if you build it, they will come") and as a result none of them are still standing. A number of the "RHIOs" out there today may suffer the same fate. We've found in our MAeHC pilot projects that the demand for HIE products and services is actually multi-dimensional and complex. While results delivery and basic electronic data exchange are minimum requirements, applications like electronic referrals tracking, quality data aggregation, and forms routing are also important to physicians. HIEs that launch without deep consideration of what physicians are actually demanding do so at their own peril.

Sunday, November 12, 2006

Every silver lining has a black cloud

Two recent surveys of clinical IT use in the US offer some good news about increased use of clinical IT among physicians. The Center for Studying Health Systems Change, and the National Center for Health Statistics, have both recently issued reports on clinical IT penetration that shows steady growth among physicians over the past five years. Before you get too excited by this trend, however, you may want to take a quick look at some of the underlying data in these reports, because there may be more bad news than good.

I find two things particularly troubling. First, the digital divide is increasing. If you look at which physicians are increasing their use, you find that growth among large practices (especially those with over 50 physicians) is masking pathetically low growth among smaller practices (10 physicians or fewer). The CSHSC study found that while physicians in large practices increased their use of clinical IT systems by 18 percentage points between 2001 and 2005, those in smaller practices increased by only 7 percentage points over the same time. And even after all this growth, the NCHS data show that if you look at all of the physicians who don’t have an EHR, 90 percent of them are in small practices (10 physicians or fewer). Yet, those smaller practices account for the vast majority of the country’s ambulatory doctors (88 percent, according to NCHS), and outpatient visits (87 percent, according to the most recent data from the National Ambulatory Care Survey).

Since most patients go to the doctors who are least likely to have an EHR, less than 1 in 5 office visits (17 percent) are with doctors who use some type of clinical IT. Thus, as bad as the digital divide is among doctors, it's even worse for patients.

The second thing I find troubling is that those physicians who report using clinical IT are using only the most rudimentary types of systems, and those systems, in turn, offer only the most rudimentary types of benefits. The macro data from NCHS makes this abundantly clear. While almost 24 percent of physicians report using an EHR, only 9 percent are using what most of us in the field would really call an EHR (ie, one that has e-prescribing, ambulatory CPOE, electronic results, and electronic documentation). If you asked what fraction are using an EHR that would pass muster by either the CCHIT or the ISO standards, I suspect that we’d have way less than 9 percent.

As another example, take the CSHSC data noted above. While there was general increase in the use of clinical IT systems, most of that change was driven by increased electronic access to guidelines and use of electronic documentation of notes. Guidelines are great, but there isn’t much evidence that access to guidelines by itself does much to improve the quality, safety, or efficiency of care. Similarly, electronic documentation is only valuable when done in a qualified EHR system (for example, one that uses structured data to trigger decision support tools), and I suspect that most of what’s reported here is little more than word processing. Good for eliminating transcription costs, but doesn’t do much for the quality of care.

By contrast, the CSHSC report shows very little growth in the use of functions that have been shown to give value, such as reminders and e-prescribing. Among small physician practices, use of such systems grew at the glacial pace of about 1 percentage point per year between 2001 and 2005. And this growth is on a surprisingly low base of adoption – only 14 percent of physicians in small practices (9 or fewer) reported using e-prescribing in 2005. Reminder generation shows somewhat higher adoption (25 to 28 percent), but similarly slow growth (about 1 percentage point per year).

The bottom line for me on these studies is that growth isn’t happening when we want it to (now), where we want it to (in the practices that see the most patients), or how we want it to (in the functions that offer the greatest value in terms of quality and efficiency). There’s a more insidious aspect to this as well. If we hype clinical IT without guiding how it gets implemented, we run the risk of having large numbers of physicians getting locked into systems that offer relatively little real benefit to patient care. At MAeHC we've found that it's easier to move physicians from paper to a qualified EHR than it is to move them from half-measure electronic systems. If current national trends continue, the tyranny of avoiding “rip and replace” will become a bigger obstacle than it already is today.

Aside from these small quibbles, letting the market take care of this seems to be working out just fine……..