There’s been a lot of debate about transparency on THCB. I believe in as full disclosure as possible about all kinds of medical data, including pay rates, utilization rates, quality indicators, etc, etc. But I also count myself among the enlightened few who realize that the individual piece-rate service level is not the place at which consumers are best qualified to make comparison judgments about the value of their care. I liken it to the computer purchaser at a corporation—sure you want to know the individual prices of the computers you buy, but what’s really important is the total cost of ownership divided by the benefit you get from them. Or take a motoring analogy, as Glen Tullman CEO of Allscripts does over at HHN on a different topic—you want to know the total cost per year for purchase, insurance, gas, repair, etc, etc for your car. You don’t care in great detail how much your mechanic charges you for an individual spark plug so that you can go to Pep boys and buy one cheaper. Instead you want to know the rough overall cost between a Yugo versus a Camry versus a Jaguar, and then within each class. That’s what the managed competition model is trying to get to.
Of course transparency does help. Joe Paduda is right when he finds the Administration’s position on not releasing Medicare physician data is too, say the least, a little odd. Unlike most of their information releases these days, this one is apparently legal and makes sense for the nation. Although I understand why the Business Roundtable is for it and equally why the AMA is no doubt against it.
Categories: Uncategorized
Actually, I’d like to see us focus on researching and discovering better ways to teach and motivate patients to comply with valid evidenced-based plans of care, as well as providing better ways to establish, evolve, disseminate, and implement such plans of care. Collaborating practitioner-researcher networks supported by new breeds of outcomes-assessment and decision-support HIT, personlized care practices, and policies creating high-fidelity healthcare systems, I contend, will accomplish this over time.
While I agree that process metrics is the best we can do because of the limitations of our current broken healthcare system, I am a firm believer that success should be measured in terms of continuous improvement: Failure means stop developing/evolving/improving; Success means always aiming for perfection, e.g., by collaborating within a community of people who are highly trained, knowledge hungry, and never satisfied. 80/20 is UNACCEPTABLE, except as a temporary intermediate objective. Getting ever closer to 100% efficacy, efficiency, satisfaction, and access/coverage is the only acceptable goal in my view.
This requires extensive outcome research using data collected in all practices/facilities. Just following certain process guidelines doesn’t necessarily mean that a particular patient is getting the most cost-effective intervention. But following a process guideline, allowing providers to deviate based on their clinical judgement, and measuring the resulting outcomes of compliance vs. deviation for particular types of patients undear particular conditions is much more useful way to obtain feedback to drive CQI.
The deal-breaker, however, is if this evidence-based knowledge is used in a punitive way that discourages honest reporting and humiliates/penalizes providers. Instead, we need to shift the paradigm to one that views healthcare as a learning industry with providers as knowledge workers who work together with researchers and patients for unimaginable achievements, without any fear that objective performance measures will harm them.
Steve
> I have a hard time seeing how performance can be
> adequately measured in smaller practices, and I am
> solo.
I think the focus for internal medicine tends to be on process metrics on the assumption that the doc can’t be responsible for (willful) non-compliance on the patient’s part. Plus, we don’t really want docs to drop non-compliant patients to make their scores go up. If the DM approach works at all, I’d like to see it enabled somehow at the family doc level: like maybe my doc could find out I’m not filling my Rx often enough, and ring me up to chew me out, uh, I mean “educate me concerning the importance of diligence with respect to my disease”. In other industries, processes like this are called Customer Relationship Management (CRM). It would be nice to move beyond process, but for the time being process is probably the best we can do except maybe in big academic centers.
t
I accept you criticism and I don’t claim to be an expert on risk-adjustment, but from what I’ve read and seen, it can be a useful, though imperfect tool, like QUALY. There are many books and articles written on the subject, and maybe there’s an expert reading this thread who can respond, but this article seems to be an objective overview of the metric — Use of risk adjustment in setting budgets and measuring performance in primary care II: advantages, disadvantages, and practicalities Following is a slightly adjusted quote about the articles conclusions:
• Use of risk adjustment could help ensure that general practices and primary care trusts are not penalized for taking on patients with complex health needs
• Risk adjustment methods may also help ensure that computerized clinical records in primary care are complete and accurate
• Risk adjustment could add to the administrative complexity of healthcare systems
• It may draw attention away from the overall level of healthcare spending
• There is no gold standard method of risk adjustment
• Use of risk adjustment in setting budgets and monitoring performance should be explored further
I recommended having providers and other stakeholders involved in establishing and approving risk-adjustment metrics because it isn’t a perfect measure. The kinds of question you raised should be heard, taken seriously, and addressed competently!
But, like just about anything else we’ve been discussing, I believe we can accomplish wondrous things by having great minds collaborate openly and honestly to derive fair and accurate ways to help providers and patients work together to continually improve clinical outcomes. And valid, reliable performance measures used to educate and reward are essential to CQI.
Of course, we need a non-punitive system in which providers feel safe to have their performance assessed and are rewarded for demonstrating improved performance, as well as a high-fidelity system in which they have the tools, time, and financial resources to deliver ever-better care.
About the cost, my partner and I were just discussing a proposal we sent to CMS about developing a low-cost software system installed at CMS that would enable providers to enter outcomes data without the need for the providers to purchase a thing. It would require some manual data entry via the Web or paper forms, however, which would require some staff time. Or the entire process could be automated if vendors sold providers EHR/CPOEs that have the outcomes data collection integrated within them.
Steve
“There’s a lot written about “risk-adjustment” methods to help account for differences in patient severity when assessing outcomes.”
There’s a lot written about the Big Bang, but we don’t really know what happened.
“I assert this can be done affordably by developing low-cost software.”
So you’re willing to buy it for us? It doesn’t benefit me to do so, unless you try and punish me for not buying it.
Two flippant responses, I know, and I am sorry. But I have a hard time seeing how performance can be adequately measured in smaller practices, and I am solo. If you look at my 15 diabetics, and I have one who is non-compliant, and I am not sure that I will compare favorably to the Lahey Clinic in Boston, when there numbers represent thousands of patients and dozens of providers.
Although, I bet it would be easy to statistically compare the Lahey Clinic to the Mayo Clinic. But that is a different story.
Dr. Hinson said: “How do you judge the quality of care provided when no two doctors take care of the same lot of patients.”
There’s a lot written about “risk-adjustment” methods to help account for differences in patient severity when assessing outcomes. Nevertheless, great care should be taken in defining outcome metrics and it should include done via collaboration among all stakeholder groups.
While the vast majority of performance measures in use are “process measures” (i.e., was procedure X performed and medication Y prescribed for condition Z), there are a number of outcome measures available, which include evaluate care quality by looking at such things as symptom alleviation, quality of life improvement, length of stay, mortality rates, cost, risk, infection rates, etc. Examples of clinical outcomes for acute care, psychiatric care, long term care, and home care at the International Quality Indicator Project (IQIP) web site at http://www.internationalqip.com/indicators.aspx Examples of clinical outcomes for hospital are JCAHO’s ORYX measures; see http://www.npic.org/Oryx_Measures.html and http://www.premierinc.com/all/informatics/services/oryx/noncoremeasures.jsp
Dr. Hinson said: “When it comes to outcomes measurements, I think doctors are also afraid not of what it will say, but of how the data will need to be generated.”
I assert this can be done affordably by developing low-cost software.
Several other issues about the validity and value of outcome measures to assess provider performance remain, however, including:
• When is a definition of quality care too narrow, e.g., by focusing on cost or symptom reduction, but not considering prevention, recurrence, coordination and continuity of care, or the patient-physician relationship?
• How do you measure quality when resources are scarce and optimal care for the community may require less than “the best” care for its individual members (e.g., delegating office nurses to perform certain activities that physicians used to do)?
• What is the best way to measure quality if outcomes are more strongly affected by patient compliance than by physician orders? This may occur, for example, if certain providers have personalities that trigger greater patient compliance, and visa versa.
• Is it poor quality care if a provider follows the recommended practice guideline, but the patient is atypical and responds poorly?
I also agree that looking a malpractice data can be problematic.
Steve
Dr. Hinson,
Thanks for that very clear explanation. I fully understand and agree with your position on the malpractice issue. The metrics explanation is also very helpful. Poor or misleading information can be worse than none, and I agree that doctors don’t need any unfunded mandates that would be unduly burdensome in either time or money.
I still would like easy access to information about education and credentials as well as pricing transparency information for common well care services, tests and procedures as well as easy access to what my insurer will pay specific providers for specific CPT-4 codes before services are rendered.
Barry writes: “I’m not sure I understand what the doctors are afraid of. Is it the measurement metrics that they consider unfair or misleading?”
I have not seen a fair metric published. How do you judge the quality of care provided when no two doctors take care of the same lot of patients. I have a young, healthy population of patients, who generally take care of themselves. As such, I have a high number of patients to goal when it comes to their blood pressure, diabetic measures, cholesterol, etc. It would not be fair, however, to compare that to another doctor’s patients, e.g., a population with less education, less income to afford the prescriptions, higher in age and complexity. Until you can factor in patient compliance, you cannot judge the physician.
When it comes to outcomes measurements, I think doctors are also afraid not of what it will say, but of how the data will need to be generated. Afraid of the costs of new technologies required by such an unfunded mandate. Afraid of the administrative burden that will come with showing this data.
Barry also mentioned a national site that would allow patients to look up doctors and their malpractice history. I sit on a credentialling board for our hospital. When someone has been sued, only the surface details of each lawsuit is summarized on the applications (you cannot expect more detail or to try and “try” the case using the narrative answer as evidence) and, as such, you really do not know the details of each case. There are so many cast-net cases out there where someone is named in a case with a bad outcome simply because their name is on the chart somewhere, and even though there was no direct patient contact. It does not matter if the case is subsequently dropped. Until it is (which can be years later), this lawsuit now follows the doctor. I do not see how there is a way to fairly make malpractice public until there is serious reform of the tort system. When we have health courts where expert judges quickly determines when a bad outcome has happened, and frivolous lawsuits will have no place, then it will make sense to publish what will be the more objective data.
By far, the most important thing, IMO, is a focus on continuous quality improvement (CQI) by providing a mechanism that helps ensure wellness/preventive, catastrophic and end-of-life care is increasingly safer, and more effective, efficient, affordable, timely, and available to everyone. This strategy would involve linking practitioners with researchers (e.g., Practice Partner Research Network) who collaborate and use HIT to perform lab and field outcome studies; establish and evolve evidence based practice guidelines; disseminate and implement the guidelines; and get computerized assistance in making diagnostic and treatment prescription decisions personalized to the particular needs of each patient, as well as managing and coordinating plans of care. It also means putting the policies and processes in place to ensure providers have the resources (money, time, staff, etc.) they need for quality improvement (i.e., create a “ high-fidelity” healthcare system). Then, as a consumer, I would have the peace of mind knowing that the healthcare system is developing and improving in a sustainable way. And when is comes to selecting a provider, I’d likely choose one that is engaged in a CQI program, than one who is not.
Consumers could also benefit from transparency if they receive meaningful feedback about the comparative “value” (aka cost-effectiveness) of the healthcare services that accounts for both the cost of treating a particular condition and the risk-adjusted outcome of such care. In this way, the value of different treatment procedures could be compared, as well. Including QUALY metrics and an indication of how the care would eat into one’s benefits, is also information useful to consumers.
One problem I have is that this can easily become punitive to providers, and provider-bashing helps no one. Transparency should not only be helpful to consumers, but should also help providers improve their performance. For example, since one factor associated with outcomes is experience, I would expect that inexperienced providers would tend to charge lower fees, thereby increasing their value metric and giving them an opportunity to learn and improve. The few providers having consistently poor outcomes, on the other hand, would use this feedback to improve their skills, or they can chose to stop performing particular procedures and focus on areas in which they perform more competently. In any case, there should be a grace period when the outcome data is shared with providers only, and tools and training should always be available to those who want to improve their performance.
After the grace period, consumers could go to a computer and select their health problem, age, gender, location, and any other required data. A list of providers in their area and the procedures they perform to treat the condition appear. Next to each procedure are a value metric and, possibly, the experience level, as well as a QUALY measure and an indication of the annual/life-time benefit consumption and out-of-pocket expense based on their coverage. Presented simply and accurately, this could be useful consumer information.
But before such transparency is possible, we have to deal with problems such as these:
• Knowing prices of healthcare services is useful only if the total cost of caring for a given condition and the effectiveness of that care is also known. Yet patients rarely have such information available to them.
• Physicians typically lack comparative information on the quality of their own care, as well as the care of other providers to whom they refer patients.
• Patients are currently in the weakest position to demand improvement in care quality. Payers, government, accrediting organizations, and professional societies are much better positioned to insist on high performance, yet are often not motivated to doing push it.
• Very sick patients incur most healthcare costs — e.g., patients with heart attacks, strokes, cancer, mental illness, fractures, and injuries — and are often treated under emergency conditions. It isn’t practical to expect them to shop for the best provider in such circumstances.
• After spending substantial time and money diagnosing a complex illness, a patient is unlikely to shop for a different provider.
• It is very difficult to get reliable data on cost and effectiveness at the individual physician level, whereas conclusions based on provider groups of 25 or greater is more valid.
In defense of providers’ perspective.
Spike is correct. Publishing historic quality data collected under providers’ expectations that it would stay private might be problematic. Why?
Making sure that the metrics are really meaningful for quality to consumers vs. whatever else health plan wanted to measure. As Barry points out, health plan metrics might have more to do with pleasing employers and managing costs than measuring actual quality of care.
It is as important to help providers improve rather than go play “gotcha” game. Simply revealing the data that providers expect to stay private is a lot like having Google turn over our searches to the Feds. They will fight this tooth and nail for very good reasons.
A more reasonable approach would be to start reporting on a set of new metrics that providers would expect to be public and will work to improve. A number of quality orgs are developing those with provider’s input.
Of course the above approach might be somewhat idealistic, but the bottom line there will need to be some fair tradeoff between the interests of consumers and providers.
Wouldn’t it be a major violation of trust (at least) with physicians to start publishing quality data which used to be confidential and was collected with the understanding that it was confidential? Maybe they could make a deal with providers and then start releasing data that was collected after that deal was struck.
If I projected myself into the doctor’s shoes, I think the answer is, yes, it would be a violation of trust. If I, as the doctor, would have no problem with releasing the data, that’s fine. I’m not sure I understand what the doctors are afraid of. Is it the measurement metrics that they consider unfair or misleading?
Issues like this speak to one of the problems with employer provided healthcare that doesn’t seem to get much publicity. That is: when the employer is writing the check for the premium, the insurance company is trying to please the employer and respond to its concerns and priorities. The key priority is keep the premium as low as possible and how the insurance company and providers interact with and relate to employees is of secondary importance at best.
If the employee owned the policy and it was clear to the insurer that the employee was paying the premium out of his or her total compensation package, it is much more likely that contracts would be structured with employee concerns in mind which, presumably, insurers would attempt to identify through surveys, focus groups, etc. This would be especially true if the employee had several insurers to choose from. Competition, transparency and trying to satisfy the customer — what a concept!
Contracts, especially hospital contracts, are negotiated on a one-by-one basis. We had to cut special deal for specialty docs in areas where there were few of that specialty all the time. Hospitals were even more specific in our contracts, with some having negotiated case rates and others negotiating differing per diems. How could the insurers make this pricing system transparent? And this was all just in behavioral health which is vastly less complicated (from a CPT/ICD-9 perspective) than medical.
A question I asked on Joe’s blog, maybe I can ask it here: Wouldn’t it be a major violation of trust (at least) with physicians to start publishing quality data which used to be confidential and was collected with the understanding that it was confidential? Maybe they could make a deal with providers and then start releasing data that was collected after that deal was struck, but it seems wrong to publish data that was collected without the providers really knowing what metrics they’re being measured on.
Barry, I think you speak for the “silent majority” of the American healthcare consumers.
Too bad they are so poorly represented in the wonkosphere. But that is why blogs are growing so rapidly – they give voices to those who could not have been heard.
I think pricing transparency depends a lot on which of the three segments of healthcare that Amy Tuteur described (well care, catastrophic or end of life) we are talking about.
In the case of well care, where she said markets can function quite nicely, I don’t see why patients should not have easy access to prices for services like: routine office visit, new patient visit, consultation of specified time length, routine physical, mammogram, etc. Also, if the doctor offers a discount from the posted price for immediate payment, make that clear. If I can learn the appropriate CPT-4 codes ahead of time, my insurance plan should also allow me to find out what it pays that particular doctor for the service in question.
It would also be helpful if patients could go to a central website when looking for a doctor to begin with to learn where they went to medical school and where they did their internship and resident training. How long have they been practicing? What specialties, if any, are they Board Certified in? How many times, if any, have they been successfully sued for malpractice?
For hospitals, the discrete services that lend themselves to pricing transparency should have it — MRI’s, etc. If Medicare rates for the 500 plus DRG codes were available, that would at least provide a general sense as to what a given procedure should cost.
I know we’ve been over this in some detail before, but I think pricing transparency, along with appropriate quality / risk / outcome information contributes to more efficient resource allocation. Second, from a personal perspective, it just makes me viscerally nervous and upset to consume expensive services with no idea ahead of time as to what my own financial responsibility will be. Even if insurance is covering most of it, it still bothers me. Maybe it’s just me and nobody else feels this way.
Hey Matthew, I am not saying you are not enlightened. Very much to the contrary. But I question the word “few” and assumption that consumers cannot be trusted to make the decisions right for them.
Regarding car vs. spark plug analogy, I think it is best to lay out all of the options with as much info as possible. Include both line items and various bundles. Line item pricing transparency would help fix the most egregious disparities, including what individuals vs. insurers pay. Yes, many people will be confused and would want to go for bundles, where it would be up to designers of plans to come up with sensible packages and offer a “volume discount”.
There are people (me included) who like to unbundle deals and pick every component on its merits. There are those who would rather trust someone to offer a package and that is fine too. I enjoyed my car’s 4 years of maintenance included with sale, but now that it has run out I shop fee-for-service.
I come from the country where there was only one brand of cheese for consumers to buy. The first reaction by people from former Soviet Union coming to US used to be getting dazed at the choice in a supermarket. But then everybody would quickly figure out what is right for them and enjoy the choice.
But you are right that from the health system design perspective we do have to watch out for adverse selection in insurance. BUT, it is only a problem when we are talking about insurance, not self-pay market.
Let the market set the price for line items and then have insurers package them into bundles right for the population.
My word, I’m supposed to be all humble now! C’mon Dmitry, give me license for a little sarcasm–you’re allowed to “respectfully disagree” about my level of “enlightenment!”
If you can get past the computer TCO piece, what about my car vs spark plug analogy? I’m not afraid of people knowing the costs, but it needs to be “what will it cost to look after this group of people for a year for a pre-defined se of services”? Otherwise as I said to grace Marie, we’ll all be competing for the cheapest foot amputation for diabetics when we forgot to price out the maintenance care the ignorance of which got them to that dires state in the first place.
I think Jack is right. We need a measure that makes sense for consumers, and their ability to make rational choices between their options. But that doesnt include allowing the slicing and dicing of the insurance market in such a way that sick pepke cannot access it and cannot make those choices.
Rule #1 for winning people over: Do not tell them they are stupid and you are part of the “enlightened few”.
Consumers do wonders to find & compare valid information whenever it is available and make the choice, right for THEM. Of course this means every kind of info (not only pricing) needs to be available and then someone has to do the work of presenting and explaining it to them. We have ways to go to get it right, no doubt about it.
TCO is just another example of voodoo economics. Typical TCO cases used in sales process are pulled out of thin air and convince very few people. Real measured TCO always depends on how exactly someone is using a product. Even then attribution of costs to a project (other than acquisition price) is often in doubt.
Be not afraid of letting markets work their magic. They do not always fail.
Total cost of ownership and quality of service are obscure metrics that no one will really understand or care about. Greater communication about the true cost and what patients think about their care are the bottom line for almost all of us. We need a consumer’s guide for healthcare services that measure the patient’s perceived value of the care they received. Let the patients tell the story, not the statisticians.
I respectfully disagree. Having been formerly involved in sales and marketing in the high-tech industry, I will say the only time “total cost of ownership” was talked about was when the discussion was intended to obscure a high price. If something was more expensive than the competitors, but there were specific feature(s) that justified the higher price than salespeople and customers had no problem breaking down and debating the relative merits of that feature(s). They had no problem questioning and understanding what was bundled and what was not. Only when the difference in price was embarassingly obvious and was only weakly defensible did the term “total cost of ownership” emerge.
The problem with transparency isn’t that it exists or doesn’t exist; the problem is that it is meaningless since it only exists for a single dimension – price. The consumer is ill-equipped to make almost ANY relevant comparisons with regard to outcomes. The only thing CDHP or HDHP have right is that the demand curve for healthcare is downward sloping – that’s axiomatic. The problem is that even with totally transparent pricing consumers can’t know what they are buying so unless the only objective is to spend less total dollars without regard to outcomes, CDHP and HDHP are a sham. (I know I’m preaching to the choir for some here.)