Eric's Blog

On August 20, 2009, The Daily Show with Jon Stewart included an interview with Betsy McCaughey, a former Lieutenant Governor of New York and the source of the assertion that the health care reform proposals currently under consideration in Congress would mandate “death panels” that would withhold medical care from senior citizens.  As is usual on The Daily Show, McCaughey came expecting a brief, largely superficial fluff interview with a comedian.  What she got was a respectful but persistent grilling by Stewart, who is one of the most acute intellects in the television industry, demanding that she produce support for her assertions from the text of the bill (the first volume of which she brought to the interview, presumably for that purpose).  Though she tried valiantly to do so, she had trouble finding the specific pages she cited, and when she did and read the passages aloud, the support was clearly lacking.  It was amusing to watch, to an extent, but also embarrassing – so much so that the medical products concern with which she had been associated, and on whose behalf she is using this canard to resist health care reform, has now terminated their relationship.

The legislation says, first of all, that Medicare will pay for end-of-life counseling for patients who request it.  Full Stop.  Elsewhere, it lists certain pieces of information that Medicare wants to collect from practitioners in order to monitor health care quality.  Full Stop.  Somehow, from these two points that do not even appear together in the legislation, McCaughey derived the assertion that they mandate patient monitoring by “death panels” with the power to grant or deny medical services.  Stewart rightly pronounced this “hyperbolic” and “dangerous”.  McCaughey, of course, is sticking to her guns, which forces us to ask whether she is as foolish as she appears, or whether she is determinedly cynical in using such hyperbole to stir resistance to the reform of the health care system that most of us know is absolutely essential.  The disruptions of town hall meetings orchestrated by lobbyists for the health insurance industry would seem to indicate the latter, but one could make a strong case for silliness, just from watching her in action in that interview.

Yet this affair touches on another issue that requires some serious thought.  The Department of Health and Human Services wants to monitor the quality of health care in a way that has not previously been possible, and to provide incentives for improving that quality based on benchmarking how health care practitioners do their jobs.  Potentially, this suggests an intrusion into the privileged communication between doctor and patient that makes many people very nervous, and rightly so.  If all the information about you can be gathered by a government agency, and if that information is in any way accessible to other parties, it is subject to abuse.  It can turn up in pre-employment background checks and result in loss of job opportunities.  Insurance companies could deny or cancel life or disability insurance coverage even if they could not deny health insurance.  Banks could deny credit by asserting too great a risk that you would become ill or die before you can pay off a loan.

HIPAA requires that patient data be kept secure, so a hospital or a medical practice has the responsibility that patient clinical records are protected from prying eyes within their systems.  But once an Electronic Health Record is transmitted from the practice or institution to any other body, the security and privacy of the data is out of their control.   As long as data is communicated among entities that are all subject to HIPAA, it is most probably secure.  But what happens when some of that data becomes available to entities not covered under HIPAA?

Data Mining refers to what used to be called “statistical analysis”.  By accessing, combining, slicing and dicing data from disparate public and private sources, it is possible to tease out confidential information on individuals.  As a simple example, suppose one has the following data sets available: aggregate pharmaceutical prescription sales dimensioned by patient demographics, but containing no individual identifying information; quality-of-care data from physicians with demographic dimensions; grocery store customer retention data that includes customer demographics and customer names; marketing information databases from warranty cards; and credit card transactions.  By combining the data from all of these sources it is possible to attach personal identification to health care records and identify patients and their medical conditions.  There are a handful of organizations that possess or have access to these kinds of information stores and make a business of doing precisely that.  Interested readers should seek out Robert O’Harrow’s “No Place to Hide”, published by the Free Press division of Simon and Schuster in 2005 for an eye-opening and scary view of industrial-scale data warehousing and data mining.

There are many good reasons that this industry exists, and we all benefit from it far more than we realize.  But it is a largely unregulated industry whose major clients are very rich, very powerful, unregulated or loosely regulated entities, including the banking and insurance industries.  The same health care quality data that will be collected by DHHS will almost certainly be demanded by the health insurance industry, and even if DHHS does not share data with Acxiom, for example, Cigna and United Health almost certainly will, thus bypassing the safeguards of HIPAA.

We all like to think that our personal privacy is guaranteed by the Constitution, though there is a body of legal opinion that says otherwise.  On the other hand, we all willingly give up much of that personal privacy when we opt into customer retention programs, or enter into a contract for insurance or other services, or apply for a loan or credit.  We now interact over the internet with so many entities via credit card that we have no idea just who knows what about us.  We can no longer control just how much information about our personal lives is actually available.  We need much stronger protections against abuse of this information than we are currently provided.

Electronic Health Records are absolutely essential to improving the quality, availability and affordability of health care services.  Less than a third of all health care practitioners currently use them, and DHHS seeks to increase that ratio significantly by 2011.  While practitioners and hospitals begin the process of adopting EHR, Congress should be considering how to prevent intrusions into patient privacy by commercial data miners, as well as abusive practices in the financial and insurance industries.

Let’s divert the public’s attention from the silliness of the Betsy McCaugheys of the world and focus on a real threat.

• Comments [1]
• Categories [Electronic Health Records, Health Care Quality]

“Meaningful use of certified electronic health records” will become effectively mandatory for health care practitioners by 2015, if not before, primarily because of the savings in the cost of claims processing for third party payers – public and private – due to eliminating the handling and storage of paper documents.  Furthermore, digitizing patient data facilitates the extraction of key performance indicators with which to develop objective measures of physician performance, therapeutic effectiveness and quality of care.

Objective measurement of practitioner performance and health care quality, like any similar objective measurement, is a matter of sorting and counting, a task at which computers excel.  First, identify a condition, an event, or an outcome related to a desired goal.  Sort the occurrences of this condition, event, or outcome into meaningful categories and count them up.  Repeat as needed until all the relevant conditions, events and outcomes have been identified, sorted and counted, and then analyze the resulting numbers to see which ones move together, or in opposite directions, or are causally related.  The resulting collection of statistics and heuristics can be used to guide decision making to control costs and improve quality.

For example, consider a group of patients who present with respiratory symptoms, some of whom are smokers.  One could extract from this the following statistics:

• Total Patients with Respiratory Symptoms
• Smokers with Respiratory Symptoms
• PerCent of Smokers to Total Patients with Respiratory Symptoms
• Additional variants on the above broken down by available demographics
  • Gender
  • Occupation
  • Ethnicity
  • Age

One can count and categorize in the same way the times the physician offers smoking cessation counseling, which measures the practitioner’s response to this condition.  The set of this and similar statistics can be collected and published in a physician scorecard that compares his achievements with the average scores for his region, state, and the total country, and with benchmark goals set by Medicare or other third parties.

If all the patient data on which such statistics are to be based is on paper, the cost to produce useful quality of care reporting is astronomical.  While medical charts do contain discrete data that could be digitized using optical scanning, the most valuable information is contained in the practitioner’s progress notes, which are free-form text entries.  Much of this information could be made discrete, but to be contained on a pre-printed paper chart would require a form many pages long, little of which would actually be used in any patient encounter.

The benefit of an EHR system is that every diagnostic observation can be defined as a discrete element, and only appear in a chart when actually needed.  The EHR implementation process includes the creation or modification of templates that the practitioner uses during a patient encounter.  A template includes a comprehensive set of observations relevant to the type of examination and conditions encountered.  It also provides for free-text entry of additional information relevant for the patient.  Templates can be designed or modified, moreover, to insure that tasks or observations required for performance goals (as set by Medicare for example, or pay-for-performance incentive plans) are addressed in every relevant patient encounter, thus encouraging achievement of higher performance and quality scores.

EHR is crucial to health care reform.  The American Renewal and Reinvestment Act of 2009 (ARRA) sets forth incentives to be provided by Medicare to encourage “meaningful use of certified electronic health records” beginning in 2011.  This is intended to give health care practitioners ample time to select and implement a certified EHR system, which takes from nine months to two years depending on the size of the practice.  While some practitioners view the incentives themselves as “disappointing”, the benefits of EHR are ultimately so compelling that they cannot be swept under the rug.

Physicians who do not take action now run several risks.  For one, those who choose not to adopt EHR at all will be penalized with lower reimbursements from third party payers.  Those who respond by not accepting any insurance at all may experience a drop in the number of their patients, whose aversion to doing battle with insurers is at least as great as the doctors’.  On the other hand, those who wait until circumstances and their business partners finally force them to adopt an EHR will probably try to get it done as a rush job.  This will mean a truncated requirements specification and selection process, but a protracted implementation process as they struggle with a hurriedly selected product that is not a good fit.  It will be a far more costly and ultimately unsatisfactory experience as a result.

Doctors have expressed dissatisfaction that neither Medicare nor private insurers will pass much of their savings on to them.  They are being asked to invest tens of thousands of dollars in computer systems, but will be given back no more than 11% of the cost savings that will be realized by third party payers.  But practitioners should recognize that their own costs will be similarly impacted once they have converted to EHR.  Many tasks and functions currently required to handle paper patient charts will vanish with the adoption of an EHR, which can save the practice enough to recover the system cost in well under five years.

Still, it is fair to say that from the point of view of an investment decision, the incentives authorized in the ARRA beginning in 2011 are not strong enough by themselves to stimulate investment in EHR.  The most compelling benefits of EHR are not monetary.  But they are considerable.  Health care professionals should recognize that by not making the switch from paper charts to EHR they penalize themselves and the rest of us as well.

• Comments [2]
• Categories [Electronic Health Records, Health Care Quality]