Category Archives: Uncategorized

How Physicians Get Paid, the Arcane Differences Matter

Nice article by Katie Jennings in Politico

Physician compensation depends critically on the Relative Value Unit (RVU) scale set by the the AMA/Specialty Society Relative Value Scale Update Committee (RUC).

This area bears continued scrutiny.  It may seem arcane, but we are talking about half a trillion dollars of public expenditure.

1) There is a chronic undersupply of primary care docs and chronic oversupply of specialists.  Does not this indicate that we are systematically underpaying PCPs and over paying specialists?  Are not there economic models for what prices should be?  How badly skewed are the AMA RVU values?

2) Why does CMS continue to accept this system?  Any way to estimate how much biases in the RVU scale are costing Medicare?

3) Congress has specified the use of these codes by law and prohibited judicial review of the RVU system.  Is the committee de facto performing a government function and therefore subject to open government regulations?  Are committee documents subject to the Freedom of Information Act?

3) Who are the members of the AMA Committee?  Do they file conflict of interest disclosures?  What conflicts have been disclosed?  Aside from the final list of CPT codes and RVUs, does the committee keep and publish minutes, notes or other material?

4) The composition of the RUC (see AMA page) is heavily biased towards specialists with only 3 members from primacy care areas (internal medicine, family practice and primary care) but 30% of all physicians are in primary care and more are needed.  The AMA sites states that composition reflects Medicare expenditures but should it not instead reflect Medicare needs?  Does CMS have any input into the composition or membership of the committee?

5) Is the AMA copyright on CPT codes enforceable?  The final physician fee schedule is the result of a federal rule making process and is published in the Federal Register.  Does not this make the fee schedule including codes and definitions the result of government work and therefore uncopyrightable?  If the copyright on CPT codes and their definitions is enforceable, is CMS giving away public property by incorporating copyright codes and definitions into their reimbursement scale?  The AMA apparently makes a considerable amount of money licensing this material.  Anyone know how much?  The CMS website specifically prohibits public disclosure of CPT codes and which inhibits public discussion of reimbursement practices.  Does this violate the First Amendment?  For the purposes of promoting public discussion of RVUs, someone might want to publish the list of codes and definitions on a public website, but to do this they would have to license the material from the AMA.

Patient Portals, Personal Medical Records and Confidentiality

With many physicians and providers moving to address Meaningful Use 2 requirements, there is a push on to get patients to use patient portals.  It is time to start paying close attention to how your provider’s portal is implemented and how this affects the confidentiality of your health data.  A colleague recently noticed that their physician’s office was using the Jardogs FollowMyHealth(TM) PMR to implement a patient information portal.  When they read the terms of service agreement, it stated that they as a patient were releasing their health records to the vendor, not the patient’s provider and therefore that use of the data was not covered by HIPAA.  In follow up, another colleague who is on the faculty at very well known school in Boston noted that the school was offering a similar service with similar terms under the guise of allowing patients the freedom to move their data between providers.  Until this discussion, he had not read the Terms and Conditions in detail and was disconcerted to learn that he had agreed to release his medical record for redistribution.

Jardogs was recently bought by Allscripts, the second largest EMR vendor in the country. Their Terms and Conditions are available on line

Several items are of concern

– Jardogs can change terms at any time without notice

– Changes and corrections you make to your PMR will NOT be transmitted back to your provider

– You give Jardogs the right to redistribute your information

– Jardogs specifically asserts that they are NOT a covered provider under HIPAA

There is a general (and reasonable) assumption on the part of patients that if a provider say “We are using a portal now.  If you want to sign up, you can get your lab results etc. on line.” that this is a HIPAA protected service.  Provider patient information sites, certainly encourage this view, see for example:

Vendors like Jardogs are offering the healthcare provider a low cost solution where the patient agrees to give them as a third party PMR the right to access the patient’s PHI and the PMR will provide portal services without all the liabilities of HIPAA/PHI as well as the potential revenue of selling advertising on the site and reselling patient data.  Obviously, the Jardogs solution will be much cheaper for the health care provider compared to implementing an internal HIPAA compliant site with all it attendant security requirements and liability.  There has been some discussion as to whether Jardogs is correct in asserting that their Terms and Conditions releases them from HIPAA  If they have an ongoing business association with your provider, they may in fact not be able to escape HIPAA despite their Terms and Conditions.  Nevertheless, Gresham’s law, the inexpensive solution is likely to win out.

To be fair, there are pluses and minuses to either solution:

Third party PMR

– Able to integrate data from multiple providers

– You control your account, it moves with you if you change providers

– Login/authentication are likely to be more convenient albeit less secure

– No guarantee of HIPAA compliant confidentiality

– Limited ability to collect damages if a breach occurs

Provider portal

– Confidentiality is guaranteed by HIPAA, and the provider faces major liability for breaches

– Only carries data from on provider.  If you see multiple providers you will need to go to multiple sites

– If you change providers, your previous providers data is unlikely to ever get incorporated into another providers EMR/portal

– Will require two factor authentication which is somewhat cumbersome to setup and maintain

To meet Meaningful Use 2 requirements, providers must demonstrate that some fraction of their patients are using portals and electronic communications.  Jardogs is advertising itself as a certified solution to MU2 requirements

There is going to be a push on the part of many providers to get patients onto portals to meet MU2 requirements.  Patients need to be aware that there is a difference between a portal operated by their provider and a PMR operated by a third party.

Ebola versus SARS

Worth looking at the way we responded to SARS vs. what is happening with Ebola

SARS timeline

The first cases of SARS in humans probably occurred in November 2002 but the WHO was not notified until February 2003.  Within a year, we had:

  • Identified a previously unknown pathogen
  • Obtained a complete genome sequence
  • Developed rapid PCR based diagnostic tests
  • Deployed these tests in the field
  • Tracked all known cases across three continents and several major cities

By May of 2004, a little over a year after the WHO was first notified, the epidemic was completely controlled and there have been no new cases since.  Note that SARS is a respiratory pathogen transmitted by aerosol.  It is far more contagious than Ebola and the epidemic rapidly spread to major urban centers.  From a public health perspective, SARS was an extraordinarily challenging threat and yet under 1,000 people worldwide died of SARS.

Ebola timeline

The first case of the 2014 Ebola epidemic probably occurred in November 2013 and the WHO was notified in March of 2014.

The point is that if we devoted anything close to the resources that were deployed in containing SARS, we could shutdown the Ebola epidemic in a matter of weeks.

The Brutal Truth We Are Not Talking About on Sovaldi

David J. States MD PhD FACMI
August 7, 2014

There is a paradox in health economics. On the one hand, the recently introduced drug Sovaldi is a wonder drug. It has a 90% success rate in curing a bad disease, hepatitis C, and it has relatively few side effects. The cost of treating is a patient with Sovaldi is high, $84,000 but much less than the cost of providing medical care for these patients which is even more expensive and can often include a liver transplant (1) . Some estimates for the lifetime medical care costs of hepatitis C run as high as $600,000. So the introduction of Sovaldi should be greatly reducing health care costs and improving the economics of Medicare and Medicaid. The paradox is that the high costs of Sovaldi threaten to bankrupt the system. Why is this?

One argument is that we are pushing forward all of the costs of future management into an expensive current day cure. But if this were the only explanation, the long term effect on health economics should be positive. We are removing $6 of future cost for every $1 we invest today. It sounds like a great bargain.

Were we ever really planning to spend the money required for the long term care of all hepatitis C patients? We perform about 6,000 liver transplants a year in the US (2), but we expect almost 120,000 prescriptions for Sovaldi to be written (1). The number of liver transplant operations performed each year is limited by the number of available donor organs as well as the capacity of surgical teams and medical centers. Even if all liver transplants were performed for hepatitis C, it would take 20 years to complete all of the necessary transplants at current capacity. In fact, hepatitis C is only one of many indications for liver transplant and not even the most common (2). We should not be counting the costs of avoiding the need for liver transplants in the benefits of prescribing Sovaldi because the reality is that we do not have the capacity or organ supply to perform all the transplants that would be required.

There are other drugs available to treat hepatitis C. The lifetime cost of these alternative drugs is comparable to that of Sovaldi (3), but all of these drugs have more serious side effects and none of them is close to being as effective as Sovaldi. Given the prospects of side effects, high co-pays of many prescription drug insurance plans and limited chance of success, most patients opt not to pursue these treatments.

The brutal truth that we are not talking about is that we were never planning to treat all of the people who currently have hepatitis C. Sovaldi is a shock to the system because the Medicare and Medicaid laws require us to provide care when it is available, and now that a safe and effective cure is available, we expect a large number of patients to avail themselves of this treatment instead of suffering inexpensively in silence.


Graduate Medical Educational – An Alternative View

David J. States, MD PhD FACMI

August 5, 2014

As a result of an aging population, the United States faces  substantial increases in demand for medical care over the coming two decades. The AAMC has argued that in the next 6 years there will be a need for 90,000 new physicians, roughly a 12% increase in the physician workforce (1), and they have argued that a dramatic increase in the number of graduate medical education (GME) positions is needed to meet this need. In fact, these estimates for the demand for medical care may significantly understate the need. Not only is the US population expanding, it will also become increasingly elderly. Further, while there are some encouraging signs that the number of obese children has stabilized and may actually be declining, more than a third of American adults are now obese and the number of obese adults continues to increase (2). These obese patients are at increased risk for arthritis, orthopedic injury, diabetes, coronary artery disease and cancer, and likely will require more medical care than historical averages.  The AAMC estimates were also made prior to the implementation of the Affordable Care Act.  The ACA has led to a dramatic increase in the number of insured patients.  These newly insured patients are not unexpectedly utilizing health care services at a greater rate because these services are now more accessible to them.

There are two problems with increasing GME slots to meet the anticipated need for additional medical services. First, we cannot increase GME slots quickly enough to meet the need. Second, while there is no dispute that we are facing increasing demand for medical services, the assumption that this can only be met by increasing the number of physicians ignores substantial changes in the organization and technology of care delivery that have occurred over the past decade and are continuing to occur.

Can we ramp up GME slots quickly enough to meet the need?

Adding additional slots to an existing GME program is not a trivial endeavor and establishing a new GME program is even more demanding. To increase the number of training slots, a program would first need to add training faculty. Faculty hiring is a slow process beginning with the development of a plan to pay for the faculty salary, advertising, interviewing candidates and making offers. Once hired, a faculty member is expected to remain with an institution for decades. Programs cannot simply say “We will need 90,000 new physicians in 6 years so let us increase GME training slots by 15,000 per year from 25,000 currently to 40,000”. And then somehow shut the system down after these new physicians were trained. The barriers to establishing new GME training programs are even greater. Accreditation alone can require years to complete. It is not unusual for a decade to pass between the decision to create a GME training program and graduating the first successful trainee.

Even if the AAMC request to add an additional 4,000 GME slots were granted and programs could expand immediately (meaning add 4,000 new slots to the incoming class for 2015), these physicians would not complete their training until 2018. In other words, we would only have 12,000 additional physicians while the anticipated need is for at least 90,000 new physicians in 2020. The fact that Congress essentially froze GME funding in 2002 and declined to increase GME slots as part of the Affordable Care Act in 2012 (3) means that we long ago passed the time when increasing GME slots would allow us to produce the numbers of fully trained physicians that will be required to meet the increase in demand for health care services of our aging, obese and newly insured population using our current health delivery approaches.  The relevant question is not “How many physicians and GME training positions do we need?” but instead, “How do we meet our rapidly increasing health care needs given a relatively fixed supply of physicians?”

Organizational and technological changes in health care

The real impact of technology on organizational productivity is rarely seen at the level of the individual worker.  Rather it is through the structural changes in the organization that technology enables. Secretaries do not type all that much more quickly on word processors than they do on typewriters. Instead, the boss now communicates by email, blogs, text messages and twitter and no longer has a secretary.  We are changing the organization and structure of health care delivery more quickly than we can ramp up GME training.  Only a few years ago, the vast majority of physicians were independently practicing professionals. Today, more than a third of hospitalists are hired as employees and the fraction of physicians in all specialties working as employees is increasing rapidly (4). Modern corporate health care organizations cannot ignore the economic imperative to achieve efficiency by replacing expensive employees with less expensive staff supported through improved automation. Rather than hire 6 physicians to staff a clinic, we will see teams with two physicians, 6 physician assistants (PA) or nurse practitioners (NP) and several lower skilled medical assistants.  Effectively, the organizational change will triple physician productivity even though much of the care is now being delivered by PAs and NPs.  We are also seeing standalone “Minute Clinics” and similar solutions where one physician medical director may supervise a network of hundreds of PAs and NPs offering limited but rapidly accessible care.  Some states are also beginning to allow MD and DO graduates who have not received GME training to practice as “assistant physicians”.  In rural areas, telemedicine and telemedicine supported PAs and NPs offer an attractive solution to address long standing needs.  Modern telecommunications, electronic medical records and clinical decision support systems are critical in making these changes possible.  Many studies have shown that patient satisfaction is determined mostly by the accessibility of care and relatively less by the credentials of the person delivering the care. While scope of practice laws currently limit the roles of NPs and PAs providers, patient demand combined with insurance company and corporate provider lobbying are likely to relax these limits in the near future, as the Institute of Medicine has recommended (5).  The market can and is adapting to meet the changes in demand.

The changing role of Graduate Medical Education

To address our rapidly increasing need for health care services, the solution is not to increase the number of GME slots or federal funding for these positions. Instead, GME programs need evolve their curricula to better match the needs of a dramatically changing physician work environment. Knowledge of organization behavior and personnel management are becoming more important for physicians as is an understanding of team dynamics and ethical behavior in a corporate environment. These are skills that have not traditionally been a major focus of graduate medical education. Similarly, it will be critical for physicians to understand information technology so that they can effectively deploy computational tools and use IT to drive organizational innovation.  We need to redirect public funding for GME to support new and innovative solutions (6).







EHRs are awful, we can do better

Medical records have always been about communication.  Communication with yourself across time, communication with other providers and yes, communication with billing and finance.  Categorical data is necessary, particularly for the latter, but communicating uncertainty and alternative interpretations has always been an important aspect of medicine, and it is very difficult in the context of a modern EMR.

Informatics is in a hole with administration because most EMRs do not increase provider productivity.  There is some evidence the EMRs increasing billing per encounter, but too often this does not offset the increase in time required for providers to complete an EMR note.  Billable encounters per provider hour is a critical administrative metric.  When informatics proposes enhancements and additions to the EMR, administration is thinking is “You have already whacked encounters per hour by 20% and now you want to add more?”

But, for all the enormous resources devoted to IT in a modern healthcare organization, a surprisingly small fraction of the compute power is actually devoted to the central task of communicating among providers about patient care.  Most of the money going to IT is about security, billing, scheduling, inventory management and other necessary but ultimately administrative tasks.  And most of the big dollars going to all those consultants is about maximizing reimbursement rather than optimizing care.  By comparison, the task of hosting a form based web interface for providers to view history and test results and to enter EMR data is a trivial task requiring only modest compute resources.

We can do better.  Much better.  Advanced natural language processing should be capable of holding an intelligent conversation with a provider about a patient.  Think Siri with a  database focused centrally on the patient at hand.  Note that this is a greatly restricted space compared to what generic Siri deals with now.  Forget stock markets, sports scores or the weather, just hold a conversation about a specific patient’s medical conditions.  Smaller scope of interest translates improved accuracy, basic multiple hypothesis testing.  This would still require a lot of compute power you say, but if it increased billable encounters per hour by 10%, it would easily be worth $10/hr per provider, and for $10/hr you can rent half a dozen maximal configured high performance compute systems on the Amazon EC2 cloud aggregating to hundreds of CPUs with hundreds of GB of RAM and terabytes of SSD disk PER PROVIDER.  That is a lot of compute power.  Are you seriously telling me we cannot implement:

Siri: How is Mrs. Doe doing this month?

Doc: She seems to be doing well.

Siri: So the UTI for which you prescribed Macrobid has resolved?

Doc: Yes.

Siri: Her blood pressure has been close to 140 over 90 on the past two visits.  Do you want to increase the dose of her Lisinopril or metoprolol?

Doc:  How much metoprolol is she on now?

Siri: Extended release 25 mg a day.

Doc: Increase to 50.

Siri: Changing metoprolol extended release to 50 mg daily.

Siri: Her glucometer dump showed a median blood sugar of 138 with no readings above 252 or below 83.  Continue her current diabetes regimen?

Doc: Yes.

Siri: Any new problems?

Doc: No.

Siri: Standard outpatient physical exam?

Doc: Yes, no concerns.

Siri: She is due for an A1c and annual diabetic eye exam.  Go ahead and schedule those?

Doc: Yes.

Siri: So we are going to bill for a moderate to high intensity visit
addressing follow up on the UTI, hypertension management and diabetes management?

Doc: Sounds good.

Siri: Here is the chart note for your review and signature.

And perhaps at some point, Dr. Siri will be speaking directly to the

We need a vision of informatics as enhancing communication and productivity.  No excuses for forcing people to work for the computer. With the compute resources available in modern environments, the computers can let the people work in the ways people are most comfortable and productive.  And yes, we can do better on billable encounters per provider hour.