Building Cost and Quality Into the Electronic Medical Record

Trends in US healthcare expenditures are financially unsustainable (1). I would like to propose two tweaks of the healthcare delivery process that may, in a small way, help rectify this problem.

Although there is a widespread impression that health information technology (HIT) will eventually “bend” the cost curve and put healthcare spending on a sustainable course, there is, as of yet, little data that convincingly supports this hypothesis (2).

Kaiser Permanente is a large, integrated healthcare delivery system which has invested heavily in HIT. George C. Halvorson, the chairman and CEO of Kaiser Permanente appears to have concluded that this investment will not solve the healthcare cost issue, when he was quoted in the New York Times (3/20/13) as stating “We think the future of health care is going to be rationing or re-engineering.”

Because HIT, as currently implemented, will probably not solve the healthcare cost problem, I would like to suggest a minor “re-engineering” of the electronic health record user interface which may help bend the cost curve.

At every office visit, the physician must make a myriad of decisions which incrementally effect the nation’s total healthcare expenditures. For example, the physician will have to decide which medicine to prescribe, and which radiology study or laboratory test to order.

In many situations, there is more than one acceptable choice. The physician’s ultimate decision will integrate their understanding of the disease process, the treatment’s side effect profile, their familiarity with the treatment options, patient preferences and many other variables.

I would suggest that every time a physician is about to order a test or a prescription, the cost of the test or prescription should be displayed to the physician. In the same vein, whenever a computer displays a test result, the cost of the test is immediately available to the reader. This information could then become an additional factor that the physician may choose to integrate (or ignore) at the moment when he/she is about to commit the patient and society (which is now paying >50% of all healthcare bills) to another healthcare expenditure. In terms of a risk/benefit analysis, I can see little downside to providing this cost information to physicians.

While the integration of cost information into the ordering process may help physicians minimize an individual patient’s medical expenses, it would also empower physicians to make treatment decisions that may promote the longterm viability of the healthcare system. While I acknowledge that some people will decry the notion that a physician has an ethical responsibility beyond the interests’ of their patient, I believe that physicians are morally culpable if they ignore the societal ramifications of their treatment decisions.

My second suggestion for “re-engineering” the healthcare delivery system would be more complicated to implement but is no less important.

I think physicians need to understand which treatments are truly “cost-effective” from a societal prospective. Toward this end, there needs to be a large number of medical-economic analyses which precisely define the “maximal acceptable daily treatment cost” for the major (or expensive) diseases.

Specifically, if a physician is treating a particular disease, how much medicine can the physician prescribe, in dollars per day, before the physician begins to consider the treatment as being “not cost effective.” This type of calculation, known as quality-adjusted life year or QALY has been used by the healthcare system in the past.

For example, diabetic patients will have fewer longterm complications and an increased life expectancy if their blood sugars are well controlled. However, each additional hypoglycemic (blood sugar lowering) medication will incrementally increase the short term healthcare expenditures. There are many hypoglycemic agents and the price of these medicines can range from a few cents a day to several dollars per day. While the selection of the “best” medication for the patient is very complicated (and beyond the scope of this article,) conspicuously missing from the physician’s thought process is an understanding of the cost-effectiveness (QALY) of the different treatment options.

I think it would be very useful if physicians knew what was the maximum daily expenditure for hypoglycemic medications which would be considered “cost-effective.”  This information could be derived from computer simulations and based on existing Medicare claims records. Once the modeling studies have been completed, an Expert Physician Panel could review the data and issue a Consensus Statement that says “Physicians should treat their diabetic patients with the most appropriate hypoglycemic agents and attempt to maintain a total daily expenditure on hypoglycemic agents that does not exceed $XXX/day”.

Clearly,  knowledge of the “maximal daily prescription expenditure” which is “cost-effective” could help physicians ensure the longterm viability of the US healthcare system and the health of the vast majority of our citizens.

Of course, if physicians don’t do this analysis ourselves, it will be imposed on society by the Federal government as a matter of fiscal necessity.


1. http://www.gao.gov/financial/citizensguide2008.pdf
2. Kellerman A, Jones S, What It Will Take To Achieve The As-Yet-Unfulfilled Promises of Health Information Technology. Health Affairs 2013;32(1):63-68.

Hayward K. Zwerling, M.D., FACP, FACE is an a board-certified internal medicine physician who specializes in endocrinology. He practices at the Lowell Diabetes & Endocrine Center in North Chelmsford, MA. Zwerling is also the president of ComChart Medical Software, LLC.

17 replies »

  1. Plenty of patients are doing that now. Is that also why patients are choosing to go to retail clinics? Besides the convenience? The utilization of those clinics has increased rapidly in the past few years.

  2. Just imagine the savings if the patient had to select what they would pay for themselves and then have to pay it.

    Nothing saves like the patient not going to the doctor in the first place.

  3. A really interesting analysis. EMR is essentially needed for the following reasons. Real Time reporting of Medical Information. Safeguarding the Patient Right to have factual information. Patient Safety is ensured by having the continuity of care enabled by the IT enabled solutions, whcih can be smartly enabled for Alerts for contra-indications or intra-reactions. Serves for rightful legitimate reimbursements from Payers. Serves as a good ground work documention against Malpractise and not the least ensures Accessibility to care with more and more Healthcare getting localised with Patients moving and shifting between service providers, the need to have accessibility to the earlier episodes of care is a must.

  4. I’ve certainly seen that. And they also don’t also have access to what is paid for with a specific patient’s benefit plan. I once saw a patient very unhappy because a physician assumed an IUD would be paid for by the health plan when it wasn’t, and the cost was surprisingly high to the patient. Another issue that comes up is when tests (lab and colonoscopies) sound like they’re “preventive” to patients (so are covered pre-deductible) but are not because the patient has an identified condition already that’s being monitored by the tests. Complex rules sometimes catch both patients and physicians off guard.

  5. Yes, patients should be involve in the decision making process and I think that physicians usually discuss the treatment options with the patient. However, the MDs are unable to tell the patient how much the test will cost because the MD does not know the cost.

  6. Your hypothesis remains to be proven. I suspect that HIT will not result in a decrease in the cost of healthcare and the comments from Kaiser’s CEO would suggest that your hypothesis was not substantiated by Kaiser, a company that invested heavily in HIT.

  7. Shouldn’t consumers be involved in these decisions? Perhaps they can help determine what they want to do to either evaluate a potential problem or to treat or prevent a potential problem. I believe the outcome is better and follow through is more consistent if they are part of decision-making. Perhaps consumers should – and they are getting used to it, I think – consider cost when making healthcare decisions.

  8. Showing the cost of the Rx or test to the physician is absolutely needed. I remember the day we made that change over 20 years ago in the VistA EHR system at the VA where I was the CIO. Within days, I was stopped by numerous physicians who said they never knew what it meant financially to the VA and to their patients when they approved a whole panel of tests or prescriptions. Behaviour began to change immediately – for the better. This is a no-brainer solution hat takes minimal effort to put in place.

  9. Great Post!

    Investing in IT solutions and maintaining the records for reference will help the physicians to contribute in controlling healthcare expenditures.

  10. Dr. Zwerling,

    Great post. There have been two interventions at Johns Hopkins along the lines of what you are suggesting. One study incorporates price of many common tests within the order entry system, and the other is a notification triggered when ordering redundant troponins. Both have shown to be cost-saving.

    As a software developer, I know that “HIT, as currently implemented, will probably not solve the healthcare cost problem.” Hospitals have neglected to invest in innovative IT solutions, and have not made it easy for physicians to develop products around workflow and usability. I’ve recently found, however, that under the guise of patient safety and, recently, cost-savings, they will listen.

  11. This is one of the best thing EHR vendors could do.

    This was also the subject of a recent AMDIS list serv debate… good timing

  12. Yes, I am aware that the “listed” price is many, many times the expected payment or the Medicare allowed charge.

    I believe that if facilities had to publish their list price, eventually, the list price will evolve until there is a realistic correlation between the listed price and the actual cost of the test.

  13. I agree that the price is a secondary issue. However, facility charges are often many multiples of what is actually paid, so posting a facility charge can be misleading.

    I was trained (in Radiology) many years ago by a professor who always emphasized the cost and likely yield of any study. And of course in those days we did a lot less imaging anyway.

    I am saddened by what I see today:
    – “Hit rate” for CT rule out PE is less than 5%
    – MRI of the L spine being wildly over ordered
    – Head CTs for kids that fall down
    – etc

    My old professor would have rolled over in his grave. And the worst offenders are the younger docs and the PAs. It does not bode well for the future.

  14. I think the selection of the price is a secondary issue.

    In the name of transparency, I think the price posted should be the facility’s charge, as this would “encourage” facilities to create a reasonable and realistic fee schedule.

  15. The idea that the cost of a test/treatment should be readily available to a physician at the time of ordering is a good one.

    The only problem is – what is the cost? Is it:
    What BCBS pays?
    What Medicare pays?
    What Medicaid pays?
    Self pay rates?