Centralized Medical Records, Behavioral Modification and Security

The currently debated “stimulus” bill has about $300 million in it for establishment of an electronic medical record system for the government. I mentioned this briefly in a previous post but think it should probably be explored in more depth.

The idea of a central repository for all medical records has great merit in an ideal world. Information would be available, no matter where a patient was located, about past medical history, allergies, current medications, and past surgical history. Since everyone is altruistic and no one would access information that was not required, the system would be wonderful in an ideal world. We don’t live in an ideal world.

A central repository of any information of value is a hackers dream. That information would be so valuable to obtain that it could be sold surreptitiously to potential employers, insurance providers, for personal motives, etc. by anyone with access and enough temptation. Miscreants could erase or modify the records for profit or for malicious reasons. The government would counter with a back up system of records. That just provides a broader target to access the information. Even without hacking through computer safeguards, the records could be compromised by individuals through bribes or other personal motives. This is a relatively common occurrence when it comes to national security issues in the United States. People with agendas or personal political philosophical differences with the government pass information to liberal publications like the Washington Post and New York Times which routinely publish classified information. Even more recently, out of one hundred forty anonymous screening tests for baseball players concerning steroids, only the name of Alex Rodriguez was leaked to the press. Someone clearly had an agenda against Alex Rodriguez. It is easy to see similar leaks for candidates for office, other sports figures, celebrities, and for personal vengeance. A history of a sexually transmitted disease, elective abortion, or potential debilitating disease could be used for extortion or political blackmail.

The breadth of the system is inherently a security and bandwidth problem. In order to be effective, the records will have to be available to any medical facility in order to care for patients wherever they present for care. That means that tens or even hundreds of thousands of facilities will have access to the system. A system that large will be similar to the AHLTA military system which is slow, unwieldy, and crashes regularly. A system with that kind of capability probably doesn’t exist outside of the military or National Security Agency anywhere in the world. An electronic health record system which is undependable would be a danger to patients as necessary patient information would not be available when required. Additionally, with the many thousands of access points to the system, there is no practical way to adequately maintain the security of the system. It is like a fence with thousands and thousands of gates, any one of which when breeched makes the entire system accessible.

As a physician, I am well aware of the tendency of health care administrators to now view patient care as a “product line”. In a group of physicians, I have once actually been instructed by a senior administrator to increase patient throughput even if it increases risk of inadequate diagnosis. The reason I bring this up is that there is a tendency for anyone under a time constraint to avoid “reinventing the wheel”. Errors in records in a central repository will propagate because the information will convey authority and go unchallenged. Errors will be like bad tattoos and go with patients everywhere.

To use the hackneyed Orwellian analogy, all health care information in the hands of the government is not a good idea. I will point out that while the government will be a problem, it may not be the biggest problem. With a records system that will inevitably leak like a sieve, it is only a matter of time before private insurance providers get their hands on the information. With that information in hand, insurance companies will begin to modify their actuarial tables depending on your private information. If they note a minor injury while surfing, skiing, sky diving, playing contact sports, riding, etc. you will have the option of stopping the activity, not being covered while doing it, or paying higher premiums. This type of focus will eventually lead to widespread financial behavioral modification. The analogous government scenario is to force behavior modification by denying benefits for activities, diet, etc. that are not approved by some group of “experts” who decide what they want you to do and not do. This is not a new argument. There was a suggestion years ago that in states with motorcycle helmet laws, if you didn’t wear a helmet, there would be no government funding for your health care should you be injured on a motorcycle. The idea was fairly popular. The same has been suggested for bicycle riding for children and cigarette smoking. The idea being that the government says out of one side of the mouth, “We aren’t forcing anyone to change their behavior” while making the behavior so financially untenable that no sane person would engage in it. It is a slippery slope that eventually leads to only government and insurance approved lifestyles being covered. Anything else leaves you on your own.

One advantage proponents will claim is the ability to gather data about treatment options and outcomes. Former Senator Daschle has already written in his book about using those data to determine which treatments will be available. In other words, the financial behavioral modification will not only be used on patients, it will be used to force physicians to become robotic in their prescribing and treatment patterns. Once again, it will be “Treat the patient in the way you see fit, but we will only pay you if you do it the way we want.” If patients cannot be treated individually, we might as well scrap the entire health care system and have everyone use a site like WebMD to treat everyone. There is no reason to talk to patients and examine them if it isn’t going to make any difference in what you can do for them.

My counter proposal, which I have actually sent to Senators Graham and DeMint, is to not have the government establish an actual computer system but rather to establish national standards for the electronic medical record. The format of the record, type of computer storage file, and transmission protocols could be specified and monitored by the Department of Health and Human Services. In the past, similar standards have been established for radio and television transmission, automobile safety, food safety, air traffic control systems, and other systems. My suggestion is that a commission or committee of overpaid top notch experts to establish standards has miniscule costs compared to establishing and building a huge overpriced ineffective system of computer networks. Additionally, in a free market economy, software developers will come out of the woodwork in droves to develop and market commercial versions of software to make new electronic records and convert existing records. To paraphrase Field of Dreams, “If there is money to be made, they will definitely come”. Having private companies supply the software will create many high-paying jobs and result in substantial tax revenue to the government, instead of costing huge amounts in government payroll. Additional jobs will be created all across the country when people are hired by health care facilities to convert existing records into the standardized electronic format. The government virtually never gets the appropriate value for its contracts, there is no reason to assume that it will on this one. Therefore, in the spirit of the American entrepreneur, to save hundreds of millions of dollars in taxpayer money and to avoid the Orwellian consequences, I believe the system I am proposing is superior to the current thinking.

Electronic Medical Records

A lot of discussion has gone on recently by the proponents of electronic medical records, including the current administration. I have discussed them before but wanted to take another look, including the views of a user. The military has already implemented an electronic system which was originally called CHCS (Composite Health Care System). When it was first introduced, it provided some valuable services despite its user unfriendliness. It allowed students and physicians to access laboratory results and radiology results without having to trek down to each department and go through a log book to retrieve them. When I was first training in medicine, that is exactly what we had to do each morning and evening. From that stand point, CHCS was a boon to efficiency. Those same tests could be ordered in the system although there was a learning curve as the names of the tests were not very intuitive. There were a lot of problems with the system. For example, if a laboratory technician entered an order from a drop down menu under the incorrect physician’s name, the results of the test would be returned to that physician and that physician would be asked to electronically sign the order. There was no way to correct the mistake. Even the administrators of the system could not fix them. Physicians were constantly receiving results and signing orders on patients they had never met. Rather than quibble on the other inconsistencies and problems with that system, let us move on.

The next iteration of the system was originally CHCS II, but has subsequently been renamed AHLTA (Armed Forces Health Longitudinal Technology Application). The AHLTA system has many of the same advantages in that no matter where you go (in theory) your drug allergies, medications, past medical history, etc. are available to a new physician to whom you present for care. That is not entirely true as I have been deployed eight times in the past five years and have never had access to the AHLTA system on any deployment. While the AHLTA folks consistently crow about the advantages of the system, they never discuss what it is like to use it. Let’s look at some of the results:

The Navy has decided that in order to improve efficiency, they want the physicians to see seventy five per cent of the productivity of major Health Maintenance Organizations (HMOs). Part of the reason is to lower costs, and part of the reason is that they are having trouble keeping enough people in the Navy to serve the patient population. In the brick and mortar hospitals, the current requirement for a General Surgeon is to see forty five clinic patients a week (based on HMO productivity). If that is only seventy five percent of what an HMO doctor sees, why should that be a problem? First, the average physician in civilian practice averages four support personnel per physician. The civilian practices know that the only person in a surgery office that makes a dime for the practice is the surgeon. Therefore, people are hired to keep the surgeon from doing anything which is not billable. Record keeping, billing, coding, telephones, communication, cleaning, education, etc. are all handled by trained assistants, freeing up the surgeon to do surgery and outpatient care for which billing can be generated. In the Navy, there are four physicians for each support person. Those people are Civil Service and we all know how much incentive the Civil Service has to work hard: none. Therefore, Navy physicians type their own notes, write their own consults, fill out their own preoperative and post operative packages, make phones call, schedule procedures, etc. I forgot to mention above that those forty five patients have to be seen in only two clinic days a week. Broken down evenly, this means twenty two patients on one clinic day and twenty three on the other.

Meanwhile back at the AHLTA system: Even after the training required to use the system, it is so cumbersome that trying to find your way through it is very time consuming and frustrating. I know how bad it is for surgeons. I can’t imagine how bad it is for primary care physicians who see such a wide variety of diagnoses. There are templates which can be customized for a specialty, which can save time but any unusual case will take a lot time to do the note. Additionally, the system can not keep up with the amount of usage so it is slow and crashes regularly. I have found the average AHLTA note to take between ten and twenty minutes to do even with a template. Therefore, twenty two times fifteen minutes equals three hundred thirty minutes. That is five and one half hours a day entering notes into AHLTA. If a surgical appointment is 20 minutes long, that means four hundred and forty minutes, or seven hours and twenty minutes a day talking to and examining patients. So far, we are up to twelve hours and fifty minutes each clinic day. Why do patients go to see a surgeon? Because they might need an operation. Assuming that only half of the patients need surgery, that means eleven preoperative packages (preoperative orders, admission paperwork, consent forms, notice to parent command, history and physical examination, etc.) need to be completed in addition to the AHLTA note for those eleven patients. Just for argument, let’s say that they only take fifteen minutes each (not realistic). That is another three hours and forty five minutes a clinic day. Now we are up to fifteen hours and thirty five minutes each clinic day. That is assuming that the physician never goes to the rest room, eats, or takes a break from paperwork. Additionally, the physician needs to round and care for inpatients which requires going to the ward or intensive care unit. Does anyone wonder why people leave for increased salaries and lots of administrative support?

Another phenomenon of the electronic record in the military is that performance evaluations have become partly based on the amount of relative value units (RVUs) billed as interpreted by the coding of each visit by the system. Subsequently, the smart physicians have learned to “game” the system and turn every visit into a coding bonanza with their templates. The coding is not based on what you do, it is based on what you write. Therefore, the smart ones learn to produce templates that over code visits by entering extraneous and unnecessary data into the AHLTA system. If you do what is considered a normal work up on a patient, you will fall way behind the coding curve. Since everyone is now forced to play the competitive coding game for advancement, even more time is spent on the computer entering redundant and irrelevant information.

Another interesting thing about the AHLTA system which has not been improved is that information, once entered is almost impossible to remove. Errors and diagnoses follow the patient around like a bad tattoo.

I wonder about the wisdom of a central repository of private medical information. I can see not only hackers but the Orwellian use of that repository for behavior modification and other abuses. It will not be hard for advocates to argue that anyone with a risky hobby (e.g., surfing or skiing) or diet should pay more for insurance or might not be looked on as favorably for employment. The proponents will argue that the privacy of the system will be guaranteed but if it is available to any physician (as it has to be to provide medical safety), it will be available to nearly anyone eventually. I believe a better alternative is for the government to establish standards for the format of electronic medical records and have them stored locally at physician offices or hospital records facilities. If records are required in another location, a secure and verified request can be made for those records and they can be instantly electronically transmitted to the requesting location. That system would also allow tracking of who requested records and for what reason. That should reduce the incentive for anyone to pry unnecessarily and would increase privacy.

Proponents of the electronic medical record argue that they will reduce medical errors by having allergies and medications available. That is intuitively true. I wonder how many errors will be produced by the lack of time to talk to and examine patients caused by the increased requirement for computer time to input data. If you want to reduce errors and provide better care, get some support personnel for the physicians so they can spend time taking care of patients and have some “scribes” enter the data into computers.