Sunday, December 26, 2010

Open Architecture Electronic Medical Record (EMR)

“In an open architecture, components have well-defined, published interfaces that allow interconnection and use in ways other than as originally implemented or intended. They allow interested parties to expand the functionality of the system without modifying existing components.”[1]
This plea could apply to the use of an off the shelf database system customized for patient records and bio-medical information. A relational database can extract correlations between the two, offering a differential diagnosis. Each specialty or student can modify the posting forms and the reporting forms to meet their own need.
Arguably, today’s practice places far too much emphasis on treatment, best evidence, and far too little on diagnosis. Many if not most diagnoses prove wrong. Correcting the record or even learning of the insufficiency remains difficult. Reasons for these discrepancies include: insurance clerk posting the most reimbursable diagnosis, making a snap diagnosis without exploring underlying or concurrent problems, failure to consider all the possibilities, lack of environmental and epidemiological information, expense and inaccessibility of current biomedical information.
A prodigious number of records accumulate over the life of a practice, more than in most clinical studies. Data mining will reveal statistically significant correlations peculiar to local population and trends. The diagnosis will evolve over time. With time, the physician can apply statistical probability to the differential diagnosis. Intriguingly, it becomes possible to correlate the patient’s genome with clinical condition and outcome. Every patient becomes a well-documented study of every event from treatment to genomics.
The practitioner may link-up with a medical school or research center for support, billing and education. Linkage provides a continuing real-time updating of medical information, diagnostic criteria and best evidence. The institution receives a significant contribution to an anonymous database that correlates genomics with clinical experience and provides a horizontal cohort study of everything else on every patient. Informed consent is no longer an issue.
In progress


[1] Deborah  Estrin and Ida Sim Open mHealth Architecture Science vol 330 5 November 2010 p759

Translational Medicine & the Health Care Debate

Translational Medicine may never see the light of day with our health-care-by-committee solution. None of the solutions put forth in the current healthcare debate take into account the rapid advances in basic medical knowledge or the need to translate this revolutionary new knowledge into hands-on clinical practice.

Researchers and medical educators coined the term Translational Medicine to encourage bio-medical research leading more directly to clinical application. Clinical practice changes slowly. New ideas engender caution if not suspicion, and for good reason. Many new ideas prove wrong a decade later. The rapidly evolving evolution in the basic science of medicine creates the need for trusted research in clinical applications and a credible translation of new knowledge into practical clinical tools. We need accelerated medical education to keep pace with the pace of bio-medical discovery. There has always been a gap between medical research and clinical medicine. Translational Medicine is a much-needed strategy to fill that gap.

The rapid advances in basic science affects as great a change in thinking as occurred when our fundamental knowledge leapt from an understanding of the anatomical structures of the human body to understanding what these organs actually did. Today’s medical science takes a quantum leap, from a traditional understanding at the cellular level, to an explosion of knowledge at the molecular level. This tsunami of information pouring out of research institutions presents a new dimension in genetics, microscopy and the science of medicine. This new world on the nano scale, unleashes a vastly expanded view of these molecular interactions. We now visualize the DNA and protein molecules directly with high-energy photon microscopes, some even capable of high-speed video imaging of actual multi-step chemical reactions.1

For example, researchers at the Cardiovascular Research Center at Massachusetts General and others, report growing ventricular heart muscle from mouse progenitor cells.2 Injecting stem cells into a damaged or weakened heart might potentially develop into life-saving and cost-reducing treatments for heart disease.

You cannot pick up a peer-reviewed bioscience journal without finding reference to basic research with the potential for wildly imaginative clinical applications. This research is not limited to just a few industrial nations but rather accelerates worldwide. Diagnostic assays and genetic probes promise bedside diagnosis in the near future.3 The immediate bioassay of the patient’s condition could have lifesaving value and economic value as well.

With this burgeoning of knowledge comes opportunity to identify solutions to heretofore-insoluble medical problems. The benefit to patients from new knowledge, however, depends on a new generation of bioengineering, clinical trials, diversity and most importantly education. Industrial age solutions will not work in this new world of technology. Medicine is not a market economy and never was. I fear that forcing industrial age free market solutions onto a scientific, academic and humanitarian infrastructure will continue to produce the lagging inequitable health care problems that we have at present --- good for big business but not for patients. Copy right, patents, and privatization of education restricts new knowledge, shared research, and graduate medical education at every turn.

Graduate education must play a major role in transmitting new knowledge to clinicians on the front line of medicine. This mission requires a high level medical institutions, education, mentoring, trust and motivation of clinicians. Both motivation and mentoring involve close two-way communication implying regional if not local involvement.

Duke University and the University of Pennsylvania medical schools each sponsor an Institute of Translational Medicine. Two new peer reviewed scientific journals trace the progress of translational medicine.4 These efforts are desperately needed. There are many obstacles. Europe is ahead of us in much of this research and in the collection of necessary bio-medical databases. 5

Conflicting Problems
Pending legislation seeks increased coverage and greater access to health care, while decreasing costs, but does not address these issues of new science or the required education to apply that knowledge clinically.
To improve quality, the stimulus package earmarks 19.6 billion for healthcare information technology (IT), 17.6 billion to promote electronic healthcare records and 2 billion for a National Coordinator for Health Information Technology.6 Whoever controls that database will control the future of medicine, including the economics, the quality and patient privacy. One of the drug companies is already offering doctors free IT for their office. The same company provides the software application for viewing and transmitting CAT scans.7 A drug company controlling protocols within patient records might not be the best idea. However, a central Office of the National Coordinator for Health Information Technology might be cumbersome as well if not tied to a broad base of academic medicine and research.

The algorithms of evidence-based medicine might be awkward when considering rapid scientific advances and regional biodiversity of the population. The NICE evidence based UK National Institute for Clinical Excellence while highly regarded may not change physician behavior as effectively as education.8 Quality improvement suggestions thus far boil down to: a medical information system, a pay for performance strategy, and public reporting of provider performance.

Obviously, planners place great faith in information systems. These strategies require fixed criteria and an analysis of the resulting medical information database. Pay for performance requires measuring performance against some fixed criteria. In a Rand analysis, the authors suggest, “Providing (public) performance information on physicians is not sufficient to change their behavior: rather a combination of education strategies might be more effective.”9

Guidelines established in Washington could, among other things, delay changes in the area of adaptation to environmental factors, the tailoring of care to the unique needs of individual patients and the implementation of new knowledge into patient care strategies.

Standards influenced by insurance companies might tend to ration care. The influence of drug companies might direct treatment toward self-serving high profit alternatives. Both will continue to exert influence on criteria by way of sponsored publications and lobbying. Among the vast plethora of medical publications, there is more miss-information out there, than there is good science --- much of it intentionally miss-leading. “There are 2.3 health care lobbyists in Washington for every member of Congress.”10 Even NIH might fail to keep pace with "best evidence" and smother the very advances in medical science they attempt to promote. Health care and its reform may be too big a challenge for central control. Vast regional differences in medical need and in patients themselves defy central control.

Treatment guidelines developed by medical schools and a few impartial multi discipline groups are enormously helpful. They are unbiased but expensive. The clinician can access vast amounts of data on a pocket PDA.11 For these or any other guidelines to be effective, however, one must have the right diagnosis.

I am concerned that we place far too much emphasis on treatment and not enough on diagnosis, differential diagnosis and interrelated problems. Although it breaks my heart, I would almost agree with the trial lawyers’ claim that far too much serious illness goes un-recognized or miss-diagnosed.

Currently the diagnosis on the insurance claim provides the basis for judging whether the doctor followed the appropriate treatment. The insurance diagnosis, however, is unreliable. Clinics often report diagnosis on insurance claims completed by an insurance clerk, often more interested in a diagnosis that justifies the level of service than what is actually on the patient record.

Rapidly evolving medical terminology not reflected in the ICDA codes, further distorts the accurate reporting of diagnosis. Even under the best of circumstances, the diagnosis is often obscure and subject to much debate at surgical, morbidity or clinical pathology conferences.

Thus, poor statistical correlation exists between reported diagnosis and the actual medical problem. This discordance works against any measurement of compliance to guidelines from a distant central location. Pay for performance begs the question, by what criteria and by whose judgment. Rationally, the best judge of good performance comes from the chiefs of service in a clinic setting. If we judge performance by data and compliance, one runs the risk of clinicians treating the guidelines not the patient. As long as pay for service and pay for diagnosis dominate the system, there is likely to be distortion of both the service and the diagnosis.

Many if not most young doctors, especially primary care, would prefer to be on salary. Many argue that salary promotes quality over quantity while fee for service favors quantity. The reporting of claims and the administrative burden alone now overshadows the advantage of fee for service to the provider. Salaries take away that administrative burden. The question is who pays the salary and who supports the institution. Surgical specialties and administrators profiting highly from the present confusion will object. These surgical specialties function well in teaching centers and medical schools where patient care remains the first priority.

Solution
An academic based public option might fund medical schools with the challenge of providing the uninsured with low cost quality medical care. Such is the history of teaching institutions until recently when with funding cuts the teaching centers behave more like private hospitals. Teaching centers have both an academic advantage and a personnel advantage, utilizing highly motivated trainees in the care of patients. Medical education and an academic approach to research and clinical care may offer the most promising solution to our healthcare dilemma.

If you were to ask, what is the best thing about American medicine? The reply would have to be our medical schools. While we might be 16th behind most of the Western World by public health criteria, we probably still rank number one in medical education. We have an impressive number of teaching instructions, widely dispersed; missing only two or three less populated states. Even those are well served by adjacent medical centers. The point being, our system of teaching institutions already serves most of the country and constitutes the best that we have. With proper funding, these institutions can take care for all that do not afford health insurance.

The cost of funding these teaching centers should be far less than the insurance solutions thus far proposed. Much of the basic research already takes place in these centers, as does graduate medical education.
This medical center option requires satellite clinics. Most medical schools provide them now. Combining Veterans medical care, covering Workman’s Compensation, and Medicaid could mean substantial savings for both the taxpayer and for employers. The economics of basing the public option on an existing infrastructure is obvious.

Solving our health care problem with a more scientific approach would generate a biomedical database of the patient population, facilitating both the research and the translation of discovery into clinical practice.
Multiple regional medical centers will better accommodate the vast regional differences in medical problems and population. Multiple regional initiatives will foster a variety of economic strategies. Multiple initiatives will likewise both: spread the risk of unworkable solutions, and increase the probability of the desirable results. Responsibility would fall to the highest levels of scientific medical leadership. This academic strategy would be a nationwide effort.

Salaried or a combination of base salary with incentive pay could better focus providers on patient care rather than quantity. State medical school employment could offer a degree of shelter from frivolous lawsuits.
Politically such public option should prove to be non-polarizing and attractive to both sides. Funding the teaching centers to cover the uninsured brings back a two-tiered system, but this time the second tier provides the better care. The disagreements on other grounds are intense, but upon medical education and the science, both sides might agree.

There are other advantages. With time, the tension between the academic and the private sector will lead to a merging of the science if not the method and a broader translation of the science into the private sector. Greed should once again give way to humanity.

Do what you can to curtail the abuses of drug12 and insurance companies, and leave the private sector in place. An expanded medical education and graduate education program will go a long way towards improving the shortcomings of our traditional insurance and fee for service system, both by way of education and competition.

The taxpayers’ money can generate a far greater return in human health by supporting research, education, and local clinics run as part of the education and translational process.

Summary
· Public option based on state medical education/hospital systems already in place.
· Accommodate the rapid changes in basic medical science.
· Correct the abuses in the present Insurance system, but leave private medicine and insurance in place.

Our people are our most vital asset. Health is an issue for our economy and our security. If we do not fix the present problems, we are both less productive and less secure. The proposed legislation goes a long way towards fixing current insurance problems and extending coverage. The current fix does not address the science of medicine nor does it provide an environment wherein the science can evolve. Translational medicine is a tool for the USA to reclaim the technology, but the science has to advance within a highly diverse academic setting. Quality must emerge from education and guidance from trusted respected mentors. Whatever healthcare solution we seek must accommodate the rapidly changing science of medicine by funding both medical research and a close connection between that research and our front line clinicians.
2,342

[1] A micrometer is 1/1,000 of a millimeter. A nanometer is 1/1,000 of a micrometer. 1 nanometer =1 x 10-6 millimeter
[2] Science vol. 326 16 Oct ’09 p426
[3] Nature, 462, 26 November 2009 p 461-462
[4] Science vol 326, 9 Oct ‘09, p205
[5] Nature vol 461, 24 Sep ’09 p448
[6] Technology Review MIT vol 112, number 3 June 2009 p47
[7] McKesson Corp SIC: 5122 Wholesale-Drugs, etc.
[8] Nature 462, 5 Nov 2009, p35 and 461, p336-339
[9] Rand Supplement: Complete Checkup, Ridgely, Adamson, Vaiana Summer 2009
[10] Alaska Journal of Commerce, Nov. 22 2009, p4 Health Care Tim Bradner
[11] The Medical Letter: Treatment Guidelines
[12] The Medical Letter vol. 51, 1324, p87: Tadalafil and Sildenafil for pulmonary hypertension cost the patient $1,060 and $1,360 respectively for a 30-day supply.

Quantum Entanglement, a conjecture

Elisabetta Collini et al[1] write of experiments with marine cryptophytes algae. These tiny deep ocean algae display quantum coherence between molecules as an adaptive mechanism for increased efficiency in capturing photons and transmitting energy for photosynthesis.

“Intriguingly, recent work has demonstrated that light absorbing molecules in some photosynthetic proteins capture and transfer energy according to quantum-mechanical probability laws instead of classical laws at temperatures up to 180 K.”[2]

This paper and a most readable summary,[3] suggest a completely new level of biological mechanisms on a whole new level of reality – a quantum leap, if you will, into a yet smaller world of particle physics.

“There are more things in Heaven and Earth Horatio than are dreamt of in your philosophy.”[4]

There seems a growing suspicion that quantum-mechanical mischief plays a role in many if not all living things. These features, however, are on so small a scale and so poorly understood that the quantum phenomena do not yet met the eye.

Medicine today struggles with translating a new found physiology on the molecular scale. The breakthrough comes from the advances in genetics and from microscopy at the atomic level. Quantum mechanics, on the other hand, may be present on a larger scale, but becomes evident only at the particle level -- where particle behavior becomes strange indeed.

Entanglement for instance, involves particles behaving identically at a distance. Efforts are under way to develop computer applications based on this phenomena of quantum entanglement. Einstein called the theory “spooky behavior at a distance” and was skeptical because it defied the laws of relativity. Experimental evidence today, however, makes entanglement a proven reality. Quantum mechanics is not yet reconciled with classical physics, nor is it yet a consideration in bio-medicine.

As a analogy, one might imagine a party of 7 girls who in the course of the evening before school is out become entangled in a metaphysical sense. In the morning, they leave school traveling home to many parts of the country. In the course of a subsequent evening one of the girls experiences an ecstatic moment. Simultaneously and instantaneously, the others experience the same – spooky indeed. However, the analogy is more complicated. Each girl might find herself spinning around in one direction or the other, but always in the same direction. Observing these girls might determine their location or their direction but not both at the same time. (About sums up my experience with the opposite sex) In observing which foot the girl was standing on, it would only be possible to offer a statistical probability. However, this is only an analogy. Entanglement occurs among electrons and protons. (Never mind Shrödinger’s cat.[5])

What intrigues me with this phenomena, is the conjecture that quantum-entanglement may play a role in the infinity of things that we do not yet and perhaps never will understand about our own biology.
Probing what we know, despite its considerable substance, hardly scratches the surface of the vast dark depth of the unknown. Bio medicine has yet to explain sleep. We cannot explain scanning logic, intuition or even the metaphor that over shadows our stilted attempts at artificial intelligence.

The Golden Plover chicks find their way from Alaska to the Fiji Islands weeks after their parents leave for the long flight south. How do hundreds of shore birds all turn instantaneously in perfect formation? One might consider this spooky behavior at a distance.

I think that the next frontier in medicine will be a quantum leap deeper into the infinitesimal -- exploring quantum entanglement at the particle level as a part of our physiology.

[1] Collini, Wong, Wilk, Curmi, Brumer and Scholes – Dep. Of Chem. Inst. for Optical Sci. and Center for Quantum Information, Univ. of Toronto, Physics and Med. Research U. of South Wales, Dep. Science and Chemistry U. of Padova, Italy
[2] Nature 463, p644, 4 Feb. 2010
[3] Photosynthesis, p614 by Grondelle and Novoderezhkin, same issue
[4] Shakespeare, “Hamlet” act I scene V
[5] Schrödinger http://en.wikipedia.org/wiki/Schr%C3%B6dinger's_cat

Thursday, December 9, 2010

Life Expectancy

https://www.cia.gov/library/publications/the-world-factbook/rankorder/2091rank.html
http://www.cdc.gov/nchs/data/nvsr/nvsr59/nvsr59_02.pdf

The Center for Disease Control today released statistics showing a drop in life expectancy for the first time since 1993. Clearly, the US has the best medical schools, the best-trained doctors and the most advanced technology, but something is obviously missing. The public health statistics tell the story, but not why. Life expectancy in 2008 dropped from 77.9 to 77.8
CDC presented 2008 as compared with 2007 counting 99% of the demographics and medical files. Six of the 15 leading causes of death actually decreased in number (Heart disease; Neoplasm; Cerebral Vascular Accident; Accidents; Diabetes and Assault)
Six others increased in number (COPD & Lower lung disease; Alzheimer’s; Influenza & Pneumonia; Nephritis, Nephrotic Syndrome, Nephrosis; Suicide; Hypertension and Renal Hypertension)
The report further reports on infant mortality, an improvement, yet a further embarrassment. 2007 rate was 6.75 deaths per 1,000 live births. In 2008 it improved by 2.4% down to 6.59. Black mortality rate is 2.3 times higher.
These figures, reflecting the health of a nation, attract the CIA’s concern as a security issue. The CIA reports that out of 224 countries, the US ranks 46th with 178 under developed countries with higher mortality rates. Life expectancy by country, also from the CIA, shows the US in 37th place, right after Cuba. Japanese can expect to live to an average of 82.6 years of age.
The CIA’s Infant mortality statistics, 2010, differ from the CDC’s,  2008, in deaths per 1,000 live births; CIA shows averages of: Angola 178, World 44, Mexico 17.84, Russia 10.32, US 6.14, Cuba 5.72,  the EU 5.61, Canada 4.99, Germany 3.95, Ireland 3.89, France 3.31 and Sweden 2.74.

Saturday, December 4, 2010

Electronic Health Record (EHR) II draft

It is easier for a doctor, not connected with a major institution, to obtain government secrets than to access current medical information. The funding for electronic health records (EHR) affords an opportunity to advance education and channel that, difficult to obtain medical information, to the medical student, the practicing doctor and the point of patient care.
Translational medicine makes an issue of researching and implementing new knowledge into clinically applicable substance. That substance has little merit until it becomes a part of clinical practice. The flood of new bio-medical information overwhelms the process of translating and delivering relevant advances to the clinician. Many EMR proposals contain substantial decision support capabilities, but neglect to feed current and changing medical information.
Vender centric programs meet the economic needs of the client and the vender. Government centric strategies stress avoiding errors, reducing cost, meeting fixed quality criteria[1] and quantifying the benefit. Drug centric solutions attempt to aid in selection of medical treatment and prevent adverse drug reactions. Hospital centric EHRs must deal with pooled data with other institutions, confidentiality, their own institutional review boards (IRB) and somehow collate data from divergent sources.[2] Hospitals and insurance companies seek control of the data as an administrative and economic strategy. Drug companies have a similar motive.
Make patient needs the first priority by centering the EMR in the computer of the individual provider while linking that database with a secure cloud database. Supply the provider with total access to total medical information. Include the decision and support capabilities. Generate a continuous differential diagnosis. Include a self-correcting statistical element. Compare and analyze the relationship between patient data and outcome.
Impediments in developing meaningful EMRs
·         Battle over final ownership of the data
·         Divergent priorities in function and goal of the EMR
·         Colossal challenge in meeting every bodies demands
Proposing that these electronic patient records (EPR) belong jointly to the patient and the primary care provider resolves a number of the impediments and problems.
·         Eliminates incompatible database problems among various institutions
·         Individual providers work with off the shelf database linked to major database provider.
·         Clinician can modify the reports and add criteria at will without upsetting the core software schema.
·         Medical school and institutional researchers can access accumulated patient data anonymously.
·         Patient genetic data correlates with clinical pathology in real-time
·         Enhances medical student education by learning the database in clinical years
·         Gains acceptance among private practice physicians by its sponsorship in the medical school and by its use among graduating physicians
·         Puts the medical school in control of the patient database
·         Puts a regional and environmental spin on the relevance of medical information
·         Promotes competition between medical schools in developing the best medical information database and operating system
·         Frees the decision support function and diagnosis from the distortion created by arbitrary regulation and economic motivation
·         Maintains political neutrality
·         Data accessible:
o   1. because in standard and widely used database program  
o   2. Because all information recorded as specific data entries and thus retrievable  
·         Informed consent no longer an issue for cohort studies
·         Use and cooperation become ubiquitous – spreads voluntarily
·         Free use provides an unwitting contract to use the information wisely
·         Errors and misdirection become immediately transparent to the medical school and a pointed direction for CME
·         Gives medical students much more of a vision of the totality of medical knowledge and a more organized way to store it.
Requirements
·         Open source and free to all providers
·         All patents and copyright in the  public domain
·         A database purchase from Major database provider, probably Oracle or IBM, unlimited users
·         Medical information in database form provided by medical school, free of copyright and cost! As an educational and CME function
·         Enough federal and state financial support
·         Initial programming of the database schema and the statistical relational rules


[1] J Am Med Inform Assoc. 2009 Sep–Oct; 16(5): 637–644 Decision Support Capabilities
[2] J Am Med Inform Assoc. 2009 Sep–Oct; 16(5): 624–630 The Shared Health Research Information Network (SHRINE)

Thursday, December 2, 2010

Bio Medical Database

I need to tell you what I am excited about. It involves the patient record, the human genome, medical information, statistics and a database program.
Scope “In April, the Department of Health and Human Services awarded a second round of grants totaling $267 million to create 28 new centers to assist health-care providers in implementing health information technology. The funds were part of the $20 billion allocated in the American Recovery and Reinvestment Act of 2009 to help doctors and hospitals make the switch from paper to electronic records.”
This is an opportunity to channel biomedical information into the hands of students and graduates and correlate that information with the patient record. 
From MIT’s Technology Review, “Electronic medical records provide vast amounts of medical information that can be combed automatically and used to ask questions…
...Scientists and physicians are now scouring the growing number of electronic medical records and genomic databases to figure out how to use this vast medical resource…”
More to the point of medical education, students and graduates alike experience a continuing need to reference current medical information. Current information is not easy to come by. It is expensive, closely held and often inaccessible. U of Michigan took a great stride in recording the lectures for students to review on their computers. Bravo, but the data is not indexed and it’s slow to access for a specific reference. In addition, there remain the student’s patient contacts and a need to correlate patient information with medical knowledge to develop a differential diagnosis.
Lists are hard to remember, but a modern user-friendly database program can retrieve a relevant list. The database’s – pardon the analogy -- left-brain contains the patient information, recorded as data points, whilst the database’s right-brain contains all available medical knowledge also as data points. The relational features of the database can easily match patient variables with diagnostic criteria, providing a differential diagnosis or list. Filing information in a database obviously compresses the data in that it eliminates repetition.
Other electronic medical records (EMR) applications stress “best evidence” treatments. While including up to date treatment options, this proposal stresses differential diagnosis.
Furthermore, a built in statistical application offers probabilities for the differential. Intriguingly, a more developed statistical attachment will correct itself in real-time based on outcome; this is the important aspect of the matter.
Everything I have read suggests great difficulty in analyzing the data -- data mining -- and that may be so on a central server, but on the student’s computer with a current state of the art database, the real-time statistical analysis should give an accurate picture of the outcome in relation to the clinical data, including the genomic information if available. As this data accumulates in the medical school server, over the professional life of the student, analysis should be much easier.  
I had some operational experience with an old Borland database while I was in practice. I got some of this schema done, but at that time, it was slow and limited. Today the computers are fast and there is virtually no limit to scalability.
It will take a couple of high-level database persons, a couple of statisticians and a couple of medical people who can pull together the full range of medical information in a database format, genetic data included. Students should be able to write the data fields for the patient record.
The resulting product should be open source and available to all clinicians willing to link to the medical school server. Both patient and medical information would flow both ways. Researchers can extract patient data anonymously. New bio-medical information can flow back the other way providing the clinician with continuous access and update of biomedical information.
 The educational advantages speak for themselves. Health and Human Services has the money. I believe that Medical Schools represent the only trusted vehicle for such a program.