Building a framework to transform health care. (Health Care Meets E-commerce).
* Open Systems
* Data Warehouses
* Evidence-based Medicine
* Predictive Modeling
* Creating Intellectual Capital
* Information Technology
PROGRESS IN MEDICAL care begins with a hypothesis posed by a physician, manager, or research scientist followed by structured clinical research trials to test its validity. Patient information Is recorded, stored in computers, and analyzed with appropriate statistical techniques to identify outcomes attributed to treatment variation and to answer such questions as:
* Will this medication improve the survival of patients with these characteristics?
* Will this surgical procedure improve function in these patients?
Treatment choices are influenced more by the physician's recent experience with patients than by clinical research results. For most physicians, medicine has a human face; it is not about populations and statistics. There are no clinical trials to tell a physician how to treat his or her patients, because they are based on other, unknown patients. But you cannot determine if one treatment is better than another based on individual cases. Using evidence-based medicine, however, physicians can apply the disciplines and results of clinical research to their practice of medicine, one patient at a time.
Duke University Medical Center leads the nation in evidence based-medical care for heart disease. Eugene Stead, MD, then Chairman of Medicine, determined in the 1970S that Duke should build an electronic medical record and data warehouse to study the treatments and outcomes of patients with heart disease. They created the Cardiovascular Databank, (1) which has evolved into a massive information system. Yearly, Duke collects more than 70,000 detailed patient records--their histories, physical examinations, laboratory studies, surgical procedures, and medical treatment outcomes.
Robert M. Califf, MD, a cardiologist on the faculty of Duke University Medical Center who directs the Cardiovascular Databank, said in 1996:
"The old system was one in which physicians studied the chemistry and physiology of how a disease worked and implemented theories based on this. Care was based on the physician 's own observations, and training was essentially an apprenticeship. But you can't decide if a therapy is working or not based on each individual case. Now we have the computer to remember and aggregate numbers. We can measure what happened to patients, and if it wasn't successful, we can change our approach. We can give multiple examples of where what we thought was right was in fact dead wrong." (2)
Predicting the future
With the advent of managed care competition and the growing importance of efficient and effective care, health care organizations will need to invest in standardized electronic data collection and sophisticated predictive modeling of patients' characteristics to determine outcomes.
Physicians are, in fact, paid to predict the future. They determine a diagnosis, a prognosis, and a course of treatment based on collecting information on a patient, understanding his or her specific condition, and applying their training and experience. Predictive modeling is a new tool that physicians can use to determine the probable outcomes of different treatments, based on the patient's condition and the highest likelihood of the desired outcomes. (3,4,5,6)
The technology for predictive modeling is maturing at a rapid rate with the advent of sophisticated mathematical techniques for pattern recognition (the new "intermaths"--regression analysis, Monte Carlo simulations, neural networks, and genetic algorithms) and parallel processing computers. James Bailey speaks to these trends eloquently in After Thought, (7) arguing that mathematics has matured from the study of place (geometry and navigation), through the study of pace (algebra and calculus), to the study of complex patterns in large data sets using computers (predictive modeling).
Bailey and other proponents of the new intermaths believe that static equations do not have the subtlety and accuracy to model complex biological processes well and to accurately predict the outcomes of medical treatment. A neural network recently proved to be best in predicting one-year mortality rates for patients with heart failure, using data from echocardiograms to make its predictions, compared to clinical judgment of cardiologists, linear discriminant analysis, and automatic heuristic methods. (8) Neural networks can accurately diagnose patients with glaucoma using data from automated visual field studies and structural data from computerized image analysis. (9) Neural networks perform well in predicting patients with acute appendicitis. (10)
Predictive modeling provides powerful tools to help clinicians evaluate patients and select treatment protocols. Eric Peterson, MD, a cardiologist at Duke Clinical Research Institute has developed models for physicians to use in their practices:
"After initial consultation with a patient complaining of chest pain, the physician can look at a model that incorporates this information and his patient's vital statistics- such as age, weight, and other health factors- to help him determine the next course of action. If the model shows that the patient might have a high likelihood of having a serious problem, the physician may then decide to skip what is typically the next step, a treadmill test, and move on to a higher level. " (11)
At each successive stage. the physician can consult a new model to help direct his decision-making. Peterson said he uses the models a great deal in his work. "Most patients in the 1990s are very involved with their care and want to participate in the decisions," he said. "They appreciate knowing the relative successes of different treatments. Knowing the specific benefits and risks of a treatment, we can make a plan that we both become very comfortable with." (11)
Clinicians and administrators need to learn the effects of medical decisions, based on systematic retrospective analysis of clinical and administrative data. Clinical data warehouses represent valuable resources for retrospective analysis for education, predictive modeling, benchmarks, outcomes improvement, marketing, and administration. (12,13)
Importance of standards
Standards for data collected across encounters and over time are critical for health care organizations if they are to pool data, establish benchmarks and norms, and share clinical data with collaborating organizations. (14) These processes, enabled by data standards, are essential for practicing evidence-based medicine and implementing an effective outcome measurement process.
The speed and skill with which an organization adjusts to changes in its business environment determines its success in the marketplace. Like organisms, the ability to learn quickly and incisively marks organizations that grow and prosper in rapidly changing times. Health care delivery systems (HCDSs) are natural laboratories for studying and improving population health and medical care. They need to invest in the continuing education of their physicians, using data and information gleaned from providing health care to millions of people--academic research trials cannot come close in the number of subjects under study. This is not only an opportunity to systematically study and improve care, but also, in the process, to develop considerable competitive advantage in the marketplace for their services.
Learning organizations and intellectual capital
In 1990, Peter Senge's extraordinarily successful book The Fifth Discipline:, The Art and Practice of the Learning Organization was published. (15) Since 1996, a number of influential books have appeared on the subject of intellectual capital. (16,17,18) Some corporations have begun to calculate and report to investors their investment in intellectual capital. (16) Intellectual capital represents the knowledge of an organization that helps it compete for business and grow profitably.
When the common stocks of companies trade at substantial premiums to their book values, the market is valuing the likely stream of future earnings from assets that are not reflected on the balance sheets. In the information age, companies are valued based on the talent they employ, their effectiveness in using that talent, their ability to learn from and satisfy their customers, and their ability to innovate and change.
"Intellectual capital is the sum of everything everybody in a company knows that gives it a competitive edge. Unlike the assets with which business people and accountants are familiar-land, factories, equipment, cash- intellectual capital is intangible. ft is the knowledge of a workforce: the training and intuition of a team of chemists who discover a billion-dollar new drug or the know-how of workmen who come up with a thousand different ways to improve the efficiency of a factory ft is the electronic network that transports information at light speed through a company, so that it can react to the market faster than its rivals. It is the collaboration- the shared learning-between a company and its customers, which forges a bond between them that brings the customer back again and again.
In a sentence: Intellectual capital is intellectual material- knowledge, information, intellectual property. experience- that can be put to use to create wealth. ft is collective brainpower. It's hard to identify and harder still to deploy effectively. But once you find it and exploit it, you win." (20)
Thomas A. Stewart, Intellectual Capital, The New Wealth of Organizations
Most health care delivery systems represent a virtual learning laboratory of hundreds of medical organizations. including thousands of physicians caring for millions of health plan members and patients. They need to establish their overall culture, rather than existing as a collection of disparate facilities. Standardized digital databases and networks--the structural capital component of intellectual capital--are needed to collect, store, and analyze data about the processes and outcomes of care for the system's member facilities and medical groups.
A model for sharing knowledge across the enterprise and with the outside world is needed to enhance the organizations value and foster learning for physicians. Physicians need the means and incentives to communicate clinical insights throughout their health care delivery system. A clinical guideline developed in a hospital in Arizona that reduces the complications of patients with diabetes should be shared throughout the system. A disease management program employed in Southern California that saves lives should be shared within all of the HCDSs and promoted to the world.
Measuring evidence-based medicine
There is a rich literature of articles and books on the measurable cost-effectiveness of evidence-based medicine and continuing education:
* Automated pharmaceutical inventory systems help hospital formulary committees use cost-effectiveness analysis, peer review, and continuing medical education to influence the purchase of drugs by hospitals and physicians prescribing habits that benefit patients and the hospital's financial condition. (2)
* Nurses offered continuing professional education relevant to their work have higher morale and work more efficiently than those deprived of such applicable instruction. (22)
* Hospital administrators, with the leadership of physician executives. can and have developed successful continuing education programs to promote cost-effective physician behavior based on the educational display of variations in their practice habits and an emphasis on evidence-based medicine. (23)
* Quality improvement programs in hospitals can reduce hospital charges, and, in the era of capitation, lead to long-term financial stability for hospitals and health care systems. (24)
* Cost-benefit analysis of educational programs for general practitioners in Sweden on prevention and treatment of depression led to measurable net savings (after expenses) of $26 million. (25)
* Inexpensive and readily available quantitative methods can give impressive information to physicians that leads to substantial changes in their drug prescription patterns, saving resources without injuring patients. (25)
* Evidence-based medicine offers strategies and techniques of observational data analysis that supports medical progress without requiring randomized clinical trials. (27) There are identifiable costs and benefits to using observational data analysis in large clinical databases to determine clinical policies, instead of relying solely on the results of randomized clinical trials.
* Many circumstances justify developing clinical policies from analysis of observational data--evidence-based medicine--in carefully developed and validated clinical data warehouses. (22)
* Evidence-based medicine can be seen as an acceptable, even necessary, limitation of clinical freedom, because it leads to practice guidelines meant to standardize and reduce the variation in clinical care.
* The costs of care must be considered in evaluating clinical policies-there are known situations where marginally beneficial services are not affordable if made available to all patients who might benefit from them. As health care systems move from fee-for-service medicine to capitation, these analyses will become more common and acceptable among clinicians. (29)
* Some of the techniques of evidence-based medicine, audits, and guidelines are being used successfully to promote clinical cost-effectiveness in the context of contracts that place the clinicians at financial risk for the care of populations of patients over time. (30) Audits of the practice habits of busy community physicians can compare their effectiveness and efficiency to standards established by 'evidence-based medicine."
* Quantitative methods can calculate the benefits to patients of changes in idiosyncratic practice habits to those supported by generally accepted clinical guidelines. Efficiency and appropriateness of clinical practice can be measured with observational data analysis. (31)
* Physicians trained in clinical care and health services research (biostatistics and epidemiology) need to lead the Implementation of practice audits and evidence-based medicine programs and intermediate between clinicians and managers to achieve optimal cost-effective care. Otherwise, the beneficial role of practice guidelines may be subverted by the administrative fiat of health care bureaucrats, whose purpose is first to reduce costs and second to improve care. (32)
Creating intellectual capital
Most health care delivery systems need to improve the operations of their facilities and affiliated medical groups by improving clinical care services through continuing education. Physicians can learn about variations in care, based on real, observational data sets (claims, laboratory results, and pharmaceutical orders).
To expedite and enhance guideline acceptance, HCDSs should:
1. Develop research that shows the need for and value of the guideline
2. Calculate the predicted outcomes from implementing the guideline
3. Measure the effects of the guideline over time
4. Teach physicians to understand the data, analyses, guidelines and how to implement them
A website is an excellent vehicle to disseminate the work and insights gleaned from data analysis for continuing education. HCDSs should:
1. Create a catapult website to educational content organized by medical specialty.
2. Identify experts in various specialties and have them moderate discussions of clinical cases posted by physicians. This will establish the health system's culture as an organization without walls, rather than as a collection of disparate facilities, and its value as a source of knowledge for its regions and constituent facilities.
3. Analyze observational data, develop tutorials to teach physicians about the findings, and post this information on the website. The analyses can be used to justify pilot studies of guidelines (results can also be posted on the website) and to promote the adoption of guidelines that lead to improved clinical and financial outcomes.
Administrative and clinical data about patients within the HCDS can be made anonymous to protect individuals' privacy and then pooled into large comparative databases so epidemiologists and health data analysts can search for measurable variation in the care of similar patients. Significant variation--once found, publicized, and addressed--is an opportunity for quality improvement. Distance learning technologies--CD-ROM, video conferencing, and the World Wide Web-are available and inexpensive to deploy, reducing the cost to disseminate the data and insights into ways to improve clinical processes throughout the health care delivery system. (1)
(2.) Duke Chronicle, February, 1996, http://www.chronicle.duke.edu/chronicle/96/02/23/01DatabaseImproves.h tml.
(3.) Nettleman, et al. Predictors of survival and the role of gender in postoperative myocardial infarction. Am J Med 1997 Nov: 103(5): 357-62.
(4.) Harris, L.M. et a1. Screening for asymptomatic deep vein thrombosis in surgical intensive care patients. J Vasc Surg 1997 Nay: 26(5): 764-9.
(5.) Lewandowski, K. et a1. High survival rate in 122 ARDS patients managed according to a clinical algorithm including extracorporeal membrane oxygenation, Intensive Care Med 1997 Aug; 23(8): 819-35.
(6.) De Sanctis, J.T. et al. Prognostic indicators in acute pancreatitis: CT vs. APACHE II. Clin Radiol 1997 Nov; 52(11): 842-8.
(7.) Bailey. James, After Thought, NY, NY: Basic Books, 1997.
(8.) Ortiz, et at, One-year mortality prognosis in heart failure: a neural network approach based on echocardiographic data, J Am Coll Cardiol 1995 Dec:26(7):1586-93.
(9.) Brigatti, L. et al. Neural networks to identify glaucoma with structural and functional measures, Yearbook of Medical Informatics. 1997. p 432-44 2.
(10.) Pesonen et at. Comparison of different neural network algorithms in the diagnosis of acute appendicitis, Yearbook of Medical Informatics, 1997, pages 443-449.
(11.) Duke Chronicle, February 1996. http://www.chronicle.duke.edu/chronicle/96/02123/OlDatabaselmproves.h tml.
(12.) Ruffin. M. The future Is here, The Physician Executive, November 1996. pp. 22 - 28.
(13,) Ruffin, M, The importance of data warehouses for physician executives, The Physician Executive, 1994 Nov;20(11):45-7.
(14.) Ruffin, M. Standardizing medical data, The Physician Executive, 1997 Sep-Oct; 23(7): 61-4.
(15.) Senge, Peter. The Fifth Discipline, the Art and Practice of the Learning Organization, NY, NY: Doubleday, 1990.
(16.) Stewart, Thomas A. intellectual Capital, The New Wealth of Organizations. NY, NY: Doubleday, 1997.
(17.) Edvinsson. Leif & Malone, Michael S. Intellectual Capital, Realizing Your Company's True Value by Finding Its Hidden Brainpower. NY, NY: Harper Business. 1997.
(18.) Brooking, Annie. Intellectual Capital. Core Asset for the Third Millennium Enterprise. Thomson Business Press. 1996.
(19.) The Skandla Group Website: http://www.skandia.se/group/comlindex.htm.
(20.) Stewart. Intellectual Capital, The New Wealth of Organizations, NY, NY: Doubleday, 1997, page ix-x.
(21.) Fins, J.J. Praxis Makes Perfect? Hastings Cent Rep. 1993 Sep.Oct:23(5):16-9.
(22.) Nolan, M. et al. Do the benefits of continuing education outweigh the costs? Br J Nurs. 1993 Mar 25-Apr 7:2(6):321-4.
(23.) Shulkin, D.J. et al. Promoting cost-effective physician behavior. Healthc Financ Manage. 1993 Jul;47(7):48, 50, 52-4.
(24.) Jones, SB. Quality improvement in hospitals: how much does It reduce healthcare costs? J Healthc Qual. 1995 Sep-Oct;17(5);11-3; quiz 13, 48.
(25.) Ruts, W. Cost-benefit analysis of an educational program for general practitioners by the Swedish Committee for the Prevention and Treatment of Depression. Acta Psychiatr Scand. 1992 Jun;85(6):457-64.
(26.) Eckert, G.M. et al. Measuring and modifying hospital drug use. Med J Aust. 1991 May 6:154(9):587-92.
(27.) Porzsolt, F. et at. Differences between evidence-based medicine and best conventional medicine. Med Kim. 1997 Sep 15;92(9):567-9.
(28.) Hornberger. J. et at. When to base clinical policies on observational versus randomized trial data. Ann Intern Med. 1997 Oct 15;127(8 Pt 2):697-703.
(29.) Hampton, JR. Evidence-based medicine. practice variations and clinical freedom. J Eval Clin Pract. 1997 Apr;3(2):123-31.
(30.) Auplish, S, Using clinical audit to promote evidence-based medicine and clinical effectiveness--an overview of one health authority's experience. J Eval Clin Pract. 1997 Feb:3(1):77.82.
(31.) O'Neill. D. et at. Central dimensions of clinical practice evaluation: efficiency, appropriateness, and effectiveness--I. J Eval Clin Pract. 1996 Feb:2(1):13-27.
(32.) Miles, A. et al. Central dimensions of clinical practice evaluation: efficiency, appropriateness. and effectiveness-II. J Eval Clin Pract. 1996 May;2(2):131-52.
Marshall de Graffenried Ruffin, Jr., MD, MPH, MBA, CPE, FACPE, is President of The Informatics Institute and Ruffin informatics, Inc. in Bethesda, Maryland, which provides clinical information systems education, consulting, and data warehousing He can be reached by calling 800/844-0922 or via email at firstname.lastname@example.org.
|Printer friendly Cite/link Email Feedback|
|Author:||Ruffin, Marshall de Graffenried|
|Date:||Jan 1, 2000|
|Previous Article:||The clinical email explosion. (Health Care Meets E-Commerce).|
|Next Article:||Are you ready for Generation X? (Changing World View).|