Student Scene
 Photo by Graham Ramsay |
Medical Records Hit Slow Going from Paper Trail to Digital Highway
Imagine a world without ATMs. This technology, visible from almost every corner in a modern city, has become an indispensable and transforming tool of modern life--convenient, useful, and ubiquitous. The health care system, on the other hand, has been slow to adopt the kind of technology that most other sectors of the economy have assimilated. Even though much of health care is increasingly high-tech, the systems on which it is based--the medical record, the ordering of prescriptions and diagnostics, and the collection and dissemination of information--remain based on paper. This fact has important consequences for the practice of medicine and the quality and cost of care.Broken Records
Today, your medical record can exist in many different forms in a variety of different locations: there is likely a version of it (in relative states of completeness), with every physician's practice where you have been a patient. Importantly, your current physician may be unable to access this information easily and in a timely fashion, potentially compromising care and at the very least wasting valuable time in the patient-doctor encounter. In a mobile economy, the lack of portability of the medical record is an anachronism.In most cases, if your physician orders some laboratory tests or writes a prescription, he hands you a slip of paper that you take to the lab or pharmacy. At each step along the way, mistakes are latent, just waiting to occur. Many of these may be innocuous and result in no harm, but mistakes do happen. In fact, medication errors are one of the easiest kind of error to prevent using electronic systems--machines have infallible memories of drug-drug interactions, dosing schedules, and patient allergies. Experts believe that there are about 8.8 million adverse drug events each year, 3 million of which are preventable. At Brigham and Women's Hospital, researchers have found that the computerized physician order entry (CPOE) systems have reduced in-hospital errors by 55 percent and serious medication errors by 88 percent. A recent study estimated the implementation of CPOE in ambulatory settings would result in a savings of $44 billion.
Off to a Slow Start
Despite the availability of effective technologies, adoption has been slow, hindered by a variety of obstacles. Today, only about 5 percent of physicians use electronic prescriptions, and only about 15 percent of hospitals in the United States have some form of CPOE. Even in systems where CPOE is available, physicians only use it for one in four prescriptions.In a recent article in Health Affairs, researchers investigated the major obstacles to implementation of the electronic medical record and found that high initial costs, uncertain benefits, and high initial physician time costs impede widespread use. The authors recommended performance-incentive payment arrangements that would reward the use of electronic medical records to improve quality.
Internist Danny Sands, HMS assistant professor of medicine at Beth Israel Deaconess Medical Center, who uses computers extensively in his practice, agrees that capital costs are hindering implementation. But beyond the direct costs, the time associated with learning how to use these systems can be a formidable barrier to adoption. Changing your practice workflow, according to Sands, is an important consideration. "How do you adapt to this kind of technology--getting notes into the system, prescribing online? This is difficult for physicians in practice to absorb, and they are not paid to do this. The system does not support the time they need to learn the new systems, to change workflow, and it does not reimburse them for loss of productivity while learning."
Despite these challenges, Sands highlights a few approaches. One is to subsidize required capital investments. A hospital system might subsidize the purchase of the technology for its physician practices, or an HMO might offer the subsidy to physician groups with which it contracts. The federal government might be involved at some level. Alternatively, professional medical organizations could take the lead. Sands describes the example of the American Academy of Family Physicians, which has arranged with several vendors to provide discounts to AAFP members.
These approaches, however, do not address the issue of the cost of transition. Sands argues that perhaps incorporating payment incentives that reward high quality care might stimulate investment in these technologies. "While electronic-based systems don't ensure high quality care, they make it easier to provide such care to patients. Studies have shown that physicians who use electronic systems do not order as many redundant tests, find the information they need faster, prescribe medications more safely, and adhere to alerts and reminders more rapidly."
Renewed Interest in IT?
A confluence of factors may finally be pushing the health care system to adopt information technologies. In the 2004 State of the Union Address, President Bush called for the adoption of the electronic medical record. Later, he reiterated the need to advance communications technology in health care: "...We can control health care costs and improve care by moving American medicine into the information age.... This would encourage the replacement of handwritten charts and scattered medical files with a unified system of computerized records. By taking this action, we would improve care and help prevent dangerous medical errors, saving both lives and money."--Erica Seiguer, a fifth-year MD-PhD student studying economics in Harvard's Doctoral Program in Health Policy
The opinions expressed in this column are not necessarily those of Harvard Medical School, its affiliated institutions, or Harvard University.
Websites of Interest
Connecting for Health: A Public-Private Collaborative Convened by the Markle FoundationThe National Health Information Infrastructure (U.S. Department of Health and Human Services)
PatientSite at Beth Israel Deaconess Medical Center