The Characteristics of Medical Innovation
Conference speakers made two important points about the characteristics of medical innovation. First, innovations in diagnostics, therapeutics, and devices are important but are not the whole story. Corresponding innovations in the health care delivery system have not taken place. David Lawrence and Jerry Grossman of the Lion Gate Management Corporation and the Kennedy School of Government both emphasized the need for innovations in the health care delivery system if the full benefits of innovation in diagnostics, therapeutics, and devices are to be achieved. Second, innovation in implanted devices and drugs follow quite different paradigms. Paul Citron of Medtronic observed that the former are much more likely to undergo improvements leading to significant cost-effectiveness improvements over time. As a result early cost-effectiveness studies for implanted devices are likely to be worst-case scenarios and could lead to premature abandonment of the technology.
MEDICAL INNOVATION SHOULD NOT BE TOO NARROWLY DEFINED
Grossman and Lawrence both emphasized that “the tools of care” have far outstripped “the tools of caring.” Innovations in diagnostics, therapeutics, and devices have moved far faster than the tools for delivering these breakthroughs. As a consequence, innovation in delivery systems is badly needed if the full benefits of innovation in diagnostics, therapeutics, and devices are to be achieved.
Lawrence observed that the vast array of medical innovations since World War II has led to a tremendous growth in the complexity of health care. The health care sector has not evolved to accommodate this complexity. In other sectors where complexity had significantly increased, sophisticated production systems have been implemented, an information technology infrastructure installed, and teamwork developed.
In medicine, these types of developments have not occurred in the health care delivery side. Production design is a foreign word. It is estimated that between 1 and 2 percent of total revenues in health care are invested in information technology—well below the level of investment in other information-rich industries. Physicians are still imbued in training with the principle of individual, professional autonomy despite the fact that most practitioners are not working in autonomous situations.
Funding information technology investments is a big problem. As McClellan commented it may be that the financial rewards for good information systems in the health care delivery industry are significantly lower than they are in other industries. Privacy concerns are also a barrier to investment in health care information systems.
Lawrence thought that there may be a role for the federal government in the development of the health care information infrastructure. He believed that Singapore might be showing the way through the creation of an investment pool for information technology experiments. He had in mind a federally sponsored investment bank that would be experiment- and innovation-driven. This bank would fund a number of major experiments and from these we would learn about how best to establish a health care information infrastructure.
DEVICES AND DRUGS ARE DIFFERENT IN THEIR COST-EFFECTIVENESS OVER TIME
Paul Citron said that the paradigms for implanted medical device innovation and drug innovation are quite different. Devices provide site-specific therapy and exhibit a direct mechanism while drugs act systematically and have an indirect mechanism of action. As a consequence, device therapy has fewer side effects than drugs. Further, devices incur a high initial cost at implantation that is amortized over the service life of the therapy, whereas the costs of drug therapy accumulate and can be substantial over the treatment period. Another key distinction is that devices undergo continuous evolutionary improvements usually with cost-effectiveness improvements while cost-effectiveness for drugs remains relatively constant.
Improvements in the cost effectiveness of devices can be intrinsic— technological improvements in the device—or they can be extrinsic—improvements in the way the technology is deployed. Examples of intrinsic
improvements are pacemaker internal current requirements and pacemaker functionality. In the 1960s and early 1970s, pacemakers were made of discrete components and required 30 micro-amps (late 1960s) and 22 micro-amps (early 1970s) to operate. Modern devices use integrated circuits and energy consumption has been reduced to about 4 micro-amps (Ohm, 1997). Thus, over the last 30 years there has been a seven-fold improvement in the internal operation of the device. In terms of pacemaker functionality, early devices stimulated the heart once a second whether it needed to be stimulated or not. Modern pacemakers are computers that constantly monitor the underlying heart beat rhythm and make adjustments as appropriate. In addition, the pacemaker stores data on what the device has done to help the cardiologist understand how the patient is progressing.
Combining intrinsic technology advances and extrinsic factors has progressively improved ICD cost-effectiveness (Stanton et al., 2000):
Around 1985, ICDs required open-chest implantation. Morbidity was about 5 percent. The batteries had a 2-year life expectancy. These first generation ICDs were judged to be marginally cost-effective at just under $50,000 per life year saved.
Shortly afterwards, the battery life was extended to 4 years, and cost effectiveness improved to just under $40,000 per life-year saved.
In the early 1990s, a paradigm shift occurred. Transvenous electrodes were developed that required less invasive surgery. Morbidity was reduced and the length of stay in the hospital was shortened. The average cost per life saved was further reduced to under $20,000.
The ICD might now be a cost-saving technology because the sensing devices built into the ICDs can now monitor and correct automatically some cardiac rhythm disorders that previously would have required a hospital visit.
Citron concluded by saying that early cost-effectiveness studies for devices are likely to present worst-case scenarios and could cause a device to be abandoned prematurely.
Ohm, O.J., and Danilovic, D. 1997. Improvements in pacemaker energy consumption and functional capability: Four decades of progress. PACE 20:2-9.
Stanton, M.S., and Bell, G.K. 2000. Economic outcomes of implantable cardioverterdefibrillators. Circulation 101:1067-1074.