Radiology as an Example of Moores Law in Healthcare - Getting to Sustainability
Looking through a past issue of Longwoods ElectronicHealthcare, I found an excellent article on “The Impact of Diagnostic Imaging Investments on the Canadian Healthcare System” by Simon Hagens, David Kwan, Craig Savage and Mark Nenadovic. I have been looking at the impact of new technologies and how they enable policy changes in healthcare systems when I came across this article. Diagnostic radiology is an excellent example of how a major new innovation has allowed for new clinical and business models which should result in a higher quality and more sustainable health care system. This short piece will comment on some of the issues raised by the EH article and add some insights and questions for future study. This piece is intended as a Longwoods Essay and the author would welcome feedback or better sources of data.
The EH article authors say: “Canada Health Infoway's investments in DI systems have had a strong, positive and measurable impact on the country's healthcare system. While the benefits discussed in this report are significant, there are many opportunities to realize even greater benefits from DI systems.” Some of the benefits specifically enumerated include: 30-40% improvement in exam turnaround, the avoidance of 10-17,000 unnecessary patient transfers, improved remote reporting and substantial cost savings in terms of films and materials and in improvement of the productivity of salaried radiology technicians. 1Each of these benefits is clear and positive. Infoway and the policy makers who created it are to be seriously congratulated for this unambiguous good news story.
Where things become less clear and more challenging for policy makers are on the productivity gains for radiologists themselves. The paper speaks of a 25-30% increase in productivity, which results in 9-11 million new examinations and “450-540 equivalent radiologists”2. More productivity is better in the abstract but for policy makers who have to pay the bills the news has some negative consequences. “Equivalent radiologists” cost the same $500,000+3 that radiologists cost. This means that the added annual bill for the productivity improvement is on the order of a quarter billion dollars of added fee for service costs. It is as if we paid Air Canada pilots by the miles flown and gave them a jet to replace their turboprop. As policy makers, how should we think about “Equivalent Radiologists”? Is this where we want to invest our scarce resources?
There are some significant second order effects as well from the introduction of this greatly improved technology, though they are more difficult to quantify. The ubiquitous nature of DI/PACS systems (as a result of Infoway-enable investments) now means that many other specialists can read their own images quickly and easily. Often these images will be “checked” by the radiologist and billed as such. Anecdotal evidence suggests that at least one downtown Toronto teaching hospital these reads happen after treatment has already occurred. So the “productivity” impact is likely to be much higher than just 25-30%.
In theory 4(see Christensen et al.) this change should open diagnostic radiology up to competition by these other qualified specialists. Cardiologists (and others) should be able if properly trained to read CT/MRI in their own field. In a competitive market these new entrants would reduce price. Currently anecdotal evidence is that many go to the US/Europe for training and are not able to read in Canada due to an inability to get privileges at hospitals. This is a restraint of trade and a competition imbalance that needs to be questioned.
A third impact is the overall numbers of self-identified radiologists increasing; perhaps due to the richness of the compensation. I believe it to be the case in one large province that total radiology bills are up over a third in one recent three year period as DI/PACS systems have come into play even though fee schedule amounts are close to flat. A group of my students at Rotman estimated current average income of Radiologists in Ontario at $708,000 per year (2009). The data are notoriously hard to come by but an average above $600k is a reliable estimate for most Canadian provinces. Increased volumes have improved service to consumers but also broken the fiscal balance in the system. Radiologist incomes are now more than double the average physician’s income. The question needs to be asked, is this how policy makers intended to spend their scarce resources?
This is not Infoway’s fault. Infoway has done its job and has been very clear about the benefits. It is up to policymakers in provincial health Ministries and health providers to manage this challenge.
There are several fundamental changes being caused by the technology-based productivity improvement as we moved from physically loading films on light boxes to full virtualization, automation and now distribution of images:
- COST: Cost to deliver the service comes down (as measured by time and other inputs).5
- GEOGRAPHY: Delocalization of service delivery. Images can be read remotely (even at home in some cases).6
- QUALITY/INDUSTRIALIZATION: The process can be industrialized and quality controlled. Appropriateness review could also be included.
- NEW COMPETITORS: Other providers read images themselves (vascular surgeons, orthopods, interventional cardiologists) and radiologists lose their monopoly position
The fee schedule negotiation approach used in most provinces does not do a good job on any of these four dimensions:
- The fee schedule will go down to a flat line but not easily go negative, even when it clearly should. Look at yesterday’s excellent Globe article on ophthalmology for another clear example that health system players have been aware of for more than a decade.
- RVU-based systems react slowly to technology changes with several years lag-time apparent in the data. We try to update more quickly but often fail and in the interim the providers are over or under rewarded.
- In Ontario (and perhaps other provinces) the fee codes do not appear to be subject to negative adjustments (perhaps due to reticence on the part of negotiators) except in the most dramatic of cases (e.g. cataract surgery)
- The fee schedule includes an assumption of physical contact and local handling
- We assume implicitly that there is local monopoly and therefore no competitive price setting structure
- There is no reward for modern process and QA review. We do not, for example, review 2% of all tests and then provide a bonus/penalty based on accuracy
- QA could also include rigorous post-hoc review of ordering practices and appropriateness of imaging practices. This could produce substantial system savings as fees are reduced for providers who do inappropriate testing
- NEW COMPETITORS:
- There appears to be an implicit non-competition agreement at the negotiating table. Other specialties may read images but Radiologists re-read them
Diagnostic radiology is now a fast-changing high tech business. It has as much to do with computer manufacture than with Marcus Welby. And like those high tech business, it has a declining cost curves at the current time and probably will for some time to come. In other high tech industries, we use fierce competitive pressures to ensure that gains are distributed away from providers and manufacturers to consumers in the form of added product quality or reduced prices. In the computer industry the approximately 20% per annum of cost reduction per CPU power (aka Moore’s law) is split 6% price reduction and 14% added features7. We don’t know how much added profit the computer makers retain but we do know that consumers see significant value every year. And these businesses have incredibly high standards of quality, QA and error correction. Every 12-year old knows that if he saves his money until next year the ipod will be cheaper and better.
How do we create this fierce “focused factory” pressure to build higher quality and sustainability in our newly high tech system? How do we manage appropriateness as we set up these diagnostic core services? How do we harvest the investments we have made through Infoway and other policy initiatives? Diagnostic Radiology should be cheaper and better next year than this year. We have invested $340 Mn of Infoway money plus provincial and hospital funds in the productivity investments needed to make it so8. These gains should not just disappear into the pockets of providers in the form of high compensation (approximately 2x the average physician). There needs to be rational, real-time adjustments made by sensible policy-makers and not just a quadrennial negotiation process.
The answer may lie in the very system we have built. By delocalizing the delivery of radiology interpretation we could allow for competition based on price and quality within the public system. We could use the DI/PACS systems and network them to allow providers to bid for volumes in a competitive sourcing arrangement. We could have flexible prices and those prices could actually come down! Quality Assurance and appropriateness review could be done on a near-real time, retrospective spot check basis to provide incentives for accuracy and rewards on that basis (with the size of the sample increasing for poor providers!). A price floor could be established so that we minimize disruption and then gradually take the floor away. Such systems are being envisioned in parts of Europe and there is already an international flow of radiology images9 and an interprovincial flow of pathology images occurring10. All of this can be done within our publicly funded system and respecting the principles of Medicare. It will certainly be “disruptive”; high tech industries change that way (as they should).
In 25 years, one can envision large industrial diagnostic centers that consolidate and accurately interpret DI and pathology images, lab tests, genomics and other diagnostics and immediately provide results for patients and their providers of care. They will compete on price and quality across Canada and will greatly improve quality and value. This trend is already underway in pathology images with the University Health Network and GE partnering to create a new center in Toronto that already serves Manitoba and Newfoundland. There is a competing center in Pittsburgh11 and there is already a relatively free flow of pathology images across borders. Lab companies have been delocalizing diagnostics for a decade and also have a declining cost curve (estimated at -3-5% per year by one industry participant). Unfortunately, they are often hampered by local regulatory barriers that need to be removed. These flows of images, samples, and tests will internationalize as well. If we are ahead of this trend, this should be a major source of new high end jobs for Canadians. We have the best healthcare in the world; why would we not export and create a huge vibrant industry?
While major central hubs will grow, we are also witnessing a decentralization of imaging and testing as a result of virtualization. Local competition will come from individual specialists who know particular body systems exceedingly well and read their own images and tests (always with retrospective quality review). They may or may not use the central hub services depending on the clinical need at the time. Christensen et al. talk about how the centralization followed by decentralization has happened in many high tech industries. Radiologists could use their expertise to screen and consult with clinicians to move into a new coaching role re: appropriate testing. Part of their training will be to communicate effectively with the medical team and help steer the team towards the most appropriate test, even refusing to do tests they deem unnecessary.
The opportunities to improve value and sustainability are very large. This is an exciting vision in which health outcomes will improve in terms of access and quality while we dramatically reduce the price charged per image. Starting the purchaser/provider split for hospital services with diagnostic purchasing is very attractive because there are multiple competing clinical and business viewpoints that should translate into a vibrant marketplace that will great improve the quality, accessibility and efficiency of diagnostics for clinicians and their patients. This gives us an achievable starting point for funding reform with well identified cost centers inside of hospitals.
A sustainable health care system is within reach if we recognize that health is increasingly a high technology service, that it can be delivered remotely (not just locally) and that for many parts of the the healthcare system services have a declining cost curve. We need to capture the benefits of this “Moore’s Law for Healthcare” and keep the benefits for payers and patients wherever we find them. Infoway has shown us that buy investing in technology we can radically and dramatically improve the cost, quality and access to radiology services. We need to apply these lessons to other diagnostic services and to other parts of health care. Breaking up the local monopoly systems allows us to improve using competition within our public system.
4 Christensen, C., J Grossman, J Hwang. 2009. The Innovator's Prescription: A Disruptive Solution For Health Care. McGraw-Hill Professional
About the Author(s)Will Falk (@willfalk) Executive Fellow in Residence, Mowat Centre for Policy Innovation, SPPG . Adjunct Professor, Rotman School of Management, University of Toronto
See additional essays and commentaries on this subject at these links:
Anton Hart wrote:
Posted 2012/05/28 at 01:43 PM EDT
Prescription for Sustainable Medical Imaging in Canada? Ask the Experts: The Radiologists
Personal Subscriber? Sign In
Note: Please enter a display name. Your email address will not be publically displayed