Monday, 15 December 2014

GE India Could Save U.S. PET/CT Scan Patients $7.5 Billion Per Year

Cancer is abnormal cell growth with the potential to spread to other parts of the body and disrupt normal functions, resulting in illness or death. Claiming over half a million lives per year, cancer is the second leading cause of death in the United States, second only to heart disease. There are over 100 known cancers effecting humans, some being more lethal than others. In all cases, early detection is crucial for improving the chances of successful treatment.

Cancer Cells, from WebMD

Cancer detection typically involves a range of methods and tests but is broken into two stages: Screening and diagnosis. Researchers have provided cancer screening guidelines to encourage regular screenings for potential symptoms of cancer, which differ for each type. Cancer symptoms might include heightened levels of proteins or other byproducts of cancer in the blood, abnormal growths, pain, bleeding, or discoloration of tissue. If potential symptoms appear, the doctor will move towards diagnosis, ordering additional tests to gather more data. These tests can be numerous and costly, and might include PET/CT scans, ultrasound, endoscopy, and MRI for imaging; lab testing on bodily fluids to detect tell-tale biomarkers; and biopsy to collect cell samples for microscopic evaluation by a Pathologist. Using data from these tests, doctors attempt to provide a definitive diagnosis and prognosis to the patient.

Many of the tests above are very expensive, but entrepreneurs in India are working on addressing an enormous and unmet domestic need for early cancer detection through innovative, low-cost product design. For example, GE Healthcare has developed the gorgeous Discovery IQ PET/CT, representing $15 million and 3 years of R&D at its center in Bangalore. Innovating within the constraints of India, they have succeeded in creating a scanner that's 40% more affordable than comparable alternatives, is modular and scalable to allow for price discrimination, offers a unified service of detection, planning, and assessment through state of the art hardware and software, and is a beautiful machine. The cost savings have come from conducting R&D in India, from redesigning a more cost-effective machine from the ground up, and from manufacturing the machine in India.

Discovery IQ PET/CT Scanner, from GE Healthcare

Currently, India only has about 120 PET/CTs operational in the country, whereas the required number of units to adequately address patient demand is over 1,000. Judging from the CEO & President of GE South Asia Terri Bresenham's remarks on Discovery IQ, GE hopes that the price elasticity of demand for PET/CT scanners in India is flexible enough that a 40% price reduction will help to fill this gap. I hope so, but it may be that more disruptive technologies like CellMax Life may ultimately be what brings early stage cancer detection to the masses of India.

Even if the Discovery IQ does not realize substantial financial returns or market share gains in India, GE and the world can still benefit tremendously because GE has created a superior, low-cost product. Producing 50-slice equivalent CT speed imaging, the highest PET axial view available on the market, the Q.SUITE platform for improved PET quantitation, and the Q.CLEAR platform for up to 2x image improvement (SNR) + 2x improvement in PET quantitation accuracy, Discovery IQ is a PET/CT that gets the job done. Certainly there are PET/CTs that may perform better on any one of these dimensions, but does it matter? For example, some PET/CTs are now capable of taking 512 slices, but since no meaningful diagnosis accuracy improvement was found from the jump between 64-slice and 128-slice scanners, 64-slice scanners generally make for the best value. In fact, according to Niharika Midha at GlobalData, the PET/CT "Hardware has reached this plateau . . . There is only so much that can be changed to the machine."

If GE India has tightly engineered Discovery IQ for maximum value (ROI), then this device may sit at the very rim of the performance plateau, including all of the right features, but not overloaded with cutting edge technology and experimental features that price it out value-conscious procurement networks.

GE Logo and Slogan from Inroads

Given that most PET/CTs cost between $2 and $3 million, the 40% price reduction for Discovery IQ puts it between $1.2 and $2.8 million for cost of acquisition. The median prices for CT scans charged by providers hover between $700 and $900, and the U.S. government estimates that 80 million scans are performed annually. The price for PET scans is about $5,000, but only 2 million are performed annually for Oncology diagnostics. Adding these up, approximately $75 billion is spent on PET/CT imaging per year in the U.S.

Of course, only a minor portion of this total spend could be chalked up to equipment cost, so if we instead look at the lowest PET/CT scan costs, and assume the imaging centers offering such low prices are highly efficient so that the overhead and profit margins per procedure are small, we might say that the 40% equipment cost savings generated by Discovery IQ could be passed on rather directly to the end payer. The lowest prices charged for CT scans seem to be around $200, and the lowest for PET scans around $1,200. Reducing these costs by 40% and multiplying by the yearly number of procedures shows a potential costs savings of nearly $7.5 billion per year.

This begs many questions. Could GE keep the cost this low in selling to PET/CT providers in the U.S.? If not, how much of the 40% cost saving be eaten up, and by what? Would GE even want to disrupt their home PET/CT market by importing Discovery IQ? Certainly short-term financial analysts held captive by Wall Streets valuation process would balk, as would the sales force taking lower per sale commissions. On the other hand, volume can make up for decreased unit margins, and such a move may gain GE significant additional market share. Also, the modularity of Discovery IQ might result in natural price discrimination, where the highest-end clients are able to pay for additional functionality, and could widen the base of the market by making a basic PET/CT available to clinics previously unable to afford one.

What Do You Think?
  • Should GE disrupt themselves by importing and commercializing Discovery IQ in the U.S.? If so, what main barriers would they face? 

Saturday, 6 December 2014

India's New Health Tablet Performs 30+ PoC Diagnostic Tests

The Swasthya Slate ('Health Tablet') is a portable diagnostic device that enables users to conduct 33 diagnostics tests with a simple box, power source, test sensors, and a smart device using the Android OS. The videos on this page give an excellent view into how the testing, analysis, and reporting process works. Of the 120m or so smart phone users in India, about 80% are using an Android OS, so the technological infrastructure is already in place for Slate to quickly scale up.
Photo Credit: Swasthya Slate
Launched in March, 2014, the Slate is the work of the Public Health Foundation of India, which also manufactures the device. The project is led by Dr. Kanav Kahol, formerly a bioinformatics professor at Arizona State University. With all sensors included, the device is being manufactured in India for only $800, a price expected to drop by 20% as it further scales. At this price, the primary target customers are health ministries, agencies, and clinics who will purchase the devices for public health management, putting them into the hands of community health workers (CHWs) who canvas the country as frontline care providers. The new diagnostic processes enabled by the device show the real value it's created, described in the following steps, and enabled by the components illustrated in the photo below:

  1. CHW visits village
  2. CHW conducts PoC diagnostic tests within minutes per individual
  3. Diagnostic box sends test results to smart device via Bluetooth
  4. Smart device uploads patient EMR to secure cloud
  5. Slate's data management software analyses data
  6. Slate's software communicates results to stakeholders via multiple channels

Photo Credit: Forbes India

The time and cost savings resulting from this process are substantial. Not only do the patients avoid costly travel for multiple tests when needed, but Slate only charged $1.25 for a panel of 10 standard biometric tests during their pilot. Referencing the table below, this is nearly 1/20th the cost of what the government was charging before at subsidized prices for the same 10 tests. Theranos, a highly disruptive lab diagnostics company featured in an earlier post, charges in the range of the government prices, but does not appear to offer Urine Glucose tests or the more manual tests like blood pressure, temperature, ECG and Malaria. And standard laboratory diagnostic test prices are sky high in comparison, as the prices from show. Slate's measurement accuracy performance is within 99.9% of the traditional path lab technologies, and they continue to publicly track their clinical evaluation data to improve the device's performance. 

How Transferable is the Slate Device to U.S. Healthcare? 

Theranos is valued at $9 billion because they are providing essentially the same laboratory diagnostics as traditional labs at a fraction of the cost (often 1/10th-1/20th), using less invasive methods, and at much faster speeds. Slate appears to provide PoC diagnostic test results even faster than Theranos using RDT strips, at another huge price discount (1/20th of Theranos's prices from the chart above), and in similarly non-invasive methods. They have also developed 18 apps so far to fulfill the true meaning of 'software as a service.' On the other hand, Slate only offers 33 tests, whereas Theranos is set to offer the full range of possible laboratory diagnostic tests, which number into the thousands. Convincing the medical establishment that the diagnostic results from Slate represent medical grade data would be difficult. EHRs have been slow to become truly personal due to hospital policies and HIPPA. Many of Slate's apps are culturally-bounded, and would need to be re-written.

Nevertheless, I am optimistic of Slate's potential to disrupt clinical laboratory diagnostic testing in the U.S. Here's why:

Firstly, there are encouraging signs that both the FDA and the medical establishment are beginning to approve diagnostic biometrics from wearables, and if from wearables, why not from Slate? At any rate, Theranos is operating as a CLIA-approved lab without FDA approval due to the unique classifications for laboratory diagnostics, and because Slate is already tracking their clinical evaluation data carefully and cheaply, this may not be such a large barrier even if their system does require approval. Secondly, apps are relatively easy to write. Thirdly, if Slate's battery of tests represent the most commonly ordered tests, then it could disrupt Theranos even as Theranos is disrupting Quest and LabCorp. After all, the device is affordable, portable, and modular enough for use by school nurses, the smallest retail clinics, and even by patients themselves. If a device like Slate becomes the go-to for routine diagnostic tests, how commonly will people need to level-up and go to a Walgreens to order a Theranos test that costs 10x the price, or to a hospital where the test will cost 200x the price? 

According to Dr. Kahol's Twitter feed, he has already received requests for the device from both emerging and developed health markets, such as Pakistan, South Africa, Malaysia, and the UK. Field workers from at least 8 different countries, representing both developed and emerging markets, are already using the device on a pilot basis, as this usage map shows.

Question for Comments: 

  • What barriers prevent Swasthya Slate from transferring to U.S. healthcare markets, and how might you recommend they reduce these?  

Monday, 1 December 2014

$100 Stents from India in the Offing

Since the USFDA approved cardiac stents in 1994, there has been tremendous growth in their use, to where 700,000 are implanted in patients annually in the U.S. alone.  Manufacturers' sales of stents reached $5.5 billion in 2012.

In the U.S. about 50% of stent implants are for acute conditions, such as unstable angina, or chest pain caused by the buildup of plaque in the arteries surrounding the heart. The other half are for elective-use patients in stable condition. This video from the Mayo Clinic illustrates Percutaneous Coronary Intervention (PCI), the procedure for which the large majority of stents are used:

The price of stents has followed the broad pattern of decline typical among innovation life cycles. Bare Metal Stents (BMSs) were first priced wholesale around $1,600 in the early 90s, but today the average cost is around $700 each. Drug-eluting Stents (DESs) were a new model introduced in the early 2000s that have helped marginally reduce potentially fatal angioplasty complications, such as blood clotting. These were originally priced around $3,000, and now cost about $1,500 each. According to Bloomberg, "Hospitals receive an average payment of about $25,000 per stent case from private insurers . . . [and] Doctors who implant stents earn a separate fee that averages about $1,000." The stent cost, then, only comprises ~5% of the total price of implantation, not to mention the ongoing costs for the patient, including blood-thinning medications, which are prescribed as a matter of course to mitigate the risk of restenosis.

Until 2001, India imported nearly 100% of stents used, although there were successful early innovation efforts at lowering the cost of stents through indigenous innovation. A network of doctors and engineers at Care Hospitals, Mediciti Hospitals, and the Society for Biomedical Technology pioneered a low-cost coil stent patented as the Kalam-Raju in 1995, crashing the prices of imported stents by several factors, and remained about 50% less expensive, costing around $250. One of their main reasons for creating a low cost stent was that the reason for the "very low number of cardiac procedures carried out in India is poor affordability arising from the high cost of imported consumables." Depending on the report, there were about 2,000 Kalam-Raju stents implanted before mesh designs replaced the coil stents.

Kalam-Raju Coil Stent, 1995
Indian innovators have generally trailed the major OEMs in stent innovations, so the design shift from mesh to scaffolding, and the biochemical shift to drug-eluting stents have not been kind to the market for indigenous stents in India. There remains a general mistrust of locally manufactured products among large, urban hospital systems, who remain the largest buyers of stents. These developments have maintained a high import price for foreign stents in India, and some reports show patients paying up to three times the import price for stents! Needless to say, the penetration of angioplasty in the Indian market remains tragically low. Of course, the high price of consumables is not the only reason for this. Another is the very low number of cardiologists in the country.

Nevertheless, it's clear that any solution to providing stents when and where they're needed in India will require low cost, high quality stents as part of the solution. Efforts to create affordable indigenous stents have continued in India, such as Relisys' Corel+C in 2007, a collaboration between Dr. Balram Bhargava, Relisys, and Germany's CINVENTION to create a non-polymer-based DES that was both safer and 50% the cost of imported DESs. Ultra efficient cardiac care centers like Narayana continue to ensure foothold markets for indigenous stents that are 'good enough.' While some may scoff at the use of 'good enough' applied to Class 3 implantable devices where quality is at a premium, we should all keep in mind that some device features are luxuries, and that having the basic version of a device that's almost as good, but that costs multiple factors less is far more valuable to a patient whose alternative is nothing at all. And the benefits of the 'best-performing' stents are quite marginal. A recent study found the rates of late stent Thrombosis to vary only marginally from one stent type to another as follows:
  • BMS stents: 1.5% 
  • Old generation DES: 1.1% 
  • New generation DES: .9%
Insured patients in the U.S. will pick the new generation DES every time for a .2% risk reduction, but the potential for disruptive innovation lives on in Indian healthcare, where only some 25% of the population even has access to some form of insurance, so that patients often haggle extensively with providers on costs for which they pay out of pocket. 

The implicit question of each post is always, 'could this device disruptively transfer into the U.S. healthcare market?' Let's compare the prices first: 
$200 is a already a low price for a stent, but the Indian OEMs give a further 25% discount to hospitals who buy directly from them, bringing the cost to $150. Furthermore, the Indian Council of Medical Research is currently conducting a study comparing the quality of imported vs. indigenous stents made by companies like Opto Circuits India, Sahajanand, Translumina Therapeutics, Vascular Concepts, and Vasmed Technologies. If they find similar levels of clinical outcomes, this would help drive domestic uptake, and further decrease the price through economies of scale, perhaps to around $100 each. 

Hospitals using fee-for-service reimbursement may desire to purchase lower cost stents, where 'cathlabs' have become valuable sources of profit at fixed reimbursement rates. Just switching to a $500 Indian DES from a $1,500 U.S. DES would save a hospital performing 2,000 implants per year $2,000,000. In time, payers will reduce reimbursement rates as low-cost stents contribute to low-cost implant procedures, further enabled through remote surgery trends and other efficiency gains. In the meantime, it may already be the case that ACOs and other specialized surgery centers in the U.S. who are incentivized to reduce costs at or above average outcomes would be interested in offering patients a choice between a $100 BMS with a 1.5% chance of Thrombosis, or a DES for $1,500 with a .9% chance of Thrombosis.

Starter Question for Comments: 
  • What's stopping low cost stents from being adopted in the U.S. today?