Takeaway: We remain Short DXCM in the Hedgeye Health Care Position Monitor.

Overview | New data, same comments from physicians

We spoke to two physicians this past week from large group practices we identified as high utilizers of continuous glucose monitoring.  Our claims data has recently expanded to include patient level detail such as diabetes type, physician, and group practice.  We've adjusted our analysis method in a couple of critical ways that seems to improve the utility of the data and provides some additional insights. 

Our experts reconfirmed our view that CGM is likely limited to the T1 and insulin intensive T2 patient populations.  Expanding into a T2 population may be possible if insurers allow intermittent use for patient training, although broad adoption is unlikely in the absence of a large study. Patient adoption increased significantly in 2018 and 2019 with the advent of G6, but in recent months has been slower. Consistent with other experts, Dexcom is 90% of T1 use, which is ~50% penetrated, while Libre claims 50% of the T2 population use, which is ~20% to 50% penetrated.  Most interesting to us was the universal praise for TNDM, the structural headwinds to switching from MDT, and that Omnipod is likely to remain a niche product.

New and existing patient trend, t1 and insulin intensive t2

CPT code 95251 is billed by physicians when they download and review a patient's CGM data.  There are limits on the billing frequency and differences in the number of times patients will have their data reviewed depending on their diabetes status.  For example, gestational diabetes is weekly while some well controlled T2s are rarely reviewed.  We've adjusted our data feed and the analysis. The trend lines up reasonably well with both physician commentary and company results.  

DXCM | Field notes on DXCM, TNDM, PODD, and ABT - totalnewT1T2

LARGE PRACTICE TRENDS

Trends among large practices have decelerated modestly on a sequential basis. Physician comments did not indicate anything other than the launch of G6 as the catalyst for growth in 2018 and 2019, but the pace has slowed.

DXCM | Field notes on DXCM, TNDM, PODD, and ABT - largegrpcgm

Field Notes

Endocrinologist #1
  • G7 - better, but not huge improvement
  • 15 days, smaller, integrated, it may expand the market
  • Expect them to take some share from Libre, but not much.  People like simple and cheap
  • Time in range is replacing A1C as the endpoint
  • Unethical not to have CGM at this point, but we are years away
  • See some use in the anti-aging market, the worried well
  • Peak CGM penetration in T2 II is 50%
  • Peak T1 penetration for CGM is 90%
  • To get use into broader population will have to be a didactic measure, patient education add on
  • "The number one task is to avoid hypoglycemia, which does not apply to T2 broadly, or even the T2 insulin intensive patient"
  • Freestyle Libre 1.0 function is similar in 2.0
  • Does not share data well and sharing is extremely important
  • Libre does not have a share function
  • alarm was great to get, but alert does not come with a reading, a blood glucose number
  • prompts you to swipe instead, which is a small but annoying thing for the patient
  • Freestyle 3 will be needed, 2.0 not a big enough deal
  • 50% of patients are "interpreted" every month, 40% measured every third month
  • G6 is preferred where cost and access is not an issue
  • Libre fits best for older, less tech savvy, alarm fatigue, simpler to use
  • 2-4% of patients quit CGM annually
  • T2 insulin patient penetration is "less than 20% of patients, maybe 15%"
  • "Some patients show me the data once a week, but you can only bill once a month"
  • I don't always submit the claim, and the patient may be seen by an educator or nurse between visits with the physician
  • 72 hours of data needed to bill
Endocrinologist # 2
  • 60-70% of T1 currently on CGM, 90% is peak
  • 40-60% of T2- insulin intensive on CGM, 60% is peak
  • "I prescribe Dexcom only... if they can have it."
  • "Some plans cover Dexcom, others don't"
  • The exclusion is an extremely anxious patient who treats too soon and :insulin stacks"
  • Libre not as accurate at the lower end of the range
  •  50% of T2 insulin patients are Libre
  • Libre 2.0 "needs to be super accurate"
  • "We use CGM for every single pregnant woman"
  •  I have never seen a patient on a basal insulin using a CGM device
  • I prescribe Tandem to everybody, ControlIQ
  • PODD, "small fraction of patients"
  • Tandem control IQ is "amazing" "incredible"
  • Tandem is the best technology
  • "closed loop"
  • Medtronic will give up a lot of share
  • Pump share currently is 70% Medtronic, 30% Tandem, but almost all new patients and patients coming out from a warranty switch to TNDM
  • Cost is a factor, insurance coverage, Warranty and Medtronic offers.
  • Expecting to see a 'big wave' of switches next year
  • 50% of T1 are on a pump, no really great reason for that to move much higher
  • PODD is cheaper and tubeless, but "I don't know where it fits"
  • For some patients tubeless is great, catheter gets kinked, change every two days is a hassle

All data available upon request. Please reach out to  with any inquiries.

Thomas Tobin
Managing Director


Twitter
LinkedIn