Memorandum

UCSF initiates six-month RCT (SUGAR) to evaluate the Livongo for Diabetes Program (cellular BGM, unlimited strips, CDE coaching) – November 15, 2018

N=300 type 2s on any medication; Control arm uses Bluetooth-enabled iHealth BGM + unlimited strips; Primary outcome: A1c reduction at 6 months; First RCT of connected BGM+CDE remote coaching

Livongo just announced the initiation of SUGARStudy to Understand Gaining Access to blood glucose Results – a six-month RCT comparing the A1c-lowering of the Livongo for Diabetes program (cellular-enabled BGM + unlimited free strips + remote CDE coaching) vs. standard diabetes care (Bluetooth-enabled BGM + unlimited free strips) in 300 type 2 adults. In short, what is the added value of Livongo’s digital and individual coaching over a connected BGM/unlimited strips without coaching? The study will take place at UCSF under the direction of PI Dr. Jenise Wong, and Livongo expects interim (3-month) and final (6-month) results to report at ADA 2019 and DTM 2019, respectively.

A1c change from baseline to six months is the primary outcome, with an interim A1c measurement at three months. Secondary outcome measures include change in fasting lipid panel at six months, percent of SMBG readings in-range/hypoglycemia/hyperglycemia and safety outcomes (DKA and severe hypoglycemia) . We would’ve also loved to see blinded CGM metrics compared for two weeks at baseline vs. two weeks at the end of the study – even in a subset of participants – though we appreciate the added cost and complexity.

SUGAR seems like a nicely-designed study with fairly broad inclusion (type 2 adults on any medications; no pumpers or CGM users), and the comparator ensures that Livongo’s brand of digital and remote CDE coaching is the isolated independent variable. We say “nearly” because the control arm is using the iHealth Bluetooth BGM, rather than the Livongo cellular-enabled meter without remote coaching. The user experience is different with the iHealth meter – which displays data on the screen or the Gluco-Smart app – which could theoretically influence SMBG behavior. This may be something to keep an eye on as the study progresses, particularly since neither arm will be instructed on how many fingersticks to perform each day.

  • This is Livongo’s first prospective RCT (that is, comparing Livongo for Diabetes to control) which is uncommon in the field to date. WellDoc (automated coaching/feedback) notably conducted an RCT way back in 2011, Virta is in the midst of a controlled (non-randomized) trial, and Omada is enrolling a DPP RCT, but mySugr, One Drop, and others in the field have not done RCTs to our knowledge. This, along with the fact that SUGAR was first posted on ClinicalTrials.gov back in 2016, speaks to the difficulty of determining the right time to conduct a laborious RCT in connected diabetes devices/digital health. By the time SUGAR is published in a journal, Livongo will likely have gone through a number of software and process iterations, though payers and providers continue to emphasize the importance of supplementing real-world outcomes data with clinical trial results. A sizable drop in A1c – which we might expect from 24/7 CDE access – could mean even greater employee and payer traction, as well as drive competitors to run similar RCTs, a win for patients across the board.

  • Livongo told us that pumpers and CGM users were excluded to keep technology use consistent across the study population. Future RCTs will be powered for outcomes in smaller subsets of members, such as those who use CGMs and pumps.

 

-- by Brian Levine, Adam Brown, and Kelly Close