493-P: Medical Assistant Health Coaching for Type 2 Diabetes in Primary Care: A Pragmatic, Cluster-Randomized Controlled Trial

Abstract

While good cardiometabolic control can reduce complications in type 2 diabetes (T2D) , many individuals do not achieve clinical targets. This real-world trial compared medical assistant health coaching (MAC) to usual care (UC) in improving outcomes among adults with T2D and ≥ 1 of the following in the past 90 days: HbA1c ≥ 8%, LDL ≥ 100 mg/dL, SBP ≥ 140 mm/Hg. This N=600 study was conducted from 2016-2021 at four cluster-randomized clinics within two diverse health systems in Southern CA: a large, private insurance-based system [n = 298, majority non-Hispanic White, English-speaking, mid/upper socioeconomic status (SES) ]; M age = 70.2 ± 12.6] and a Federally-Qualified Health Center (n = 302, majority Hispanic, Spanish-speaking, lower SES; M age = 56.9 ± 12.0) . MAC and UC received diabetes care as usual in primary care, with referral to diabetes education and/or specialists as indicated. MAC also received in-clinic and telephone-based health coaching for up to one year (1-14 sessions; M = 5.2 ± 2.8) . Baseline mean HbA1c = 8.8 ± 2.2%, LDL = 101.8 ± 44.8 mg/dL, and SBP = 137.9 ± 19.4 mm/Hg. From baseline to 12-months, the percentage of individuals meeting quality metrics increased in both groups: HbA1c < 8% (MAC: 74.8 to 79.9%; UC: 64.0 to 80.4%) , LDL < 100 mg/dL (MAC: 75.7 to 84.9%; UC: 72.6 to 82.4%) , SBP < 140 mm/Hg (MAC: 75.0 to 86.9%; UC: 80.2 to 86.4%) . When outcomes were analyzed continuously using multi-level modeling, both MAC and UC demonstrated statistically significant improvements on all clinical outcomes; however, the rate of improvement was largely equivalent across groups (time-by-group ps > .05) . Due to the pragmatic nature of this trial, there was marked variation (by design) in intervention duration and intensity across participants, and in clinical and sociodemographic characteristics of participants across health systems. Additional analyses examining under what conditions and/or for whom the MAC intervention may prove more/less effective will be presented. Disclosure A.L.Fortmann: Consultant; Ascensia Diabetes Care, Employee; Dexcom, Inc. A.Philis-tsimikas: Advisory Panel; Bayer AG, Novo Nordisk, Research Support; Lilly Diabetes, National Institute of Diabetes and Digestive and Kidney Diseases, Novo Nordisk, Viking Therapeutics. T.Clark: None. S.R.Spierling bagsic: None. H.Sandoval: None. K.L.Savin: None. S.Roesch: None. L.Gallo: None. Funding NIH/NIDDK 1R18DK104250