Summary: In a mixed-methods comparison nested within the PREPARE 4 CGM hybrid type-3 study, 76 Colorado primary care practices self-selected into either practice-led, self-paced continuous glucose monitoring (CGM) implementation (46 practices) or referral to a virtual CGM implementation service (30 practices). The presence of a diabetes care and education specialist (DCES) was the single characteristic distinguishing the groups: 35% of self-paced practices versus 0% of virtually supported practices had a DCES (p<.001). This is an associational, descriptive finding about implementation choice, not a test of CGM clinical effectiveness.
PICO Summary
| Element | Detail |
|---|---|
| Population | 76 primary care practices in Colorado, USA, enrolled in the PREPARE 4 CGM hybrid type-3 effectiveness-implementation study and interested in implementing CGM for patients with diabetes. Mixed (multiple) methods design: baseline characteristic comparison plus semi-structured interviews. |
| Intervention | Practice-led, self-paced implementation using the American Academy of Family Physicians Transformation in Practice Series (TIPS) CGM modules (n=46 practices that chose this arm). |
| Comparison | Referral to a virtual CGM implementation service (virCIS) staffed by a primary care multidisciplinary team, providing CGM initiation and data-interpretation support for at least six months (n=30 practices that chose this arm). |
| Outcome | Only one of 38 compared baseline characteristics differed between arms: presence of a DCES. Among self-paced practices, 16/46 (35%) had a DCES; among virtually supported practices, 0/30 (0%) had a DCES (chi-square X²(1, N=62)=11.046, p<.001). No 95% CI, effect size, ARR or NNT were reported (descriptive comparison, no clinical outcome). All practices with a DCES chose self-paced implementation; among practices without a DCES, the choice was split roughly evenly. |
Expert Commentary
This study is best read as a descriptive, hypothesis-generating account of how primary care practices choose to implement CGM, rather than as evidence that any strategy improves glycaemic or patient outcomes. The headline signal is clean and biologically plausible: every practice that employed a diabetes care and education specialist elected to run CGM implementation itself, whereas none of the practices choosing virtual support had such a specialist on staff. The reported association is strong and statistically robust, but it rests on a single distinguishing variable out of thirty-eight examined, in 76 self-selected practices within one US state, which limits both precision and generalisability. The central limitation is that arm allocation was by practice choice, not randomisation, so the groups differ in unmeasured ways and no causal or comparative-effectiveness claim can be supported; the data describe correlation between staffing and preference, nothing more. No clinical endpoint, such as time-in-range or HbA1c, was assessed here. Can I use this with my patients? Not directly at the bedside, but it is useful for clinic leaders and commissioners: if a practice already has a diabetes educator, supporting in-house CGM rollout is reasonable, while practices without one may benefit from a virtual implementation service to access multidisciplinary expertise. The work was investigator-led and registered, with no commercial CGM-manufacturer sponsorship evident. Future randomised work should test whether matching strategy to practice staffing actually translates into better patient outcomes.
References
Wiggins KT, Hall TL, Jortberg B, Dickinson WP, Dickinson LM, Parascando JA, et al. Primary care practices’ choice of implementation strategy for continuous glucose monitoring for patients with diabetes: a multiple methods study within a larger hybrid type-3 effectiveness-implementation study. BMC Prim Care. 2025;26(1):195. doi:10.1186/s12875-025-02903-0
