**Background:** Diabetes is a leading cause of premature mortality, blindness, renal failure, amputations, and cardiovascular events worldwide. Despite evidence-based guidelines, substantial gaps in quality of care persist. Healthcare systems increasingly invest in quality improvement (QI) programmes, which typically consist of multiple strategies targeting patients, healthcare professionals, and/or system-level changes. This Cochrane review updates previous reviews (2006: 58 trials; 2012: 142 trials) using novel meta-analytical techniques to estimate the effectiveness of individual QI strategies and their combinations.
**Methods:** The authors searched CENTRAL, MEDLINE, Embase, CINAHL, ClinicalTrials.gov, and WHO ICTRP to 4 June 2019 (with a top-up search to 23 September 2021). They included randomized trials assessing QI programmes in outpatient settings for adults with diabetes. QI strategies were classified using an adapted Cochrane EPOC taxonomy: system-targeted (case management [CM], team changes [TC], electronic patient registry [EPR], facilitated relay [FR], continuous quality improvement [CQI]), provider-targeted (audit and feedback [AF], clinician education [CE], clinician reminders [CR], financial incentives [FI]), and patient-targeted (patient education [PE], promotion of self-management [PSM], patient reminders [PR]). Patient-targeted strategies required at least one provider- or system-targeted strategy. Two independent reviewers screened and extracted data. The authors modelled the association of each QI strategy with outcomes using hierarchical multivariable meta-regression in a Bayesian framework. For continuous outcomes (HbA1c, SBP, LDL-C), baseline interaction models were used (thresholds based on median baseline values: HbA1c 8.3%, SBP 136 mmHg, LDL-C 107 mg/dL). QI strategies were qualitatively ordered into Top, Middle, and Bottom tiers based on relative magnitude of effect.
**Key Results:** 553 trials (428 patient-randomized, 125 cluster-randomized) with 412,161 participants were included. 66% involved type 2 diabetes only; 50% of participants were female; median age was 58.4 years; mean follow-up was 12.5 months. HbA1c was the most commonly reported outcome (89% of trials). The most frequently evaluated QI strategies were PE (50%), PSM (45%), and CM (39%); the least frequent were AF (6%), FI (3%), and CQI (2%). The median number of QI strategies in multicomponent programmes was 3.
Four QI strategies (CM, TC, PE, PSM) were consistently identified as 'Top' across most outcomes. For HbA1c in populations with higher baseline (>8.3%), CM was associated with a reduction of -0.27% (95% CrI -0.39 to -0.15), PE with -0.17% (95% CrI -0.30 to -0.05), and EPR with -0.17% (95% CrI -0.33 to -0.02). For lower baseline HbA1c (≤8.3%), PSM was associated with -0.14% (95% CrI -0.25 to -0.06), PR with -0.14% (95% CrI -0.25 to -0.03), and CE with -0.13% (95% CrI -0.24 to -0.01).
Combinations of the three most effective strategies yielded: PR+PSM+CE: HbA1c reduction of -0.41% (95% CrI -0.61 to -0.22) when baseline HbA1c <8.3%; CM+PE+EPR: HbA1c reduction of -0.62% (95% CrI -0.84 to -0.39) when baseline HbA1c >8.3%; PE+TC+PSM: SBP reduction of -2.14 mmHg (95% CrI -3.80 to -0.52) when baseline SBP <136 mmHg; CM+TC+PSM: SBP reduction of -4.39 mmHg (95% CrI -6.20 to -2.56) when baseline SBP >136 mmHg; TC+PE+CM: LDL-C reduction of -5.73 mg/dL (95% CrI -7.93 to -3.61) when baseline LDL <107 mg/dL; TC+CM+CR: LDL-C reduction of -5.52 mg/dL (95% CrI -9.24 to -1.89) when baseline LDL >107 mg/dL.
For screening outcomes assuming 50% baseline screening, PE+PR+TC was estimated to yield 83% (95% CrI 62% to 94%) for retinopathy screening, and PE+TC+Other yielded 88% (95% CrI 70% to 95%) for foot screening.
Certainty of evidence was generally low to very low due to risk of bias concerns (51% of studies had high risk of bias for incomplete outcome data) and imprecision.
**Clinical Implications:** Multicomponent QI programmes combining effective strategies can achieve meaningful population-level improvements in diabetes care. Effects approximate the minimal clinically important difference for pharmaceutical interventions in glycaemic control (0.3-0.4% HbA1c reduction). The authors recommend that those planning diabetes QI programmes consider including one of team changes or case management AND one of promotion of self-management or patient education. The specific choice should depend on the targeted population and local resources. The review will be maintained as a living systematic review.