Effectiveness of Interactive Self‐Management Interventions in Individuals With Poorly Controlled Type 2 Diabetes: A Meta‐Analysis of Randomized Controlled Trials

Abstract

Aims: To identify, assess, and summarize available scientific evidence on the effectiveness of interactive self-management interventions on glycemic control and patient-centered outcomes in individuals with poorly controlled type 2 diabetes. Methods: Major English and Chinese electronic databases including Medline, EMBASE, Cochrane Central Register of Controlled Trials, China National Knowledge Infrastructure, and WanFang Data were searched to identify randomized controlled trials that reported the effectiveness of interactive self-management interventions in individuals with poorly controlled type 2 diabetes (glycated hemoglobin [HbA1c] 7.5% or 58 mmol/mol), from inception to June 2015. Data extraction and risk-of-bias assessment were performed by two reviewers independently. Meta-analysis was performed using Review Manager 5.3. Results: A total of 16 trials with 3,545 participants were included in the meta-analysis. Interactive self-management interventions could have a beneficial effect in individuals with poorly controlled type 2 diabetes in reducing HbA1c (mean difference: −0.43%, 95% CI: −0.67% to −0.18%), improving diabetes knowledge (standardized mean difference [SMD]: 0.30, 95% CI: 0.03 to 0.58), enhancing self-efficacy (SMD: 0.29, 95% CI: 0.14 to 0.44), and reducing diabetes-related distress (SMD: −0.21, 95% CI: −0.39 to −0.04). Self-management interventions supported with theory and structured curriculum showed desirable results in glycemic control. The behavioral change techniques, including providing feedback on performance, problem-solving, and action planning, were associated with a significant reduction in HbA1c. Linking Evidence to Action: Individuals with poorly controlled type 2 diabetes could benefit from interactive self-management interventions. Interventions targeting patients with poorly controlled diabetes, those who are at the greatest risk of developing complications, should be prioritized. Our findings indicate that providing feedback on performance, problem-solving, and action planning are promising behavioral change techniques specifically for individuals with poor glycemic control.

Examining the theory‐effectiveness hypothesis: A systematic review of systematic reviews

“…Of the remaining 6,173 articles, the titles and abstracts were screened for eligibility and 141 full texts were obtained for a full assessment. Nine reviews met all criteria (Alageel, Gulliford, McDermott, & Wright, ; Avery et al ., ; Cheng et al ., ; Cradock et al ., ; Dombrowski, Sniehotta, Avenell, & Coyne, ; Goodwin, Ostuzzi, Khan, Hotopf, & Moss‐Morris, ; Huttunen‐Lenz, Song, & Poland, ; Lara, Evans, et al ., ; Samdal, Eide, Barth, Williams, & Meland, ). In total, 132 texts were excluded (see Figure and Appendix S4 for excluded papers).…”

“…Theory use was reported by 47% (85 out of 183) of studies included in the nine systematic reviews (see Table for an overview of individual review results). A range of populations were studied and included individuals diagnosed with type 2 diabetes (Avery et al ., ; Cheng et al ., ; Cradock et al ., ), individuals diagnosed with cardiovascular disease (Goodwin et al ., ; Huttunen‐Lenz et al ., ), obese individuals who were over 40 years old (Samdal et al ., ), obese individuals with comorbidities (Dombrowski et al ., ), healthy adults (Alageel et al ., ), and individuals of retirement age (54–70 years; Lara, Evans, et al ., ). Three reviews examined dietary and PA interventions (Cradock et al ., ; Dombrowski et al ., ; Samdal et al ., ), two reviewed multiple behaviour change interventions (Alageel et al ., ; Goodwin et al ., ), and the other reviews investigated dietary interventions (Lara, Evans, et al ., ); PA interventions (Avery et al ., ); diabetes self‐management interventions (Cheng et al ., ); and cardiac rehabilitation interventions for smoking behaviour (Huttunen‐Lenz et al ., ).…”

“…Table summarizes review outcomes used to examine associations between theory base and effectiveness, which consisted of 16 separate tests of 11 different outcomes across the nine reviews. These included plasma HbA1c levels in individuals with type 2 diabetes (Avery et al ., ; Cheng et al ., ; Cradock et al ., ), PA levels (Avery et al ., ; Samdal et al ., ), body weight (Alageel et al ., ; Dombrowski et al ., ), dietary behaviour (Lara, Evans, et al ., ; Samdal et al ., ), BMI (Alageel et al ., ), smoking cessation‐related outcomes (Huttunen‐Lenz et al ., ), smoking prevalence and coronary heart disease mortality (Goodwin et al ., ), and other cardiovascular disease‐related outcomes, such as systolic and diastolic blood pressure, and serum total cholesterol levels (Alageel et al ., ). Intervention duration ranged from 1 month to 5 years.…”

“…Follow‐up times within the included studies ranged from 1 month to 9 years. When analysing data to determine effectiveness of theory‐based interventions, eight reviews appeared to pool all relevant data for different lengths of follow‐up periods (Alageel et al ., ; Avery et al ., ; Cheng et al ., ; Cradock et al ., ; Dombrowski et al ., ; Goodwin et al ., ; Huttunen‐Lenz et al ., ; Lara, Evans, et al ., ) and one review assessed effectiveness at specific time points – ‘short term’ (≤6 months) and ‘long term’ (≥12 months; Samdal et al ., ). Publication dates of studies included in the reviews ranged from 1974 to 2016.…”

“…All but two reviews used a standardized quality assessment tool to examine risk of bias – the reviews which did not were secondary analyses (Avery et al ., ; Lara, Evans, et al ., ). Eight reviews reported the intent to conduct a test for homogeneity and adopting a random‐effects model when there was significant heterogeneity (Alageel et al ., ; Avery et al ., ; Cheng et al ., ; Cradock et al ., ; Dombrowski et al ., ; Goodwin et al ., ; Lara, Evans, et al ., ; Samdal et al ., ). Five reviews provided a clear characteristics table displaying participant, intervention, and outcome characteristics for their included studies (Alageel et al ., ; Cheng et al ., ; Cradock et al ., ; Goodwin et al ., ; Samdal et al ., ).…”