Effectiveness of Integrated Diabetes Care Interventions Involving Diabetes Specialists Working in Primary and Community Care Settings: A Systematic Review and Meta-Analysis

Abstract

Introduction: Evidence that integrated diabetes care interventions can substantially improve clinical outcomes is mixed. However, previous systematic reviews have not focussed on clinical effectiveness where the endocrinologist was actively involved in guiding diabetes management.

Methods: We searched EMBASE, COCHRANE, MEDLINE, SCOPUS, CINAHL, Google Scholar databases and grey literature published in English language up to 25 January 2021. Reviewed articles included Randomised Controlled Trials (RCTs) and pre-post studies testing the effectiveness on clinical outcomes after ≥6 months intervention in non-pregnant adults (age ≥ 18 years) with type 1 or type 2 diabetes mellitus. Two reviewers independently extracted data and completed a risk of bias assessment. Appropriate meta-analyses for each outcome from RCTs and pre-post studies were performed. Heterogeneity was assessed using the I² statistic and Cochran's Q and publication bias assessed using Doi plots. Studies were not pooled to estimate the cost-effectiveness as the cost outcomes were not comparable across trials/studies.

Results: We reviewed 4 RCTs and 12 pre-post studies. The integrated care model of diabetes specialists working with primary care health professionals had a positive impact on HbA1c in both RCTs and pre-post studies and on systolic blood pressure, diastolic blood pressure, total cholesterol and weight in pre-post studies. In the RCTs, interventions reduced HbA1c (-0.10% [-0.15 to -0.05]) (-1.1 mmol/mol [-1.6 to -0.5]), versus control. Pre-post studies demonstrated improvements in HbA1c (-0.77% [-1.12 to -0.42]) (-8.4 mmol/mol [-12.2 to -4.6]), systolic blood pressure (-3.30 mmHg [-5.16 to -1.44]), diastolic blood pressure (-3.61 mmHg [-4.82 to -2.39]), total cholesterol (-0.33 mmol/L [-0.52 to -0.14]) and weight (-2.53 kg [-3.86 to -1.19]). In a pre-post study with no control group only 4% patients experienced hypoglycaemia after one year of intervention compared to baseline.

Conclusions: Integrated interventions with an active endocrinologist involvement can result in modest improvements in HbA1c, blood pressure and weight management. Although the improvements per clinical outcome are modest, there is possible net improvements at a holistic level.

Keywords: clinical outcome; cost-effectiveness; diabetes mellitus; integrated health care systems; multidisciplinary care; primary health care.

Figure 1

PRISMA flowchart diagram for studies selection (based on four-phase PRISMA flowchart diagram).

Figure 2

Forest plots for randomised controlled trials clinical outcomes. The results are expressed in WMD – weighted mean difference with 95% confidence intervals.

Figure 3

Forest plots for pre-post studies clinical outcomes. The results are expressed in WMD – weighted mean difference with 95% confidence intervals.