Screening strategies for adults with type 2 diabetes mellitus: a systematic review protocol

There is limited evidence on whether screening for type 2 diabetes mellitus affects health outcomes. A recent systematic review of randomised clinical trials found only one trial that met their inclusion criteria; therefore, current guidelines for screening interventions for type 2 diabetes mellitus are based on expert opinions and best practice rather than synthesised evidence. This systematic review seeks to collate evidence from non-randomised studies to investigate the effect of screening for adults with type 2 diabetes on outcomes including diabetes-related morbidity, mortality (all-cause and diabetes-related) and harms.

This systematic review will follow Effective Practice and Organisation of Care (EPOC) guidelines for the synthesis of non-randomised studies. We will search PubMed/MEDLINE, Scopus, Web of Science, CINAHL, Academic Search Premier and Health Source Nursing Academic (from inception onwards). We will include non-randomised trials, controlled before-after studies, interrupted time-series studies, repeated measures studies and concurrently controlled prospective cohort studies. The primary outcome will be diabetes-related morbidity (microvascular complications of diabetic retinopathy, nephropathy or neuropathy or macrovascular complications of non-fatal myocardial infarction, peripheral arterial disease or non-fatal stroke). The secondary outcomes will be mortality (all-cause and diabetes-related) and harms of screening strategies to patients (including psychological harms or adverse events following treatments) or to health care system (including resource allocation for false-positives or overdiagnosis). Two reviewers will independently screen all citations and full-text articles. Data will be abstracted by one reviewer and checked by a second. The risk of bias of individual studies will be appraised using the ROBINS-I tool. GRADE will be used to determine the quality of the scientific evidence. If feasible, we will conduct random effects meta-analysis where appropriate. If necessary, analyses will be conducted to explore the potential sources of heterogeneity (e.g. age, sex, socio-economic status, rural versus urban or low-middle income versus high-income country). We will disseminate the findings via publications and through relevant networks.

The protocol outlines the methods for systematically reviewing and synthesising evidence of screening strategies for type 2 diabetes mellitus and their effect on health outcomes associated with the disease. The potential impact of this systematic review is improved evidence-informed decision-making for policies and practice for screening of type-2 diabetes.

PROSPERO CRD42020147439.