The authors concluded that insulin therapy in type 2 diabetes should start with adding a daily injection to oral blood glucose-lowering agents. There was no conclusive evidence about the best approach to use for subsequent management. A lack of information about review methods and failure to assess study quality mean these conclusions may require a degree of caution.

To assess the most effective insulin regimen for type 2 diabetes.

PubMed and EMBASE were searched from 2000 to April 2008. Search terms were reported. The search was limited to studies in English.

Randomised controlled trials (RCTs) of insulin with or without oral hypoglycaemic therapy for adults with type 2 diabetes were eligible for inclusion, provided they had at least 10 participants in each arm and at least two months’ follow up. Outcomes were required to include glycaemic control (glycosylated haemoglobin (HbA1c), post-prandial plasma glucose and/or fasting blood glucose) and hypoglycaemic events (as defined in the primary study). Trials of inhaled insulin, pump therapy, amylin agonists and glucagon-like peptide-1 analogues were excluded.

Studies in the review varied widely in their inclusion criteria as to participant age, body mass index and glycaemic values. Many participants were insulin naïve; others were using insulin at baseline (with or without oral glucose-lowering agents). Many different types of insulin were investigated (such as human, analogue, long- and short-acting, premix, preprandial and basal-bolus). Individual premix and basal-bolus regimens varied widely. The studies compared different regimens (with or without oral glucose-lowering agents) with each other or with oral glucose-lowering agents alone. Adverse events were reported in the review: these comprised hypoglycaemic events and weight gain. Study duration ranged from 12 weeks to six years. Almost all the studies were sponsored by pharmaceutical companies.

Potentially eligible studies were retrieved by one reviewer based on study title and abstract (if available). The authors stated neither how the papers were selected for the review nor how many reviewers performed the selection.

[A: Study selection, data extraction and assessment was performed by the first author, but in case of any doubt about inclusion or assessment the second author was consulted.]

The authors did not state that they assessed validity.

Data on outcomes of interest were extracted from individual studies. Also, data on any positive or negative events reported in the studies were extracted.

The authors stated neither how the data were extracted for the review nor how many reviewers performed the data extraction.

[A: Study selection, data extraction and assessment was performed by the first author, but in case of any doubt about inclusion or assessment the second author was consulted.]

Studies were grouped by intervention and were combined by narrative synthesis, supported by tables. The interpretation was weighted by patient years.

Seventy-eight studies were included in the review (n=18,288, range 24 to 826 participants; two to 4,956 patient-years).

Combinations of insulin with oral glucose-lowering agents achieved glycaemic control similar to or better than monotherapy insulin, with (in general) less weight gain and lower insulin requirements (13 RCTs).

Rosiglitazone or pioglitazone achieved similar glycaemic control to long-acting insulin, with fewer hypoglycaemic events; both were given in combination with other oral glucose-lowering agents (four RCTs).

Longer-acting insulin analogues (glargine and detemir) achieved similar glycaemic control to NPH insulin, with fewer hypoglycaemic events (14 RCTs); in most of these RCTs both arms also received oral glucose-lowering agents.

In nine of 10 relevant RCTs, premix insulin achieved better glycaemic control than long-acting insulin (with or without oral glucose-lowering agents in both arms); in most of these studies there were more hypoglycaemic events in the premix arms.

Analogue premix insulin achieved glycaemic control equal to or better than human premix, without increased hypoglycaemia rates or weight gain (eight RCTs). No conclusions could be reached about basal-bolus regimens.

Other results were reported in the review.

Insulin therapy in type 2 diabetes should start with adding a daily injection to oral blood glucose-lowering agents. There was no conclusive evidence about the best approach to use for subsequent management

The objectives and inclusion criteria of the review were clear and relevant sources were searched for studies. Only two databases were used, the search was limited by language and it did not appear that specific attempts were made to retrieve unpublished studies, so it was possible that some studies were missed. It did not appear that steps were taken to minimise the risk of reviewer error and bias by having more than one reviewer independently make decisions about review processes; potentially eligible studies were selected by a single reviewer and the processes used for final study selection and for data extraction were not described. It did not appear that study validity was systematically assessed. These factors make it difficult to assess the reliability of the findings. The decision not to pool studies statistically appeared appropriate, given the heterogeneity between them. The studies with most patient-years were emphasised in the interpretation of results, but no effect sizes or measures of statistical significance were reported. This made it difficult to evaluate the clinical significance of the findings. In view of the lack of information about review methods and the failure to assess study quality, the authors’ conclusions may require a degree of caution.

Practice: The authors stated that metformin and/or sulphonylurea should continue when insulin therapy commenced and that the addition of rosiglitazone to a combination of two oral glucose-lowering agents may be a reasonable alternative to starting insulin. For a patient not on maximally tolerated doses of oral glucose-lowering agents, the clinician could consider adding insulin instead of an oral glucose-lowering agent. More than one insulin injection daily may be required. Analogue premixes may be particularly suitable for patients able to reach a low HbA1c.

Research: The authors stated that large multicentre RCTs with long follow up and relevant outcomes measures were needed to determine the optimum insulin regimen for type 2 diabetes.

Sanofi-Aventis.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.