This review found that intravitreal injections of bevacizumab resulted in short-term improvements in visual acuity in patients with primary diabetic macular oedema. These benefits did not extend beyond 12 weeks. The review was based on a small number of studies but was well conducted and the authors' cautious conclusion is likely to be reliable.

To compare intravitreal bevacizumab injections to macular photocoagulation or combination of intravitreal bevacizumab and intravitreal triamcinolone acetonide injections in patients with primary diabetic macular oedema.

MEDLINE, The Cochrane Library, EMBASE, LILACS, TRIP and ClinicalTrials.gov databases and Google Scholar were searched to December 2009 for relevant published studies. Search terms were reported. Reference lists of the included studies and abstracts from meetings of the Association for Research in Vision and Ophthalmology from 2005 onwards were checked for additional studies. The reviewers contacted experts in retinal diseases for information on ongoing studies. There were no language restrictions.

Randomised controlled trials (RCTs) in patients with primary diabetic macular oedema that compared intravitreal bevacizumab injections to either macular photocoagulation or combination treatment of intravitreal bevacizumab and intravitreal triamcinolone acetonide injections were eligible for inclusion. The primary outcome was visual acuity and follow-up was required to be at least six weeks. Secondary outcomes were central macular thickness and adverse events.

The included studies were conducted in Iran and the United States. Where reported, the mean age of included patients ranged from 57 to 62 years. From 47.4% to 61% of patients were male. Doses of intravitreal bevacizumab ranged from 1.25mg to 2.5mg. Macular photocoagulation was the comparator treatment in all the included studies; combination treatment of intravitreal bevacizumab and intravitreal triamcinolone injections was also used in most of the trials.

Two reviewers independently performed the study selection.

Two reviewers independently assessed methodological quality using the five-point Jadad scale of randomisation, blinding and treatment of withdrawals and drop-outs. Jadad scores of 3 or more were judged to be of high quality. A risk of bias assessment was conducted to evaluate selection bias, performance bias, detection bias, attrition bias and recall bias.

Data were extracted by two independent reviewers to calculate mean differences and 95% confidence intervals (CI) leading to summary estimates in visual acuity.

Weighted mean differences (WMD) and 95% CIs were calculated using a random-effects model. Statistical heterogeneity was assessed using Ι². The reviewers performed sensitivity analyses on the basis of study quality (Jadad score >3). Potential for publication bias was evaluated by visual inspection of funnel plots.

Four studies (445 patients, 492 eyes) were included in the review. Sample sizes ranged from 97 to 129 patients. Two studies scored 2 and single studies scored 4 and 5 points on the Jadad scale. The studies with lower scores did not use appropriate randomisation or double-blinding. Follow-up ranged from 12 to 36 weeks. Attrition ranged from 0% to 17%.

At six weeks follow-up, statistically significant benefits were associated with treatment with intravitreal bevacizumab in visual acuity compared to macular photocoagulation (WMD -0.09, 95% CI -0.15 to -0.02; Ι²=3%; four trials) and intravitreal bevacizumab combined with triamcinolone acetonide (WMD -0.07, 95% CI -0.14 to 0.00; Ι²=0%; three trials). There were no significant differences in visual acuity for the comparison of intravitreal bevacizumab with other treatments at 12 weeks follow-up (three trials).

Statistically significant reductions in central macular thickness were observed in patients treated with intravitreal bevacizumab at six weeks follow-up compared to patients who received macular photocoagulation (WMD -30.36, 95% CI -54.12 to -6.60; Ι²=0%; three trials). Differences between intravitreal bevacizumab and macular photocoagulation were not significant at 12 weeks follow-up (three trials). There were no differences between intravitreal bevacizumab and combination treatment of intravitreal bevacizumab plus triamcinolone acetonide at six weeks follow-up (three trials) and 12 weeks follow-up (four trials).

No injection-related adverse events were reported. One patient developed endophthalmitis. Cases in other studies of mild anterior chamber reaction and ocular hypertension resolved spontaneously after one week.

Sensitivity analyses found significant benefits of intravitreal bevacizumab in visual acuity compared to all alternative treatments at 12 weeks follow-up for two studies with scores of at least 3 on the Jadad scale (WMD -0.09, 95% CI -0.16 to -0.02; Ι²=0%).

There were indications that intravitreal bevacizumab injections were effective in improving visual acuity in patients with primary diabetic macular oedema at six weeks post-treatment but benefits were no longer present at 12 weeks follow-up.

The review addressed a clearly defined question. Criteria for the inclusion of studies were stipulated and were reproducible. Several appropriate databases were searched without language restrictions and attempts were made to identify unpublished studies. Steps were taken at each stage of the review process to minimise reviewer error and bias. Methodological quality was assessed and standards varied. There were some discrepancies between text and tables for the reporting of duration of follow-up. The authors' decision to combine the results in meta-analyses appeared justified and appropriate sensitivity analyses were conducted to explore potential variations in the results. The authors correctly acknowledged the limitations of the review for the small number of identified studies, lack of information on lens types of the investigated eyes, use of short-term follow-up and imbalances in baseline visual acuity and central macular thickness values. Potential for publication bias was discussed.

The review was based on a small number of studies but was well conducted and the authors' cautious conclusion is likely to be reliable.

Practice: The authors stated that clinicians should continue to follow clinical guidelines to use macular photocoagulation as first-line treatment in patients with diabetic macular oedema and that treatment providers should discuss benefits and risks of treatment options for eye care.

Research: The authors stated that further research was required to determine optimal dosing regimes for the treatment of eyes with diabetic macular oedema with intravitreal bevacizumab.

Not stated.