Author, date and country | Patient group | Study type (level of evidence) | Outcomes | Key results | Study Weaknesses |
---|---|---|---|---|---|
Blaivas et al 2002 USA | 61 patients with visual loss and ocular pathology including trauma. CT scan used as gold standard. Ultrasound performed by Emergency physicians who have had minimal training of 1 h lecture and 1 h hands on. | prospective observational study | detection of ocular pathology with ultrasound | 8 patients out of 61 with RD. 100% sensitivity and specificity | small number. convenience sample. no kappa value given. Power not calculated as no previous studies available. |
Yoonessi et al 2010 USA | 48 patients with less than 48 hour symptoms of ocular pathology. gold standard was set as final diagnosis from ophthalmologist. ultrasound performed by 15 different emergency physicians of different skill and experience levels | triple blinded prospective observational study | detetction of retinal detachment | 18 out of 48 patients had RD. Sensitivity 100% (95% CI 78% to 100), Specificity 83% (95% CI 65% to 94%) | small sample size. very wide exclusion criteria.no kappa value given. wide confidence intervals |
Shinar et al 2011 USA | 92 patients convenience sample with ocular pathology. gold standard was final diagnosis by ophthalmologist. 31 different emergency physicians performed ultrasound after minimal training for ocular ultrasound | prospective observational study | detection of retinal detachment | 29 out of 92 patients with RD. Sensitivity 6.5% (95% CI 82% to 100%), specificity 92.0% (95% CI 68% to 95%) | convenience sampling. wide confidence interval. no kappa value. power of study not calculated |
Jacobsen et al 2015 USA | charts of 109 patients with suspected retinal detachment reviewed retrospectively. gold standard was final diagnosis by ophthalmologist. Reviewers were blinded to results of emergency ultrasound. | retrospective study | detection of RD with ocular ultrasound | 34 out of 109 patients had RD. sensitivity 91% (95% CI 76-98) specificity 96% (95% CI 89-99) | retrospective study. one EP performed 35% of the ultrasound causing potential inflation bias. Ophthalmologist diagnosing as gold standard not blinded to EP ultrasound results |
Vrablik et al 2014 USA | A systematic review and meta-analysis. 7771 unique citation found. 78 papers selected for full text review. 4 papers assesed for quality. 3 deemed to have low risk of bias | SYstematic review and Meta-analysis | accuracy of bedside ultrasound for detection of RD | range under the curve 0.943 to 1.0. summary area under the curve 0.957 | total number of enrolled patients in 3 trials is 201. referral or spectrum bias could not be excluded. |