Comparison between clinical and ultrasound findings in patients with vitreous hemorrhage.

PURPOSE: To ascertain the causes of vitreous hemorrhage and to determine the accuracy of ultrasound (U/S) in these cases, based on the degree of agreement between ultrasound and clinical findings. METHODS: A chart review of 96 consecutive patients (106 eyes) with dense vitreous hemorrhage who underwent A- and B-scan U/S by one examiner between June 1996 and June 1999. U/S records were evaluated to determine the presence and exact distribution of areas of retinal detachment and the presence of posterior vitreous detachment, retinal tear, intraocular foreign body, or choroidal detachment. Clinical information was obtained from the medical records after the vitreous hemorrhage was reabsorbed or following vitreous surgery. Clinical and U/S findings were compared. False-positive and False-negative rates for U/S were calculated based on clinical findings. RESULTS: In 37 eyes (35%) the vitreous hemorrhage was because of proliferative diabetic retinopathy and in 33 eyes (31%) because of ocular trauma. The false-positive rate for retinal detachment (retinal detachment by U/S without clinical confirmation) was 18.9% (seven of 37 eyes). Retinal tears were diagnosed and localized accurately in only four of nine eyes (44%). CONCLUSIONS: The most common cause of vitreous hemorrhage was proliferative diabetic retinopathy, followed by ocular trauma. U/S correctly diagnosed all cases of retinal detachment, but less than 50% of retinal tears. A total of 18.9% of the eyes were falsely diagnosed as having retinal detachment. U/S is an effective diagnostic tool in patients with vitreous hemorrhage.