Diagnostic pars plana vitrectomy report of a 21-year retrospective study.

PURPOSE

To review the experience of diagnostic pars plana vitrectomies (PPV).

METHODS

The authors reviewed 405 consecutive diagnostic PPV's performed between November 1973 and October 1994.

RESULTS

Diagnostic vitrectomy was performed in 215 (53%) of 405 eyes for suspected endophthalmitis. Of those 215 cases, acute inflammation was confirmed in 62 (28.8%), 60 (27.9%) had microbial organisms present and 36 (16.7%) were culture-positive. Microbial organisms were observed microscopically in 31 (20%) of 156 patients suspected of postoperative endophthalmitis. Of those 31 cases, 23 (74%) were gram-positive, eleven (37%) of 30 eyes had organisms associated with glaucoma filtering procedures and 20 (16%) of 126 eyes had organisms with non-filtering procedures. The pooled percentage of eyes that developed postoperative endophthalmitis as a complication during the period July 1990 thru June 1994 is 5 (0.046%) out of a heterogeneous group of 10,898 cases operated on at the Wilmer Eye Institute for cataract, glaucoma, corneal transplant, pars plana vitrectomy and retinal detachment. Bacteria were identified microscopically in 6 (18%) of 34 post-traumatic cases. Microbial organisms were identified in 23 (92%) of 25 cases with an endogenous infection. Patients with endogenous infections had the most fungal infections, and the majority were in males. Neoplasms were diagnosed in 58 (14%) of the 405 cases. The most common neoplasm was ocular lymphoma 42 (72%), 69% of which were in females. Only 42 (48.3%) of 87 patients clinically suspected of having ocular lymphoma, actually had ocular lymphoma. Those negative for lymphoma were significantly older (67.4 +/- 10 years) compared to those with lymphoma (60.4 +/- 14 years) (P = 0.01).

CONCLUSION

Diagnostic PPV has proved to be valuable in confirming and establishing various clinical diagnoses.