patients transported from satellite clinics to a nongovern-mental hospital. Outcomes were assessed in a systematicsample of patients operated on in 1996–1997 by follow upat home, and average costs by provider were derived fromactual expenditures during the year. Almost half thepatients operated on in government camps were dissatisfiedwith the outcome (34/70, 49%). More than one third wereblind in the operated eye after the surgery (25/70, 36%).User satisfaction was higher with other providers (medicalcollege hospital 82%; nongovernment hospital 85%) andfewer patients remained blind. Camps were a low-costoption, but the poor outcomes reduced their cost-effective-ness to US $97 per patient. The state medical collegehospital was least cost-effective, at US $176 per patient,and the nongovernmental hospital was the most cost-effective at US $54 per patient. The authors concludedthat the government of India should review its policy forgovernment camp surgery and consider alternatives, suchas transporting patients to better permanent facilities.India and other developing countries should monitoroutcomes in cataract surgery programs, as well as through-put.—Nancy J. Newman

*Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L35QA, UK; e-mail: a.j.singh@liv.ac.uk

● Air bags and ocular injuries. Stein JD, Jaeger EA,*Jeffers JB. Tr Am Ophthalmol Soc 1999;97:59–82.

THE AUTHORS PERFORMED A RETROSPECTIVE STUDY OF

ocular injuries associated with air bag deployment.They identified 44 articles between 1991 and 1998 thatdescribed 97 patients. Corneal abrasions occurred in 49%of occupants, hyphemas in 43%, vitreous or retinal hem-orrhages in 25%, and retinal tears or detachment in 15%.The globe was ruptured in 10 patients. Patients involved inhigher-speed accidents (over 30 mph) sustained a greaterpercentage of vitreous or retinal hemorrhages and trau-matic cataracts, whereas those at slower speeds were moreprone to retinal tears or detachments. In a subset of 14patients with serious ocular injuries, the impact speed of 11patients was recorded at 30 mph or less. Slower speed maybe a risk factor for some ocular injuries. Occupant heightwas not a significant factor. National statistics confirm thatair bags reduce fatalities in motor vehicle accidents.However, children sitting in the front seat without a seatbelt and infants in passenger-side rear-facing car seats areat risk for fatal injury. The authors concluded that air bagscombined with seat belts are an effective means of reduc-ing injury and death in adults during motor vehicleaccidents. However, this study documented a wide varietyof ocular injuries associated with air bag deployment.—Thomas J. Liesegang

*Wills Eye Hospital, Ninth & Walnut Streets, Philadelphia, PA 19107.

● Indeterminate melanocytic proliferations of the con-junctiva. Grossniklaus HE, Margo CE, Solomon AR. TrAm Ophthalmol Soc 1999;97:156–183.

THE AUTHORS TESTED THE HYPOTHESIS THAT A SUBSET

of conjunctival melanocytic proliferations cannot bereproducibly classified as benign, malignant, or indetermi-nate. Three groups of excisional biopsy specimens ofconjunctival melanocytic proliferations were evaluated byfive ophthalmic pathologists. These groups included le-sions that were considered by the authors to representbenign (group 1, n 5 5), malignant (group 2, n 5 5), andindeterminate (group 3, n 5 5) melanocytic proliferations.The panel classified the same sections in all three groups ina randomized, masked fashion, first without and then witha clinical history of patient age, sex, and race. The kappastatistic (k) was used to quantify the degree of agreementamong observers. There was strong concordance amongthe panel for both group 1 (benign, k 5 0.76) and group2 (malignant, k 5 0.70) melanocytic proliferations. Therewas no concordance of the panel for group 3 (indetermi-nate) lesions (k 5 20.045). The concordance for groups 1and 2 and lack of concordance for group 3 lesions wereindependent of knowledge of clinical history of age, sex,and race. The authors concluded that a subset of melano-cytic proliferations of the conjunctiva exists that cannot bereproducibly classified by pathologists as benign, malig-nant, or indeterminate.—Thomas J. Liesegang

*Montgomery Ophthalmic Pathology Laboratory, BT 428, Emory EyeCenter, 1365 Clifton Rd, Atlanta, GA 30322.

● Acute hydrops in the corneal ectasias: associated fac-tors and outcomes. Grewal S, Laibson PR, Cohen EJ,Rapuano CJ. Tr Am Ophthalmol Soc 1999;97:187–198.

ALL PATIENTS WITH ACUTE HYDROPS SEEN BY A REFER-

ral cornea service between June 1996 and December1998 were reviewed. Twenty-one patients (22 eyes) withacute hydrops were seen. Nineteen patients had keratoco-nus, two had pellucid marginal degeneration, and one hadkeratoglobus. Twenty-one (95%) of 22 eyes had seasonalallergies, and 20 (91%) of 22 eyes had allergy-associatedeye-rubbing behavior. Six (27%) of 22 had a diagnosis ofDown syndrome. Six patients were able to identify atraumatic inciting event: vigorous eye rubbing in four andtraumatic contact lens insertion in two. The affected arearanged from 7% to 100% of the corneal surface area andwas related to disease duration and final visual acuity.Three serious complications were observed: a leak, aninfectious keratitis, and an infectious keratitis and coinci-dental neovascular glaucoma. Various medical therapiesdid not differ significantly in their effect on outcome, andultimately four (18%) of 22 patients underwent penetrat-ing keratoplasty. Best-corrected visual acuity was equal to

AMERICAN JOURNAL OF OPHTHALMOLOGY702 MAY 2000