Volume 16 Number 1 / February 2012 e9

excluded because they were sent for a medical or ophthalmic con-dition requiring a dilated fundus examination (Diabetes, JRA, Neuro-fibromatosis, Glaucoma, Trauma).Results: Of 190 patients analyzed, cycloplegic examination providedadditional information in only 4 patients, 2 preverbal children, 1 withnon-visually significant congenital cataract\1mm, one with refrac-tive amblyopia from moderate astigmatism, 1 child with moderate hy-peropia who was put into glasses because parents noted esotropiaand was later found to have AET and 1 child with increased cuppingbut normal IOP. Although now being examined yearly, this patient isfelt unlikely to have glaucoma.Discussion: A normal plusoptiX, combined with normal alignment/motility evaluation and visual acuity was found to have a 98% nega-tive predictive value for ophthalmic pathology including significantrefractive error.Conclusions: Dilating drops are traumatic for many children. The plu-soptiX and an abbreviated eye examination may obviate the need fordilation in select children.

031 Comparison of inferior oblique myectomy to recession for thetreatment of superior oblique palsy. Elias I. Traboulsi, ReechaSachdeva, Paul Rychwalski, Andreas MarcottyPurpose: To compare inferior oblique (IO) myectomy to recession forthe treatment of superior oblique (SO) palsy.Methods: A retrospective chart review identified all patients with SOpalsy who underwent IO weakening procedures. Patients were ex-cluded if IO muscle surgery was combined with other vertical musclesurgery and if follow up was less than 4 weeks. Good outcome pa-rameters were: improvement of deviation in primary gaze, resolutionof diplopia in primary and reading gazes, and correction of head tilt tothe contralateral side. We analyzed subsets of subjects with largepreoperative vertical deviations (.20D) and those with small to mod-erate deviations (\20D). Comparisons were made using Wilcoxonrank sum and the Fisher exact tests.Results: Eighty-seven eyes of 86 patients met the study criteria.Forty-five eyes underwent myectomy, and 42 underwent recessionof the IO. Those patients who underwent myectomy compared to re-cession had less postoperative hypertropia (HT) in primary gaze (P50.002) and were more likely to meet criteria for success as measuredby #4D residual HT (P 5 0.056). This difference was pronounced (P5 0.005) in patients with preoperative small to moderate deviations,but statistical significance was lost in patients with large preopera-tive deviations. There was no difference between the two groupswith respect to resolution of diplopia or anomalous head tilt.Discussion: IO weakening is effective in the treatment of SO palsy.Our findings support myectomy as more effective than recession inimproving HT in primary gaze independent of preoperative measure-ments; this difference is more pronounced in individuals with small tomoderate preoperative hyperdeviations.Conclusions: Simple inferior oblique myectomy is an excellent surgi-cal modality for the management of patients with SO palsy andmild tomoderate vertical deviations.

032 Development and validation of a standardised tool for thereporting of retinal findings in abusive head trauma. PatrickWatts, Wai S. Ng, Zoe Lawson, Alison Kemp, Sabine MaguirePurpose: To develop and validate a robust standardized reporting toolfor describing retinal findings in children with suspected abusivehead trauma (AHT).Methods: Seven pediatric ophthalmologists and seven residents as-sessed a series of 105 high quality Retcam images of 21 eyes from

Journal of AAPOS

AHT cases with varying degrees of retinal hemorrhage and associ-ated findings. The pediatric ophthalmologists performed repeatedassessments of the randomized images. The images were observedsimultaneously with standardized display settings. Retinal findingswere recorded on a predefined assessment proforma, derived fromthe results of a recent systematic review of retinal findings in AHT.Inter and intra-observer variability were assessed by Free-MarginalMulti-rater Kappa (FMK), Intra-class Correlation Coefficient (ICC) andConcordance Coefficient (CC).Results: Almost perfect inter-observer agreement was observed forresidents and pediatric ophthalmologists in assessing the presenceand number of fundus haemorrhages (ICC 0.912 and 0.866). Bothgroups achieved very high inter-observer agreement for location ofhemorrhages (CC 0.860 [0.769-0.950], 0.849 [0.743-0.955]). Substantialagreement was observed by both in assessing size of hemorrhage(CC 0.732 [0.640-0.824], 0.762 [0.669-0.854]), moderate agreement forhemorrhage morphology (CC 0.532 [0.485-0.579], 0.521 [0.451-0.592])and other findings (CC 0.480 [0.419-0.541], 0.594 [0.497-0.691]). Intra-observer agreement for pediatric ophthalmologists was almostperfect (Free-marginal kappa 0.90-1.00) for presence of fundus hae-morrhage.Discussion: We have developed and validated a standardized, clini-cal reporting tool for ophthalmic findings in suspected AHT, whichhas excellent inter and intra-observer agreement among consultantspecialists and traineesConclusions:We suggest that the use of this tool will improve clinicalreporting in AHT.

033 Objective detection of amblyopia and strabismus . . . not justscreening for risk factors. Susan E. Yanni, Reed M. Jost, Cynthia L.Beauchamp, David R. Stager, Sr., Eileen E. BirchBackground: To date, preschool vision screening for strabismus and/or amblyopia involves assessment of risk factors, including refractiveerror and stereoacuity. Most children who are positive for risk fac-tors never develop strabismus or amblyopia. The Pediatric VisionScanner (PVS) is the first objective, automated screening devicethat directly assesses bifoveal fixation. This study was undertakento assess the ability of the PVS to identify patients with strabismusand/or amblyopia in a pediatric ophthalmology office setting.Methods: A total of 114 children (2-6 years) were tested with the PVS,which produced a pass, refer right eye, refer left eye, or refer bilat-eral recommendation. Children were also tested with the SureSightVision Screener and the Randot Preschool Stereoacuity Test(RPST). A comprehensive pediatric ophthalmic examination servedas the gold-standard.Results: 59 of 114 children had strabismus and/or amblyopia. Sensi-tivity of the PVS was 98% (95% CI, 90-100%), significantly higher thanthe SureSight (sensitivity5 58%) and the RPST (sensitivity5 73%). Inthis office setting, with high prevalence of strabismus and amblyopia,overall specificity was good and comparable among the 3 tests (74%-88%). However, the SureSight had a significantly larger number offalse positives due to refractive error than the PVS (13% vs 6%, P\ 0.04).Discussion: The PVS identified children with amblyopia or strabis-mus with high sensitivity, outperforming the SureSight and the RPST.Conclusions: Preschool vision screening may be more efficient witha device that directly identifies strabismus and amblyopia. The PVSprovides rapid, accurate identification of children with amblyopiaand/or strabismus early in life, when they are most amenable totreatment.