Establishment And Primary Application Of The Paediatric Open-Globe Ocular Trauma Score

Purpose:Open globe injury is full-thickness eyeball laceration caused by sharp or blunt objects, which includes penetrating injury, perforating injury, intraocular foreign body injury, globe rupture, combined injury. Sharp objects can cause penetrating injury with a full-thickness entrance wound and can cause perforating with a full-thickness entrance and exit wound. Blunt objects can cause globe rupture with a full-thickness eyeball wound. Intraocular foreign body injury implies retained foreign object in the eye. Open globe wounds may destruct eye structures seriously and exhibit a poorer visual outcome, of which 43% are in children younger than 18 years. Paediatric open-globe ocular trauma is more complex than adult because of immature visual function, fragile ocular structure and intense inflammatory response.Ocular trauma as a leading cause of monocular visual disability and noncongenital blindness, not only impairs the visual function, but also effects all-round healthy development, resulting in heavy economic burden. The lack of safety consciousness and unclear clinical history delay the timely treatment. Proper surgical treatment can improve visual prognosis and so predicting visual outcome be of significant clinical value.The ocular trauma score (OTS) has been applied in various ocular trauma and become an important indicator of prognosis evaluation. Because of dependent on initial visual acuity and RAPD of OTS, Paediatric ocular trauma score (POTS) is proposed to predict visual outcome in paediatric penetrating eye injuries. They have a certain limitation and the trauma score can’t still be calculated in some open-globe eye injuries without initial visual acuity and RAPD.In this study, the analysis of related prognostic factors is performed, the predictive ability of OTS and POTS is evaluated, and the difference between them is compared. On this basis, a prognostic model named pediatric open-globe trauma score(POGTS) is proposed for more paediatric ocular trauma. More clinical information of paediatric open-globe ocular trauma is collected prospectively to investigate the POGTS prediction efficiency and compared the differences among the three ocular trauma models.Methods:1. We conducted a retrospective review of all children (age?15years) with open globe injury that was hospitalized in the Department of Ophthalmology, the First Affiliated Hospital and The Children’s Hospital of Zhejiang University, and was followed up for more than 6 months and until stable visual acuity. Univariate chi-square analysis was used to ascertain which clinical variables were related to poor visual outcome and good visual outcome.2. OTS can be calculated in open-globe injuries patients with clinical data on necessary variables. POTS can be calculated in paediatric penetrating eye injuries. Receiver operating characteristic (ROC) curves were constructed to evaluate the predictive ability of OTS and POTS for visual prognosis. And hence, the prognostic effectiveness of OTS and POTS for final visual outcome can be compared in the paediatric penetrating eye injuries with complete clinical data.3. According to OTS and POTS, significant predictors relating to visual outcome are assigned values and then pediatric open-globe trauma score is established. In patients who are uncooperative to initial vision test, the equation was used to obtain the trauma score as followed:2×(age+zone)+corresponding pathologies.4. After establishment of POGTS, a prospective study on open globe injury patients is conducted and OTS, POTS and POGTS are calculated separately, if possible. To investigate the relationship between the POGTS and likelihood of visual outcomes, we compare the score differences between the good visual outcome and poor visual outcome groups. The likelihood of the visual outcome being NLP, LP/HM, 1/200-19/200,20/200-20/50, and≥20/40 is showed in each OTS category. Receiver operating characteristic (ROC) curves are performed to display discriminatory ability of POGTS for good visual outcome and poor visual outcome. AUROC are performed to compare the predictive ability among POGTS, POTS and OTS.Results:1. The retrospective study comprised 113 children with open globe injury. The mean injury age was 7.2±4.2 years. Of these children,73 patients were male, and 40 patients were female, which yield a male-to-female ratio of 1.8:1. Of these eyes,98 (86.7%) had penetrating eye injuries and 15 (13.3%) had globe ruptures, which included 5 IOFB injuries. Most injury-causing instruments are sharp objects such as scissor, nail,needle and so on. Most injuries (85.8%) located in zone Ⅰ, followed by zone Ⅱ (10.6%) and zone Ⅲ (3.5%). The treatments can improve the visual acuity and make the number of patients with visual acuity< 20/200 decreased. Univariate analysis showed that the parameters correlative with poor visual outcome were poor initial vision (P=0.010), posterior entrance site (P< 0.0001), trauma type (P=0.001), retinal detachment (P< 0.0001), endophthalmitis (P=0.016), RAPD(P<0.0001), hyphema (P< 0.0001), vitreous haemorrhage (P< 0.0001), wound length(P< 0.0001). There are some relevant variables of good visual outcome, which included initial vision (P< 0.0001), entrance site (P< 0.0001), iris prolapse (P=0.003), hyphema (P=0.007) and wound length (P< 0.022).2. OTS was calculated in 91 patients (80.53%) whose clinical data were completely available. For poor visual outcome, AUROC of OTS is 0.824 (SE 0.009, P=0.000<0.05,95%CI 0.806-0.843). For good visual outcome, AUROC of OTS is 0.740 (SE 0.007, P=0.000<0.05,95%CI 0.726-0.754).98 children who suffered from penetrating eye injuries have POTS:AUROC is 0.864 (SE 0.007, P=0.000<0.05,95%CI 0.850-0.878) for poor visual outcome and is 0.642 (SE 0.008, P=0.000<0.05,95%CI 0.626-0.659) for good visual outcome. OTS and POTS are calculated in 79 penetrating eye injuries with complete clinical data (69.91%). OTS and POTS have exact prediction of poor visual outcome (AUROC 0.887 vs.0.875), and meanwhile have relatively poor prediction of good visual outcome (AUROC 0.692 vs.0.646). And meanwhile, their predictive power was similar for poor visual outcome (Z=1.129, P=0.259) and the predictive value of POTS is poorer than OTS for good visual outcome (Z=4.07, P< 0.05).3. The parameters of POGTS included:initial visual acuity, age, trauma type and entrance site, iris prolapse, hyphema, vitreous haemorrhage, wound length, retinal detachment, endophthalmitis. On the basis of OTS and POTS, the significant predictors were assigned and POGTS was established.4. There are significant differences in POGTS between poor visual outcome patients and others(32.73±24.12 vs.68.43±17.15, P< 0.05). The POGTS is higher in good vision patients than others (73.75±12.31vs.50.74±26.15, P< 0.05). With the higher the scores, the more likelihood of the good visual outcome. There are no significant difference of POTS between poor visual outcome patients and others (43.75±16.52 vs.60±22.58, P> 0.05). And so, POTS isn’t calculated for these patients. For these open-globe injury patients, AUROC of POGTS is 0.878 (SE 0.007,.P=0.000< 0.05, 95%CI 0.864-0.891) for poor visual outcome and is 0.674 (SE 0.011, P=0.000< 0.05,95%CI 0.653-0.694) for good visual outcome.For the open-globe injury patients with complete clinical information, AUROC of POGTS is 0.878 (SE 0.007, P=0.000< 0.05,95%CI 0.864-0.891) for poor visual outcome. In open-globe injury patients with complete clinical data, OTS and POGTS were compared. Their AUROC for poor visual outcome were 0.869 (SE 0.010, P=0.000< 0.05,95%CI 0.849-0.890) and 0.905 (SE 0.006, P=0.000< 0.05, 95%CI 0.892-0.918), and for good visual outcome were 0.708 (SE 0.011, P=0.000 < 0.05,95%CI 0.686-0.730) and 0.703 (SE 0.011, P=0.000< 0.05,95%CI 0.681-0.725). The similar predictive ability of them was discovered for good visual outcome (Z=0.32, P=0.75), The predictive ability of POGTS is better than OTS for poor visual outcome (Z=3.09, P< 0.05).Conclusions:1. In penetrating eye injuries with complete clinical data, the predictive power of OTS and POTS is similar for poor visual outcome. On the whole, they had strong prediction power of poor visual outcome and relative poor prediction power of good visual outcome. Although OTS and POTS have some limits, they still were effective prognostic tools for some paediatric ocular trauma.2. In this study, POGTS was first proposed and applied in the prognostic evaluation of paediatric open-globe ocular trauma. The POGTS diagnosis correlated with final visual outcome. With The higher the scores, the rate of good visual outcome was significantly increased. Compared to OTS, POGTS is better in prognosis of poor visual outcome and is similar in predictive ability. If the POGTS is greater than 67.5, the final visual acuity must be better than 20/200. If the POGTS is less than 80, the final visual acuity better than 20/40 may be very unlikely. On the whole, POGTS could help us in evaluating the visual outcome of paediatric open-globe ocular trauma.