方法：临床纳入原发性婴幼儿型青光眼患儿172例257眼。对患儿的临床资料进行分析，同时于术后对患儿进行临床随访，观察术后眼压变化情况，以术后眼压升高且使用1或2种局部降压药无效时作为手术失败的标准。采用多元Logistic回归分析导致小梁切开术失败的危险因素，采用Cox回归分析能够影响手术成功时间的相关因素以及各时点手术成功率曲线的描记。

结果：患儿中位数随访时间为30.7(8.4～57.9)mo，随访率为89.02%。多元Logistic回归分析显示，术前眼压、角膜透明度是导致小梁切开术失败的高危因素(P<0.05)； Cox回归生存分析显示，角膜透明度属于与手术成功时间相关的因素(P<0.05)。

结论：术前角膜透明度与眼压相比，更能够稳定的反映出婴幼儿型青光眼病情的严重程度。通过对患儿进行术前角膜透明度评分以及眼压水平检测，能够作为预测术后眼压控制效果。

METHODS: The clinical data of 172 infants(257 eyes)with primary glaucoma was analyzed. After the follow-up visits, the change of postoperative intraocular pressure was observed. Upon the rising intraocular pressure, the operation was considered to be failure as long as 1 or 2 kinds of drugs were taken to ineffectively reduce the pressure. Based on the multivariate Logistic regression analysis, the risk factors for the failure of trabeculotomy were analyzed. Through Cox regression analysis, the influence factor of successful operation and the curve for success rate of operation at each time point was drawn.

RESULTS: The median time of follow-up visits was 30.7(8.4～57.9)mo and follow-up rate was 89.02%. Through the multivariate Logistic regression analysis, the preoperative intraocular pressure and corneal transparency were the high risk factors for the failure of trabeculotomy(P<0.05); through Cox regression analysis, the corneal transparency was the influence factor for the successful time of trabeculotomy(P<0.05).

CONCLUSION: Compared with preoperative intraocular pressure, the corneal transparency can reflect the severity of primary infantile glaucoma in a securer way. The scores of preoperative corneal transparency and detection of intraocular pressure can predict the control effect of postoperative intraocular pressure.