黔东某高校新生高度近视风险预测模型的建立

Establishment of a prediction model for high myopia among freshmen at a university in Qiandong

目的探讨大学生高度近视的主要影响因素,建立高度近视的风险模型,为大学生视力健康分级管理提供依据。方法采用整体抽样方法,于2022年9月对本校大一新生进行视力筛查、视力健康问卷调查,通过logistic回归分析筛选出高度近视的影响因素,运用R语言rms程序进行计算,建立人群高度近视Nomogram风险预测评分模型。结果1776名调查对象中近视检出率为84.45%,其中高度近视占46.80%。logistic回归分析结果显示:每天白天户外活动时间≥1.5 h(OR=0.197,95%CI:0.138~0.281)、每天总是坚持做眼保健操(OR=0.225,95%CI:0.152~0.332)是高度近视的保护因素(P<0.01);女性(OR=3.097,95%CI:2.378~4.033)、每天课后读写时间≥3.0h(OR=2.524,95%CI:1.477~4.311)、每天玩手机时间≥3.0h(OR=7.080,95%CI:4.434~11.305)、生源地是城镇(OR=3.062,95%CI:2.349~3.990)是高度近视的危险因素(P<0.01)。风险预测模型的ROC曲线分析:AUC值为0.992(95%CI:0.988~0.996)。结论该校新生高度近视发病率高,模型预测准确度和区分度较高,可量化高度近视的发生风险。

Objective To explore the main factors influencing high myopia among college students,establish the risk model of high myopia,and provide the basis for the management of visual health in college students.Methods Using the overall sampling method,a vision screening and vision health questionnaire survey was conducted on freshmen in our college in September 2022,and the factors influencing high myopia were screened out by logistic regression analysis.The R language rms program was used to establish a Nomogram prediction score model for the risk of high myopia in this population.Results The detection rate of myopia among 1776 respondents was 84.45%,of which 46.80%had high myopia.Logistic regression analysis of the factors related to high myopia showed that duration of outdoor activity≥1.5 hours during the day(OR=0.197,95%CI:0.138-0.281)and doing eye exercises every day(OR=0.225,95%CI:0.152-0.332)were protective factors for high myopia(P<0.01).Female sex(OR=3.097,95%CI:2.378-4.033),continuous writing for≥3.0 hours after college(OR=2.524,95%CI:1.477-4.311),daily mobile phone usage time≥3.0 hours(OR=7.080,95%CI:4.434-11.305),and original residence in a town(OR=3.062,95%CI:2.349-3.990)were risk factors for high myopia(P<0.01).The ROC curve of the risk prediction model was obtained,and the AUC was 0.992(95%CI:0.988-0.996).Conclusion The incidence rate of high myopia among the freshmen in this college is high,and the model has high predictive accuracy and discriminatory reliability in quantifying the risk of high myopia.