【作者】 谭晖；

【导师】 汪玲； 钱序； 王克利；

【作者基本信息】 复旦大学 ， 流行病与卫生统计学， 2010， 博士

【摘要】 研究背景近几十年来,我国中小学生近视呈现检出率高、发病年龄提前、低视力检出率增加的流行特点。近视已成为影响儿童青少年健康的三大主要问题之一。我国1992年提出学生近视综合防治方案,并运用行政力量,使之成为学校卫生工作的重要部分。实施近二十年来,效果尚待评估。本研究采用任务分解框架,分析学生近视防制策略执行中的不足,并将干预重点定位于学生视力保健行为的改变。通过文献评阅,发现视力保健行为内涵较以往有所扩展,且缺乏有效的测量方法,致使目前学生近视行为影响因素研究和行为干预研究缺乏可比性。据此,本研究采用理论构建法和因素分析法,编制“儿童视力保健行为评定量表”,并应用该量表分析上海市小学生的行为特点；在此基础上试图运用健康行为生态学理论分解行为的多维影响因素及进行干预探索。研究目的基于学生近视防治策略执行效果的分析,将干预重点定位于学生视力保健行为改变;编制、应用“儿童视力保健行为评定量表”分析上海市小学生视力保健行为流行特点及与视力减弱的关联特征；运用健康行为生态学理论分解视力保健行为的多维影响因素,探讨上海市小学生视力保健行为干预方案和实施效果。研究方法1、上海市学生近视干预策略研究应用Delphi法,经3轮专家咨询,集中相关专业、行业30名专家的意见,探索大城市中小学生近视干预策略和对策,并进行现场实施的可行性和可接受性研究。采用知情人访谈方法,依据任务分解框架分析干预对策执行中的薄弱环节。2、“儿童视力保健行为评定量表”的编制编制该量表的主要目的在于为分析学生视力保健行为的特点及行为干预的评估提供量化的测评工具。主要采用理论构建法和因子分析法编制问卷量表,以小学生视力保健行为为基点,通过查阅有关文献和专家访谈、函询,定义行为内容,运用理论构建法编制核心行为框架及题项库,经两次预测试形成含65项题条的“初始问卷”。形成的“初始问卷”经585名受试对象的测试,及探索性因素分析,编制出含31项题条的同质量表;对840名受试量表数据进行测量学的一系列检验,形成“儿童视力保健行为评定量表”。3、上海市小学生视力保健行为现况调查应用“儿童视力保健行为评定量表”,对上海市2个中心城区、2个近郊城区1 1431名学生进行横断面调查,描述性分析小学生视力保健行为的流行特点；采用典型相关分析、多分类Logistic回归模型分析视力保健行为与视力的关联结构及关联强度；应用聚类分析方法分析学校间上述关联的聚集性,为开展基于学校的学生视力保健行为干预提供依据。4、上海市徐汇区小学生视力保健行为干预探索在现况分析基础上,运用健康行为生态学模型分解行为的多维影响因素,并设计干预方案,自2008年9月在徐汇区小学实施。采用多次重复测量的准实验设计方法,以徐汇区5所小学737名2008级学生作为干预组,长宁区5所小学575名2008级学生作为平行对照组,对3个学期的测量数据应用混合效应模型分析干预对学生行为改变的影响。主要结果1、中小学生近视干预策略经3轮Delphi专家咨询,分析专家在有效性、可行性、可接受性三个方面的评分结果,筛选出一级干预策略4方面,二级干预对策16条。一级干预策略包括建立政策支持环境、改善学习生活条件、培养正确用眼习惯、视力保护和筛查制度。分析16条对策发现与1992年制订、目前仍在实施的学生近视防治措施基本吻合,有新的研究发展的三级治疗措施在第一轮咨询中就全部被筛除。2、干预对策执行效果分析对区教育体卫条线、区CDC眼防条线负责人、学校分管校长、卫生老师、班主任进行个人访谈,访谈结果按照干预策略的一级条目和二级条目,从任务落实、监督考核、执行效果主观评价和存在困难及解决关键几个方面整理分析,认为：政策支持环境已较成熟；学习环境由政府主导正在逐步改善；视力筛查制度由市眼病防治中心主导也在逐步完善；但用眼习惯培养尚存在执行人不清、无监督无考核、效果不明的现状。3、“儿童视力保健行为评定量表”在查阅文献、专家访谈的基础上,定义小学生视力保健核心行为包括：视近活动、视近间歇休息、户外活动、视近姿势、视近环境选择、生活习惯、眼保健操等6个方面。运用理论构建法和因素分析法形成含31项题条的“量表”,并经信度检验、效度检验和区分效应的实证分析,其结构良好,各项指标达到了测量学的要求,具有较好的鉴别力和实证效应,可以作为小学生视力保健行为的评价工具。形成的“儿童视力保健行为评定量表”包含、9个维度31项题条,9个维度分别为眼放松行为、用眼距离、阅读环境选择、持续用眼、户外活动、读写坐姿、饮食习惯、眼保健操和握笔姿势。4、上海市小学生视力保健行为流行特点小学生视力保健行为表现出四种类型的年级分布特点：(1)相反类型,包括“眼放松行为”和“持续用眼”。在量表中“眼放松行为”反映的是采用各种技能松弛眼睛的行为；“持续用眼”反映的是持续用眼不采取休息的行为。其年级特点表现为“眼放松行为”随年级增加逐渐增多,而“持续用眼”与此相反,随年级增加逐渐降低。(2)U型类型或称为两头翘类型,包括“用眼距离”、“握笔姿势”和“读写坐姿”,即通常讲的“一拳一尺一寸”。表现为1、5年级得分较高,中间2-4年级较低。(3)“户外活动”类型：表现为1年级较多,余皆不足。反映了户外活动习惯没有养成。(4)“眼保健操”类型：随年级增加逐步改善,5年级最好。视近活动分解为读写活动和视屏活动。读写活动时间随年级增加而增长,其中2年级增长较快；视屏活动时间出现阶段性增长,2年级以看电视时间增加为主,4年级以使用电脑、PSP时间增加为主。5、视力保健行为与学生视力的关联性分析采用多因素方差分析,分解年级对变异的影响,结果发现：(1)视力好坏与眼放松行为、用眼距离、持续用眼和握笔姿势有关,行为得分越低视力越差。(2)视力好坏与户外活动时间、视屏娱乐时间不仅存在关联,也存在相互的影响。总体看,户外活动时间越少视力越差,但视力低于4.5(小数记录0.3)的1、2年级学生组出现双休日补偿性户外活动增加；此外,视力越差学生,视屏娱乐时间越少,也与传统的观念相悖。采用典型相关分析法对视力与视力保健行为的关联结构分析发现：(1)“用眼距离”更多地影响裸眼视力；(2)“持续用眼”更多地影响近视屈光度数；(3)视近活动强度对裸眼视力和近视屈光度数均有较大影响。多分类Logistic回归分析发现：(1)眼放松行为、用眼距离、握笔姿势、户外活动是视力降低的弱保护因素,其作用大小不受视力降低的幅度影响,即OR在轻、中、重度视力低常组中基本相同；(2)读写时间、女性、早期写字行为是视力降低的弱危险因素,其0R在三组中也基本稳定；(3)遗传和年级的影响远大于上述两个方面因素,且视力越低其作用越明显。学校间的聚类分析发现,学校间存在着两个大的关联层次。在高视力低常率的学校,同时也存在着低户外活动和高持续用眼；在低视力低常率的学校,同时也存在着高户外活动和低持续用眼。两个层次中如果合并有学习日视近活动强度增加,其视力低常率将在本层内有小幅度增加。6、视力保健行为干预探索研究运用健康行为生态学模型,按个人、班级、家庭、学校及学区几个层面,从影响行为的倾向因素、促成因素、强化因素分析制订干预措施；结合小学生视力保健行为及视力低常的流行特点,进行全程干预和阶段性强化的干预设计。强化阶段实施主体为学校,目标对象包括1、3年级。1年级的干预重点行为是端正读写姿势和休息眼睛技能,其干预措施覆盖上述的所有层次。3年级干预重点行为是增加户外活动、减少视屏视近活动、休息眼睛、读写姿势,从个人、班级、家庭内层设计。全程干预主要从生态模型的外层设计,包括区政府、区教育局的政策导向,督导机制建设等。干预自2008年9月实施,目标人群为2008级学生。采用多次重复测量的准实验设计方法观察干预效果。应用混合效应模型分析目前已监测的3次行为数据,发现干预在阅读环境选择、眼保健操、握笔姿势3个维度上有效。干预实施中督导机制和教师培训机制尚有待完善。结论1、目前的学生近视防治综合干预策略中,用眼习惯养成教育存在执行人不清、无监督无考核、效果不明的问题。2、视力保健行为具有一定结构性,编制的“儿童视力保健行为评定量表”可以作为小学生视力保健行为的测评工具。3、小学生视力保健行为具有年级特点,表现为四种类型；不同的视近活动也呈现出年级特点。可作为干预设计时的参考。4、视近活动强度和视力保健行为的某些维度与视力降低具有关联,并呈现出一定的关联结构。户外活动、持续用眼与视力降低的关联在学校间具有聚集性,并受视近活动强度的影响。5、运用健康行为生态学模型设计实施的行为干预目前仅在阅读环境选择、眼保健操、握笔姿势3个维度上有效；干预实施中的师资培训和督导机制尚待完盖更多还原

【Abstract】 BackgroundNearly several decades, the phenomenon of higher near-sight prevalence, lower age incidence and higher near-sight seriousness prevailed in primary and high school in China. Near-sight problem acts as one of three key health issues which negatively impacted child and adolescent health development in China.In order to tackle student’s near sight problem, several administration departments worked together and developed a comprehensive near-sight prevention and treatment outline in 1996. The outline was then integrated into routine school health work required by government administration departments. The practice effects have not yet been evaluated since then.This study systemically examined the practice issues, especially shortages, and then focused key prevention measures on student vision care behaviors changes. After references review, we found the contents of vision care behaviors have got certain developments in past several years. However, due to the lack of scale measures, it’s difficult for us to make the comparability between/among different studies on near-sight issues. So, we developed the "Child Vision Care Behaviors Scale" (CVCBS) based on the logical and factors analysis ways. The epidemiological characters of vision care behaviors in Shanghai primary schools were then measured by the CVCBS. Thus, we tried to understand the multiple influence factors on vision care behaviors by health behavior ecology theories, and proposed targeted intervention strategies. AimBased on effect analysis of the near-sight prevention and treatment outline, to focus intervention on student’s vision care behavior changes; To develop "Child Vision Care Behaviors Scale" (CVCBS) and study the epidemiological characters of vision care behaviors in Shanghai primary schools, and correlation between near-sight and vision care behaviors; To understand multiple influence factors on vision care behaviors from health behavior ecology theories, and propose intervention strategies, then evaluate intervention effects.MethodsThe Delphi’s method was adopted in the beginning of this study. After three rounds expert consultancy, we gathered 30 experts’ideas and proposed strategies and tactics on near-sight intervention in metropolitan primary and middle schools. The feasibility and acceptability of strategies and tactics were also studied. By personal interview, we systemically examined shortages of the Outline practice.The aim of CVCBS was to analyze the characters of student’s vision care behaviors and provide measuring tools for behavior intervention evaluation. Based on the main questionnaire development ways, i.e. logical way and factors analysis way, we reviewed references, held focus group discussion meetings and consulted experts via mail. We carefully defined behaviors contents and drafted the prototype scale after two tests. Then,585 students were measured by the prototype. After analysis and refine, another 840 students participated the Scale examination, and finally the“Child Vision Care Behaviors Scale”(CVCBS) was developed. 11431 primary school students from 2 urban districts and 2 suburb districts took the investigation adopted by CVCBS. The investigation showed main epidemiological characters of vision care behaviors in Shanghai primary schools. Analysis methods such as Canonical Correlation, Multinomial Logistics were used to understand the correlation strength and structure between vision care behavior and sight. The Hierarchical Cluster method helped to know the correlation cluster between schools, and then provided proof or signs to intervention strategies.On the basis of sectional investigation, health behavior ecology model was applied to clarify the multiple influence factors on vision care behaviors. Then, an intervention proposal was developed and put into practice in XuHui District primary schools since September, 2008. In the study, 737 students (2008 grade) from 5 primary schools in XuHui District classified as intervention group, and 575 students (2008 grade) from 5 primary schools in ChangNing District as controlled group. The study lasted 3 semester terms. The Mixed Models method was used to explain impacts from behavior intervention.ResultsBy 3 rounds Delphi’s consultancy, 4 aspects of first level intervention strategies and 16 means of second level intervention strategies were screened out after results analysis on expert views on intervention strategies validity, feasibility and acceptability. The first level intervention strategies embody setting up policy supporting ambiance, improving study condition, cultivating correct vision care habits, and formulating vision protection and near-sight screening regulation. It’s glad to understand that 16 means of second level intervention strategies basically coincide with the practicing Outline developed in 1996. Some third level intervention strategies proposed recently were all ticked out in the first round Delphi’s consultancy. The results from focus group discussion and personal interviews participated by district education administrators, optical diseases prevention and control center chiefs, school principals, school sanitary teachers, class teachers showed that: the policy supporting ambiance is relatively formulated, the school environment and study condition is improving under government investment and support, and the sight screening regulation is polished under the direction from municipal optical diseases prevention and control center. However, in the respect of cultivating vision care habits, we found that whose responsibility was not clear, and there’s no supervision and inspection, and so the results were uncertain.After reference review and Delph’s consultancy, core vision care behaviors in primary school were defined as following 6 aspects: nearwork, nearwork intermission rest, outdoors activity, nearwork gesture, nearwork environment condition, life habit, and vision care exercise. The CVCBS (include 9 aspects and 31 items) developed by logical way and factors analysis way performed very well on credibility test, validity test and other examinations, met strict metrology requirements and could act as basic evaluation tool for primary school student vision care behaviors. The detail 9 aspects refer to:vision relax behaviors, vision distance, reading and studying environment, vision duration, outdoors activities, reading and writing gesture, diet habits, vision care exercise, and pen-holding gesture.Four types of grade characters were found in the vision care behaviors study. Type 1, "opposite":reflected on "vision relax behaviors" and "vision duration". In the CVCBS, "vision relax behaviors" refers to any behaviors to try to relax vision or sight; and "vision duration" refers to no rest behavior during vision time. The results showed more "vision relax behaviors" and less "vision duration" with increase of students’grade level. Type 2, "U type":reflected on "vision distance", "pen-holding gesture" and "reading and writing gesture" aspects. The results showed that grade 1 and grade 5 got better marks and others lower marks. Type 3, "outdoors activities":Grade 1 had more outdoors activities than any other grades. Type 4, "vision care exercise":good high marks with high grades.We defined nearwork as two parts, one was reading and writing activities, and another was screen watching activities (TV watch, PSP game play, and computer screen watch). With high grade, we found more times on reading and writing activities, especially in grade 2. As to screen watching activities, more TV watch time in grade 2, and more computer and PSP play time in grade 4.Multi-factors analysis suggested that, (1) vision status correlated with vision relax behaviors, vision distance, vision duration and pen-holding gesture, lower marks worse vision status; (2) vision status correlated with outdoors activity time and screen watching time, and both interact. Generally, less outdoor activities time worse vision status. However, we found from this study that grade 1 and 2 students with vision status less than 4.5 had more compensated outdoor activities during weekend. And more, less screen watching time among worse vision status students. That’s contradictory with common sense.Canonical Correlation analysis suggested that, (1) vision distance more impact vision status; (2) vision duration more impact refractive status; and (3) nerwork strength more impact both of vision status and refractive status.Multinomial Logistics analysis suggested that, (1) vision relax behaviors, vision distance, pen-holding gesture and outdoors activities were weak protection factors for vision status, and its impacts not disturbed by the status of vision decrease (in other words, there’s no difference among light, medium and severe vision decrease groups); (2) reading and writing time, female and early writing behavior were weak risk factors for vision status; (3) heredity and grade factors have more impacts than factors mentioned above, especially in severe vision decrease group.Hierarchical Cluster analysis suggested that schools with high incidence of severe vision decrease, had less outdoors activities and more vision duration; vice versa.Based on health behavior ecology model, we developed intervention strategies for behaviors predisposing, enabling, and reinforcing factors standing on personal, class, family, school and community levels. Combined with vision care behaviors and high incidence, we formulated whole process intervention and stage intensified means. The whole process intervention focused on outer ambiance from the point of health behavior ecology model, including related district government administration policy support, supervision and inspection system development. The stage intensified means focused on school, and the target group was grade 1 and grade 3. In grade 1, behavior intervention included cultivating correct reading and writing gesture and vision relax skills, focus on all levels. In grade 3, behavior intervention included more outdoors activities, less screen watching and nearwork, sight rest, good reading and writing gesture, focus on personal, class and family level.The intervention was implemented since September,2008. The study targeted student group who attended school in 2008 (grade 2008). Up to now, we had 3 times behavior investigation, and Mixed Model analysis found that, intervention had good impact on reading environment condition, vision care exercise and pen-holding gesture. Conclude(1) In current vision care intervention, the responsibility of cultivating good vision care habits was not clear, and there’s no supervision and inspection, and so the intervention results were uncertain.(2) The "Child Vision Care Behaviors Scale" (CVCBS) was formulated, and could act as a tool for vision care behavior measurement in primary school.(3) Four type of grade characters were existed in vision care behaviors. Targeted intervention strategies should seriously consider those characters.(4) There’re correlation between vision decrease and nearwork strength and some certain vision care behaviors. And there’re school cluster on the correlation between vision decrease and outdoors activities, vision duration.(5) Behavior intervention developed from health behavior ecology suggested certain good results on reading environment condition, vision care exercise and pen-holding gesture.更多还原