Flanagan P, Coll C G, Andreozzi L, Riggs S
Department of Pediatrics, Women and Infants' Hospital, Providence, USA.
Arch Pediatr Adolesc Med. 1995 Apr;149(4):451-5. doi: 10.1001/archpedi.1995.02170160105016.
To determine the degree to which knowing certain characteristics about young high-risk families can help distinguish those families most likely to maltreat their children from those families at lower risk of maltreating their children.
Observational cohort from which the following predictor variables were gathered when infants were 2 months old: maternal age, depressive symptoms, childrearing attitudes, social support, and living situation (with or apart from related adults). Families were followed up for 24 months to identify the occurrence of maltreatment.
An urban, socioeconomically disadvantaged cohort of teenage mothers and their infants attending a hospital-based special primary care clinic for teen mothers and their infants.
All full-term infants and mothers enrolled into the clinic in 1990 participated in the study. This included 47 mother-infant pairs enrolled when infants were 2 months of age. Forty-five of these pairs were available for follow-up when infants were 24 months of age.
Maltreatment defined as any incident that prompted investigation by the state child protective agency and was found to be a substantiated case of maltreatment by that agency.
Maltreatment occurred in 15 of 45 families before the child's second birthday. Discriminate function analysis produced a model that correctly classified 13 of 15 maltreating mothers and misclassified one of 30 non-maltreating mothers. Stepwise analysis revealed that living situation was by far the strongest predictive variable (R2 = 7).
Maltreatment was a predictable outcome within this extremely high-risk cohort. Living apart from related adults was the strongest risk factor associated with maltreatment. This easily obtainable piece of information may be an important risk marker for practitioners, social service personnel, and others working with this very-high-risk population. It may allow early supportive interventions that might prevent maltreatment.
确定了解年轻高危家庭的某些特征在多大程度上有助于区分极有可能虐待孩子的家庭与虐待孩子风险较低的家庭。
观察性队列研究。当婴儿2个月大时收集以下预测变量：母亲年龄、抑郁症状、育儿态度、社会支持和生活状况（与相关成年人同住或分开居住）。对这些家庭进行了24个月的随访，以确定虐待行为的发生情况。
一个城市中社会经济条件不利的青少年母亲及其婴儿队列，这些母亲和婴儿在一家以医院为基础的青少年母亲及其婴儿特殊初级保健诊所就诊。
1990年登记到该诊所的所有足月婴儿和母亲均参与了这项研究。其中包括47对母婴，婴儿2个月大时登记入组。当婴儿24个月大时，其中45对母婴可供随访。
虐待定义为任何促使州儿童保护机构进行调查且该机构认定为确凿虐待案件的事件。
在孩子两岁生日前，45个家庭中有15个发生了虐待行为。判别函数分析得出一个模型，该模型正确分类了15名虐待母亲中的13名，错误分类了30名非虐待母亲中的1名。逐步分析显示，生活状况是迄今为止最强的预测变量（R2 = 7）。
在这个极高风险队列中，虐待是一个可预测的结果。与相关成年人分开居住是与虐待相关的最强风险因素。这条易于获取的信息可能是从业者、社会服务人员以及其他与这一极高风险人群打交道的人的重要风险指标。它可能有助于开展早期支持性干预措施，从而预防虐待行为。
