Pandya Nirav K, Baldwin Keith D, Wolfgruber Hayley, Drummond Denis S, Hosalkar Harish S
Department of Orthopaedic Surgery, Children's Hospital of Philadelphia, Philadelphia, USA.
J Pediatr Orthop B. 2010 Nov;19(6):535-41. doi: 10.1097/BPB.0b013e32833ce424.
Child abuse is a serious problem affecting the pediatric population, which has tremendous medical and social implications. There exists no gold standard test to diagnose child abuse, and as a result, clinicians are often in a difficult position of both protecting the child adequately, and respecting the rights and privacy of families. Upper extremity fractures are also common injuries seen in the emergency room, and a need exists to differentiate accidental from nonaccidental etiologies in young children. The purpose of this study was to produce an algorithm-based on statistical analysis that would allow clinicians to differentiate between humerus fractures stemming from abuse versus accidental trauma. We hypothesized that accidental humerus fractures in pediatric patients under the age of 4 years can be accurately distinguished from child abuse using a combination of history, physical exam findings, radiographic findings, and age. We searched our institutions Suspected Child Abuse and Neglect and trauma databases for nearly a decade. We identified 36 (representing 39 humerus fractures) patients in whom the etiology of their humerus fracture was abuse, and compared that group with 95 patients (representing 95 humerus fractures) in whom accidental trauma was the etiology. Univariate and multivariate statistical analysis techniques were applied to determine factors important in the diagnosis of child abuse given a humerus fracture. Univariate analysis found that location of fracture, polytrauma, age, prior injury, and history were factors that are important in the diagnosis of child abuse. Our multivariate analysis found that age above 18 months, physical and/or radiographic evidence of prior injury, and suspicious history were found in greater frequency in the group of patients experiencing abusive humerus fractures. In conclusion, based on our statistical analysis and earlier studies we developed an algorithm that clinicians can use to guide judgment and refer to social services when encountered with a young child presenting to the emergency room with a humerus fracture.
虐待儿童是一个严重影响儿童群体的问题,具有重大的医学和社会影响。目前不存在诊断虐待儿童的金标准测试,因此,临床医生常常处于既要充分保护儿童,又要尊重家庭权利和隐私的困境中。上肢骨折也是急诊室常见的损伤,需要区分幼儿意外和非意外病因。本研究的目的是基于统计分析生成一种算法,使临床医生能够区分因虐待与意外创伤导致的肱骨骨折。我们假设,结合病史、体格检查结果、影像学检查结果和年龄,可以准确区分4岁以下儿科患者的意外肱骨骨折与虐待性骨折。我们在近十年间搜索了本机构的疑似虐待和忽视儿童数据库以及创伤数据库。我们确定了36名(代表39例肱骨骨折)肱骨骨折病因是虐待的患者,并将该组与95名(代表95例肱骨骨折)病因是意外创伤的患者进行比较。应用单变量和多变量统计分析技术来确定在肱骨骨折情况下对诊断虐待儿童重要的因素。单变量分析发现,骨折部位、多发伤、年龄、既往损伤和病史是诊断虐待儿童的重要因素。我们的多变量分析发现,在遭受虐待性肱骨骨折的患者组中,18个月以上的年龄、既往损伤的体格和/或影像学证据以及可疑病史的出现频率更高。总之,基于我们的统计分析和早期研究,我们开发了一种算法,临床医生在遇到肱骨骨折的幼儿到急诊室就诊时可使用该算法来指导判断并转介至社会服务部门。