CLINICAL REPORT

Evaluating Children With Fractures for Child PhysicalAbuse

abstractFractures are common injuries caused by child abuse. Although theconsequences of failing to diagnose an abusive injury in a childcan be grave, incorrectly diagnosing child abuse in a child whose frac-tures have another etiology can be distressing for a family. The aim ofthis report is to review recent advances in the understanding of frac-ture specificity, the mechanism of fractures, and other medical dis-eases that predispose to fractures in infants and children. Thisclinical report will aid physicians in developing an evidence-based dif-ferential diagnosis and performing the appropriate evaluation whenassessing a child with fractures. Pediatrics 2014;133:e477–e489

INTRODUCTION

Fractures are the second most common injury caused by child physicalabuse; bruises are the most common injury.1 Failure to identify aninjury caused by child abuse and to intervene appropriately mayplace a child at risk for further abuse, with potentially permanentconsequences for the child.2–4 Physical abuse may not be consideredin the physician’s differential diagnosis of childhood injury becausethe caregiver may have intentionally altered the history to conceal theabuse.5 As a result, when fractures are initially evaluated, a diagnosisof child abuse may be missed.3 In children younger than 3 years, asmany as 20% of fractures caused by abuse may be misdiagnosedinitially as noninflicted or as attributable to other causes.3 In addition,fractures may be missed because radiography is performed beforechanges are obvious or the radiographic images are misread ormisinterpreted.2 However, incorrectly diagnosing physical abuse in a childwith noninflicted fractures has serious consequences for the child andfamily. To identify child abuse as the cause of fractures, the physicianmust take into consideration the history, the age of the child, the locationand type of fracture, the mechanism that causes the particular type offracture, and the presence of other injuries while also considering otherpossible causes.

DIFFERENTIAL DIAGNOSIS OF FRACTURES

Trauma: Child Abuse Versus Noninflicted Injuries

Fractures are a common childhood injury and account for between 8%and 12% of all pediatric injuries.6–8 In infants and toddlers, physical

Emalee G. Flaherty, MD, Jeannette M. Perez-Rossello, MD,Michael A. Levine, MD, William L. Hennrikus, MD, and theAMERICAN ACADEMY OF PEDIATRICS COMMITTEE ON CHILDABUSE AND NEGLECT, SECTION ON RADIOLOGY, SECTION ONENDOCRINOLOGY, and SECTION ON ORTHOPAEDICS, and theSOCIETY FOR PEDIATRIC RADIOLOGY

KEY WORDfractures

ABBREVIATIONSCML—classic metaphyseal lesionsCPR—cardiopulmonary resuscitationCT—computed tomographyOI—osteogenesis imperfect

This document is copyrighted and is property of the AmericanAcademy of Pediatrics and its Board of Directors. All authorshave filed conflict of interest statements with the AmericanAcademy of Pediatrics. Any conflicts have been resolved througha process approved by the Board of Directors. The AmericanAcademy of Pediatrics has neither solicited nor accepted anycommercial involvement in the development of the content ofthis publication.

The guidance in this report does not indicate an exclusivecourse of treatment or serve as a standard of medical care.Variations, taking into account individual circumstances, may beappropriate.

www.pediatrics.org/cgi/doi/10.1542/peds.2013-3793

doi:10.1542/peds.2013-3793

All clinical reports from the American Academy of Pediatricsautomatically expire 5 years after publication unless reaffirmed,revised, or retired at or before that time.

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2014 by the American Academy of Pediatrics

PEDIATRICS Volume 133, Number 2, February 2014 e477

FROM THE AMERICAN ACADEMY OF PEDIATRICS

Guidance for the Clinician inRendering Pediatric Care

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

abuse is the cause of 12% to 20% offractures.9 Although unintentionalfractures are much more commonthan fractures caused by child abuse,the physician needs to remain awareof the possibility of inflicted injury.Although some fracture types arehighly suggestive of physical abuse,no pattern can exclude child abuse.10,11

Specifically, it is important to recog-nize that any fracture, even fracturesthat are commonly noninflicted inju-ries, can be caused by child abuse.Certain details that can help thephysician determine whether a frac-ture was caused by abuse ratherthan unintentional injury include thehistory, the child’s age and de-velopmental stage, the type and lo-cation of the fracture, the age of thefracture, and an understanding of themechanism that causes the particu-lar type of fracture. The presence ofmultiple fractures, fractures of dif-ferent ages or stages of healing, de-lay in obtaining medical treatment,and the presence of other injuriessuspicious for abuse (eg, coexistinginjuries to the skin, internal organs,or central nervous system) shouldalert the physician to possible childabuse.

Child’s Age and Development

The physician should consider thechild’s age and level of development.Approximately 80% of all fracturescaused by child abuse occur in chil-dren younger than 18 months,12 andapproximately one-quarter of fracturesin children younger than 1 year arecaused by child abuse.1,9,13–15 Physicalabuse is more likely to be the cause offemoral fractures and humeral frac-tures in children who are not yetwalking compared with children whoare ambulatory,15–18 and the percent-age of fractures caused by abusedeclines sharply after the child beginsto walk.9,19,20

Fracture Specificity for Abuse

Fractures With High Specificity forAbuse

As shown in Table 1, certain fractureshave high specificity for or strong as-sociation with child abuse, particularlyin infants, whereas others may haveless specificity.21 Rib fractures ininfants, especially those situated post-eromedially, and the classic meta-physeal lesions of long bones, havehigh specificity for child abuse. Frac-tures of the scapula, spinous process,and sternum also have high specificityfor abuse but are uncommon.

Rib fractures are highly suggestive ofchild abuse. Most abusive rib fracturesresult from anterior-posterior com-pression of the chest. For this reason,rib fractures are frequently found ininfants who are held around the chest,squeezed, and shaken. Rib fractureshave high probability of being causedby abuse.15,17,21 The positive predictivevalue of rib fractures for child abuse inchildren younger than 3 years was95% in one retrospective study.22 Otherless common causes of rib fractures ininfants include significant traumasustained during childbirth or a motorvehicle crash as well as minor trauma

in infants who have increased bonefragility.23–25

Cardiopulmonary resuscitation (CPR)has been proposed as a cause of ribfractures, but conventional CPR with 2fingers of 1 hand rarely causes frac-tures in children.26,27 Recent recom-mendations that CPR be performedusing 2 hands encircling the rib cagehave raised concerns that this tech-nique might cause rib fractures. Ananalysis of infants who were discov-ered during autopsy to have rib frac-tures and had received 2-handedchest compressions antemortemsuggested that 2-handed CPR is as-sociated with anterior-lateral ribfractures of the third to sixth ribs.28

In this small study, no posterior ribfractures were observed. The frac-tures in these infants were alwaysmultiple, uniformly involved thefourth rib, and were sometimes bi-lateral. Additional research is neededto examine the relationship betweenthe 2-handed CPR technique and ribfractures.

Classic metaphyseal lesions (CMLs)also have high specificity for childabuse when they occur during thefirst year of life.21,29 CMLs are themost common long bone fracturefound in infants who die with evi-dence of inflicted injury.30 CMLs areplanar fractures through the pri-mary spongiosa of the metaphysis.These fractures are caused whentorsional and tractional shearingstrains are applied across the met-aphysis, as may occur with vigorouspulling or twisting of an infant’s ex-tremity.31 Fractures resemblingCMLs radiographically have beenreported after breech delivery32 andas a result of treatment of club-foot.33

Depending on the projection of theradiograph, CMLs can have the ap-pearance of a corner or a bucket-handle fracture. Acute injuries can

TABLE 1 Specificity of radiologic findings ininfants and toddlers19

High specificitya

CMLsRib fractures, especially posteromedialScapular fracturesSpinous process fracturesSternal fractures

Moderate specificityMultiple fractures, especially bilateralFractures of different agesEpiphyseal separationsVertebral body fractures and subluxationsDigital fracturesComplex skull fractures

Common, but low specificitySubperiosteal new bone formationClavicular fracturesLong-bone shaft fracturesLinear skull fractures

a Highest specificity applies in infants.

e478 FROM THE AMERICAN ACADEMY OF PEDIATRICS by guest on June 15, 2018www.aappublications.org/newsDownloaded from

be difficult to visualize radiographi-cally. CMLs commonly heal withoutsubperiosteal new bone formation ormarginal sclerosis. They can healquickly and be undetectable on plainradiographs in 4 to 8 weeks.31

Fractures With Moderate Specificityfor Abuse

Although many children who havebeen abused will have only a singlefracture,34 the presence of multiplefractures, fractures of different agesand/or stages of healing, and com-plex skull fractures have moderatespecificity for physical abuse. In ad-dition, epiphyseal separations, verte-bral body fractures, and digitalfractures have moderate specificityfor abuse. The presence of multiplefractures or fractures of differentages can be signs of bone fragilitybut should also evoke considerationof child abuse. Besides the predictivevalue of the particular pattern offractures, many other factors, suchas the history and the child’s age,must be considered when de-termining whether the injury wasinflicted.

Common Fractures With LowSpecificity for Child Abuse

Long bone fractures (other than CMLs),linear skull fractures, clavicle fractures,and isolated findings of subperiostealnew bone formation have low speci-ficity for child abuse. In contrast, thesingle long bone diaphyseal fracture isthe most common fracture patternidentified in abused children.1,13,34

An understanding of the extent andtype of load that is necessary to causea particular long bone fracture canhelp to determine whether a specificfracture is consistent with the injurydescribed by the caregiver.35,36 Trans-verse fractures of the long bones arecaused by the application of a bendingload in a direction that is perpendicular

to the bone, whereas spiral fracturesare caused by torsion or twisting ofa long bone along its long axis. Obli-que fractures are caused by a combi-nation of bending and torsion loads.37

Torus or buckle fractures are the re-sult of compression from axial loadingalong the length of the bone. Althoughearlier studies suggested that spiralfractures should always raise suspi-cion for child abuse,12 more recentstudies do not show that any partic-ular fracture pattern can distinguishbetween abuse and nonabuse withabsolute certainty.16,38

Falls are common in childhood.39 Shortfalls can cause fractures, but theyrarely result in additional significantinjury (eg, neurologic injury).11,40–42 Ina retrospective study of short falls,parents reported that 40% of thechildren before 2 years of age hadsuffered at least 1 fall from a height ofbetween 6 inches and 4 feet. Approx-imately one-quarter of these childrensuffered an injury; bruises were themost common injury observed.43

The femur, humerus, and tibia are themost common long bones to be in-jured by child abuse.1,34 Femoralfractures in the nonambulatory childare more likely caused by childabuse, whereas these fractures inambulatory children are most com-monly noninflicted.10,16,43–45

Certain femur fractures may occur asa result of a noninflicted injury in youngchildren. Several studies have demon-strated that a short fall to the knee mayproduce a torus or impacted transversefracture of the distal femoral meta-diaphysis.46,47 Oblique distal femur met-aphyseal fractures have been reportedin children playing in a stationary ac-tivity center, such as an Exersaucer(Evenflo, Picqua, OH).48

In both ambulatory and nonambulatorychildren, under some circumstances,falls on a stairway can cause a spiralfemoral fracture. For example, a fall

down several steps and landing with 1leg folded or twisted underneatha child can lead to excessive torsionalloading of the femur and a spiralfracture.46 In ambulatory children,noninflicted femoral fractures havebeen described in children who fellwhile running or who fell and landedin a split-leg position.43

A fracture of the humeral shaft in a childyounger than 18 months has a highlikelihood of having been caused byabuse.15,49,50 In contrast, supracondylarfractures in ambulatory children areusually noninflicted injuries resultingfrom short falls.15

Physicians should also be aware ofa particular mechanism reported toproduce a noninflicted spiral-obliquefracture of the humerus in 1 casereport.51 When the young infant wasrolled from the prone position to thesupine while the child’s arm is ex-tended, the torsion and stress placedon the extended arm appeared tocause a spiral-oblique fracture of themidshaft of the humerus.

Linear skull fractures of the parietalbone are the most common skullfracture among young children, usu-ally children younger than 1 year.13 Ashort fall from several feet ontoa hard surface can cause a linear,nondiastatic skull fracture.19,52 Themajority of linear skull fractures arenot inflicted.53 By contrast, complex orbilateral skull fractures are typical ofnonaccidental trauma.

Syndromes, Metabolic Disorders,Systemic Disease

Preexisting medical conditions andbone disease may make a child’sbones more vulnerable to fracture.Some conditions may manifest skele-tal changes, such as metaphyseal ir-regularity and subperiosteal newbone formation. These entities shouldbe considered in the differential di-agnosis of childhood fractures.

PEDIATRICS Volume 133, Number 2, February 2014 e479

FROM THE AMERICAN ACADEMY OF PEDIATRICS

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

Osteogenesis Imperfecta

Osteogenesis imperfecta (OI) is a het-erogeneous family of diseases, usuallycaused by heterozygous mutations ofthe genes COL1A1 and COL1A2,54 butmutations in these and other genescan cause autosomal recessive formsof OI. The COL1A1 and COL1A2 genesencode the chains of type I collagen,which forms the structural frameworkof bone. Although it is a genetic dis-order, many children have de novomutations or autosomal-recessive dis-ease and no family history of bonefragility. In addition, the presentation ofthe disease within affected membersof the same family can be quite vari-able. Phenotypic expression of thedisease depends on the nature of themutation, its relative abundance at-tributable to mosaicism, and its ex-pression in target tissues.55 Sometypes of OI involve reduced productionof collagen, and the symptoms resolveor lessen after puberty.56 Table 2 liststhe various signs and symptoms thatcan be present in a case of OI.

The diagnosis of OI is often suggestedby a family history of fractures, shortstature, blue sclera, poor dentition,and radiographic evidence of low bonedensity or osteopenia. The fracturesare most commonly transverse innature, occurring in the shafts of the

long bones. It is unusual to have mul-tiple long bone fractures or rib frac-tures, particularly in infancy, withoutother clinical and radiographic evi-dence of OI.57,58

OI has been misdiagnosed as childabuse.59 On the other hand, OI is oftensuggested as the cause of fractures inchildren who have been abused. Iffractures continue to occur whena child is placed in a protective envi-ronment, a more thorough evaluationfor an underlying bone disease isneeded. Child abuse is more commonthan OI,60 and children with OI andother metabolic or genetic conditionsmay also be abused.61,62

Preterm Birth

Preterm infants have decreased bonemineralization at birth, but after thefirst year of life, bone density nor-malizes.63,64 Osteopenia of prematurityhas been well described as a compli-cation in low birth weight infants.65

Infants born at less than 28 weeks’gestation or who weigh less than 1500g at birth are particularly vulnerable.Osteopenia of prematurity is multi-factorial. Infants are also at risk ifthey receive prolonged (for 4 or moreweeks) total parenteral nutrition, havebronchopulmonary dysplasia, and/orhave received a prolonged course ofdiuretics or steroids.66 Osteopeniacommonly presents between 6 and 12weeks of life. Osteopenia of pre-maturity can be ameliorated if infantsare monitored closely and receive thenutritional and mineral supplementa-tion initiated in the NICU.

Fractures associated with osteopenia ofprematurity usually occur in the firstyear of life.67 Rib fractures are typicallyencountered incidentally, whereas longbone fractures commonly present withswelling of the extremity. Osteopenia ofprematurity can be associated withrickets, and in such cases, meta-physeal irregularities may be present.

Although osteopenia of prematuritymay make the infant more vulnerableto fracture, preterm infants are also atan increased risk of abuse.68

Vitamin D Deficiency Rickets

Suboptimal vitamin D concentrationsand rickets have been proposed ascauses of fractures in infants.69 VitaminD insufficiency in otherwise healthyinfants and toddlers is common. Ap-proximately 40% of infants and tod-dlers aged 8 to 24 months in an urbanclinic had laboratory evidence of vita-min D insufficiency (serum concen-trations of 25-hydroxyvitamin D of ≤30ng/mL).70 Prolonged breastfeedingwithout vitamin D supplementationwas a critical factor that placed theseinfants at risk, although increased skinpigmentation and/or lack of sunlightexposure may also have contributed.Rickets is characterized by de-mineralization, loss of the zone ofprovisional calcification, wideningand irregularity of the physis, andfraying and cupping of the meta-physis.71 Despite the high prevalenceof vitamin D insufficiency in infantsand toddlers, rickets is uncommon.72

The claim that vitamin D deficiency orinsufficiency causes skeletal lesionsthat lead to the incorrect diagnosis ofchild abuse in infants is not sup-ported in the literature. A systematicclinical, laboratory, and radiologicassessment should exclude thatpossibility.73–75 Schilling et al foundno difference in serum concen-trations of 25-hydroxyvitamin D inyoung children with fractures suspi-cious for abuse and noninflictedfractures.76 Vitamin D insufficiencywas not associated with multiplefractures, in particular rib fracturesor CMLs, the high specificity indica-tors of abuse. Perez-Rossello et alstudied radiographs of 40 healthyolder infants and toddlers with

TABLE 2 Characteristics of OsteogenesisImperfecta

Fragile bones with few, some, or many of thefollowing findings:Poor linear growthMacrocephalyTriangular-shaped faceBlue scleraeHearing impairment as a result of otosclerosisHypoplastic, translucent, carious, late-erupting,

or discolored teethEasy bruisabilityInguinal and/or umbilical herniasLimb deformitiesHyperextensible jointsScoliosis and/or kyphosisWormian bones of the skullDemineralized bones

e480 FROM THE AMERICAN ACADEMY OF PEDIATRICS by guest on June 15, 2018www.aappublications.org/newsDownloaded from

vitamin D insufficiency and deficiencyand concluded that radiographic ra-chitic changes were uncommon andvery mild. In this population, thereported fracture prevalence waszero.72

In a study of 45 young children withradiographic evidence of rickets,investigators found that fractures oc-curred only in those infants and tod-dlers who were mobile.77 Fractureswere seen in 17.5% of the children,and these children were 8 to 19months of age. The fractures involvedlong bones, anterior-lateral and lat-eral ribs, and metatarsal and meta-physeal regions. The metaphysealfractures occurred closer to the di-aphysis in the background of floridmetaphyseal rachitic changes anddid not resemble the juxtaphysealcorner or bucket handle pattern ofthe CML. In infant fatalities in whichabuse is suspected, rachitic changesappear to be rare histologically.78

Osteomyelitis

Osteomyelitis in infants can present asmultiple metaphyseal irregularitiespotentially resembling CMLs.79 Typically,the lesions become progressively lyticand sclerotic with substantial sub-periosteal new bone formation. Othersigns of infection are often present,such as fever, increased erythrocytesedimentation rate, elevated C-reactiveprotein concentration, and elevatedwhite blood cell count.

Fractures Secondary toDemineralization From Disuse

Any child with a severe disability thatlimits or prevents ambulation can be atrisk for fractures secondary to disusedemineralization, even with normalhandling.80,81 The fractures are usuallydiaphyseal rather than CMLs. Often,these fractures occur during physicaltherapy and range-of-motion exercises.It can be difficult to distinguish between

inflicted and noninflicted fractures oc-curring in these children. At the sametime, children with disabilities are at anincreased risk of being maltreated.82–84

When multiple or recurrent fracturesoccur in a disabled child, a trial changein caregivers may be indicated to de-termine whether the fractures can beprevented. This is an extreme in-tervention and should be reserved forunusual circumstances.63

Scurvy

Scurvy is caused by insufficient intakeof vitamin C, which is important for thesynthesis of collagen. Although raretoday because formula, human milk,fruits, and vegetables contain vitaminC, scurvy may develop in older infantsand children given exclusively cowmilk without vitamin supplementationand in children who eat no foodscontaining vitamin C.85–87 Althoughscurvy can result in metaphysealchanges similar to those seen withchild abuse, other characteristic bonechanges, including osteopenia, in-creased sclerosis of the zones ofprovisional calcification, dense epiph-yseal rings, and extensive calcificationof subperiosteal and soft tissue hem-orrhages, will point to the diagnosisof scurvy.

Copper Deficiency

Copper plays a role in cartilage for-mation. Copper deficiency is a rarecondition that may be complicated bybone fractures. Preterm infants areborn with lower stores of copper thanterm infants, because copper is ac-cumulated at a faster rate during thelast trimester.88 Copper insufficiencymay be observed in children withsevere nutritional disorders, for ex-ample, liver failure or short gut syn-drome.89 This deficiency is not likelyto be observed in full-term childrenyounger than 6 months of age orpreterm infants younger than 2.5

months of age, because fetal copperstores are sufficient for this lengthof time. In addition, human milk andformula contain sufficient copper toprevent deficiency. Psychomotor re-tardation, hypotonia, hypopigmentation,pallor, and a sideroblastic anemia aresome of the characteristic findings ofcopper deficiency in infants. Radiologicchanges that should lead to furtherevaluation for possible deficiencyinclude cupping and fraying of themetaphyses, sickle-shaped metaphysealspurs, significant demineralization, andsubperiosteal new bone formation.

Menkes Disease

Menkes disease, also known asMenkes kinky hair syndrome, is a rarecongenital defect of copper metabo-lism.90 Menkes disease is an X-linkedrecessive condition and occurs only inboys. Although it has many of thefeatures of dietary copper deficiency,anemia is not associated with Menkesdisease. Metaphyseal fragmentationand subperiosteal new bone forma-tion may be observed on radiographs,and the findings may be difficult todistinguish from fractures caused byabuse.91 Other signs of Menkes dis-ease include sparse, kinky hair, cal-varial wormian bones, anterior ribflaring, failure to thrive, and de-velopmental delay. A characteristicfinding is tortuous cerebral vessels.Intracranial hemorrhage can occur inMenkes disease but has not beenreported in infants with copper de-ficiency.

Systemic Disease

Chronic renal disease affects bonemetabolism because children withchronic renal disease may developa metabolic acidosis that interfereswith vitamin D metabolism. Chronicrenal disease can cause renal osteo-dystrophy resulting in the same ra-diographic changes as nutritional

PEDIATRICS Volume 133, Number 2, February 2014 e481

FROM THE AMERICAN ACADEMY OF PEDIATRICS

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

rickets. Because chronic liver disease(eg, biliary atresia) interferes with vi-tamin D metabolism, such children maybe at an increased risk of fractures.Fanconi syndrome, hypophosphatasia,hypophosphatemic (vitamin D resistant)rickets, hyperparathyroidism, and renaltubular acidosis also cause clinicalvariants of rickets.

Temporary Brittle Bone DiseaseHypothesis

Physicians should be aware of alter-native diagnoses that are unsupportedby research but are sometimes sug-gested when an infant has unexplainedfractures. In 1993, Paterson proposedthat some infants may be born withbones that are temporarily morefragile or vulnerable to fracture in thecontext of normal handling, which hecalled “temporary brittle bone dis-ease.”92 Paterson suggested thatsome trace element deficiency, suchas copper or a transient collagen im-maturity, caused the disease but pro-vided no scientific data that confirmedhis hypotheses and offered no specifictest that confirmed temporary brittlebone disease.61 Subsequent studies didnot support his hypotheses, and hiscase analysis has been refuted.57,93–95

Miller hypothesized that temporarybrittle bone disease is a result of fetalimmobilization or intrauterine con-finement that leads to transient boneloss or osteopenia.96,97 In support ofhis hypothesis, he reported that 95%of 21 infants with multiple unexplainedfractures had decreased fetal move-ments, according to their mothers.97,98

Although he used bone densitometry ineach patient as a basis for his con-clusions, none of the patients had hadbone densitometry performed at thetime of the fracture. The testing wasperformed 8 to 21 weeks later, and noinfants were tested before 5 months ofage. In addition, bone densitometrystandards have not been established

for infants. He relied on the mother’shistory of decreased fetal movementsand provided no independent mea-surements of those movements. Pala-cios and Rodriguez found no evidencethat oligohydramnios affects bone massof the fetus, probably because fetalmovement is only restricted in the lasttrimester of pregnancy by oligohy-dramnios and because the mechanicalloading on the bones stimulating boneformation is conserved.99

Medical Evaluation

History of Present Illness

It is essential to obtain a detailed his-tory to determine how an injury oc-curred. If an injury in a nonverbal childwas witnessed, the caregiver should beable to provide details about the child’sactivity and position before an injuryand the child’s final position and loca-tion after the injury occurred.46 Verbalchildren with concerning fracturesshould be interviewed apart fromcaregivers and ideally by a professionalwho is skilled in forensic interviewing.

A comparison of the histories providedby caregivers of children with non-inflicted femoral fractures and bycaregivers of children whose injurieswere caused by abuse is instructive.When an injury was caused by abuse,the caregiver provided either no his-tory of an injury or related a history ofa low-energy event. By contrast, 29% ofthe caregivers of children with non-inflicted injuries provided some high-energy explanation, such as a motorvehicle collision or that the child fellfrom a height.16 Most of the low-energy mechanisms provided for thenoninflicted injuries involved falls in-cluding stair falls and siblings landingon the femur during play.16,46

The child’s response to the event mayalso provide important clues aboutthe etiology. The majority of childrenwith long bone fractures will have

some swelling, pain, or other signs,such as decreased use of the ex-tremity, suggesting a fracture.100,101

Some children, however, will haveminimal external signs of injury.102

The absence of any history of injury,a vague description of the event,a delay in seeking care, the absenceof an explanation for an injury par-ticularly in a nonambulatory child, oran inconsistent explanation shouldincrease the physician’s concern thatan injury was caused by child abuse(see Table 3).13,16

Past Medical History

The past medical history is importantand should include details about themother’s pregnancy. If the child wasborn preterm, the infant’s bone min-eral content may be reduced, and theinfant may be at risk for fracture. Ahistory of total parental nutrition,hepatobiliary disease, diuretic ther-apy, hypercalciuria, or corticosteroidsmay make the bones of a low birthweight infant even more vulnerable tofracture. In addition, chronic diseases,such as renal insufficiency or meta-bolic acidosis, malabsorption, cere-bral palsy or other neuromusculardisorders, genetic diseases that affectskeletal development, or any illnessthat limits mobility, may affect bonestrength. A thorough dietary historyand history of medications that can

TABLE 3 When Is a Fracture Suspicious forChild Abuse?

• No history of injury• History of injury not plausible—mechanismdescribed not consistent with the type offracture, the energy load needed to cause thefracture, or the severity of the injury

• Inconsistent histories or changing historiesprovided by caregiver

• Fracture in a nonambulatory child• Fracture of high specificity for child abuse (eg,rib fractures)

• Multiple fractures• Fractures of different ages• Other injuries suspicious for child abuse• Delay in seeking care for an injury

e482 FROM THE AMERICAN ACADEMY OF PEDIATRICS by guest on June 15, 2018www.aappublications.org/newsDownloaded from

predispose to fractures are impor-tant. The physician should inquireabout previous injuries includingbruises and determine the child’s de-velopmental abilities, because chil-dren who are not yet mobile are muchmore likely to have fractures causedby abuse.

Family History

A family history of multiple fractures,early-onset hearing loss, abnormallydeveloped dentition, blue sclera, andshort stature should suggest thepossibility of OI.

Social History

The physician should obtain a com-plete psychosocial history, includingasking who lives in the home and whohas provided care for the child. Thehistory should inquire about intimatepartner violence, substance abuseincluding drugs and alcohol, mentalillness, and previous involvement withchild protective services and/or lawenforcement.

Physical Examination

The child should have a comprehensivephysical examination, and the growthchart should be carefully reviewed.Abnormal weight may suggest neglector endocrine or metabolic disorders.Any signs or symptoms of fractures,such as swelling, limitation of motion,and point tenderness should bedocumented. The physician should doa complete skin examination to lookfor bruises and other skin findingsbecause bruises are the most commoninjury caused by child abuse. Themajority of children with fractures donot have bruising associated with thefracture; the presence or absence ofsuch bruising does not help to de-termine which fractures are caused bychild abuse.103,104 Bruising in a childwho is not yet cruising or bruising inunusual locations, such as the ears,

neck, or trunk should raise suspicionfor child abuse.105,106 The child shouldbe examined for other injuries causedby child abuse, in addition to signs ofother medical conditions associatedwith bone fragility. Blue sclerae areseen in certain types of OI. Sparse,kinky hair is associated with Menkesdisease. Dentinogenesis imperfecta isoccasionally identified in older chil-dren with OI.

Laboratory Evaluation

The clinical evaluation should guidethe laboratory evaluation. In childrenwith fractures suspicious for abuse,serum calcium, phosphorus, and al-kaline phosphatase should be reviewed,although alkaline phosphatase may beelevated with healing fractures. Thephysician should consider checkingserum concentrations of parathyroidhormone and 25-hydroxyvitamin D, aswell as urinary calcium excretion (eg,random urinary calcium/creatinineratio) in all young children withfractures concerning for abuse, butthese levels should certainly beassessed if there is radiographic ev-idence of osteopenia or metabolicbone disease. Screening for abdomi-nal trauma with liver function studiesas well as amylase and lipase con-centration should be performed whensevere or multiple injuries are iden-tified. A urinalysis should be per-formed to screen for occult blood.Serum copper, vitamin C, and ceru-loplasmin concentrations should beconsidered if the child is at risk forscurvy or copper deficiency and hasradiographic findings that includemetaphyseal abnormalities.

If OI is suspected, sequence analysisof the COL1A1 and COL1A2 genes thatare associated with 90% of cases ofOI as well as other genes associatedwith less common autosomal-recessive forms of OI may be moresensitive than biochemical tests of

type I collagen and may identify themutation to guide testing of otherfamily members.107 Some of the lesscommon forms of OI are OI types IIBand VII, CRTAP; OI type VI, FKBP10; OItype VIII, LEPRE1; OI type IX, PPIB; OI typeX, SERPINH1; OI type XI, SP7; OI type XII,SERPINF1; and OI type XIII, BMP1. DNAsequencing can be performed usinggenomic DNA isolated from peripheralblood mononuclear cells or even sa-liva, whereas the biochemical analysisof type I collagen requires a skin bi-opsy. Doing both DNA analysis andskin biopsy is not indicated in mostcases. Consultation with a pediatricgeneticist may be helpful in decidingwhich children to test and which testto order.108

Imaging Approach

Children younger than 2 years withfractures suspicious for child abuseshould have a radiographic skeletalsurvey to look for other bone injuriesor osseous abnormalities.109 Addi-tional fractures are identified in ap-proximately 10% of skeletal surveys,with higher yields in infants.110 Skeletalsurveys may be appropriate in somechildren between ages 2 and 5 years,depending on the clinical suspicionof abuse. If specific clinical findingsindicate an injury at a particularsite, imaging of that area shouldbe obtained regardless of the child’sage.

The American College of Radiology hasdeveloped specific practice guidelinesfor skeletal surveys in children.111

Twenty-one images are obtained, in-cluding frontal images of the appen-dicular skeleton, frontal and lateralviews of the axial skeleton, and obli-que views of the chest. Oblique viewsof the chest have been shown to in-crease the sensitivity, specificity, andaccuracy of the identification of ribfractures.112 A full 4 skull series shouldbe obtained if there are concerns of

PEDIATRICS Volume 133, Number 2, February 2014 e483

FROM THE AMERICAN ACADEMY OF PEDIATRICS

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

head injury. Computed tomography (CT)3-dimensional models are valuableadjuncts to the radiographs and havethe potential to replace the skull se-ries.113 This has not been studied sys-tematically in this context, however.Because lateral views of the extremitiesincrease yield, some authors suggestthat these views be included in theimaging protocol.114 Fractures may bemissed if the guidelines are not fol-lowed or if the images are of poorquality.115 A repeat skeletal surveyshould be performed approximately 2to 3 weeks after the initial skeletalsurvey if child abuse is strongly sus-pected.109,116 The follow-up examinationmay identify fractures not seen on theinitial skeletal survey, can clarify un-certain findings identified by the initialskeletal survey, and improves bothsensitivity and specificity of the skeletalsurvey.116,117 In one study, 13 of 19fractures found on the follow-up ex-amination were not seen on the initialseries.116 The number of images on thefollow-up examination may be limitedto 15 views by omitting the views of theskull, pelvis, and lateral spine.118

Radiography may assist in assessingthe approximate time when an injuryoccurred because long bone fracturesheal following a particular sequence.119

If the healing pattern is not consistentwith the explanation provided, theaccuracy of the explanation should bequestioned.

Bone scintigraphy may be used tocomplement the skeletal survey butshould not be the sole method ofidentifying fractures in infants. Al-though it has high overall sensitivity, itlacks specificity for fracture detectionand may fail to identify CMLs and skullfractures.109,119,120 Scintigraphy doeshave high sensitivity for identifying ribfractures, which can be difficult todetect before healing. In toddlers andolder children, the use of bone scin-tigraphy or skeletal survey depends on

the specific clinical indicators ofabuse.109

Because brain injuries are often occult,head imaging should be considered forany child younger than 1 year witha fracture suspicious for abuse.121 Im-aging studies may help clarify whetherthe child has been abused, providefurther support for a diagnosis of childabuse, and identify other injuries thatrequire treatment. Additional imagingmay be needed if the child has signs orsymptoms of chest, abdominal, or neckinjury.

Chest CT can identify rib fractures thatare not seen on chest radiographs.122

CT is particularly useful in detectinganterior rib fractures and rib frac-tures at all stages of healing—earlysubacute, subacute, and old fractures.Although CT may be more sensitive inidentifying these injuries, a chest CTexposes the child to significantly moreradiation than chest radiography. Ev-ery effort should be made to reducechildren’s exposure to radiation whileat the same time considering the riskto the child if abuse is not identi-fied.123 Therefore, selective applicationof this technique in certain clinicalsettings is appropriate.

Other modalities may become availablein the future that will provide moreaccurate identification of skeletal inju-ries. Whole-body short tau inverse re-covery imaging, a magnetic resonanceimaging (MRI) technique, may identifyrib fractures not recognized on theradiographic skeletal survey.124 Ina study of 21 infants with suspectedabuse, whole-body MRI at 1.5-Tesla wasinsensitive in the detection of CMLsand rib fractures. In some cases,whole-body MRI identified soft tissueedema and joint effusions that led tothe identification of skeletal injurieswith additional radiographs.125 Bonescintigraphy with 18F-sodium fluoridepositron emission tomography (18F-NaFPET) bone scan may be useful in cases

of equivocal or negative skeletal sur-veys when there is high clinical suspi-cion of abuse. If available, a 18F-NaFpositron emission tomography bonescan has better contrast and spatialresolution than 99mTc-labeled methy-lene diphosphonate.120

Although bone densitometry by dual-energy x-ray absorptiometry is usefulto predict bone fragility and fracturerisk in older adults, interpretation ofbone densitometry in children andadolescents is more problematic.126 Inadults, bone densitometry is inter-preted using T scores, which describethe number of SDs above or below theaverage peak bone mass for a gender-and race-adjusted reference group ofnormal subjects. Because peak bonemass is not achieved until approxi-mately 30 years of age, in children,z scores must be used to express bonedensity, because z scores express thechild’s bone mineral density as a func-tion of SDs above or below the averagefor an age- and gender-matched normcontrol population.127 In addition, be-cause bone size influences dual-energyx-ray absorptiometry, z scores mustalso be adjusted for height z scores.128

The International Society for ClinicalDensitometry recommends that thediagnosis of osteoporosis in childhoodshould not be made on the basis of lowbone mass alone but should also in-clude a clinically significant historyof low-impact fracture. The recom-mendations currently apply to chil-dren 5 years and older, althoughreference data are available for chil-dren as young as 3 years.129,130 Un-fortunately, there are limited referencedata for the young, nonverbal childwho is most at risk for suffering frac-tures caused by child abuse.

Evaluation of Siblings

Siblings, especially twins, and otheryoung household members of childrenwho have been physically abused

e484 FROM THE AMERICAN ACADEMY OF PEDIATRICS by guest on June 15, 2018www.aappublications.org/newsDownloaded from

should be evaluated for maltreat-ment.131 In a study of 795 siblings in400 households of a child who hadbeen abused or neglected, all sib-lings in 37% of households and somesiblings in 20% of households hadsuffered some form of maltreat-ment.132 In this study, which includedall manifestations of maltreatment,siblings were found to be more atrisk for maltreatment if the indexchild suffered moderate or severemaltreatment. In addition to a carefulevaluation, imaging should be consid-ered for any siblings younger than 2years, especially if there are signs ofabuse.

DIAGNOSIS

When evaluating a child with a frac-ture, physicians must take a carefulhistory of any injury event andthen determine whether the mechan-ism described and the severity andtiming are consistent with the injuryidentified (see Table 3).133 They mustconsider and evaluate for possible di-agnoses in addition to other signs orsymptoms of child abuse. A carefulevaluation for other injuries is im-portant because the presence of ad-ditional injuries that are associatedwith child abuse increases the likeli-hood that a particular fracture wasinflicted.16,43 It is important to re-member that even if a child has anunderlying disorder or disability thatcould increase the likelihood ofa fracture, the child may also havebeen abused because children withdisabilities and other special healthcare needs are at increased risk ofchild abuse.83,84 Physicians shouldkeep an open mind to the possibility ofabuse and remember that childabuse occurs in all socioeconomic

groups and across all racial and ethnicgroups. Many of these diagnoses arecomplex. If a physician is uncertainabout how to evaluate an injury or ifthey should suspect a fracture wascaused by child abuse, they shouldconsult a child abuse pediatricianor multidisciplinary child abuse teamto assist in the evaluation, particu-larly if the child is nonambulatoryor younger than 1 year of age.134 Incertain circumstances, the physicianwill need to consult an orthopedist,endocrinologist, geneticist, or othersubspecialists.

All US states, commonwealths, andterritories have mandatory reportingrequirements for physicians andother health care providers whenchild abuse is suspected. Physiciansshould be aware of and comply withthe reporting requirements of theirstate. Typically, the standard formaking a report is when the reporter“suspects” or “has reason to believe”that a child has been abused orneglected. Sometimes determiningwhether that “reasonable belief” or“reasonable suspicion” standard hasbeen met can be nuanced and com-plex. The physician should keep inmind that incontrovertible proof ofabuse or neglect is not required bystate statutes, and there may becases in which it is reasonable toconsult with a child abuse pediatri-cian about whether a report shouldbe made.

LEAD AUTHOREmalee G. Flaherty, MD, FAAP

COMMITTEE ON CHILD ABUSE ANDNEGLECT, 2012–2013Cindy W. Christian, MD, FAAP, ChairpersonJames E. Crawford-Jakubiak, MD, FAAP

Emalee G. Flaherty, MD, FAAPJohn M. Leventhal, MD, FAAPJames L. Lukefahr, MD, FAAPRobert D. Sege MD, PhD, FAAP

LIAISONSHarriet MacMillan, MD – American Academy ofChild and Adolescent PsychiatryCatherine M. Nolan, MSW, ACSW – Administrationfor Children, Youth, and FamiliesLinda Anne Valley, PhD – Centers for DiseaseControl and Prevention

STAFFTammy Piazza Hurley

SECTION ON RADIOLOGY EXECUTIVECOMMITTEE, 2012–2013Christopher I. Cassady, MD, FAAP, ChairpersonDorothy I. Bulas, MD, FAAPJohn A. Cassese, MD, FAAPAmy R. Mehollin-Ray, MD, FAAPMaria-Gisela Mercado-Deane, MD, FAAPSarah Sarvis Milla, MD, FAAP

STAFFVivian Thorne

SECTION ON ENDOCRINOLOGYEXECUTIVE COMMITTEE, 2012–2103Irene N. Sills, MD, FAAP, ChairpersonClifford A. Bloch, MD, FAAPSamuel J. Casella, MD, MSc, FAAPJoyce M. Lee, MD, FAAPJane Lockwood Lynch, MD, FAAPKupper A. Wintergerst, MD, FAAP

STAFFLaura Laskosz, MPH

SECTION ON ORTHOPEDICS EXECUTIVECOMMITTEE, 2012–2013Richard M. Schwend, MD, FAAP, ChairpersonJ. Eric Gordon, MD, FAAPNorman Y. Otsuka, MD, FAAPEllen M. Raney, MD, FAAPBrian A. Shaw, MD, FAAPBrian G. Smith, MD, FAAPLawrence Wells, MD, FAAPPaul W. Esposito, MD, USBJD Liaison

STAFFNiccole Alexander, MPP

PEDIATRICS Volume 133, Number 2, February 2014 e485

FROM THE AMERICAN ACADEMY OF PEDIATRICS

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

REFERENCES

1. Loder RT, Feinberg JR. Orthopaedic inju-ries in children with nonaccidentaltrauma: demographics and incidencefrom the 2000 kids’ inpatient database[published correction appears in JPediatr Orthop. 2008;28(6):699]. J PediatrOrthop. 2007;27(4):421–426

2. Jenny C, Hymel KP, Ritzen A, Reinert SE,Hay TC. Analysis of missed cases of abu-sive head trauma [see comment; pub-lished correction appears in JAMA. 1999;282(1):29]. JAMA. 1999;281(7):621–626

3. Ravichandiran N, Schuh S, Bejuk M, et al.Delayed identification of pediatric abuse-related fractures. Pediatrics. 2010;125(1):60–66

4. Skellern C, Donald T. Suspicious childhoodinjury: formulation of forensic opinion. JPaediatr Child Health. 2011;47(11):771–775

5. O’Neill JA, Jr, Meacham WF, Griffin JP,Sawyers JL. Patterns of injury in thebattered child syndrome. J Trauma. 1973;13(4):332–339

6. Gallagher SS, Finison K, Guyer B, Good-enough S. The incidence of injuriesamong 87,000 Massachusetts childrenand adolescents: results of the 1980–81Statewide Childhood Injury PreventionProgram Surveillance System. Am J Pub-lic Health. 1984;74(12):1340–1347

7. Spady DW, Saunders DL, Schopflocher DP,Svenson LW. Patterns of injury in children:a population-based approach. Pediatrics.2004;113(3 pt 1):522–529

8. Rennie L, Court-Brown CM, Mok JYQ,Beattie TF. The epidemiology of fracturesin children. Injury. 2007;38(8):913–922

9. Leventhal JM, Martin KD, Asnes AG. In-cidence of fractures attributable to abusein young hospitalized children: resultsfrom analysis of a United States database.Pediatrics. 2008;122(3):599–604

10. Schwend RM, Werth C, Johnston A. Femurshaft fractures in toddlers and youngchildren: rarely from child abuse. JPediatr Orthop. 2000;20(4):475–481

11. Hennrikus WL, Shaw BA, Gerardi JA. Inju-ries when children reportedly fall froma bed or couch. Clin Orthop Relat Res.2003; (407):148–151

12. Worlock P, Stower M, Barbor P. Patterns offractures in accidental and non-accidentalinjury in children: a comparative study. BrMed J (Clin Res Ed). 1986;293(6539):100–102

13. Leventhal JM, Thomas SA, Rosenfield NS,Markowitz RI. Fractures in young children.Distinguishing child abuse from un-

intentional injuries. Am J Dis Child. 1993;147(1):87–92

14. Leventhal JM, Larson IA, Abdoo D, et al.Are abusive fractures in young childrenbecoming less common? Changes over 24years. Child Abuse Negl. 2007;31(3):311–322

15. Kemp AM, Dunstan F, Harrison S, et al.Patterns of skeletal fractures in childabuse: systematic review [review]. BMJ.2008;337:a1518

16. Hui C, Joughin E, Goldstein S, et al. Fem-oral fractures in children younger thanthree years: the role of nonaccidental in-jury. J Pediatr Orthop. 2008;28(3):297–302

17. Pandya NK, Baldwin K, Wolfgruber H,Christian CW, Drummond DS, Hosalkar HS.Child abuse and orthopaedic injury pat-terns: analysis at a level I pediatrictrauma center. J Pediatr Orthop. 2009;29(6):618–625

18. Coffey C, Haley K, Hayes J, Groner JI. Therisk of child abuse in infants and toddlerswith lower extremity injuries. J PediatrSurg. 2005;40(1):120–123

19. Kleinman PK. The spectrum of non-accidental injuries (child abuse) and itsimitators. In: Hodler J, Zollikofer CL,Schulthess GK, eds. Musculoskeletal Dis-eases 2009–2012. Milan, Italy: SpringerItalia; 2009:227–233

20. Clarke NMP, Shelton FRM, Taylor CC, KhanT, Needhirajan S. The incidence of frac-tures in children under the age of 24months—in relation to non-accidentalinjury. Injury. 201243(6):762–765

21. Kleinman PK. Diagnostic Imaging of ChildAbuse. 2nd ed. St. Louis, MO: Mosby; 1998

22. Barsness KA, Cha E-S, Bensard DD, et al.The positive predictive value of rib frac-tures as an indicator of nonaccidentaltrauma in children [see comment]. JTrauma. 2003;54(6):1107–1110

23. Bulloch B, Schubert CJ, Brophy PD, JohnsonN, Reed MH, Shapiro RA. Cause and clinicalcharacteristics of rib fractures in infants.Pediatrics. 2000;105(4). Available at: www.pediatrics.org/cgi/content/full/105/4/E48

24. Kleinman PK, Schlesinger AE. Mechanicalfactors associated with posterior ribfractures: laboratory and case studies.Pediatr Radiol. 1997;27(1):87–91

25. Bixby SD, Abo A, Kleinman PK. High-impacttrauma causing multiple posteromedialrib fractures in a child. Pediatr EmergCare. 2011;27(3):218–219

26. Feldman KW, Brewer DK. Child abuse,cardiopulmonary resuscitation, and ribfractures. Pediatrics. 1984;73(3):339–342

27. Spevak MR, Kleinman PK, Belanger PL,Primack C, Richmond JM. Cardiopulmo-nary resuscitation and rib fractures ininfants. A postmortem radiologic-pathologic study. JAMA. 1994;272(8):617–618

28. Matshes EW, Lew EO. Two-handed cardio-pulmonary resuscitation can cause ribfractures in infants. Am J Forensic MedPathol. 2010;31(4):303–307

29. Kleinman PK, Perez-Rossello JM, NewtonAW, Feldman HA, Kleinman PL. Prevalenceof the classic metaphyseal lesion ininfants at low versus high risk for abuse.AJR Am J Roentgenol. 2011;197(4):1005–1008

30. Kleinman PK, Marks SC, Jr, Richmond JM,Blackbourne BD. Inflicted skeletal injury:a postmortem radiologic-histopathologicstudy in 31 infants. AJR Am J Roent-genol. 1995;165(3):647–650

31. Kleinman PK. Problems in the diagnosis ofmetaphyseal fractures. Pediatr Radiol.2008;38(suppl 3):S388–S394

32. O’Connell A, Donoghue VB. Can classicmetaphyseal lesions follow un-complicated caesarean section? PediatrRadiol. 2007;37(5):488–491

33. Grayev AM, Boal DK, Wallach DM, Segal LS.Metaphyseal fractures mimicking abuseduring treatment for clubfoot [see com-ment]. Pediatr Radiol. 2001;31(8):559–563

34. King J, Diefendorf D, Apthorp J, NegreteVF, Carlson M. Analysis of 429 fractures in189 battered children. J Pediatr Orthop.1988;8(5):585–589

35. Pierce MC, Bertocci G. Injury bio-mechanics and child abuse. Annu RevBiomed Eng. 2008;10:85–106

36. Pierce MC, Bertocci GE, Vogeley E, MorelandMS. Evaluating long bone fractures in chil-dren: a biomechanical approach with il-lustrative cases. Child Abuse Negl. 2004;28(5):505–524

37. Pierce MC, Bertocci G. Fractures resultingfrom inflicted trauma: assessing injuryand history compatibility. Clin PediatrEmerg Med. 2006;7(3):143–148

38. Rex C, Kay PR. Features of femoral frac-tures in nonaccidental injury. J PediatrOrthop. 2000;20(3):411–413

39. Haney SB, Starling SP, Heisler KW, OkwaraL. Characteristics of falls and risk of in-jury in children younger than 2 years.Pediatr Emerg Care. 2010;26(12):914–918

40. Nimityongskul P, Anderson LD. The likeli-hood of injuries when children fall out ofbed. J Pediatr Orthop. 1987;7(2):184–186

e486 FROM THE AMERICAN ACADEMY OF PEDIATRICS by guest on June 15, 2018www.aappublications.org/newsDownloaded from

41. Lyons TJ, Oates RK. Falling out of bed:a relatively benign occurrence. Pediatrics.1993;92(1):125–127

42. Hansoti B, Beattie T. Can the height of fallpredict long bone fracture in childrenunder 24 months? Eur J Emerg Med. 2005;12(6):285–286

43. Thomas SA, Rosenfield NS, Leventhal JM,Markowitz RI. Long-bone fractures inyoung children: distinguishing accidentalinjuries from child abuse. Pediatrics.1991;88(3):471–476

44. Loder RT, O’Donnell PW, Feinberg JR. Epi-demiology and mechanisms of femurfractures in children. J Pediatr Orthop.2006;26(5):561–566

45. Baldwin K, Pandya NK, Wolfgruber H,Drummond DS, Hosalkar HS. Femur frac-tures in the pediatric population: abuseor accidental trauma? Clin Orthop RelatRes. 2011;469(3):798–804

46. Pierce MC, Bertocci GE, Janosky JE, et al.Femur fractures resulting from stair fallsamong children: an injury plausibilitymodel. Pediatrics. 2005;115(6):1712–1722

47. Haney SB, Boos SC, Kutz TJ, Starling SP.Transverse fracture of the distal femoralmetadiaphysis: a plausible accidentalmechanism. Pediatr Emerg Care. 2009;25(12):841–844

48. Grant P, Mata MB, Tidwell M. Femur frac-ture in infants: a possible accidental eti-ology. Pediatrics. 2001;108(4):1009–1011

49. Pandya NK, Baldwin KD, Wolfgruber H,Drummond DS, Hosalkar HS. Humerusfractures in the pediatric population: analgorithm to identify abuse. J PediatrOrthop B. 2010;19(6):535–541

50. Strait RT, Siegel RM, Shapiro RA. Humeralfractures without obvious etiologies inchildren less than 3 years of age: when isit abuse? Pediatrics. 1995;96(4 pt 1):667–671

51. Hymel KP, Jenny C. Abusive spiral frac-tures of the humerus: a videotaped ex-ception. Arch Pediatr Adolesc Med. 1996;150(2):226–227

52. Laskey AL, Stump TE, Hicks RA, Smith JL.Yield of skeletal surveys in children ≤18months of age presenting with isolatedskull fractures. J Pediatr. 2013;162(1):86–89

53. Wood JN, Christian CW, Adams CM, RubinDM. Skeletal surveys in infants with iso-lated skull fractures. Pediatrics. 2009;123(2). Available at: www.pediatrics.org/cgi/content/full/123/2/e247–e252

54. Byers PH, Steiner RD. Osteogenesisimperfecta. Annu Rev Med. 1992;43:269–282

55. Wallis GA, Starman BJ, Zinn AB, Byers PH.Variable expression of osteogenesisimperfecta in a nuclear family isexplained by somatic mosaicism for a le-thal point mutation in the alpha 1(I) gene(COL1A1) of type I collagen in a parent. AmJ Hum Genet. 1990;46(6):1034–1040

56. Barsh GS, Byers PH. Reduced secretion ofstructurally abnormal type I procollagenin a form of osteogenesis imperfecta.Proc Natl Acad Sci USA. 1981;78(8):5142–5146

57. Sprigg A. Temporary brittle bone diseaseversus suspected non-accidental skeletalinjury. Arch Dis Child. 2011;96(5):411–413

58. Greeley CS, Donaruma-Kwoh M, VettimattamM, Lobo C, Williard C, Mazur L. Fractures atdiagnosis in infants and children withosteogenesis imperfecta. J Pediatr Orthop.2013;33(1):32–36 doi:10.1097/BPO.1090-b1013e318279c318255d

59. Singh Kocher M, Dichtel L. Osteogenesisimperfecta misdiagnosed as child abuse.J Pediatr Orthop B. 2011;20(6):440–443

60. Gahagan S, Rimsza ME. Child abuse orosteogenesis imperfecta: how can we tell?Pediatrics. 1991;88(5):987–992

61. Ablin DS, Sane SM. Non-accidental in-jury: confusion with temporary brittlebone disease and mild osteogenesisimperfecta. Pediatr Radiol. 1997;27(2):111–113

62. Knight DJ, Bennet GC. Nonaccidental injuryin osteogenesis imperfecta: a case report.J Pediatr Orthop. 1990;10(4):542–544

63. Jenny C; Committee on Child Abuse andNeglect. Evaluating infants and youngchildren with multiple fractures. Pediat-rics. 2006;118(3):1299–1303

64. Backström MC, Kuusela A-L, Mäki R. Met-abolic bone disease of prematurity. AnnMed. 1996;28(4):275–282

65. Naylor KE, Eastell R, Shattuck KE, Alfrey AC,Klein GL. Bone turnover in preterminfants. Pediatr Res. 1999;45(3):363–366

66. Harrison CM, Johnson K, McKechnie E.Osteopenia of prematurity: a nationalsurvey and review of practice. Acta Pae-diatr. 2008;97(4):407–413

67. Amir J, Katz K, Grunebaum M, YosipovichZ, Wielunsky E, Reisner SH. Fractures inpremature infants. J Pediatr Orthop. 1988;8(1):41–44

68. Bugental DB, Happaney K. Predicting in-fant maltreatment in low-income families:the interactive effects of maternal attri-butions and child status at birth. DevPsychol. 2004;40(2):234–243

69. Keller KA, Barnes PD. Rickets vs. abuse:a national and international epidemic.Pediatr Radiol. 2008;38(11):1210–1216

70. Gordon CM, Feldman HA, Sinclair L, et al.Prevalence of vitamin D deficiency amonghealthy infants and toddlers. Arch PediatrAdolesc Med. 2008;162(6):505–512

71. Greenspan A. Orthopedic Imaging: APractical Approach. 4th ed. Philadelphia,PA: Lippincott Williams & Wilkins; 2004

72. Perez-Rossello JM, Feldman HA, KleinmanPK, et al. Rachitic changes, deminer-alization, and fracture risk in healthyinfants and toddlers with vitamin Ddeficiency. Radiology. 2012;262(1):234–241

73. Slovis TL, Chapman S. Vitamin D in-sufficiency/deficiency—a conundrum.Pediatr Radiol. 2008;38(11):1153

74. Slovis TL, Chapman S. Evaluating the dataconcerning vitamin D insufficiency/deficiency and child abuse. PediatrRadiol. 2008;38(11):1221–1224

75. Jenny C. Rickets or abuse? Pediatr Radiol.2008;38(11):1219–1220

76. Schilling S, Wood JN, Levine MA, LangdonD, Christian CW. Vitamin D status inabused and nonabused children youngerthan 2 years old with fractures. Pediat-rics. 2011;127(5):835–841

77. Chapman T, Sugar N, Done S, Marasigan J,Wambold N, Feldman K. Fractures ininfants and toddlers with rickets. PediatrRadiol. 2010;40(7):1184–1189

78. Perez-Rossello JM, McDonald AG, RosenbergAE, Ivey SL, Richmond JM, Kleinman PK.Prevalence of rachitic changes in deceasedinfants: a radiologic and pathologic study.Pediatr Radiol. 2011;41(suppl 1):S57

79. Ogden JA. Pediatric osteomyelitis andseptic arthritis: the pathology of neonataldisease. Yale J Biol Med. 1979;52(5):423–448

80. Whedon GD. Disuse osteoporosis: physio-logical aspects. Calcif Tissue Int. 1984;36(suppl 1):S146–S150

81. Presedo A, Dabney KW, Miller F. Fracturesin patients with cerebral palsy. J PediatrOrthop. 2007;27(2):147–153

82. Westcott H. The abuse of disabled chil-dren: a review of the literature. Child CareHealth Dev. 1991;17(4):243–258

83. Sullivan PM, Knutson JF. The associationbetween child maltreatment and dis-abilities in a hospital-based epidemiolog-ical study. Child Abuse Negl. 1998;22(4):271–288

84. Sullivan PM, Knutson JF. Maltreatmentand disabilities: a population-based epi-demiological study. Child Abuse Negl.2000;24(10):1257–1273

PEDIATRICS Volume 133, Number 2, February 2014 e487

FROM THE AMERICAN ACADEMY OF PEDIATRICS

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

85. Olmedo JM, Yiannias JA, Windgassen EB,Gornet MK. Scurvy: a disease almost for-gotten. Int J Dermatol. 2006;45(8):909–913

86. Larralde M, Santos Muñoz A, Boggio P, DiGruccio V, Weis I, Schygiel A. Scurvy ina 10-month-old boy. Int J Dermatol. 2007;46(2):194–198

87. Nishio H, Matsui K, Tsuji H, Tamura A,Suzuki K. Immunohistochemical study oftyrosine phosphorylation signaling in theinvoluted thymus. Forensic Sci Int. 2000;110(3):189–198

88. Shaw JC. Copper deficiency and non-accidental injury. Arch Dis Child. 1988;63(4):448–455

89. Marquardt ML, Done SL, Sandrock M,Berdon WE, Feldman KW. Copper de-ficiency presenting as metabolic bonedisease in extremely low birth weight,short-gut infants. Pediatrics. 2012;130(3).Available at: www.pediatrics.org/cgi/con-tent/full/130/3/e695–e698

90. Tümer Z, Møller LB. Menkes disease. Eur JHum Genet. 2010;18(5):511–518

91. Bacopoulou F, Henderson L, Philip SG.Menkes disease mimicking non-accidentalinjury [published correction appears inArch Dis Child. 2009;94(1):77]. Arch DisChild. 2006;91(11):919

92. Paterson CR, Burns J, McAllion SJ. Osteo-genesis imperfecta: the distinction fromchild abuse and the recognition of a vari-ant form [see comment]. Am J Med Genet.1993;45(2):187–192

93. Chapman S, Hall CM. Non-accidental injuryor brittle bones. Pediatr Radiol. 1997;27(2):106–110

94. Paterson CR. Temporary brittle bone dis-ease: fractures in medical care. ActaPaediatr. 2009;98(12):1935–1938

95. Paterson CR. Temporary brittle bone dis-ease: the current position. Arch Dis Child.2011;96(9):901–902

96. Miller ME. The lesson of temporary brittlebone disease: all bones are not createdequal. Bone. 2003;33(4):466–474

97. Miller ME. Temporary brittle bone disease:a true entity? Semin Perinatol. 1999;23(2):174–182

98. Miller ME, Hangartner TN. Temporarybrittle bone disease: association with de-creased fetal movement and osteopenia.Calcif Tissue Int. 1999;64(2):137–143

99. Palacios J, Rodriguez JI. Extrinsic fetalakinesia and skeletal development:a study in oligohydramnios sequence.Teratology. 1990;42(1):1–5

100. Rivara FP, Parish RA, Mueller BA. Extremityinjuries in children: predictive value of

clinical findings. Pediatrics. 1986;78(5):803–807

101. Taitz J, Moran K, O’Meara M. Long bonefractures in children under 3 years of age:is abuse being missed in emergency de-partment presentations? J Paediatr ChildHealth. 2004;40(4):170–174

102. Farrell C, Rubin DM, Downes K, Dormans J,Christian CW. Symptoms and time tomedical care in children with accidentalextremity fractures. Pediatrics. 2012;129(1). Available at: www.pediatrics.org/cgi/content/full/129/1/ e128–133

103. Peters ML, Starling SP, Barnes-Eley ML,Heisler KW. The presence of bruising as-sociated with fractures. Arch PediatrAdolesc Med. 2008;162(9):877–881

104. Valvano TJ, Binns HJ, Flaherty EG, Leon-hardt DE. Does bruising help determinewhich fractures are caused by abuse?Child Maltreat. 2009;14(4):376–381

105. Sugar NF, Taylor JA, Feldman KW; PugetSound Pediatric Research Network.Bruises in infants and toddlers: those whodon’t cruise rarely bruise. Arch PediatrAdolesc Med. 1999;153(4):399–403

106. Pierce MC, Kaczor K, Aldridge S, O’Flynn J,Lorenz DJ. Bruising characteristics dis-criminating physical child abuse fromaccidental trauma. Pediatrics. 2010;125(1):67–74

107. Shapiro JR, Sponsellor PD. Osteogenesisimperfecta: questions and answers. CurrOpin Pediatr. 2009;21(6):709–716

108. Marlowe A, Pepin MG, Byers PH. Testingfor osteogenesis imperfecta in cases ofsuspected non-accidental injury. J MedGenet. 2002;39(6):382–386

109. American Academy of Pediatrics; Sectionon Radiology; American Academy of Pedi-atrics. Diagnostic imaging of child abuse.Pediatrics. 2009;123(5):1430–1435

110. Duffy SO, Squires J, Fromkin JB, Berger RP.Use of skeletal surveys to evaluate forphysical abuse: analysis of 703 consecu-tive skeletal surveys. Pediatrics. 2011;127(1). Available at: www.pediatrics.org/cgi/content/full/127/1/e47–e52

111. American College of Radiology. ACR Prac-tice Guidelines for Skeletal Surveys inChildren. 2011. Available at: www.acr.org/∼/media/ACR/Documents/PGTS/guidelines/Skeletal_Surveys.pdf. Accessed May 5, 2013

112. Ingram JD, Connell J, Hay TC, Strain JD,MacKenzie T. Oblique radiographs of thechest in nonaccidental trauma. EmergRadiol. 2000;7(1):42–46

113. Prabhu SP, Newton AW, Perez-RosselloJM, Kleinman PK. Three-dimensionalskull models as a problem-solving tool

in suspected child abuse. Pediatr Radiol.2013;43(5):575–581

114. Karmazyn B, Duhn RD, Jennings SG, et al.Long bone fracture detection in suspectedchild abuse: contribution of lateral views[published online ahead of print October6, 2011]. Pediatr Radiol. doi: 10.1007/s00247-011-2248-3

115. van Rijn RR. How should we image skeletalinjuries in child abuse? Pediatr Radiol.2009;39(suppl 2):S226–S229

116. Kleinman PK, Nimkin K, Spevak MR, et al.Follow-up skeletal surveys in suspectedchild abuse. AJR Am J Roentgenol. 1996;167(4):893–896

117. Bennett BL, Chua MS, Care M, KachelmeyerA, Mahabee-Gittens M. Retrospective reviewto determine the utility of follow-up skele-tal surveys in child abuse evaluationswhen the initial skeletal survey is normal.BMC Res Notes. 2011;4(1):354

118. Harlan SR, Nixon GW, Campbell KA, HansenK, Prince JS. Follow-up skeletal surveysfor nonaccidental trauma: can a morelimited survey be performed? PediatrRadiol. 2009;39(9):962–968

119. Mandelstam SA, Cook D, Fitzgerald M,Ditchfield MR. Complementary use of ra-diological skeletal survey and bone scin-tigraphy in detection of bony injuries insuspected child abuse. Arch Dis Child.2003;88(5):387–390, discussion 387–390

120. Drubach LA, Johnston PR, Newton AW,Perez-Rossello JM, Grant FD, Kleinman PK.Skeletal trauma in child abuse: detectionwith 18F-NaF PET. Radiology. 2010;255(1):173–181

121. Rubin DM, Christian CW, Bilaniuk LT,Zazyczny KA, Durbin DR. Occult head injuryin high-risk abused children. Pediatrics.2003;111(6 pt 1):1382–1386

122. Wootton-Gorges SL, Stein-Wexler R, WaltonJW, Rosas AJ, Coulter KP, Rogers KK.Comparison of computed tomography andchest radiography in the detection of ribfractures in abused infants. Child AbuseNegl. 2008;32(6):659–663

123. Brody AS, Frush DP, Huda W, Brent RL;American Academy of Pediatrics Sectionon Radiology. Radiation risk to childrenfrom computed tomography. Pediatrics.2007;120(3):677–682

124. Stranzinger E, Kellenberger CJ, BraunschweigS, Hopper R, Huisman TAGM. Whole-bodySTIR MR imaging in suspected child abuse:an alternative to skeletal survey radiog-raphy? Eur J Radiol Extra. 2007;63(1):43–47

125. Perez-Rossello JM, Connolly SA, NewtonAW, Zou KH, Kleinman PK. Whole-body MRI

e488 FROM THE AMERICAN ACADEMY OF PEDIATRICS by guest on June 15, 2018www.aappublications.org/newsDownloaded from

in suspected infant abuse. AJR Am JRoentgenol. 2010;195(3):744–750

126. Bachrach LK, Sills IN; Section on Endocri-nology. Clinical report—bone densitome-try in children and adolescents.Pediatrics. 2011;127(1):189–194

127. Khoury DJ, Szalay EA. Bone mineral den-sity correlation with fractures in non-ambulatory pediatric patients. J PediatrOrthop. 2007;27(5):562–566

128. Zemel BS, Kalkwarf HJ, Gilsanz V, et al.Revised reference curves for bone min-eral content and areal bone mineraldensity according to age and sex for blackand non-black children: results of thebone mineral density in childhood study. J

Clin Endocrinol Metab. 2011;96(10):3160–3169

129. Gordon CM, Baim S, Bianchi M-L, et al;International Society for Clinical Densi-tometry. Special report on the 2007 Pedi-atric Position Development Conference ofthe International Society for Clinical Den-sitometry. South Med J. 2008;101(7):740–743

130. Henderson RC, Lark RK, Newman JE, et al.Pediatric reference data for dual x-rayabsorptiometric measures of normalbone density in the distal femur. AJR Am JRoentgenol. 2002;178(2):439–443

131. Lindberg DM, Shapiro RA, Laskey AL, PallinDJ, Blood EA, Berger RP; ExSTRA Inves-

tigators. Prevalence of abusive injuries insiblings and household contacts of phys-ically abused children. Pediatrics. 2012;130(2):193–201

132. Hamilton-Giachritsis CE, Browne KD. Aretrospective study of risk to siblings inabusing families. J Fam Psychol. 2005;19(4):619–624

133. Asnes AG, Leventhal JM. Managing childabuse: general principles. Pediatr Rev.2010;31(2):47–55

134. Banaszkiewicz PA, Scotland TR, MyerscoughEJ. Fractures in children younger than age1 year: importance of collaboration withchild protection services. J Pediatr Orthop.2002;22(6):740–744

PEDIATRICS Volume 133, Number 2, February 2014 e489

FROM THE AMERICAN ACADEMY OF PEDIATRICS

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2013-3793 originally published online January 27, 2014; 2014;133;e477Pediatrics 

PEDIATRIC RADIOLOGYENDOCRINOLOGY, SECTION ON ORTHOPAEDICS and the SOCIETY FOR CHILD ABUSE AND NEGLECT, SECTION ON RADIOLOGY, SECTION ON

Hennrikus, and the AMERICAN ACADEMY OF PEDIATRICS COMMITTEE ON Emalee G. Flaherty, Jeannette M. Perez-Rossello, Michael A. Levine, William L.

Evaluating Children With Fractures for Child Physical Abuse

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/133/2/e477including high resolution figures, can be found at:

Referenceshttp://pediatrics.aappublications.org/content/133/2/e477#BIBLThis article cites 125 articles, 32 of which you can access for free at:

Subspecialty Collections

ubhttp://www.aappublications.org/cgi/collection/child_abuse_neglect_sChild Abuse and Neglectson_prevention_subhttp://www.aappublications.org/cgi/collection/injury_violence_-_poiInjury, Violence & Poison Preventionhttp://www.aappublications.org/cgi/collection/section_on_radiologySection on Radiologyogyhttp://www.aappublications.org/cgi/collection/section_on_endocrinolSection on Endocrinologyabuse_and_neglecthttp://www.aappublications.org/cgi/collection/committee_on_child_Committee on Child Abuse and Neglecthttp://www.aappublications.org/cgi/collection/current_policyCurrent Policyfollowing collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlin its entirety can be found online at: Information about reproducing this article in parts (figures, tables) or

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on June 15, 2018www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2013-3793 originally published online January 27, 2014; 2014;133;e477Pediatrics 

PEDIATRIC RADIOLOGYENDOCRINOLOGY, SECTION ON ORTHOPAEDICS and the SOCIETY FOR CHILD ABUSE AND NEGLECT, SECTION ON RADIOLOGY, SECTION ON

Hennrikus, and the AMERICAN ACADEMY OF PEDIATRICS COMMITTEE ON Emalee G. Flaherty, Jeannette M. Perez-Rossello, Michael A. Levine, William L.

Evaluating Children With Fractures for Child Physical Abuse

http://pediatrics.aappublications.org/content/133/2/e477located on the World Wide Web at:

The online version of this article, along with updated information and services, is

ISSN: 1073-0397. 60007. Copyright © 2014 by the American Academy of Pediatrics. All rights reserved. Print the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois,has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

by guest on June 15, 2018www.aappublications.org/newsDownloaded from