The Journal of the Arkansas Medical Society Med Journal Jan 2020 | Page 18
Table 3: Index Fracture
Characteristics
Fracture Type N=276 (%)
Skull 166 (60.1)
Femur 43 (15.6)
Humerus 34 (12.3)
Rib 10 (3.6)
Clavicle 10 (3.6)
Radius 8 (2.9)
Ulna 4 (1.4)
Tibia 10 (3.6)
Fibula 2 (0.7)
Scapula 1 (0.4)
abuse status change from no suspicion to suspi-
cion after the pathway was completed. Addition-
al fractures were identified in only three children
(1.8 %) with a final diagnosis of no suspicion of
abuse while additional fractures were identified
in 29 (27%) of those with a final diagnosis of sus-
picion of abuse (p < .0001).
Bruising at any site was found to be a posi-
tive predictor for final suspicion of abuse (OR
3.022; 95% CI 1.589-5.750; p = 0.0007). Age of 9-12
months was found to be a negative predictor for
suspicion of abuse (OR 0.276; 95% CI 0.116-0.659;
p = 0.0047). Type of insurance was found to have
no relationship to a final suspicion of abuse (OR
1.34; 95% CI 0.514-3.503; p = 0.0561). When skull
fracture was compared to all other types of index
fractures, it was found to be a negative predictor
for suspicion of abuse (OR 0.386; 95% CI 0.223-
0.668; p = 0.0007). There was one case (0.3%) of
bone disease detected during the evaluation. The
mother of this patient was known to have Osteo-
genesis Imperfecta. There were no cases in which
isolated lab abnormalities resulted in a suspicion
of abuse.
Discussion and Conclusions
Use of the infant skeletal trauma pathway re-
sulted in the identification of additional fractures
in 11.6 % of our cases. This is similar to the yield
from other studies using a pre-test suspicion of
abuse in order to complete a skeletal survey. 5 If
pre-test suspicion for abuse is an effective strate-
gy in identifying which cases to evaluate, a higher
yield of positive skeletal surveys in the “suspi-
cious” group would be expected. Our results sup-
port that infants presenting with any fracture are
a specifically high-risk group even if there are no
other indicators for abuse on initial evaluation.
166 • The Journal of the Arkansas Medical Society
In 22% of the cases, there was a change in final
determination of suspicion of abuse, with 18%
from suspicion to no suspicion of abuse and 4%
from no suspicion to suspicion of abuse. There-
fore, use of this objective clinical pathway may
have utility in both “ruling-in” and “ruling-out”
a suspicion of abuse. However, it should be em-
phasized that a negative skeletal survey does
not eliminate the possibility for abuse and pro-
viders should not be falsely reassured by a neg-
ative skeletal survey if the index injury alone is
suspicious. The 4% of cases in which the clin-
ical impression changed from no suspicion to
suspicion of abuse represent possible missed
opportunities to identify abuse if the decision
to obtain a skeletal survey were based solely on
abuse-suspicion. Type of insurance was not as-
sociated to final suspicion for abuse and using an
objective pathway may reduce bias in evaluation.
Bruising at any site was predictive of final sus-
picion of abuse. This reinforces the need to have a
low threshold to evaluate infants, with bruising as
potential “sentinel injuries” with specific empha-
sis on bruises on the torso, neck, face, head and
ears. 4 Infants are at particularly high risk to have
occult injuries related to physical abuse since they
do not localize pain well on exam making it diffi-
cult to distinguish between a “fussy” infant and
one who has been injured. Additionally, the types
of events that are often inflicted by a frustrated
caregiver include forceful squeezing of the chest
and yanking/jerking of extremities which can re-
sult in rib and metaphyseal fractures, which do
not typically have overlying deformity or bruising.
A thorough evaluation at the time an infant pres-
ents with what may be a subtle “sentinel” injury
provides an opportunity to recognize and address
a situation in which the child will be at risk for fur-
ther and possibly more serious injury if returned
to the same environment without intervention. 10
The study was limited by its retrospective
nature, so it was not possible to identify the spe-
cific factors that went into initial and final deter-
mination of suspicion of abuse versus no abuse.
Skeletal survey results may not have been the
only factor that could have been considered. We
also do not know the final outcome of investi-
gations, and clinical suspicion may not accu-
rately reflect the final determination of abuse or
accidental injury. There was no comparison of
the 9-12-month age group to younger infants in
terms of skeletal survey yield because of the low
numbers in each group.
Adaptation of a skeletal survey pathway for
particular injuries, as opposed to suspicion for
abuse, has the potential to reduce bias during
evaluation and to increase identification of phys-
ical abuse. Given that there was a positive skele-
tal survey for more than one of every 10 infants
evaluated with a fracture at ACH and 4% of cases
changed from a determination of not suspicious
to suspicious after the skeletal survey was ob-
tained, our results support the current practice
of obtaining a skeletal survey in all infants with a
fracture regardless of suspicion for abuse.
References
1. Kemp A, Dunstan F, Harrison S, et al. Patterns
of skeletal fractures in child abuse: system-
atic review. BMJ 2008;337:a1518.
2. Leventhal J, Martin K, Asnes A. Incidence
of fractures attributable to abuse in young
hospitalized children: Results from anal-
ysis of United States database. Pediatrics
2008;122:599-604.
3. Flaherty E, Perez-Rosello J, Levine M, et al.
Evaluating children with fractures for child
physical abuse. Pediatrics 2014;133:e477.
4. Glick JC, Lorand MA, Bilka KR. Physical
abuse of children. Pediatrics in Review
2016;37(4):146-156.
5. Duffy S, Squires J, Fromkin J, Berger R. Use
of skeletal surveys to evaluate for physical
abuse: analysis of 703 consecutive skeletal
surveys. Pediatrics 2011;127:e47-52.
6. Wood J, Feudtner C, Medina S, et al. Varia-
tion in occult injury screening for children
with suspected abuse in selected US chil-
dren’s hospitals. Pediatrics 2012;130:853.
7. Lane W, Rubin D, Monteith R, et al. Ra-
cial differences in the evaluation of pe-
diatric fractures for physical abuse. 2002
JAMA;288:1603-1609.
8. Higginbotham N, Lawson K, Gettig K, et al.
Utility of a child abuse screening guideline in
an urban pediatric emergency department. J
Trauma Acute Care Surg 2014;76:871-877.
9. Louwers E, Korfage I, Affourtit M, et al. Ef-
fects of systematic screening and detection
of child abuse in emergency departments.
Pediatrics 2012;130:457.
10. Deans K, Thackeray J, Askegard-Giesmann,
et al. Mortality increases with recurrent epi-
sodes of nonaccidental trauma in children. J
Trauma Acute Care Surg 2013;75:161-165. •
www.ArkMed.org