ORIGINAL ARTICLE
Missed Opportunities to Diagnose Child Physical Abuse Elizabeth L. Thorpe, MD,* Noel S. Zuckerbraun, MD, MPH,* Jennifer E. Wolford, DO, MPH,† and Rachel P. Berger, MD, MPH† Objectives: This study aimed to determine the incidence of missed opportunities to diagnose abuse in a cohort of children with healing abusive fractures and to identify patterns present during previous medical visits, which could lead to an earlier diagnosis of abuse. Methods: This is a retrospective descriptive study of a 7-year consecutive sample of children diagnosed with child abuse at a single children’s hospital. Children who had a healing fracture diagnosed on skeletal survey and a diagnosis of child abuse were included. We further collected data for the medical visit that lead to the diagnosis of child abuse and any previous medical visits that the subjects had during the 6 months preceding the diagnosis of abuse. All previous visits were classified as either a potential missed opportunity to diagnose abuse or as an unrelated previous visit, and the differences were analyzed. Results: Median age at time of abuse diagnosis was 3.9 months. Fortyeight percent (37/77) of the subjects had at least 1 previous visit, and 33% (25/77) of those had at least 1 missed previous visit. Multiple missed previous visits for the same symptoms were recorded in 7 (25%) of these patients. The most common reason for presentation at missed previous visit was a physical examination sign suggestive of trauma (ie, bruising, swelling). Missed previous visits occurred across all care settings. Conclusions: One-third of young children with healing abusive fractures had previous medical visits where the diagnosis of abuse was not recognized. These children most commonly had signs of trauma on physical examination at the previous visits. Key Words: abuse, skeletal survey, healing fractures (Pediatr Emer Care 2014;30: 771–776)
M
ore than 120,000 children were victims of child physical abuse in the United States in 2010.1 Child physical abuse can be difficult to diagnose, particularly in infants and young children.2,3 It is estimated that between 8% and 31% of children diagnosed with physical abuse have been previously evaluated by a medical professional.4–7 A previous study determined that children who died of child abuse had been evaluated by a physician during the year preceding their fatal abuse.4 Likewise, children with abusive head trauma have been shown to have had previous recent evaluations by physicians; however, abuse was not recognized.7 Although these studies do not clearly determine if these previous visits represented missed opportunities to diagnose abuse, the presenting complaint frequently included signs and symptoms of injury such as facial bruising and fussiness, which
From the Divisions of *Pediatric Emergency Medicine, and †Child Advocacy Department of Pediatrics, Children’s Hospital of Pittsburgh, University of Pittsburgh Medical Center, Pittsburgh, PA. Disclosure: The authors declare no conflict of interest. Reprints: Rachel Berger, MD, MPH, Children’s Hospital of Pittsburgh, UPMC Child Advocacy Center, 4401 Penn Avenue, Floor 2 Pittsburgh, PA 15224 (e‐mail: [email protected]). This study was supported by the Summer Student Research Training Program of Pediatric Clinical and Translational Research Center, Children’s Hospital of Pittsburgh, University of Pittsburgh Medical Center. Copyright © 2014 by Lippincott Williams & Wilkins ISSN: 0749-5161
Pediatric Emergency Care • Volume 30, Number 11, November 2014
can suggest abuse. Tools for early diagnosis of child abuse, well before brain injury or deaths occur, are needed. Among children identified with fractures on skeletal survey, a large proportion (79%) have a healing fracture.8 As the presence of healing fracture(s) demonstrates previous trauma, this suggests that there may have been a previous medical encounter at which time abuse might have been recognized. An opportunity for an earlier diagnosis of abuse may exist if we can identify patterns associated with previous visits in patients diagnosed with abusive trauma. We chose to focus our study on children diagnosed with abuse who had healing fractures on skeletal survey. The primary objective of this study was to identify the incidence of previous examinations by a medical provider in a cohort of children diagnosed with healing abusive fracture(s). The secondary objective of this study was to identify factors present at the previous examinations that may increase the likelihood of a diagnosis of physical abuse.
METHODS Study Design This was a retrospective, descriptive chart review approved by the University of Pittsburgh Institutional Review Board with a waiver of informed consent requirements. Children were included in the study if (1) they underwent a skeletal survey because of concern for physical abuse at the Children’s Hospital of Pittsburgh of the University of Pittsburgh Medical Center (UPMC) between April 1, 2002, and March 31, 2009; (2) the skeletal survey demonstrated at least one healing fracture; and (3) they were given a diagnosis of physical abuse. Healing fractures were identified based on the attending radiologist report. Skull fractures, although recognized as difficult to date, for this article, were defined as healing if they did not have associated soft tissue swelling. The skeletal surveys performed at the Children’s Hospital of Pittsburgh of UPMC meets the AAP guidelines for skeletal survey.9 The diagnosis of abuse was made by the Children’s Hospital of Pittsburgh of UPMC Child Protection Team. The use of the expertise of a child protection team to determine whether an injury is attributable to abuse is a commonly used standard.7,8,10
Data Collection All subjects were patients at the Children’s Hospital of Pittsburgh of UPMC, which is a large referral center for Pennsylvania, West Virginia, eastern Ohio, and southern New York and has an electronic medical record that includes information from associated community hospitals and primary care offices. Each subject was interviewed by the child protection team at the time of abuse diagnosis. The child protection team routinely collects a detailed medical history, including previous medical encounters, from caretakers and primary care physicians, and reviews the medical record. Each subject’s medical record was reviewed for the following information related to the visit at which the skeletal survey was www.pec-online.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
771
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Thorpe et al
performed: sex, age at the time of skeletal survey, reason for the presentation to the study hospital, general appearance and dermatologic examination, skeletal survey and head computed tomography (CT) results, subsequent clinical course, and presence of previous medical visits. Race and socioeconomic data were not collected.
making the determination of a missed or unrelated previous visit. P < 0.05 and a 95% confidence interval were considered significant.
Data Collected for Visit at Which Abuse Was Diagnosed
During the 7-year study period, 1466 children underwent a skeletal survey for concern for physical abuse (Fig. 1); 100 (6.8%) had healing fractures. Of the 100 subjects with healing fractures, 77% (77/100) were diagnosed with abuse and were therefore eligible for the current study. Of the 77 patients included in the study, 37 patients (48%) had previous visits, of whom 25 patients (32%) had missed opportunities for the diagnosis of child abuse. The median age of the 77 subjects at the time of abuse diagnosis was 3.9 months (range, 0.5–70.3 months). Eighty-eight percent were less than 2 years of age. Fifty-one percent were male. Thirty percent of the subjects required admission to an intensive care unit at the time the abuse was diagnosed; 60% were admitted to a surgical or medical floor, and 10% were discharged from the emergency department. Ten percent (8/77) of the study subjects died in the intensive care unit.
We collected the following variables: reason for presentation, general appearance, dermatologic examination, and radiological studies. The variable “reason for presentation to the study hospital” was developed during an iterative process that has been previously described and validated.8 The 9 choices for “reason for presentation to the study hospital” were history of trauma, signs suggestive of trauma (eg, bruising), request to evaluate for abuse, nonspecific symptoms (eg, irritability, vomiting), apparent life-threatening event, seizure, increased occipitofrontal head circumference, respiratory distress, and other. Data about each subject’s general appearance and dermatologic examination were also collected. General appearance was categorized as “well appearing,” “fussy but consolable,” “fussy, not consolable,” or “severe neurologic compromise. ” Dermatologic examination was classified as “normal,” “bruising,” “swelling,” “both bruising and swelling,” or “other traumatic abnormality” (eg, torn frenulum). Results of the skeletal survey were classified as “normal” or “abnormal.” When fractures were identified, they were classified by location (rib, upper extremity, lower extremity, clavicle, or skull) and by type (healing, acute, or both). Head CT results were classified as “normal,” “skull fracture,” “intracranial hemorrhage,” or “both skull fracture and intracranial hemorrhage.”
Data Collection for Previous Visits A previous visit was defined as any visit to a health care provider other than a well child care visit within the 6 months (180 days) before the diagnosis of abuse. Previous studies looking at missed abuse have included health care visits up to 365 days before the abuse visit, with most studies using 160 to 190 days as the upper limit.4,6,7 Identifying previous visits by using the child protection team history in combination with the medical record of a large tertiary care children’s hospital has also been used in previous studies.6,7 The following data were collected for each previous visit: time between previous visit and abuse diagnosis, visit location (primary care physician [PCP], general emergency department [GED], pediatric emergency department [PED], or subspecialist), reason for presentation (the same selection of 9 validated reasons used earlier), general appearance, dermatologic examination, visit interventions (laboratory data, imaging, or prescriptions), and discharge diagnosis.
Classification of “Missed Visits”
RESULTS Demographic Features
Clinical Characteristics at the Time Abuse Was Diagnosed The reasons for presentation to the hospital at the time of abuse diagnosis were signs suggestive of trauma (43%), history of trauma (20%), nonspecific symptoms (10%), request to evaluate for abuse (9%), apparent life-threatening event (9%), seizure (8%), and increased occipitofrontal head circumference (1%). On general examination at the time of abuse diagnosis, 46% (35/77) of the subjects were well appearing, 27% (21/77) were fussy but consolable, 9% (7/77) were fussy and not consolable, and 18% (14/77) had severe neurologic compromise. On dermatologic examination, 74% (57/77) had signs of trauma; 33% (25/77) had bruising, 27% (21/77) had swelling, and 9% (7/77) had bruising and swelling. The remaining 5% (4/77) had epistaxis, a torn frenulum, palpable chest wall crepitus, and mouth bleeding of unclear etiology.
Skeletal Survey Results Of the 77 subjects, 49 (64%) had only healing fractures, and 28 (36%) had both acute and healing fractures. Thirty-nine percent of the subjects had healing fractures in more than 1 body location. Healing fractures occurred in the following locations: rib (61%), lower extremity (40%), upper extremity (30%), and skull (25%). Of the 28 subjects with acute fractures, 64% were lower extremity fractures, 43% were skull fractures, 28% were upper extremity fractures, 14% were rib fractures, and 7% were clavicle fractures. Percentages add up to greater than 100 for both healing and acute fractures because many patients had more than 1 type of fracture.
All previous visits were reviewed and classified independently by 3 authors (E.L.T., R.P.B. and N.S.Z.) to determine whether they were a “missed previous visit” or an “unrelated previous visit.” A “missed previous visit” was defined as a “missed or probable missed diagnosis either of an abusive fracture or of another type of physical abuse.” All visits that were not classified as a “missed previous visit” were classified as “unrelated previous visit.”
Eighty-six percent (66/77) of the subjects had a head CT performed at the time of abuse diagnosis; 48% (32/66) had a normal head CT finding, 26% (17/66) had intracranial hemorrhage, 18% (12/66) had a skull fracture, and 8% (5/66) had both hemorrhage and skull fracture.
Statistical Analysis
Previous Visits by Subject
SPSS 19.0 (SPSS, Chicago, IL) was used for all analyses. Descriptive statistics were used to describe the study population. A κ statistic was performed to assess interrater reliability in
Previous visits were identified in 48% (37/77) of the subjects, and 20% (16/77) had 2 or more previous visits. Previous visits were classified as “missed” or “unrelated” (Fig. 2).
772
www.pec-online.com
Head CT
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Missed Opportunities to Diagnose Physical Abuse
FIGURE 1. Patient inclusion/exclusion and missed visits by number of subjects.
Thirty-three percent (25/77) of the subjects had at least one missed previous visit (Tables 1 and 2). The mean (SD) time between the missed previous visit and the time of abuse diagnosis was 24.0 (24.5) days, with a range of 1 to 90 days. The κ statistic for assessment of interrater reliability for assessing visits as missed or unrelated was 0.94, demonstrating strong agreement.
Location of Previous Visits As some subjects had more than 1 previous visit, a total of 74 previous visits exist for the 37 subjects. Of the 74 previous visits, 57% (42) were classified as “missed previous visits” and 43% (32) were classified as “unrelated previous visits.”
FIGURE 2. Missed and unrelated visits by number of visits. © 2014 Lippincott Williams & Wilkins
Of the 42 missed previous visits, 43% (18) occurred at a PCP, 36% (15) at a GED, 17% (7) at a PED, and 5% (3) at a subspecialist. The most common reason for presentation for the missed previous visits was a sign on physical examination suggestive of trauma (ie, bruising, swelling) (41%, 17/42). Other reasons included nonspecific symptoms (31%), history of trauma (14%), request to evaluate for abuse (5%), respiratory distress (5%), increased occipitofrontal head circumference (2%), and other (2%). General appearance was recorded for 45% (33/74) of the previous visits. General appearance was compared between missed and unrelated previous visits. For 60% (6/10) of missed previous visits, children were documented to be “well appearing” compared with 83% (19/23) of children with unrelated previous visits. There was no statistical difference between groups. Dermatologic examination was recorded for 50% (37/74) of the previous visits. Comparison was made between missed and unrelated previous visits with regard to dermatologic examination. For 40% (6/15) of missed previous visits, children were documented to have a normal dermatologic examination finding compared with 95% (21/22) of children with unrelated previous visits. There was no statistical difference between groups. Two children (subjects 6 and 7, Table 1) with missed previous visits subsequently died. Subject 6 was seen at a PED for vomiting without diarrhea, was given intravenous fluids, and was discharged. Nineteen days later, the subject presented in cardiac arrest. Head CT showed subdural hemorrhages, and postmortem skeletal survey demonstrated multiple healing fractures. Subject 7 was seen at 2 months of age by the PCP for subconjunctival hemorrhages and was diagnosed with constipation and www.pec-online.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
773
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Thorpe et al
TABLE 1. Subjects With 1 Missed PV, n = 17
ID Age, mo
Days After Visit First PV Location
Presentation
1
10
First PV 9 days
PCP PCP
Ecchymosis on chest Ecchymosis, not using arm
2
5
First PV 1 day
PED PED
3
0.6
First PV
GED
25 d
GED
Cough, congestion Call back for final read on chest x-ray Respiratory distress, absence of rales, rhonchi, or wheezes, father intoxicated Fall off couch while home alone with father Unwitnessed fall, eye ecchymosis and eyebrow laceration Soft tissue swelling of scalp Developmentally delayed, decreased arm movement, no mechanism Referred by PCP, still lack of mechanism for fx Vomiting, no diarrhea Cardiac arrest
4
3
First PV
GED
5
36
90 d First PV
PED PCP
1 day
Ortho
6
6
First PV 19 d
PED PED
7
1.8
8
1.3
First PV 4d First PV 2d
PCP PED GED GED
9
3
First PV
GED
2d
GED
10
5
First PV 2d
PCP PED
11
1.8
First PV
PCP
22 d
GED
First PV
PED
42 d
PED
First PV
GED
1d
PCP
First PV 2d First PV 1d First PV 3d First PV 10 d
PCP GED PCP PED PCP PCP PCP PED
12
21
13
12
14
3
15
8
16
1.8
17
1.8
Subconjunctival hemorrhages Seizures, intubated on arrival Soft tissue swelling of scalp Soft tissue swelling of scalp, ear ecchymosis Ecchymosis to forehead, cheek, abdomen; irritability Call back for final read on chest x-ray Soft tissue swelling of scalp Soft tissue swelling of scalp, irritability Subconjunctival hemorrhages, ecchymosis on wrist and jaw Arm swelling Eye swelling and bruising, unwitnessed fall Leg swelling, no history of trauma Developmentally delayed, subconjunctival hemorrhage, blood in ear Sibling disclosed abuse during visit Irritability, vomiting alone Apparent life threatening event Irritability Swelling of arm, bruise on cheek Irritability Swollen thigh Irritability, abdominal pain, no fever Seizures, facial bruising
Evaluation
Diagnosis
None Acute rib fx, acute and healing extremity fxs Chest x-ray, preliminary read negative Healing rib fxs RSV washing negative
Unknown Child abuse Viral illness Child abuse RSV bronchiolitis
Acute femur fx, acute tibia fx, skull fx, healing rib fxs Laceration repair
Child abuse
Skull fx and healing extremity fxs Bilateral acute wrist fxs, referral to orthopedics Bilateral wrist fxs, healing femur fx, ecchymosis, bite mark Normal urine, intravenous fluids Acute and healing extremity fxs, subdural hemorrhage None Healing rib fx, subdural hemorrhage None Skull fx, intracranial hemorrhage, multiple healing rib fx Chest x-ray, preliminary negative
Child abuse Fracture
Laceration
Healing rib fxs, skull fx and subdural hemorrhage None Healing rib fxs, skull fx, healing extremity fx Negative head CT and negative bleeding workup Acute humerus fx, healing tibia fx, healing rib fx CT of orbits negative for fracture, admitted for intravenous antibiotics Acute tibia fx and healing tibia fx None
Healing tibia fx
Child abuse Gastroenteritis Child abuse/death Constipation Child abuse/death Unknown Child abuse Unknown Child abuse Reflux Child abuse Unknown Child abuse Periorbital cellulitis Child abuse Unknown
Child abuse
None Unknown Healing rib fx and subdural hemorrhage Child abuse None Unknown Healing fx of humerus Child abuse Simethicone “Gassy” infant Healing femur fx Child abuse Antibiotics and Zantac Acute otitis media/ reflux Healing rib fx, healing clavicle fxs, Child abuse healing lower extremity fxs
Fx indicates fracture; Ortho, orthopedic outpatient visit; PV, previous visit; RSV, respiratory syncytial virus.
774
www.pec-online.com
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Missed Opportunities to Diagnose Physical Abuse
TABLE 2. Patients With 2 or More Missed PVs, n = 8
ID Age, mo
Days After PV First PV Location
Presentation
18
1.5
First PV 7d 8d 23 d 30 d
PCP GED GED PCP GED
Ecchymosis on hands Unwitnessed fall, irritability Chest crepitus, irritability, fever Soft tissue swelling to head Irritability, fever
19
0.75
First PV 53 d
PED PED
54 d 55 d First PV 4d 24 d
GED PCP GED PCP GED
First PV 7d 30 d 37 d 44 d First PV 1d 44 d
PCP PCP PCP PCP PCP GED Ortho PCP
Irritability, vomiting alone Irritability, seizure vs spell, vomiting alone Irritability, vomiting alone Irritable, OFC 97% Holding leg “funny” Holding leg “funny” Coma, respiratory distress, facial bruising Irritability Irritability Soft tissue swelling to chest Soft tissue swelling to chest Soft tissue swelling to chest Unwitnessed fall, not using arm Supracondylar fx Swollen knee, no mechanism
55 d
Ortho
Healing tibia fx
First PV
GED
42 d
GED
44 d
PED
Right arm swelling, no mechanism Right arm swelling, no mechanism, no fever Seeking second opinion on arm swelling, no fever
First PV 2d
GED GED
4d 12 d First PV 53 d
GED PED PED Ortho
Fall from parent’s lap Called back for possible missed fx on SS Called back again Referral for CPT opinion Swollen arm, delay in seeking care Hip fracture, no mechanism
77 d 82 d
ID PED
Osteomyelitis PICC line dislodged
20
1
21
1
22
22
23
24
25
30
2.5
24
Evaluation
Diagnosis
None Normal abdominal x-ray finding Normal blood and urine testing finding None Healing rib fxs, healing extremity fxs, skull fxs Normal abdominal x-ray finding None
Unknown Colic Viral syndrome Normal head Child abuse
None Healing rib fxs, subdural hemorrhage Normal lower extremity x-ray findings US of hips normal Healing rib fxs, healing extremity fxs, skull fx and subdural hemorrhage None None None None Healing rib fxs, healing extremity fxs Supracondylar fx, referred to ortho Treated with sling Healing fx of tibia, referred to orthopedic surgeon Healing supracondylar fx, healing radius fx, healing tibia fx, healing metatarsal fx Acute proximal humerus fx
Gastroenteritis Child abuse Unknown Unknown Child abuse
Plain films only, admitted for 24 h of intravenous antibiotics Bruises on body, acute fx through healing right arm fx, additional extremity healing fxs Skull fx, SS preliminary read negative Confusion over read on SS, discharge
Osteomyelitis
Repeat SS, persistent confusion, discharge Healing rib fx and skull fx Supracondylar fx Hip fracture, development of osteomyelitis in arm None Unexpected healing clavicle fx on chest x-ray
Gastroenteritis Gastroenteritis
Colic Reflux Unknown Unknown Child abuse Fx Fx Fx Child abuse Fx
Child abuse
Skull fx Unknown Unknown Child abuse Fracture Fracture/osteomyelitis Osteomyelitis Child abuse
CPT indicates child protection team; fx, fracture; ID, infectious disease outpatient PV; OFC, orbitofrontal cortex; Ortho, orthopedic outpatient PV; SS, skeletal survey; US, ultrasound.
prescribed karo syrup. Four days later, the subject presented with seizures and altered mental status. Evaluation demonstrated healing rib fractures and intracranial hemorrhage. This subject died 13 days later.
DISCUSSION This study represents the first evaluation of missed opportunities to diagnose physical abuse in children diagnosed with physical abuse and healing fractures. We show that one third of the subjects with abusive healing fractures had previous missed opportunities to diagnose abuse. These missed visits occurred in multiple locations © 2014 Lippincott Williams & Wilkins
(emergency departments, primary care offices, and subspecialty practices), and signs of trauma were often noted on physical examination. Thirty-three percent of the subjects in this cohort had a previous visit, which was felt to be a missed opportunity to identify abuse. This is consistent with and higher than other studies of possible missed abuse reported to date in the literature.4–7,11 Jenny et al7 demonstrated that 31% of children diagnosed with abusive head trauma had been seen by physicians after the abuse, but the diagnosis was not recognized. King et al showed that 30% of children who died of child abuse had health care visits for reasons www.pec-online.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
775
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Thorpe et al
other than routine well child care within the year before their death. In a study by Ravichandiran et al, 21% (54/258) of children referred to a child abuse team with suspected abusive fractures had had at least 1 previous physician visit, which in hindsight, had been a missed opportunity to identify the fracture and/or to identify that the fracture was the result of abuse. The most common reason for presentation at missed previous visits was a “sign suggestive of trauma,” specifically bruising or swelling (Tables 1 and 2). Similar to our study, several studies have shown that bruising and/or swelling in premobile children is concerning for abuse and should prompt complete evaluation for injuries.12–18 The 2007 AAP guidelines on the evaluation of child abuse recommend that physicians pay close attention to any bruises, burns, bites, or lacerations and specifically encourages the careful evaluation of skin injuries in infants who cannot cruise.12 Following these evidenced-based guidelines on the skin evaluation of young infants could substantially reduce additional abusive injuries to the same child. Of note, at the time of abuse diagnosis, less than half of subjects in our study were classified as well appearing although a majority had signs of trauma on examination. For missed visits, more than half of the patients were also felt to be well appearing. This demonstrates that children can be well appearing despite having abuse injuries and that clinicians need to maintain a high suspicion for abuse in all children, regardless of appearance. Of the 25 subjects with missed previous visits, 7 (subjects 1, 8, 10, 19, 20, 21, and 23 in Tables 1 and 2) had the same presentation on 2 or more occasions. Almost one quarter of the children having multiple missed opportunities to diagnose abuse combined with the most common missed presentation being one known to be associated with physical abuse suggests that education of providers regarding common signs and symptoms of child physical abuse alone has not yet been successful. Two patients who died had missed previous visits, and many of the patients with missed previous visits had additional morbidity as a result of the missed opportunity to diagnose abuse. In addition to the morbidity associated with healing fractures, more than one third of the patients were admitted to the intensive care unit, and 34% had an intracranial bleed at the time of abuse diagnosis. This underscores the importance of early detection of abuse in the prevention of the morbidity and mortality associated with repeated exposure to a violent environment. Some missed previous visits were for nonspecific symptoms such as vomiting, fever, and irritability. Although these symptoms are seen in a myriad of common benign pediatric conditions, they are also seen in abused children. Classifying visits for nonspecific symptoms as possible missed visits, in a cohort of abused children, is supported by the literature.4,7 The difficulty for clinicians lies in discerning when these symptoms indicate potentially serious pathology. Although all infants with nonspecific symptoms do not need to undergo a full evaluation for abuse, child abuse should be included in the differential diagnosis, particularly when the infant has had repeated medical visits for these symptoms.
Limitations This study has several limitations. First, data were collected retrospectively, and therefore, some variables such as “reason for presentation to the hospital” required assessment by the data extractor. This information could bias the assignment of previous visit categories (missed vs unrelated). We attempted to minimize this limitation in 2 ways. First, we used a single extractor for all data collection. Second, we had 3 authors separately categorize the extracted data, and the high ĸ value demonstrates internal consistency. Missing data may have also biased results. Although there were no missing data for the visits at which abuse was diagnosed,
776
www.pec-online.com
there were missing data for some previous visits. It is possible that more visits would have been classified as missed previous visits if more information were available, suggesting a possible underestimation of missed previous visits.
CONCLUSIONS In summary, our study found that nearly one third of children with abusive healing fractures had at least 1 previous visit in a variety of clinical settings, which may have been a missed opportunity to diagnose abuse. The most common reason for presentation at these previous visits was a “sign suggestive of trauma,” specifically bruising or swelling. Our study adds to the growing literature indicating that premobile infants with bruising and swelling should be carefully evaluated for physical abuse. Although abuse may be difficult to recognize in the patient with nonspecific symptoms such as vomiting, our data support previous literature demonstrating that physical abuse should remain on the differential, particularly in those with multiple visits. REFERENCES 1. U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau. Available at: https://www.acf.hhs.gov. Accessed November 5, 2012. 2. Alexander R, Crabbe L, Sato Y, et al. Serial abuse in children who are shaken. Am J Dis Child. 1990;144:58–60. 3. Taitz J, Moran K, O’Meara M. Long bone fractures in children under 3 years of age: is abuse being missed in emergency department presentations? J Paediatr Child Health. 2004;40:170–174. 4. King WK, Kiesel EL, Simon HK. Child abuse fatalities: are we missing opportunities for intervention? Pediatr Emerg Care. 2006;22:211–214. 5. Oral R, Blum KL, Johnson C. Fractures in young children: are physicians in the emergency department and orthopedic clinics adequately screening for possible abuse? Pediatr Emerg Care. 2003;19:148–153. 6. Ravichandiran N, Schuh S, Bejuk M, et al. Delayed identification of pediatric abuse-related fractures. Pediatrics. 2010;125:60–66. 7. Jenny C, Hymel KP, Ritzen A, et al. Analysis of missed cases of abusive head trauma. JAMA. 1999;281:621–626. 8. Duffy SO, Squires J, Fromkin JB, et al. Use of skeletal surveys to evaluate for physical abuse: analysis of 703 consecutive skeletal surveys. Pediatrics. 2011;127:e47–e52. 9. Diagnostic imaging of child abuse. Pediatrics. 2009;123:1430–1435. 10. Berger RP, Dulani T, Adelson PD, et al. Identification of inflicted traumatic brain injury in well-appearing infants using serum and cerebrospinal markers: a possible screening tool. Pediatrics. 2006;117:325–332. 11. Oral R, Yagmur F, Nashelsky M, et al. Fatal abusive head trauma cases: consequences of medical staff missing milder forms of physical abuse. Pediatr Emerg Care. 2008;24:816–821. 12. Kellogg ND. Evaluation of suspected child physical abuse. Pediatrics. 2007;119:1232–1241. 13. Feldman KW. The bruised premobile infant: should you evaluate further? Pediatr Emerg Care. 2009;25:37–39. 14. Sheets LK, Leach ME, Koszewski IJ, et al. Sentinel injuries in infants evaluated for child physical abuse. Pediatrics. 2013;131(4):701–707. 15. Pierce MC, Kaczor K, Aldridge S, et al. Bruising characteristics discriminating physical child abuse from accidental trauma. Pediatrics. 2010;125:67–74. 16. Pierce MC, Smith S, Kaczor K. Bruising in infants: those with a bruise may be abused. Pediatr Emerg Care. 2009;25:845–847. 17. Maguire S, Mann MK, Sibert J, et al. Are there patterns of bruising in childhood which are diagnostic or suggestive of abuse? A systematic review. Arch Dis Child. 2005;90:182–186. 18. Sugar NF, Taylor JA, Feldman KW. Bruises in infants and toddlers: those who don’t cruise rarely bruise. Puget Sound Pediatric Research Network. Arch Pediatr Adolesc Med. 1999;153:399–403.
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Missed Opportunities to Diagnose Child Physical Abuse Elizabeth L. Thorpe, MD,* Noel S. Zuckerbraun, MD, MPH,* Jennifer E. Wolford, DO, MPH,† and Rachel P. Berger, MD, MPH† Objectives: This study aimed to determine the incidence of missed opportunities to diagnose abuse in a cohort of children with healing abusive fractures and to identify patterns present during previous medical visits, which could lead to an earlier diagnosis of abuse. Methods: This is a retrospective descriptive study of a 7-year consecutive sample of children diagnosed with child abuse at a single children’s hospital. Children who had a healing fracture diagnosed on skeletal survey and a diagnosis of child abuse were included. We further collected data for the medical visit that lead to the diagnosis of child abuse and any previous medical visits that the subjects had during the 6 months preceding the diagnosis of abuse. All previous visits were classified as either a potential missed opportunity to diagnose abuse or as an unrelated previous visit, and the differences were analyzed. Results: Median age at time of abuse diagnosis was 3.9 months. Fortyeight percent (37/77) of the subjects had at least 1 previous visit, and 33% (25/77) of those had at least 1 missed previous visit. Multiple missed previous visits for the same symptoms were recorded in 7 (25%) of these patients. The most common reason for presentation at missed previous visit was a physical examination sign suggestive of trauma (ie, bruising, swelling). Missed previous visits occurred across all care settings. Conclusions: One-third of young children with healing abusive fractures had previous medical visits where the diagnosis of abuse was not recognized. These children most commonly had signs of trauma on physical examination at the previous visits. Key Words: abuse, skeletal survey, healing fractures (Pediatr Emer Care 2014;30: 771–776)
M
ore than 120,000 children were victims of child physical abuse in the United States in 2010.1 Child physical abuse can be difficult to diagnose, particularly in infants and young children.2,3 It is estimated that between 8% and 31% of children diagnosed with physical abuse have been previously evaluated by a medical professional.4–7 A previous study determined that children who died of child abuse had been evaluated by a physician during the year preceding their fatal abuse.4 Likewise, children with abusive head trauma have been shown to have had previous recent evaluations by physicians; however, abuse was not recognized.7 Although these studies do not clearly determine if these previous visits represented missed opportunities to diagnose abuse, the presenting complaint frequently included signs and symptoms of injury such as facial bruising and fussiness, which
From the Divisions of *Pediatric Emergency Medicine, and †Child Advocacy Department of Pediatrics, Children’s Hospital of Pittsburgh, University of Pittsburgh Medical Center, Pittsburgh, PA. Disclosure: The authors declare no conflict of interest. Reprints: Rachel Berger, MD, MPH, Children’s Hospital of Pittsburgh, UPMC Child Advocacy Center, 4401 Penn Avenue, Floor 2 Pittsburgh, PA 15224 (e‐mail: [email protected]). This study was supported by the Summer Student Research Training Program of Pediatric Clinical and Translational Research Center, Children’s Hospital of Pittsburgh, University of Pittsburgh Medical Center. Copyright © 2014 by Lippincott Williams & Wilkins ISSN: 0749-5161
Pediatric Emergency Care • Volume 30, Number 11, November 2014
can suggest abuse. Tools for early diagnosis of child abuse, well before brain injury or deaths occur, are needed. Among children identified with fractures on skeletal survey, a large proportion (79%) have a healing fracture.8 As the presence of healing fracture(s) demonstrates previous trauma, this suggests that there may have been a previous medical encounter at which time abuse might have been recognized. An opportunity for an earlier diagnosis of abuse may exist if we can identify patterns associated with previous visits in patients diagnosed with abusive trauma. We chose to focus our study on children diagnosed with abuse who had healing fractures on skeletal survey. The primary objective of this study was to identify the incidence of previous examinations by a medical provider in a cohort of children diagnosed with healing abusive fracture(s). The secondary objective of this study was to identify factors present at the previous examinations that may increase the likelihood of a diagnosis of physical abuse.
METHODS Study Design This was a retrospective, descriptive chart review approved by the University of Pittsburgh Institutional Review Board with a waiver of informed consent requirements. Children were included in the study if (1) they underwent a skeletal survey because of concern for physical abuse at the Children’s Hospital of Pittsburgh of the University of Pittsburgh Medical Center (UPMC) between April 1, 2002, and March 31, 2009; (2) the skeletal survey demonstrated at least one healing fracture; and (3) they were given a diagnosis of physical abuse. Healing fractures were identified based on the attending radiologist report. Skull fractures, although recognized as difficult to date, for this article, were defined as healing if they did not have associated soft tissue swelling. The skeletal surveys performed at the Children’s Hospital of Pittsburgh of UPMC meets the AAP guidelines for skeletal survey.9 The diagnosis of abuse was made by the Children’s Hospital of Pittsburgh of UPMC Child Protection Team. The use of the expertise of a child protection team to determine whether an injury is attributable to abuse is a commonly used standard.7,8,10
Data Collection All subjects were patients at the Children’s Hospital of Pittsburgh of UPMC, which is a large referral center for Pennsylvania, West Virginia, eastern Ohio, and southern New York and has an electronic medical record that includes information from associated community hospitals and primary care offices. Each subject was interviewed by the child protection team at the time of abuse diagnosis. The child protection team routinely collects a detailed medical history, including previous medical encounters, from caretakers and primary care physicians, and reviews the medical record. Each subject’s medical record was reviewed for the following information related to the visit at which the skeletal survey was www.pec-online.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
771
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Thorpe et al
performed: sex, age at the time of skeletal survey, reason for the presentation to the study hospital, general appearance and dermatologic examination, skeletal survey and head computed tomography (CT) results, subsequent clinical course, and presence of previous medical visits. Race and socioeconomic data were not collected.
making the determination of a missed or unrelated previous visit. P < 0.05 and a 95% confidence interval were considered significant.
Data Collected for Visit at Which Abuse Was Diagnosed
During the 7-year study period, 1466 children underwent a skeletal survey for concern for physical abuse (Fig. 1); 100 (6.8%) had healing fractures. Of the 100 subjects with healing fractures, 77% (77/100) were diagnosed with abuse and were therefore eligible for the current study. Of the 77 patients included in the study, 37 patients (48%) had previous visits, of whom 25 patients (32%) had missed opportunities for the diagnosis of child abuse. The median age of the 77 subjects at the time of abuse diagnosis was 3.9 months (range, 0.5–70.3 months). Eighty-eight percent were less than 2 years of age. Fifty-one percent were male. Thirty percent of the subjects required admission to an intensive care unit at the time the abuse was diagnosed; 60% were admitted to a surgical or medical floor, and 10% were discharged from the emergency department. Ten percent (8/77) of the study subjects died in the intensive care unit.
We collected the following variables: reason for presentation, general appearance, dermatologic examination, and radiological studies. The variable “reason for presentation to the study hospital” was developed during an iterative process that has been previously described and validated.8 The 9 choices for “reason for presentation to the study hospital” were history of trauma, signs suggestive of trauma (eg, bruising), request to evaluate for abuse, nonspecific symptoms (eg, irritability, vomiting), apparent life-threatening event, seizure, increased occipitofrontal head circumference, respiratory distress, and other. Data about each subject’s general appearance and dermatologic examination were also collected. General appearance was categorized as “well appearing,” “fussy but consolable,” “fussy, not consolable,” or “severe neurologic compromise. ” Dermatologic examination was classified as “normal,” “bruising,” “swelling,” “both bruising and swelling,” or “other traumatic abnormality” (eg, torn frenulum). Results of the skeletal survey were classified as “normal” or “abnormal.” When fractures were identified, they were classified by location (rib, upper extremity, lower extremity, clavicle, or skull) and by type (healing, acute, or both). Head CT results were classified as “normal,” “skull fracture,” “intracranial hemorrhage,” or “both skull fracture and intracranial hemorrhage.”
Data Collection for Previous Visits A previous visit was defined as any visit to a health care provider other than a well child care visit within the 6 months (180 days) before the diagnosis of abuse. Previous studies looking at missed abuse have included health care visits up to 365 days before the abuse visit, with most studies using 160 to 190 days as the upper limit.4,6,7 Identifying previous visits by using the child protection team history in combination with the medical record of a large tertiary care children’s hospital has also been used in previous studies.6,7 The following data were collected for each previous visit: time between previous visit and abuse diagnosis, visit location (primary care physician [PCP], general emergency department [GED], pediatric emergency department [PED], or subspecialist), reason for presentation (the same selection of 9 validated reasons used earlier), general appearance, dermatologic examination, visit interventions (laboratory data, imaging, or prescriptions), and discharge diagnosis.
Classification of “Missed Visits”
RESULTS Demographic Features
Clinical Characteristics at the Time Abuse Was Diagnosed The reasons for presentation to the hospital at the time of abuse diagnosis were signs suggestive of trauma (43%), history of trauma (20%), nonspecific symptoms (10%), request to evaluate for abuse (9%), apparent life-threatening event (9%), seizure (8%), and increased occipitofrontal head circumference (1%). On general examination at the time of abuse diagnosis, 46% (35/77) of the subjects were well appearing, 27% (21/77) were fussy but consolable, 9% (7/77) were fussy and not consolable, and 18% (14/77) had severe neurologic compromise. On dermatologic examination, 74% (57/77) had signs of trauma; 33% (25/77) had bruising, 27% (21/77) had swelling, and 9% (7/77) had bruising and swelling. The remaining 5% (4/77) had epistaxis, a torn frenulum, palpable chest wall crepitus, and mouth bleeding of unclear etiology.
Skeletal Survey Results Of the 77 subjects, 49 (64%) had only healing fractures, and 28 (36%) had both acute and healing fractures. Thirty-nine percent of the subjects had healing fractures in more than 1 body location. Healing fractures occurred in the following locations: rib (61%), lower extremity (40%), upper extremity (30%), and skull (25%). Of the 28 subjects with acute fractures, 64% were lower extremity fractures, 43% were skull fractures, 28% were upper extremity fractures, 14% were rib fractures, and 7% were clavicle fractures. Percentages add up to greater than 100 for both healing and acute fractures because many patients had more than 1 type of fracture.
All previous visits were reviewed and classified independently by 3 authors (E.L.T., R.P.B. and N.S.Z.) to determine whether they were a “missed previous visit” or an “unrelated previous visit.” A “missed previous visit” was defined as a “missed or probable missed diagnosis either of an abusive fracture or of another type of physical abuse.” All visits that were not classified as a “missed previous visit” were classified as “unrelated previous visit.”
Eighty-six percent (66/77) of the subjects had a head CT performed at the time of abuse diagnosis; 48% (32/66) had a normal head CT finding, 26% (17/66) had intracranial hemorrhage, 18% (12/66) had a skull fracture, and 8% (5/66) had both hemorrhage and skull fracture.
Statistical Analysis
Previous Visits by Subject
SPSS 19.0 (SPSS, Chicago, IL) was used for all analyses. Descriptive statistics were used to describe the study population. A κ statistic was performed to assess interrater reliability in
Previous visits were identified in 48% (37/77) of the subjects, and 20% (16/77) had 2 or more previous visits. Previous visits were classified as “missed” or “unrelated” (Fig. 2).
772
www.pec-online.com
Head CT
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Missed Opportunities to Diagnose Physical Abuse
FIGURE 1. Patient inclusion/exclusion and missed visits by number of subjects.
Thirty-three percent (25/77) of the subjects had at least one missed previous visit (Tables 1 and 2). The mean (SD) time between the missed previous visit and the time of abuse diagnosis was 24.0 (24.5) days, with a range of 1 to 90 days. The κ statistic for assessment of interrater reliability for assessing visits as missed or unrelated was 0.94, demonstrating strong agreement.
Location of Previous Visits As some subjects had more than 1 previous visit, a total of 74 previous visits exist for the 37 subjects. Of the 74 previous visits, 57% (42) were classified as “missed previous visits” and 43% (32) were classified as “unrelated previous visits.”
FIGURE 2. Missed and unrelated visits by number of visits. © 2014 Lippincott Williams & Wilkins
Of the 42 missed previous visits, 43% (18) occurred at a PCP, 36% (15) at a GED, 17% (7) at a PED, and 5% (3) at a subspecialist. The most common reason for presentation for the missed previous visits was a sign on physical examination suggestive of trauma (ie, bruising, swelling) (41%, 17/42). Other reasons included nonspecific symptoms (31%), history of trauma (14%), request to evaluate for abuse (5%), respiratory distress (5%), increased occipitofrontal head circumference (2%), and other (2%). General appearance was recorded for 45% (33/74) of the previous visits. General appearance was compared between missed and unrelated previous visits. For 60% (6/10) of missed previous visits, children were documented to be “well appearing” compared with 83% (19/23) of children with unrelated previous visits. There was no statistical difference between groups. Dermatologic examination was recorded for 50% (37/74) of the previous visits. Comparison was made between missed and unrelated previous visits with regard to dermatologic examination. For 40% (6/15) of missed previous visits, children were documented to have a normal dermatologic examination finding compared with 95% (21/22) of children with unrelated previous visits. There was no statistical difference between groups. Two children (subjects 6 and 7, Table 1) with missed previous visits subsequently died. Subject 6 was seen at a PED for vomiting without diarrhea, was given intravenous fluids, and was discharged. Nineteen days later, the subject presented in cardiac arrest. Head CT showed subdural hemorrhages, and postmortem skeletal survey demonstrated multiple healing fractures. Subject 7 was seen at 2 months of age by the PCP for subconjunctival hemorrhages and was diagnosed with constipation and www.pec-online.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
773
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Thorpe et al
TABLE 1. Subjects With 1 Missed PV, n = 17
ID Age, mo
Days After Visit First PV Location
Presentation
1
10
First PV 9 days
PCP PCP
Ecchymosis on chest Ecchymosis, not using arm
2
5
First PV 1 day
PED PED
3
0.6
First PV
GED
25 d
GED
Cough, congestion Call back for final read on chest x-ray Respiratory distress, absence of rales, rhonchi, or wheezes, father intoxicated Fall off couch while home alone with father Unwitnessed fall, eye ecchymosis and eyebrow laceration Soft tissue swelling of scalp Developmentally delayed, decreased arm movement, no mechanism Referred by PCP, still lack of mechanism for fx Vomiting, no diarrhea Cardiac arrest
4
3
First PV
GED
5
36
90 d First PV
PED PCP
1 day
Ortho
6
6
First PV 19 d
PED PED
7
1.8
8
1.3
First PV 4d First PV 2d
PCP PED GED GED
9
3
First PV
GED
2d
GED
10
5
First PV 2d
PCP PED
11
1.8
First PV
PCP
22 d
GED
First PV
PED
42 d
PED
First PV
GED
1d
PCP
First PV 2d First PV 1d First PV 3d First PV 10 d
PCP GED PCP PED PCP PCP PCP PED
12
21
13
12
14
3
15
8
16
1.8
17
1.8
Subconjunctival hemorrhages Seizures, intubated on arrival Soft tissue swelling of scalp Soft tissue swelling of scalp, ear ecchymosis Ecchymosis to forehead, cheek, abdomen; irritability Call back for final read on chest x-ray Soft tissue swelling of scalp Soft tissue swelling of scalp, irritability Subconjunctival hemorrhages, ecchymosis on wrist and jaw Arm swelling Eye swelling and bruising, unwitnessed fall Leg swelling, no history of trauma Developmentally delayed, subconjunctival hemorrhage, blood in ear Sibling disclosed abuse during visit Irritability, vomiting alone Apparent life threatening event Irritability Swelling of arm, bruise on cheek Irritability Swollen thigh Irritability, abdominal pain, no fever Seizures, facial bruising
Evaluation
Diagnosis
None Acute rib fx, acute and healing extremity fxs Chest x-ray, preliminary read negative Healing rib fxs RSV washing negative
Unknown Child abuse Viral illness Child abuse RSV bronchiolitis
Acute femur fx, acute tibia fx, skull fx, healing rib fxs Laceration repair
Child abuse
Skull fx and healing extremity fxs Bilateral acute wrist fxs, referral to orthopedics Bilateral wrist fxs, healing femur fx, ecchymosis, bite mark Normal urine, intravenous fluids Acute and healing extremity fxs, subdural hemorrhage None Healing rib fx, subdural hemorrhage None Skull fx, intracranial hemorrhage, multiple healing rib fx Chest x-ray, preliminary negative
Child abuse Fracture
Laceration
Healing rib fxs, skull fx and subdural hemorrhage None Healing rib fxs, skull fx, healing extremity fx Negative head CT and negative bleeding workup Acute humerus fx, healing tibia fx, healing rib fx CT of orbits negative for fracture, admitted for intravenous antibiotics Acute tibia fx and healing tibia fx None
Healing tibia fx
Child abuse Gastroenteritis Child abuse/death Constipation Child abuse/death Unknown Child abuse Unknown Child abuse Reflux Child abuse Unknown Child abuse Periorbital cellulitis Child abuse Unknown
Child abuse
None Unknown Healing rib fx and subdural hemorrhage Child abuse None Unknown Healing fx of humerus Child abuse Simethicone “Gassy” infant Healing femur fx Child abuse Antibiotics and Zantac Acute otitis media/ reflux Healing rib fx, healing clavicle fxs, Child abuse healing lower extremity fxs
Fx indicates fracture; Ortho, orthopedic outpatient visit; PV, previous visit; RSV, respiratory syncytial virus.
774
www.pec-online.com
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Missed Opportunities to Diagnose Physical Abuse
TABLE 2. Patients With 2 or More Missed PVs, n = 8
ID Age, mo
Days After PV First PV Location
Presentation
18
1.5
First PV 7d 8d 23 d 30 d
PCP GED GED PCP GED
Ecchymosis on hands Unwitnessed fall, irritability Chest crepitus, irritability, fever Soft tissue swelling to head Irritability, fever
19
0.75
First PV 53 d
PED PED
54 d 55 d First PV 4d 24 d
GED PCP GED PCP GED
First PV 7d 30 d 37 d 44 d First PV 1d 44 d
PCP PCP PCP PCP PCP GED Ortho PCP
Irritability, vomiting alone Irritability, seizure vs spell, vomiting alone Irritability, vomiting alone Irritable, OFC 97% Holding leg “funny” Holding leg “funny” Coma, respiratory distress, facial bruising Irritability Irritability Soft tissue swelling to chest Soft tissue swelling to chest Soft tissue swelling to chest Unwitnessed fall, not using arm Supracondylar fx Swollen knee, no mechanism
55 d
Ortho
Healing tibia fx
First PV
GED
42 d
GED
44 d
PED
Right arm swelling, no mechanism Right arm swelling, no mechanism, no fever Seeking second opinion on arm swelling, no fever
First PV 2d
GED GED
4d 12 d First PV 53 d
GED PED PED Ortho
Fall from parent’s lap Called back for possible missed fx on SS Called back again Referral for CPT opinion Swollen arm, delay in seeking care Hip fracture, no mechanism
77 d 82 d
ID PED
Osteomyelitis PICC line dislodged
20
1
21
1
22
22
23
24
25
30
2.5
24
Evaluation
Diagnosis
None Normal abdominal x-ray finding Normal blood and urine testing finding None Healing rib fxs, healing extremity fxs, skull fxs Normal abdominal x-ray finding None
Unknown Colic Viral syndrome Normal head Child abuse
None Healing rib fxs, subdural hemorrhage Normal lower extremity x-ray findings US of hips normal Healing rib fxs, healing extremity fxs, skull fx and subdural hemorrhage None None None None Healing rib fxs, healing extremity fxs Supracondylar fx, referred to ortho Treated with sling Healing fx of tibia, referred to orthopedic surgeon Healing supracondylar fx, healing radius fx, healing tibia fx, healing metatarsal fx Acute proximal humerus fx
Gastroenteritis Child abuse Unknown Unknown Child abuse
Plain films only, admitted for 24 h of intravenous antibiotics Bruises on body, acute fx through healing right arm fx, additional extremity healing fxs Skull fx, SS preliminary read negative Confusion over read on SS, discharge
Osteomyelitis
Repeat SS, persistent confusion, discharge Healing rib fx and skull fx Supracondylar fx Hip fracture, development of osteomyelitis in arm None Unexpected healing clavicle fx on chest x-ray
Gastroenteritis Gastroenteritis
Colic Reflux Unknown Unknown Child abuse Fx Fx Fx Child abuse Fx
Child abuse
Skull fx Unknown Unknown Child abuse Fracture Fracture/osteomyelitis Osteomyelitis Child abuse
CPT indicates child protection team; fx, fracture; ID, infectious disease outpatient PV; OFC, orbitofrontal cortex; Ortho, orthopedic outpatient PV; SS, skeletal survey; US, ultrasound.
prescribed karo syrup. Four days later, the subject presented with seizures and altered mental status. Evaluation demonstrated healing rib fractures and intracranial hemorrhage. This subject died 13 days later.
DISCUSSION This study represents the first evaluation of missed opportunities to diagnose physical abuse in children diagnosed with physical abuse and healing fractures. We show that one third of the subjects with abusive healing fractures had previous missed opportunities to diagnose abuse. These missed visits occurred in multiple locations © 2014 Lippincott Williams & Wilkins
(emergency departments, primary care offices, and subspecialty practices), and signs of trauma were often noted on physical examination. Thirty-three percent of the subjects in this cohort had a previous visit, which was felt to be a missed opportunity to identify abuse. This is consistent with and higher than other studies of possible missed abuse reported to date in the literature.4–7,11 Jenny et al7 demonstrated that 31% of children diagnosed with abusive head trauma had been seen by physicians after the abuse, but the diagnosis was not recognized. King et al showed that 30% of children who died of child abuse had health care visits for reasons www.pec-online.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
775
Pediatric Emergency Care • Volume 30, Number 11, November 2014
Thorpe et al
other than routine well child care within the year before their death. In a study by Ravichandiran et al, 21% (54/258) of children referred to a child abuse team with suspected abusive fractures had had at least 1 previous physician visit, which in hindsight, had been a missed opportunity to identify the fracture and/or to identify that the fracture was the result of abuse. The most common reason for presentation at missed previous visits was a “sign suggestive of trauma,” specifically bruising or swelling (Tables 1 and 2). Similar to our study, several studies have shown that bruising and/or swelling in premobile children is concerning for abuse and should prompt complete evaluation for injuries.12–18 The 2007 AAP guidelines on the evaluation of child abuse recommend that physicians pay close attention to any bruises, burns, bites, or lacerations and specifically encourages the careful evaluation of skin injuries in infants who cannot cruise.12 Following these evidenced-based guidelines on the skin evaluation of young infants could substantially reduce additional abusive injuries to the same child. Of note, at the time of abuse diagnosis, less than half of subjects in our study were classified as well appearing although a majority had signs of trauma on examination. For missed visits, more than half of the patients were also felt to be well appearing. This demonstrates that children can be well appearing despite having abuse injuries and that clinicians need to maintain a high suspicion for abuse in all children, regardless of appearance. Of the 25 subjects with missed previous visits, 7 (subjects 1, 8, 10, 19, 20, 21, and 23 in Tables 1 and 2) had the same presentation on 2 or more occasions. Almost one quarter of the children having multiple missed opportunities to diagnose abuse combined with the most common missed presentation being one known to be associated with physical abuse suggests that education of providers regarding common signs and symptoms of child physical abuse alone has not yet been successful. Two patients who died had missed previous visits, and many of the patients with missed previous visits had additional morbidity as a result of the missed opportunity to diagnose abuse. In addition to the morbidity associated with healing fractures, more than one third of the patients were admitted to the intensive care unit, and 34% had an intracranial bleed at the time of abuse diagnosis. This underscores the importance of early detection of abuse in the prevention of the morbidity and mortality associated with repeated exposure to a violent environment. Some missed previous visits were for nonspecific symptoms such as vomiting, fever, and irritability. Although these symptoms are seen in a myriad of common benign pediatric conditions, they are also seen in abused children. Classifying visits for nonspecific symptoms as possible missed visits, in a cohort of abused children, is supported by the literature.4,7 The difficulty for clinicians lies in discerning when these symptoms indicate potentially serious pathology. Although all infants with nonspecific symptoms do not need to undergo a full evaluation for abuse, child abuse should be included in the differential diagnosis, particularly when the infant has had repeated medical visits for these symptoms.
Limitations This study has several limitations. First, data were collected retrospectively, and therefore, some variables such as “reason for presentation to the hospital” required assessment by the data extractor. This information could bias the assignment of previous visit categories (missed vs unrelated). We attempted to minimize this limitation in 2 ways. First, we used a single extractor for all data collection. Second, we had 3 authors separately categorize the extracted data, and the high ĸ value demonstrates internal consistency. Missing data may have also biased results. Although there were no missing data for the visits at which abuse was diagnosed,
776
www.pec-online.com
there were missing data for some previous visits. It is possible that more visits would have been classified as missed previous visits if more information were available, suggesting a possible underestimation of missed previous visits.
CONCLUSIONS In summary, our study found that nearly one third of children with abusive healing fractures had at least 1 previous visit in a variety of clinical settings, which may have been a missed opportunity to diagnose abuse. The most common reason for presentation at these previous visits was a “sign suggestive of trauma,” specifically bruising or swelling. Our study adds to the growing literature indicating that premobile infants with bruising and swelling should be carefully evaluated for physical abuse. Although abuse may be difficult to recognize in the patient with nonspecific symptoms such as vomiting, our data support previous literature demonstrating that physical abuse should remain on the differential, particularly in those with multiple visits. REFERENCES 1. U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau. Available at: https://www.acf.hhs.gov. Accessed November 5, 2012. 2. Alexander R, Crabbe L, Sato Y, et al. Serial abuse in children who are shaken. Am J Dis Child. 1990;144:58–60. 3. Taitz J, Moran K, O’Meara M. Long bone fractures in children under 3 years of age: is abuse being missed in emergency department presentations? J Paediatr Child Health. 2004;40:170–174. 4. King WK, Kiesel EL, Simon HK. Child abuse fatalities: are we missing opportunities for intervention? Pediatr Emerg Care. 2006;22:211–214. 5. Oral R, Blum KL, Johnson C. Fractures in young children: are physicians in the emergency department and orthopedic clinics adequately screening for possible abuse? Pediatr Emerg Care. 2003;19:148–153. 6. Ravichandiran N, Schuh S, Bejuk M, et al. Delayed identification of pediatric abuse-related fractures. Pediatrics. 2010;125:60–66. 7. Jenny C, Hymel KP, Ritzen A, et al. Analysis of missed cases of abusive head trauma. JAMA. 1999;281:621–626. 8. Duffy SO, Squires J, Fromkin JB, et al. Use of skeletal surveys to evaluate for physical abuse: analysis of 703 consecutive skeletal surveys. Pediatrics. 2011;127:e47–e52. 9. Diagnostic imaging of child abuse. Pediatrics. 2009;123:1430–1435. 10. Berger RP, Dulani T, Adelson PD, et al. Identification of inflicted traumatic brain injury in well-appearing infants using serum and cerebrospinal markers: a possible screening tool. Pediatrics. 2006;117:325–332. 11. Oral R, Yagmur F, Nashelsky M, et al. Fatal abusive head trauma cases: consequences of medical staff missing milder forms of physical abuse. Pediatr Emerg Care. 2008;24:816–821. 12. Kellogg ND. Evaluation of suspected child physical abuse. Pediatrics. 2007;119:1232–1241. 13. Feldman KW. The bruised premobile infant: should you evaluate further? Pediatr Emerg Care. 2009;25:37–39. 14. Sheets LK, Leach ME, Koszewski IJ, et al. Sentinel injuries in infants evaluated for child physical abuse. Pediatrics. 2013;131(4):701–707. 15. Pierce MC, Kaczor K, Aldridge S, et al. Bruising characteristics discriminating physical child abuse from accidental trauma. Pediatrics. 2010;125:67–74. 16. Pierce MC, Smith S, Kaczor K. Bruising in infants: those with a bruise may be abused. Pediatr Emerg Care. 2009;25:845–847. 17. Maguire S, Mann MK, Sibert J, et al. Are there patterns of bruising in childhood which are diagnostic or suggestive of abuse? A systematic review. Arch Dis Child. 2005;90:182–186. 18. Sugar NF, Taylor JA, Feldman KW. Bruises in infants and toddlers: those who don’t cruise rarely bruise. Puget Sound Pediatric Research Network. Arch Pediatr Adolesc Med. 1999;153:399–403.
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
