1.3.1

Lack of initial complaint

Shaking or any kind of infant abuse is rarely the cause triggering the initial medical consultation of the infant. In fact, it is only in less than 5% of cases that a person will finally admit, during the investigation, to shaking the infant to resuscitate him or her . Furthermore, in 70 to 97% of the cases, any history of trauma or injury, even minor ones, will be denied . The sociolegal pediatric clinic of the Sainte-Justine University Hospital in Canada has more than 20 years of experience and has performed over 400 clinical exploratory examinations for potential AHT. In less than 10% of the cases, clinical evaluation was initiated following alleged shaking episode usually because one of the parents was accusing the other of abuse and shaking the baby. These exploratory examinations were always negative and no validation or invalidation was possible. In two cases, the shaking was observed in a hospital setting by a healthcare professional, in one case the ongoing investigation had already documented evidence of previous abuse (shaken baby), whereas in the other case the investigation post-incident was negative.

1.3.2

Age and sex

The majority of studies indicate the age and sex of their subjects. Regarding age, Table 1 suggests that most abused children are younger than 12 months and the mean age is under 6 months. The median is between 3 and 5 months. There are more male victims than female, for obscure reasons, for this kind of physical abuse. Several hypotheses were brought up to try and explain this element, but none has been scientifically validated to this day.

1.3.3

Initial signs and symptoms

Very few studies analyzed the initial signs and symptoms, e.g., the ones that bring the infant to a medical consultation. In regards to this, four studies are interesting, the ones from Ludwig and Warman, 1984 , King 2000 , Mireau 2005 and PCSP 2008 . The first three ones are retrospective studies, whereas the last one is more prospective in its nature.

The study from Ludwig and Warman, 1984 was based on children who, from 1977 to 1982, were the object of a report for suspected child abuse from the Philadelphia children hospital to Child Protective Services. Out of the 1250 medical files revised, only the cases with “shaken-only symptoms”, e.g. 20 cases, were kept; all children with other types of abuse such as burns, bruises or fractures were excluded. Most times the parents came to the hospital because their infant was lethargic (80% of cases) or had difficulty breathing (60% of cases).

In his retrospective study, King 2000 compiled 364 reported cases between 1988 and 1998. These cases came from 11 university hospitals spread all over Canada and offering advance pediatric care. This study included all children under the age of 5 presenting head traumas with suspicion of SBS, with or without impact, that were reported by healthcare teams to the child protective services. The most frequently encountered clinical symptoms were seizures (45% of the cases) and decreased alertness (43% of the cases).

Mireau 2005, in his thesis , focused on all cases of subdural hematoma (SDH) in children under the age of 3 hospitalized in the neurosurgery ward of the Necker Hospital for Sick Children in Paris (Hôpital Necker–Enfants-Malades de Paris) between 1994 and 2004. However, all cases of SDH caused by a validated accidental head trauma or a well-documented pathology were excluded. Finally, 404 cases were kept where SDH seemed to have been the result of abusive head trauma. The most common initial clinical symptoms were seizures (73.3% of cases) and hypotonia (65% of cases).

The Canadian Pediatric Surveillance Program (CPSP) from the Canadian Pediatric Society (CPS) compiled 220 cases of abusive head trauma between 2005 and 2008. This study sent each month to each Canadian pediatrician, member of the CPS, a form to be filled out to indicate if he or she had in the last month treated a child under the age of 14 for head or brain trauma with a suspicion of physical abuse or neglect and had reported that case to Child Protective Services. In Canada, according to local laws, any professional suspecting child abuse has the obligation to report the case to the Child Protective Services of the county or province. Once the case is reported, in a confidential manner, the CPSP sent a detailed questionnaire that the pediatrician had to fill out. Thus 394 cases were reported for the study. Ninety-two were reported twice, meaning by more than one pediatrician simultaneously, 51 cases did not meet the inclusion criteria and in 31 cases the additional information was too tenuous to be included in the final data analysis. In these series also, the most common initial symptoms were decreased alertness (36% of cases), traumatic injuries of the soft tissues (30% of cases), followed by seizures (29% of cases).

Table 2 sums up the main signs and symptoms reported in each of the four studies listed. Obviously, we cannot be sure that the signs and symptoms meant the same to all authors. However, in average, validated or suspected seizures as well as decreased alertness seemed to be the most frequently encountered symptoms without being present in all cases. Even if they are non-specific, they point towards a potential neurological lesion. Still, in most cases, more than 50% of children seemed to initially show non-neurological symptoms such as respiratory dysfunction, hypotonia, vomiting, apnea or irritability.

Curiously, only Ludwig and Warman and Mireau reported the state of the fontanel which isn’t closed in most children because of their young age (68% of cases according to Mireau). The two authors found a bulging or over tense fontanel in 55% of cases for Ludwig and Warman and 68% for Mireau . This telltale sign is very relevant since it directly points to a potential brain injury and should require additional imaging exams. Being aware of this sign will completely change the differential diagnosis and orientation of the clinical examination when faced with an infant presenting other non-specific symptoms such as vomiting, irritability or lethargy and other symptoms that can be commonly found in the ER or in a walk-in consultation.

It is also very important to underline that in more than one third of the cases, trauma-related skin lesions were found at the first medical evaluation of these infants. Several authors, such as Labbé and Caouette 2001 and Wedgwood 1990 , correlated the significance of bruises to age and reported that, in non-walking children they most often are sign of child abuse. According to Atwal 1998 , who studied 24 cases of death by inflicted brain trauma, 70% of the children had bruises on their skin. Most often, the bruises were found on the cheeks, forehead, back, buttocks and more rarely on the limbs. This same author also reported that the aspect of the lesions were more symptomatic of hitting than grabbing. Greenes and Schutzman 1999 , in a study on head trauma in infants, noticed that 93% of newborns that had bruises on the scalp also had associated brain injuries.

The signs and symptoms are rarely isolated, but only Mireau 2005 studied their concomitant association to other symptoms as presented in Table 3 . These associations, even if there are non-specific, do orientate towards a neurological origin rather than a respiratory or gastrointestinal one. It seems here quite clear that in AHT cases vomiting, which is an extremely frequent primary complaint in newborns, is associated to other signs and symptoms that should guide the physician towards additional neurological exams and thus, lead eventually to the discovery of suspicious brain injuries.

According to Newton 2006 , death can also be added to the list of telltale clinical conditions for AHT. In fact, a certain proportion of children with shaken baby syndrome are brought to the hospital dead or agonizing as reported by Carty and Pierce 2002 . Just like in Sudden Infant Death Syndrome (SIDS), many AHT infants do not show any external signs of injuries, this is why an autopsy and a complete investigation surrounding the circumstances of death are necessary to correctly identify the cause of death, AAP 2008 . It is difficult to estimate the number and percentage of cases since often, they are not included in the retrospective studies and the diagnosis was based on data from various authorities as well as coroners, outside of a clinical setting ( ).

The initial telltale signs and symptoms, i.e. the ones for which infants, later identified as AHT victims, are brought to the medical consultation can be grouped into four categories:

• •
  

  non-specific of head or brain injury: loss of appetite, vomiting, irritability, sleepiness and death;

  
  
• •
  

  suggesting head or brain lesions: decreased alertness, seizures or seizure-like episodes, tense or bulging fontanel;

  
  
• •
  

  suggesting head or brain traumatic injuries: bruise or signs of head contusion, known history of minor head trauma;

  
  
• •
  

  suggesting physical abuse or violence: numerous bruises on the body of a non-walking infant.

In fact, as reported by Jenny et al. 1999 and Case et al. 2001 when known history of trauma, even minor ones, is unavailable and first signs or symptoms are non specific, some aspects of an exhaustive clinical examination should point to AHT, such as fontanel examination, looking for bruises on the skin, especially on the head or face or anywhere on the body of a non-walking infants. These last two elements, looking for bruises and palpation of the fontanel should systematically be included in any basic clinical examination of an infant brought to the medical consultation, regardless of the other alleged signs or symptoms. All suspicious cases should get brain imaging especially simple (without infusion) CT-scan which seems to be the imaging exam of choice.

1.3.4

Symptoms that could negatively impact the diagnosis

Several authors such as Carty and Pierce 2002, Jenny et al. 1999 and Oral et al. 2008 , reported that a certain number of their validated AHT cases could have been diagnosed earlier during a previous medical consultation than the one where their condition was accurately diagnosed. Only Jenny et al. 1999 presented a systematic study. The author studied retrospectively all the cases of AHT in infants under the age of 3 that were referred to their pediatric unit within the University Hospital from 1990 to 1995. Out of the 232 cases studied, 173 were included in the final study. In three cases, the children were older than 3, 38 cases were accidental and 13 could not be included due to the non-availability of relevant medical charts. Fifty-four cases (31.2%) were not correctly diagnosed during the first medical evaluation. The diagnosis was made after a mean of 2.8 medical consultations and a 7-day lapse. The authors were able to establish that errors in diagnosis were more common in young newborns (mean age 180 days) than in older ones (mean age 278 days). The diagnosis was missed in 37% of children of Caucasian origin versus 19% in African-American children and in 40% of cases in married couples opposed to 19% of separated parents.

In regards to the signs and symptoms, 65% of the children who were not properly diagnosed were alert and awake during the first consultation, 82% did not present any respiratory dysfunctions, 56% had vomiting symptoms essentially and 65% were irritable. The statistical analysis tends to show that four factors or variables have a significant predictive value, i.e. the presence of:

• •
  

  respiratory disorders;

  
  
• •
  

  seizures;

  
  
• •
  

  traumatic injury of the face or scalp;

  
  
• •
  

  separated parents.

The authors estimated that the lack of one of these four variables meant that the baby had only 20% chances for AHT to be diagnosed at the first consultation.

An exhaustive study of each medical chart showed that four out of five deaths could have been prevented in the group of 54 children where the diagnosis was missed, whereas for the 15 other children (28%), they would probably not have been submitted to another shaking episode.

In 14 and 10 cases respectively, the reported wrong diagnoses were viral gastroenteritis and accidental head trauma and in another 20 cases, it was infectious pathology. In two cases, a spinal tap with blood was erroneously attributed to a traumatic procedure.

In order to reduce the rate of missed diagnoses, the authors recommend, as mentioned above, that for all infants a physician should systematically look for skin bruises, evaluate the state of the fontanel and assess the head perimeter. They also recommend an eye funduscopic examination with pharmacological pupil dilation if necessary.

To conclude, the signs and symptoms most likely to distract the physician from the proper diagnosis are the ones belonging to the “non specific of head or brain injury” group: e.g. loss of appetite, vomiting, irritability and sleepiness which are so frequently the main reason for medical consultation. Most often, they result from benign infectious pathologies rather than traumas. Faced with the lack of family or social risk factors or the notion of a benign traumatic event, they can distract the physician from a neurological diagnosis.

1.5.1

Risk factors related to the abuser

Most times, the abuser is a man (68,5% Starling, 1995; 70% Starling, 2000; 72% King, 2000 under the age of 30 (mean: 22 years [14–46 years], Showers, 1994; median: 22 years [16–46 years], Starling, 2008) having a privileged relationship with the victim, either being the father of the child or the mother’s boyfriend (44% biological father, 23% boyfriend: Showers, 1994; 37% biological father, 20.5% boyfriend: Starling et al. 1995; 44% biological father, 20% boyfriend: Sinal, 2000; 33% biological father, 13% boyfriend: King, 2000) . Caretakers are also potential abusers. In fact, in the series of 127 cases examined by the team of Starling 1995 , they amounted for more than 20% of the abusers. Similar results were found in the retrospective study of the medical files conducted by Sinal and his team in 2000 . This category of unrelated abusers is rarely discriminated in the studies on social and family risk factors, which could really have created a bias in their analysis.

1.5.2

Risk factors related to the child

In most studies, the victim is most often a boy. This data has not been clearly explained yet but should be mentioned since it contradicts the most recent Canadian data on the incidence of physical abuse for the 0–3 years of age range . According to this large Canadian study, more girls (57%) would be the object of physical abuse corroborated after a report to Child Protection Services.

Even though AHT is often seen in children under the age of 2 and especially under the age of 1, most of the victims are under the age of 6 months (62.8% Mireau 2005; 65% Goldberg 2000; King, 2000) . Ethnicity was suggested for the first time as a risk factor by Guthkelch in 1971. He suggested that white populations (in the United Kingdom) were more prone to shake their babies as a disciplinary method, probably under the assumption that shaking is more socially acceptable than slapping. This hypothesis was also studied by Sinal and Ball as well as Brenner and his team in 1987. In fact, their results suggested that there was a higher occurrence of Caucasian victims (17/24 Sinal 1987; 8/12 p < 0.008 Brenner 1987) . Due to some methodological biases, mainly on the type of population going to the targeted hospitals, Sinal and his team (2000) focused on this controversial issue by using a more rigorous methodology and this time, did not find a significant difference ( p = 0.0895) between a group of South Carolina Caucasian children and another group with South Carolina children from different racial origins. In the series of 80 cases of abusive head trauma in North Carolina reported by Keenan et al. in 2003 , race/ethnicity (non–European American) was considered a significant risk factor for inflicted traumatic brain injury (TBI) in regards to the general population. However, race or ethnicity was not significantly correlated to AHT compared to accidental brain trauma. Data suggest that the variable “race or ethnicity” is a risk factor stemming from other social markers that put the child at a higher risk of injuries. On the other hand, in the analysis of 173 cases of false negative diagnosis of AHT Jenny in his 1999 study reported that the cases that were less likely to be recognized by medical professionals were young Caucasian infants. In the same line, in 2002, Lane and his team concluded that more children from African American or Hispanic ethnicity were more likely to be evaluated and reported for suspicion of child abuse. This observation, if true, could challenge the lack of significant difference between ethnic groups and could show also the bias linked to racial preconceptions.

Several studies show higher AHT rates for some pregnancy-related conditions or delivery-related conditions than those usually found in the general population. Mireau 2005 reported an 11% rate of pathological pregnancies and an 11% rate of premature births. Concordet and Bonnaure 2000 unveiled 55% of complicated pregnancies and 17% of premature births. A similar rate of premature births (15.6%) was shown by Keenan et al. 2003 . Minns and Brown 2005 found a 29% rate of premature births < 37 weeks in a sample of 129 Scottish medical files, amounting to the higher rate found in the literature. King 2000 reported a rate of 16% for pathological pregnancies and 21% for premature births and Matsuura 2002 showed a rate of 13.8% for premature births and 22% of pregnancy disorders on 72 medical files classified as SBS in the state of Hawaii. Multiple births were also found to be more frequent than in the general population: Minns and Brown 2005 found 9%, Concordet and Bonnaure 2000 7%, Keenan et al. 2003 5.3%, whereas Mireau 2005 found a 5% rate in his sample of 404 cases of children treated at the Necker Hospital Center in Paris.

Some studies have also identified the birth rank of the child in the family to show that in at least half of the cases, the victim was the mother’s first born (51% Mireau 2005; 58% Concordet and Bonnaure 2000; > 50% Goldberg 2000; 92% Bonnier 1995, Keenan et al., 2003 5.3%) .

1.5.3

Risk factors related to parents

The mean age of parents at the time of the physical abuse seems to be much lower than the general population at the birth of their first child. According to a sample group in Minns and Brown’s study in 2005 , mothers had a mean age of 22 years (SD 5 years) and fathers/partners had a mean age of 26 years (SD 5.8 years) whereas in the series from Mireau in 2005 , mothers and fathers had a respective mean age of 28 and 32. By comparing the age of mothers at the birth of their child, Keenan et al. 2003 and his team unearthed a significant difference between the age of mothers in the general North Carolina population ( p < 0.001) and the ones of mothers of children victims of AHT, the latter being significantly younger. Unfortunately, most studies on parents’ age only offer diffuse data not allowing for a precise comparison. However, the young age of parents is often mentioned by authors in the literature and noted as a potential risk factor.

Several incomplete and heterogeneous data do show that parents or abusers are in most cases involved with social issues or emotional disorders. The Scottish series of 129 victims reports 25% of cases with a history of drug or alcohol abuse. Starling 2008 mentions the history of family violence in almost half (40.7%) of the identified abusers. Minns and Brown 2005 only report 25% of family violence in the victims’ parents, but a 41% rate of violence inflicted at a young age. In a Canadian study on 364 cases, King 2000 found lower rates, i.e. 12% for a history of family violence, 7% of violence at young age and 9% of substance abuse and addictions in the abusers. Minns and Brown 2005 reported 32% of present or past mental disorders whereas Concordet and Bonnaure reported a 16% in their 2000 study , so did Goldberg and Maurey-Forguy 2000 and a 4% rate for King’s study in 2000 .

The psychosocial analyses on two series of French infants by Concordet and Bonnaure 2000 and Goldberg and Maurey-Forguy 2000 underline the major distress in the parent-child relationship, lack of family, friends or social support and presence of stress factors (e.g. problems in the parenting couple, work problems, moving house). These studies were essentially descriptive and it was quite complicated to extract precise data. However, the presence of this type of factors seem more important in the first series ( n = 69) than in the second one ( n = 106) for no apparent reason. King, in his 2000 study on Canadian parents, highlighted several factors suggesting that social isolations and lack of parenting skills were involved. Mireau 2005 noted that parents did not have a proper understanding of the needs and normal behaviors of their child.

1.5.4

Risk factors related to the social context

The recent study by Minns in 2008 offers a detailed description of the socioeconomic context, over a 10-year period, for 25 cases of Abusive Head Trauma (AHT) from the Scottish region of Lothian. This series represents probably all cases from this region over the given period. The authors demonstrated that most cases (64%) came from the most impoverished neighborhoods. For their evaluation, they used the Scottish Index of Multiple Deprivation (SIMD), which unveiled that 76% of cases came from the poorest areas as far as education, skills and social abilities were concerned, 72% in areas with the highest rates of crimes, 68% in areas with the poorest health, 60% of cases in areas with the lowest incomes, 52% in area with the poorest quality of housing, 48% in areas with high rates of unemployment. But paradoxically, 76% of these 25 cases who lived in these poor and socially challenged neighborhoods had access to many more services, e.g. transportation, healthcare, schools, shops or postal services. King’s study in 2000 reported that poverty affected 28% of the cases, ranking this factor second in the list of risk factors related to family and social context.

These studies contradict the results and hypotheses from the previous study that showed that AHT did not have a higher incidence in poorer populations. For example, Concordet and Bonnaure 2000 reported that most parents had a stable job, 81% of mothers were working. For Goldberg and Maurey-Forguy 2000 ), most parents had a secondary or even higher level of education. The middle classes were the most represented with a good number of upper classes as well. Mireau 2005 showed that the parents in his study most often did not have any major difficulties or social issues.

To sum up, the data available to this day are incomplete and sometimes contradictory, but suggest that the most likely victims of AHT are boys, firstborns, under the age of 6 months, born prematurely after a difficult or multiple pregnancy, living with parents with a past or present history of abuse (alcohol, drugs, family violence) that are unaware of how to manage the needs of their child. Abuse can also have no correlation to the socioeconomic, cultural or ethnic context. Furthermore, if it is safe to assume that a child victim of AHT has more chances to belong to that group, it is wrong to think that most children with these characteristics are victim of this type of abuse. Thus, to establish the proper diagnosis, it is essential to focus more on the clinical, biological and imaging data.

2.3.1

Absence de plainte initiale

Le secouement, ou toute autre forme de sévices à l’enfant, est rarement le motif qui amène initialement l’enfant en consultation médicale. Effectivement, ce n’est que dans moins de 5 % des cas, qu’une personne reconnaîtra finalement en cours d’investigation avoir secoué l’enfant dans le but de le réanimer . De même dans 70 à 97 % des cas, toute histoire d’un traumatisme quelconque, même mineur, demeurera niée . À la clinique de pédiatrie sociojuridique du CHU Sainte-Justine, en 20 ans, après plus de 400 explorations cliniques pour possibilité de traumatisme crânien non accidentel de l’enfant, dans moins de 10 % des cas, l’exploration fut initiée en raison d’une histoire de secouement rapportée par un des deux parents de l’enfant, ce dernier accusant l’autre d’en être l’auteur. Ces explorations se sont toujours avérées négatives et n’ont pas permis de confirmer ou infirmer les faits. En deux occasions, le secouement a été observé en milieu hospitalier par un professionnel de santé, dans un cas l’investigation, déjà en cours, avait documenté des évidences de secouement antérieur, alors que dans l’autre, l’investigation post-événement demeura négative.

2.3.2

Âge et sexe

La majorité des études indiquent l’âge et le sexe des sujets répertoriés ( Tableau 1 ).

Tableau 1

Âge et sexe des enfants victimes d’un traumatisme crânien non-accidentel.

Auteurs	Nombre de cas	Âge moyen (mois)	Âge médian (mois)	Moins de 6 mois (%)	Moins de 12 mois (%)	Garçons (%)
Ludwig 1984	20	5,8	3	75	–	70
Duhaime et al., 1987	48	7,85		–	–	65
King 2000	364	–	4,6	–	–	56
Ettaro et al., 2004	89	–	–	–	87	53
Mireau 2005	404	5,4		62,8	94,7	72,7
Talvick 2006	26	3,9	–	–	100	77
CPSP 2008	220		5	58	75	58

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Analysis of return to competition and repeat rupture for 298 anterior cruciate ligament reconstructions with patellar or hamstring tendon autograft in sportspeople Effects of a five-week intensive and multidisciplinary spine-specific functional restoration program in chronic low back pain patients with or without surgery Describing and developing the field of competence in Physical and Rehabilitation Medicine (PRM) in Europe – preface to a series of papers published by the Professional Practice Committee of the PRM section of the Union of European Medical Specialists (UEMS) Diagnosis and treatment of osteoporosis in spinal cord injury patients: A literature review Early post-stroke physical conditioning in hemiplegic patients: A preliminary study Mycetoma (Madura foot): A case report

Stay updated, free articles. Join our Telegram channel

Join