Shaken Baby Syndrome

The list of injuries associated with non-accidental trauma, sustained during infancy, such as retinal hemorrhage intracranial and musculoskeletal injuries are generally known as shaken baby syndrome (SBS). SBS has also been referred to as: whiplash-shaken infant syndrome, shaken impact syndrome, infant shaken impact syndrome, non-accidental or abusive head injury (Dias et al. 2005). Regardless of what it is called the impact on a young infant can be quite severe, even fatal. 

SBS occurs when a child is taken by the torso, leg, or arm and shaken at an angular movement repeatedly. This acceleration and deceleration motion causes the brain to rotate within the baby’s skull and apply high pressure of gravitational forces on the brain (Deputy 2003; Tsao et al. 2002; Macdonald & Helfrich 2001). Due to the weight of an infant’s head, the weak neck muscles, the thick skull wall, the soft tissue of the brain, the lack of myelination, and its higher water content this violent shaking make an infant’s brain extremely vulnerable to damage (Lancon et al. 1998; Lewin 2008; Showers 1992). The force at which an infant is shaken causes blood vessels to be broken resulting in bleeding within the brain leading to further damage (Carbaugh 2004; Macdonald & Helfrich 2001). SBS is typically seen within the first year of life; however, cases have been reported up to the age of three (Duhaime et al. 1998; Lancon et al. 1998; Tsao et al. 2002).

Risk Factors & Incidence

An infant’s frequent and inconsolable crying is believed to the most common risk factor and reason for induced SBS. An infant will typically cry 1.5 to 3 hours a day but an excessive amount of inconsolable crying is called “colic”. The frustration and anger felt by the caregivers during these periods of “colic” are often sited at the main trigger for shaking an infant (Carbaugh 2004; Goulet et al. 2009; Lewin 2008). Occurrences of SBS tends to increase between 2.5 to 4 months when the period of “colic” is at its highest (Goulet et al. 2009). Due to the undeveloped anatomy of an infant, they are at an elevated risk of long-term disabilities, impairments, injury and even death (Gutierrez et al. 2004). SBS deaths account for 13-50% all non-accidental pediatric deaths recorded (Dias et al. 2005; Goulet et al. 2009; Lancon et al. 1998). Severe neurological impairments and physical disabilities are recorded in 50-75% the survival population (Dias et al. 2005; Goulet et al. 2009). Additional risk factors for SBS include: being male, difficult temperaments, born premature, low-birth weight, special needs and medically fragile (Carbaugh 2004; Lewin 2008).

Diagnosis/Clinical Findings

As the clinical symptoms of SBS are non-specific, each case tends to vary in how it presents clinically. Minor symptoms of SBS can be mistaken for other childhood illnesses which lead to the challenges of recognizing SBS. Common clinical signs include irritability, seizures, impaired consciousness, bulging fontanelle or forehead, respiratory arrest, inability to focus their eyes, breathing abnormalities, vomiting, lethargy, impaired consciousness, constipation, poor feeding, apnea and muscle weakness (Altimier 2008; Carbaugh 2004; Duhaime et al. 1998; Lewin 2008). The damage of SBS continues to impact the infant’s life, long after the attack. Any Infant who has been the victim of SBS should undergo a complete assessment and immediate attention (Gutierrez et al. 2004).

Assessments such as: skull radiography, CT scan findings, ophthalmologic examinations and MRI examination findings are used to diagnosis SBS (see Table below) (Coody et al. 1994; Duhaime et al. 1998). CT scans have been shown to be superior when viewing the damage to the infant’s brain.  Intracranial hemorrhage, hairline fractures, and compressions in the skull are all visible on the CT scan (Coody et al. 1994). In infants where CT findings are not definitive, an MRI has been shown to be useful in detecting extra axial hemorrhages (Duhaime et al. 1998). There is a direct relationship between the force at which the infant is shaken and the amount of hemorrhaging that has occurred within the brain. 

A common finding in SBS, which has been reported in 65 to 95% of the population, is retinal hemorrhages (RH) (Duhaime et al. 1998). The amount of force it takes to cause RH is unknown. Duhaime et al., (1998) have noted that although RH is not specific to SBS, the appearance of “severe bilateral retinal hemorrhage with retinal folds or detachments” is.

Table: Diagnostic Findings in Shaken Baby Syndrome (Carbaugh 2004 pg 110)


Findings consistent with SBS

Computerized Tomography (CT)

Subdermal haemorrhages
Subarachnoid haemorrhage
Cerebral contusions
Cerebral edema
Subtle skull fractures
Ventricular enlargement
Brain atrophy
Hypodense areas

Magnetic Resonance Imaging (MRI)

Subdural haematomas
Subarachnoid haemorrhages
Cerebral contusions
Cerebral edema
Subtle skull fractures
Ventricular enlargement
Brain atrophy
Hypodense areas
Intra parenchymal lesions
Chemical state changes of haemoglobin that substantiate repeated injuries

Skeketal Survey

Long bone injury
Traction fractures (corner and bucket handle
Periosteal stripping
Metaphyseal fracture
Shaft fracture long bone bruising
Skull fractures
Rib fractures
Fractures in various stage of healing

Ophthalraologic Examination

Retinal haemorrhages
Vitreous haemorrhage
Retinal detachment
Anisocoria (unequal pupils)
Orbital and lid ecchymosis
Subconjunctival haemorrhage
Disconjugate eye movements
Sixth nerve palsy
Disruptions of contents of the eye
Optic nerve haemorrhage
Optic nerve scleral junction haemorrhage
Orbital fat haemorrhage

Other Findings

Bloody cerebrospinal fluid
Mildly to moderately anemic
Clotting dysfunction
Elevated transaminase levels (indicates occult liver injury)


Evidence on the types of treatment used in the SBS population is scarce and requires more research to be completed. Discussion of treatment of consequences resulting from SBS include if the infant is unconsciousness at acute care the use of “intubation, ventilation, fluid resuscitation and anticonvulsant therapy” (Duhaime et al. 1998) may be used. Aggressive treatment of an infant with poor prognosis outcomes has been questioned since most cases result in death despite continuous management (Duhaime et al. 1998). 

Long-Term Outcomes

SBS can result in many possible long-term consequences such as “permanent brain damage, visual impairments, developmental delays, disabilities and motor impairments, paralysis, eye damage, hearing loss, blindness, decreased movement from spastic muscles seizures and even death” (Carbaugh 2004). Outcomes of SBS include: death in approximately 20 percent of infants; a variety of permanent disabilities and impairments including permanent brain injury; mild learning disorders; cerebral palsy; blindness; developmental retardation; seizures; paralysis; or SBS may result in a permanent vegetative state (Duhaime et al. 1998; Gutierrez et al. 2004).

Individual Studies

Table: Impact of SBS on the Developing Brain


The cases of 404 SBS children were reviewed by Bourgeois et al. (2008), to examine seizure disorder and related outcomes. 296 patients presented with seizures, 135 had motor deficits and 303 presented with a bulging fontanelle. 73% were found to have epileptic seizures with 50% of these patients displaying multiple types of seizures. Behavioral problems were found to be closely associated with seizures in 96% of patients. Poor prognosis for the outcome of SBS children was found to be associated with the presence of seizure activity.

King et al. (2003) reviewed the charts of 364 patients to determine the associated characteristics of SBS and related outcomes. Clinical features of SBS included: seizures (45%), decreased consciousness (43%) and respiratory difficulties (34%). Retinal haemorrhages which were found to be associated with neurological deficits, subdural haemorrhages and even death were present in 76% of patients. Several patients (85% of the 364) were found to need ongoing multidisciplinary care due to the fact that these infants had sustained a moderate to severe disability or were in a vegetative state.

84 SBS cases were selected to be contacted about their long term outcomes in the study conducted by Duhaime et al. (1998). Only 14 individuals were used in this study with a mean follow up from acute injury of 10 years. One patient died 5 years post injury but up until that time was in a vegetative state. Six patients were severely disabled, two patients were moderately disabled and 5 patients presented with good outcomes. Associations were found between the acute stage of SBS and their long-term outcomes. Five patients who were unresponsive at acute care turned out to be in a vegetative state or severely disabled long term. At acute care, six patients required intubation lead all patients to develop severe or moderate disabilities. Children who presented with a bilateral or unilateral SDH on their CT scan were not found to have a severe disability at time of follow-up. Overall, children who were found to be unresponsive, have a unilateral or bilateral SDH and were intubated while in acute care, had a much poorer prognosis for long term outcomes. 


Permanent severe neurological outcomes and disabilities are often the result of shaken baby syndrome.


Ophthalmological Outcomes

Retinal hemorrhages (RH) have been found in 65-95% of the population and are potential outcomes of SBS (Duhaime et al. 1998). Two theories exist on the etiology of RH: either due to retinal venous obstruction from increase ICP or traction between vitreous and retina when an infant’s head is accelerating or decelerating (Kivlin et al. 2000). RH is a physical presentation that will warrant a head scan and thereby decrease the likelihood of SBS going undetected and prevent the infant from potential further abuse (Kivlin et al. 2000).  

Individual Studies

Table: Ophthalmologic Findings Post Shaken Baby Syndrome (SBS)


Kivlin et al. (2000) noted that RH was found in 79% of the entire study population (n=123). Of those patients who had a detected RH, bilateral RH was found more frequently (68%) than unilateral RH (11%). Of the 36 patients who died, ophthalmologic examination revealed a lack of visual response in 35, poor papillary response in 26 patients, and RH was noted in 34 patients. Study authors suggest the presence of these symptoms led to the reported fatal outcome for those children with SBS (Kivlin et al. 2000).

Of the 30 cases reviewed by McCabe & Donahue (2000) subdural haemorrhages were detected in 21 patients and were more common than intracerebral (n=11) and subarachnoid (n=10) haemorrhages. Seizure activity occurred in 67% of the study population. All 8 patients who died displayed nonreactive pupils to visuals and 6 of the patients who died had a midline shift detected. It was determined that lack of visual response and a midline shift are indicators of fatal outcomes among SBS. As well, visual responsiveness was deemed to be determinate on the level of severity brain damage.

In a case series of 10 patients, Mills (1998) studied how ophthalmologic examinations of SBS individuals determined future outcomes. Of note: intraretinal haemorrhages were found in all patients but was not found to be significantly associated with fatal outcomes; a lack of visual response was found in 4 patients, of which 3 died later; and circular perimacular retinal folds and peripheral retinoschisis was also found in all 3 individuals who died. Moreover, in a case series of 14 children in Wilkinson et al. (2014), the severity of RH predicted the severity of acute neurological outcomes. Other predictors of severe neurological injury included vitreous or subhyaloid hemorrhages.


There is Level 4 evidence that the lack of visual response in infants with shaken baby syndrome at ophthalmologic examinations may lead to fatal outcomes.

There is Level 4 evidence that the presence of poor papillary response, the presence of a RH, a midline shift, circular perimacular retinal folds and peripheral retinoschisis may also lead to fatal outcomes among individuals with shaken baby syndrome. 


Ophthalmologic examinations can be used to reveal the severity of brain injury resulting from SBS.


Education & Prevention

The “Don’t Shake the Baby” project implemented by Showers (1992) provided a base for future SBS educational programs to build off. “Don’t Shake the Baby” project included an information card providing tips on how to calm a crying infant and a response card to be filled out by the parents. It was concluded that there needed to be more education provided to parents on the dangers of shaking infants and on how to properly care for a crying infant (Showers 1992).

The prevention of SBS appears to be related to the education received by parents either prior to delivery at routine office visits, during prenatal visits, prenatal classes and post-delivery prior to leaving the hospital (Walls 2006).

Individual Studies

Table: Educational Programs Designed to Reduce the Risk of SBS


A recurring intervention for educating mothers on SBS and the care needed to prevent such injury is called PURPLE. Each letter in PURPLE represents a characteristic of crying in infants that is troublesome for caregivers, potentially leading to SBS (Fujiwara et al. 2012). Overall, the PURPLE intervention improved maternal knowledge in infant crying compared to an infant safety information control group (Barr, Barr, et al. 2009; Barr, Rivara, et al. 2009; Fujiwara et al. 2012). Improvement in knowledge on infant shaking was conflicting between studies. Two studies reported no statistical significant improvement compared to the control group (Barr, Rivara, et al. 2009; Fujiwara et al. 2012) whereas another study reported significant improvement (Barr, Barr, et al. 2009). However, 95% of mothers post-intervention were reported to have a perfect score on knowledge of infant shaking following the PURPLE intervention (Reese et al. 2014).

The improvement in knowledge of crying and shaking behaviours moderately translated into a change in behaviours following the PURPLE intervention. Mothers reported walking away from infant with inconsolable crying more than the control group (Barr, Rivara, et al. 2009; Fujiwara et al. 2012), and more mothers shared this information with other caregivers (Barr, Barr, et al. 2009; Barr, Rivara, et al. 2009; Fujiwara et al. 2012). Two studies reported significant differences in PURPLE compared to control in the sharing of information on the dangers of SBS with other caregivers (Barr, Barr, et al. 2009; Barr, Rivara, et al. 2009) whereas one study reported no significant difference (Fujiwara et al. 2012). Overall, only 41% of mothers shared information that they had learned in the intervention with other care providers for their child (Reese et al. 2014). The most common reason for lack of sharing information was due to low perceived risk of infant shaking under the other caregiver’s watch. However the majority of perpetrators that cause SBS are not mothers (Reese et al. 2014), and therefore the PURPLE intervention should be modified to increase the sharing of learned knowledge and to involve other caregivers.

Other educational programs have been evaluated for informing caregivers of the effects of SBS and SBS prevention. All of the programs that were implemented were short-term with little participant involvement. One program administered a brochure on the steps to take when their child was crying (Bechtel et al. 2011). Most programs used a brochure and short video (8-11min) combination (Altman et al. 2011; Deyo et al. 2008; Dias et al. 2005) and one used a talk from a pediatrician (Simonnet et al. 2014). There is little time consumption for participants in such programs and therefore are attractive and easily implemented. Programs that are administered within the hospital and provided through a healthcare professional are effective in translating the dangers of shaking an infant (Altman et al. 2011; Bechtel et al. 2011; Deyo et al. 2008; Dias et al. 2005; Simonnet et al. 2014).

Although all aforementioned studies improve caregiver knowledge on SBS, only a few evaluated the change in incidence of SBS following the educational interventions. Future research is warranted to determine the outcomes of educational interventions on rate of SBS.


There is Level 1a evidence that PURPLE intervention program is effective to reduce maternal knowledge of infant crying and shaken baby syndrome, compared to an infant safety control group.

There is Level 4 evidence supporting the role of education programs for informing the caregivers of shaken baby syndrome and its detrimental effects.


The PURPLE intervention for shaken baby syndrome increases knowledge of crying and effects of shaken baby syndrome, as well as behaviours of walking away during infant inconsolable crying.

Education programs on infant crying and safety are effective to inform parents on the dangers of shaken baby syndrome. 




Overall the pediatric literature for ABI is sparse. It is imperative to determine preventative measures and interventions to reduce potential long-term detrimental effects of an ABI. Many of the interventions that were evaluated in this review have been studied and are effective in the adult population. Children have different developmental trajectories compared to adults, and there is even a developmental difference when comparing a young child to an adolescent. Therefore, practitioners and other health professionals cannot rely on studies within the adult population for treatments for children following an ABI. Further research is warranted in order to be able to make recommendations for the pediatric population.