Subdural hematomas in the neonate are unusual, with a difficult delivery accounting for most cases. As obstetric methods have improved, the incidence of this complication has declined, making subdurals even less common.1 Automobile accidents or assaults involving the mother are an even rarer cause of fetal subdural hematomas and are nearly always associated with skull fracture.2 One study reported a high incidence of subdural hematomas without skull fracture in stillborn fetuses of Samoan mothers. Although domestic assault of the mothers was considered, this could not be confirmed in any case.3 There is only one case report of a newborn infant with subdural hematomas secondary to trauma without a skull fracture.4 In this case, the mother reported being beaten and immediately noticed an absence of fetal movements. A full-term infant was then delivered via cesarean section and had a left subdural hematoma that, on aspiration, was acute in nature. We report here a case of a 34-week premature infant with bilateral chronic subdural hematomas of various ages secondary to maternal abuse. The infant remained in utero for a significant period of time after the trauma occurred. The case is unusual in light of the absence of skull fracture or other injury and the amount of time elapsed between injury and delivery.
Baby girl K was the 2138-g product of a 34-week gestation born by cesarean section to a 25-year-old gravida 7, para 3, aborta 3 African-American woman. The mother's past medical history was significant for chronic hypertension. High blood pressures were noted during the pregnancy and the mother was treated with Aldomet until 2 weeks before delivery. The pregnancy was otherwise uncomplicated. The mother's prenatal screens were: blood type A negative/anti-D antibody positive, rubella immune, Veneral Disease Research Laboratory test negative. Gonorrhea and chlamydia cultures were negative. Hepatitis B antibody was negative, as was a group B streptococcus screen done 26 days before delivery. The mother received incomplete prenatal care and denied substance abuse.
On the day before delivery, the mother noted decreased fetal movements and was admitted to the obstetrics unit. Her physical examination was unremarkable with the exception of several healed linear scars on her arms. A biophysical profile was done which was scored 8 out of 10. The mother's blood pressure was 170/85 and she was started on hydralazine and Aldomet. A repeat profile the next day was scored 2 out of 10. Because of the worsening profile, the decision was made to deliver the child via cesarean section. The amniotic sac was artificially ruptured at delivery, producing a normal amount of clear amniotic fluid. At delivery, the patient was blue with poor tone and respiratory distress. She was successfully resuscitated with positive pressure ventilation for 30 seconds and was noted to be more vigorous. The Apgar scores were 6/10 at 1 minute and 8/10 at 5 minutes. Grunting and flaring were noted. After 10 minutes of life, the infant continued to grunt and flare and was noted to have a room air pulse oximetry of 88 percent. Because of her prematurity and respiratory distress, the patient was transferred to the neonatal intensive care unit.
When admitted to the intensive care unit, it was noted that the respiratory distress and oxygen requirement had resolved. Baby girl K's physical examination at that time was significant for a weight of 2138-g (25th percentile), length of 44.5 cm (50th percentile), and head circumference of 33 cm (greater than the 95th percentile). Her pulse was 160 beats per minute and her respirations were 41 breaths per minute. Blood pressure was 55/33. She was noted to have a large anterior fontanel, large posterior fontanel, and widely split cranial sutures. The remainder of the examination was normal.
The initial laboratory work was significant for an O negative blood type with a negative antibody screen. The hematocrit was 23.2% (0.232) and the reticulocyte count 22.3% (0.223). The total bilirubin was 3.5 mg/dL (60 umol/L) with a direct fraction of .4 mg/dL (6.8 umol/L). The partial thromboplastin time was normal at 37 seconds as was the prothrombin time at 12.5 seconds. The fibrinogen level was 254 mg/dL (2.5 g/L) and the D-dimers were slightly elevated at 2–4 mg/L. Electrolytes and urine toxicology screen were normal.
Because of her anemia, an umbilical venous catheter was placed and an exchange transfusion was performed which increased the hematocrit to 43% (0.43). A head ultrasound performed on the first day of life revealed large bilateral subdural hematomas (Fig 1). Computed tomography and magnetic resonance imaging of the cranium were then performed. These scans revealed large bilateral subdural hematomas of varying ages with significant compression of the underlying brain (Figs 2 and 3). Neurosurgery was consulted and the hematomas were surgically drained. A coagulopathy evaluation, including assays of von Willebrandt's factor, Factor XIII, and platelet function was normal.
Upon further questioning, baby girl K's mother admitted that the child's father had physically assaulted her numerous times throughout the pregnancy. She described approximately 8 beatings during the 6 months that she knew she was pregnant. The beatings usually involved her being struck in the face and chest as well as being shoved to the ground. She was also frequently kicked from behind. The mother stated that she was only struck in the abdomen on one occasion, though on one other occasion she struck her abdomen on a piece of furniture as she fell after being pushed. She also describes having vaginal bleeding and abdominal pain throughout the pregnancy that were frequently brought on by a beating. The mother stated that this infant had fewer fetal movements than her other children throughout the pregnancy. She also stated that she was beaten and pushed to the ground, landing on her abdomen, just before seeking medical care for decreased fetal movements. It was this visit that resulted in delivery of the infant.
The subdural hematoma is venous in origin. Its pathogenesis in the fetus is hypothesized to be a tearing of the bridging veins between the brain and the dura due to rotational movement of the brain within the cranium.5 This tearing is more likely to occur in the fetus because of several factors. First, the fetus' head is large and its neck muscles weak. This allows more rotational movement with angular acceleration.5 Second, the subarachnoid space is larger in neonates and fetuses, allowing the brain more freedom to move within the cranium.5 Finally, the fetal brain has a higher water content that increases its mass.6 This allows it to develop more momentum when acceleration is applied. These unique conditions make the fetus more susceptible to developing subdural hematomas and allow the development of subdural hematomas without other injury.
Using this model, we believe that our patient developed multiple hemorrhages over time due to the external force applied from the beatings. These forces caused acceleration of the fetal cranium within the amniotic sac that was then suddenly reversed by impact with either a maternal structure such as the pelvis or a stationary object such as a piece of furniture when the mother fell. While direct blows to the abdomen could certainly cause such injuries, the history does not indicate that this occurred. Given the varying ages of the hemorrhages seen on computed tomographic scan, the intracranial pressure that eventually caused the cessation of fetal movements was cumulative, occurring over enough time to allow a compensatory reticulocytosis to develop in response to the blood loss. We further postulate that the physics associated with our patient's injuries are identical to those responsible for the shaken baby syndrome recognized in the child abuse literature. This syndrome is defined by the presence of retinal hemorrhages and subarachnoid or subdural hemorrhages in an infant with little if any evidence of trauma. We submit, then, that our patient is actually a shaken-impact fetus, at risk for the neurologic sequelae thereof, including seizures, cerebral palsy, hydrocephalus, and mental retardation. With the injury occurring at points much earlier in fetal brain development than has been previously recorded, actual outcome cannot be predicted.
Recognizing the overall rarity of neonatal subdural hematomas, we feel that it is important to consider the diagnosis of the shaken-impact fetus when confronted by a newborn with this condition. Indeed, even when an initial history is negative, further questioning may make the diagnosis, as it did in this case. Conversely, a careful history should also aim to uncover other causes of fetal trauma that may lead to subdural hematomas before accusations are made.
- Received April 1, 1996.
- Accepted July 29, 1996.
Address all correspondence to: Robert P. Stephens, MD, 7575 Northcliff Avenue, Suite 301, Brooklyn, OH 44144.
- Copyright © 1997 American Academy of Pediatrics