Author, date and country | Patient group | Study type (level of evidence) | Outcomes | Key results | Study Weaknesses |
---|---|---|---|---|---|
Dias et al, 2004, USA | 215 Patients with head injury >2-year-old, GCS >12 and at least one of following: more than brief LOC, significant amnesia, >1 episode of vomiting, persistent lethargy | Prospective cohort | Follow-up phone call at 1 day post-discharge | No abnormalities on CT scan. Two children returned due to persistent vomiting, repeat scans normal in both cases | Poor follow-up rate. Telephone follow-up abandoned after 2 months |
Mitchell et al, 1994, USA | 401 Patients <18 years with GCS >12 who were admitted for observation following head injury; 218 of the children had a CT scan, the rest of the children only had skull x-ray. 51 Patients had GCS 15, normal neuro exam and normal radiological exam | Retrospective review. | Follow-up to document occurrence and duration of symptoms | No adverse events in the subset of patients with normal GCS, neuro exam and radiological investigations | Small subset applicable to question. Radiological examination was CT or plain x-ray or both. Limited follow-up |
Davies et al, 1995, USA | 400 Patients <18 years with GCS >12 and negative CT scan following head injury | Retrospective case-series study. | Readmission or death within 1 month of head injury | Four patients were readmitted. One patient taking coumarin was found to have a subdural haematoma and required surgery. One child was admitted with a symptomatic haemorrhagic contusion. One child was readmitted following a seizure but the repeat scan was normal. The fourth child was admitted with ‘concussion symptoms’ and admitted for observation but not rescanned | Only 78% had a GCS of 15 on admission. Of the four readmitted patients the original GCS is not stated; 76 (18%) of the normal CT scans were subsequently read as abnormal in the formal report. Follow-up relied on a statewide database |
Dahl-Grove et al 1995 USA | 62 Patients with closed head injury <18 years, GCS >12 and normal cranial CT scan | Retrospective review. | ED and hospital course, radiographic studies and follow-up | 56/62 Patients had GCS 15 on attendance. No patients developed any neurological deterioration | Small study. Criteria for cranial CT scan and subsequent admission could not be determined by chart review. No long-term follow-up |
Schunk et al 1996 USA | 313 Patients with head injury <18 years, GCS 15, no focal neurological deficit and cranial CT scan | Retrospective review. | Results of CT scans, disposition and outcome | Of the 300 patients with no intracranial injury on CT 159 were discharged and 141 were admitted. Three discharged patients re-attended, none had a new ICI on repeat scan. Of the 141 admitted patients, five had repeat scan: two normal, one small contusion–managed conservatively, two showed skull fractures | Study centre is major referral centre for paediatric trauma so may be selection bias. No departmental protocol for obtaining a CT scan. Poor documentation of symptoms noted. No formal follow-up |
Roddy et al, 1998, USA | 62 Patients with head injury <16 years, normal neurological exam, normal CT scan | Retrospective review. | Deterioration in CNS exam, new CT findings, need for prolonged (>24 h) hospital stay | No child developed any significant CNS sequelae | Small study. Few infants included. Author suggests that non-verbal children should be excluded from safe discharge. No long-term follow-up |
Mandera et al, 2000, Poland | 166 Patients <18 years, with head injury and GCS >12 on admission | Retrospective review. | Subsequent deterioration | 95 (57%) Patients had GCS of 15 on attendance. Only 28 (17%) had normal CT scan. Six patients with a normal initial scan had a haematoma present on a subsequent scan | High proportion of children had neurological symptoms at presentation. High proportion of abnormal CT scans and need for surgical intervention as tertiary referral centre |
Adams et al, 2001, USA | 1033 Patients <18 years with head injury and GCS of 15. 386 Patients had a cranial CT | Retrospective review. | Length of stay and any required procedures or complications | No neurosurgical intervention required in any of the patients | No breakdown of study results. No follow-up. Focal neurology not mentioned |
Ros and Ros, 1989, USA | 73 Patients <18 years with a history of minor head injury GCS >12, and a normal CT scan | Retrospective review | Complications/adverse events | No neurological complications and no re-attendances following discharge | Small, retrospective study |
Spencer et al, 2003, USA | 197 Patients with head injury, age <14 years and GCS 15, no focal neurological signs on exam and normal CT scan | Retrospective review | Delayed inhospital complications defined as focal neurological deficit, intracranial bleeding, worsening mental status or recurrent seizures | None of the 197 patients had delayed complications resulting from their head injury. Five patients had persistent symptoms: three headache, two vomiting. One patient had repeat scan due to vomiting but this was normal | No long-term follow-up. Included patients with concomitant trauma |
Holmes et al, 2011, USA | 13 543 Patients <18 years presenting after head injury with GCS >13 and normal CT scan to 25 centres over a 2-year period | Prospective cohort study | Need for repeat imaging and neurosurgical intervention | Of 11 058 patients discharged from the ED, 197 (2%) had a subsequent CT or MRI scan, five of these showed an abnormality and no children needed neurosurgical intervention. Of the 2485 patients admitted to hospital, 137 (6%) received a subsequent CT or MRI scan, 16 or these were reported as abnormal and no children required neurosurgical intervention | Only 79% of discharged patients followed up successfully by phone or letter although medical record and morgue review was carried out for the remainder |