Table 1
| Author, date and country | Patient group | Study type (level of evidence) | Outcomes | Key results | Study weaknesses |
|---|---|---|---|---|---|
| *Our calculation. | |||||
| Gorman DF, 1987, England1 | 5768 head injuries in all age groups | Retrospective case note review | Presence of skull fracture | More common in vomiting children (p<0.005) | Not specific to children |
| 6685 head injuries in all age groups | Prospective patient study | 7% of all patients vomited | Skull fracture is only a proxy outcome for intracranial problems | ||
| 25.7% of patients with skull fracture vomited | |||||
| Hugenholtz H, et al, 1987, Canada2 | 96 children (GCS 13–15) <16 years | Prospective consecutive case series retrospective study of case notes over the previous two years | Presence of skull fracture with GCS >12 | No difference | Small sample size |
| 29 children (GCS 8–12) <16 years | Presence of skull fracture GCS 8–12 | Less common in vomiting children | Skull fracture is only a proxy outcome for intracranial problems | ||
| Chan KH, et al, 1989, Hong Kong3 | 12 072 paediatric head injury cases <16 years | Retrospective case note review | Probability of IC complication with impaired conciousness + skull fracture + | 62% if vomiting v 74% if not vomiting | Retrospective audit. |
| Development of intracranial complications manifested during the first 48 hours of injury | Probability of IC complication with normal consciousness + no skull fracture + | 0.08% if vomiting v 0.14% if vomiting | Identification of risk factors is dependant on accurate documentation (which is unlikely) | ||
| Probability of IC complication with impaired conciousness + no skull fracture + | 12% if vomiting v 18% if no vomiting | ||||
| Probability of IC complication with no impaired conciousness + skull fracture + | 1% if vomiting v 2% if no vomiting | ||||
| Ando S, et al, 1992, Japan4 | 147 patients with head injury, all ages analysed by age group | Prospective cohort study | Presence of skull fracture | No difference between children vomiting and not vomiting | Small study |
| Presence of IC haematoma on CT | No difference between children vomiting and not vomiting | Results not specific to paediatric patients | |||
| Dietrich AM, et al, 1993, USA5 | 324 consecutive trauma patients in an urban childrens hospital requiring CT. Mean age 7.1 years | Prospective cohort study | Risk of IC haematoma age <2 | 76/191 patients with no IC lesion had vomited | Small cohort, low event rate |
| 10/36 patients with IC lesion had vomited | |||||
| Risk of IC haematoma age >2 | 12/39 patients with no IC lesion had vomited | ||||
| 0/3 patients with IC lesion had vomited | |||||
| Duus BR, 1993, Denmark6 | 1876 patients mean age 27.5 (19.9 years) | Retrospective case note review | Presence of IC complication | 1.2% if vomiting v 0.2% if not vomiting | Intracranial complication not defined. Retrospective All age groups |
| Schunk JE, et al, 1996, USA7 | 508 patients aged <18 undergoing CT for head trauma. 179 excluded for decreased GCS, depressed skull space, bleeding diathesis or developmental delay | Retropsective case note review | Abnormal CT findings | 5.5% if vomiting v 3.4% if not vomiting | No protocol for CT request, inclusion based on physician request. |
| Referral bias (major trauma centre) | |||||
| Arienta C, et al, 1997, Italy8 | 10 000 patients with head injury aged between 6 and 95 years (median age 31) | Prospective cohort study | Abnormal CT result | 4 of 213 patients with single episode of vomiting had abnormal CT result | Not specific to the paediatric population |
| 6 of 14 patients with repeated vomiting had an abnormal CT result | Low event rate | ||||
| Hsiang JN, et al, 1997, Hong Kong9 | 1360 patients with mild head injury older than 11 years of age | Prospective cohort study | Radiographic abnormailty in GCS 13 group | 4 patients with vomiting v 11 patients with no vomiting (p=1) | Not specific to paediatric population |
| Radiographic abnormality in GCS 14 group | 8 patients with vomiting v 16 patients with no vomiting (p=0.68) | ||||
| Radiographic abnormailty in GCS 15 group | 30 patients with vomiting v 93 with no vomiting (p=0.924) | ||||
| Miller EC, et al, 1997, United States10 | 2143 patients of all ages with a history of head injury within 2 hours of arrival at the emergency department | Prospective cohort study | Abnormal CT | 15% if vomiting v 5% if not (p<0.001) | Not specific to paediatric population |
| 20% if nauseous v 9% if not (p<0.001) | |||||
| Quayle KS, et al, 1997, USA11 | 322 consecutive paediatric patients with head injury | Prospective cohort study | Odds ratio for vomiting predicting intracranial injury | 1.51 (95% CI=0.67, 3.37) | Non-trivial injuries excluded. Resultant event rate for IC injury is therefore increased. Not all patients had the gold standard investigations |
| All patients had radiography and CT | Postive predictive value for vomiting predicting intracranial injury | 10.9% | |||
| Negative predictive value for vomiting predicting intracranial injury | 92.5% | ||||
| Nee P, et al, 1999, UK12 | 5416 consecutive patients with head injury, over one year period | Prospective cohort study | Incidence of vomiting in children | 12% | Skull fracture is only a proxy outcome for intracranial problems. |
| Sensitivity of detecting skull fracture if child and vomiting | 33.3% | Methods suggest that additional follow up data were collected, but it is not reported. | |||
| Specificity of detecting skull fracture if child and vomiting | 93.3% | ||||
| Likelihood ratio for child and vomiting* | 4.9 | ||||
| Brown FD, et al, 2000, UK13 | 563 patients aged 0–13 with minor head injury presenting to a paediatric A+E | Prospective cohort study | Incidence of vomiting | 15.8% | Only minor head injury patients included. Not all patients were radiographed or scanned. Very few patients with significant intracranial pathology |
| Incidence of skull fracture | <1% | ||||
| Incidence of skull fracture + vomiting | 0% | ||||
