BACKGROUND. Epistaxis in childhood is common but unusual in the first years of life. Oronasal blood has been proposed as a marker of child abuse.
METHODS. We performed a retrospective review of all hospital notes of children in the Lothian region of Scotland who were <2 years of age and in whom facial blood had been recorded over a 10-year period.
RESULTS. There were 77173 accident and emergency department attendances with 58059 admissions during the 10-year study period in children <2 years of age; 16 cases of nose bleed and 3 cases of hemoptysis were recorded. All cases of hemoptysis were associated with significant bouts of coughing and respiratory infections. Epistaxis in 8 cases was associated with visible trauma and in 4 cases with thrombocytopenia (secondary to malignancy in 3). In 2 cases, an associated apparent life-threatening event was described, and in 2 cases there was a coincident upper respiratory tract infection. Review of previous and subsequent history suggested 7 cases of “accidental” injury that might have been caused by abuse. These cases are described here. All children who presented with this problem to the accident and emergency department had been admitted for observation or management.
CONCLUSIONS. Epistaxis is rare in the accident and emergency department and hospital in the first 2 years of life and is often associated with injury or serious illness. The investigation of all cases should involve a pediatrician with expertise in child protection.
The finding of blood on the face of a young infant must be alarming for a parent. The differential diagnosis contains serious and life-threatening conditions and has also been associated with child abuse.1–3 Little has been published about the epidemiology of nose bleeds or bleeding from the mouth in early life, although it is regarded as rare.4 This study reports the epidemiology of oronasal bleeding in those who presented to the accident and emergency department (AED) and/or were admitted at a large district general hospital for children that also provides specialty services for southeast Scotland.
All AED attendances and admissions at the Royal Hospital for Sick Children have had their symptoms and diagnoses coded for the last 12 years (according to the International Classification of Diseases, 10th Revision [ICD-10]). Medical charts within a 10-year period (April 1, 1996, to March 31, 2006) were reviewed to find diagnostic codes that the hospital coding clerks said that they could have used to record the presence of facial blood. These codes were R04 (epistaxis/nasal bleeding/nose bleeds), R04.2 (hemoptysis), R04.8 (pulmonary hemorrhage, older than newborn), and K13.7 (other and unspecified lesions of oral mucosa).
The notes of all patients <2 years of age and with residence in Lothian, Scotland, were reviewed independently by 3 pediatricians, 2 of whom were lead pediatricians in child protection of many years standing. Data were extracted about the incident of the hemorrhage (if confirmed) and the reasons for previous and/or subsequent admissions, and the date of the last hospital contact was noted. A cause for the bleeding was allocated (usually the cause recorded on the index admission). For cases in which there was delay in consulting medical services after an acute injury or a witness at the time of presentation considered the story to be incompatible or variable (phrases used concurrently in the notes are in quote marks in Table 1), abuse was considered possible. In addition, each of us independently made a decision as to whether abuse was possible (footnote “a” in Table 1) on the basis of both the event and the previous and subsequent history. If only 1 or no pediatrician felt that abuse was possible, abuse was said to be unlikely; if 2 or 3 pediatricians independently were concerned about possible abuse, abuse was said to be possible.
In each case we made a decision with the information available at the time on whether (by current standards) a multiagency discussion would have been appropriate during the index admission.
The total population for the Lothian area of Scotland is 790000, of whom 17000 are <2 years of age. During the 10-year review period, there were 350000 AED attendances and 207289 admissions from the Lothian region of southeast Scotland: 38215 AED attendances and 34590 admissions were of children <1 year old, and 38958 AED attendances and 23469 admissions were of children between the ages of 1 and 2 years. It is unlikely that infants with this problem in the Lothian region would be seen outside the Royal Hospital for Sick Children.
There were 888 attendances or admissions during this time for the R04 ICD-10 code, which includes epistaxis, pulmonary hemorrhage, and hemoptysis. There were 62 attendances or admissions for lesions in the mouth (ICD-10 code K13.7). Of the children and infants who were <2 years old, there were only 30 cases, and all were admitted to the hospital. The AED and inpatient notes of all these children were evaluated. Nine cases referred from outside the region were excluded (because we had no denominator information), which left 21 cases from within the Lothian region. Three cases of hemoptysis occurred during the second year of life (at ages 378, 499, and 502 days of life); in each case the hemoptysis occurred in an infant who had been to their general practitioners in the recent past with chronic and paroxysmal coughing. These cases were accepted as described, and there seemed to be nothing, even in retrospect, that would suggest further problems. No case of pulmonary hemorrhage outside the neonatal period occurred in a Lothian infant. Two Lothian cases coded as K13.7 were found, both in patients with malignancy who were receiving chemotherapy. The lesions were described in the notes as either mucositis or ulcers, and in neither case was there associated bleeding.
There were 16 cases of epistaxis: 8 in children <1 year old and 8 in children between 1 and 2 years old. The detail of each case is shown in Table 1. Eleven of 16 cases had a cause allocated at the time (2 secondary to an upper respiratory tract infection), and for 4 children with recurrent epistaxis a cause was never established (cases 3, 6, 9, and 16) even after ear, nose, and throat (ENT) referral. In 8 cases there was obvious associated injury. All injuries were accepted as accidental at the time of their presentation, although in 4 cases (1, 3, 7, and 14) child abuse was considered by at least 1 staff member. Review of the notes for delay in presentation, or subsequent history of poorly explained trauma, suggested that 5 injuries might have been associated with physical abuse (cases 3, 7, 8, 14, and 15). Two other children without obvious injury at presentation may also have been abused (cases 1 and 16), and 1 (case 13) was thought after review by the 2 child protection experts to be a possible case of fabricated or induced illness. Four infants had thrombocytopenia at the time of the nose bleed (1 with idiopathic thrombocytopenic purpura and 3 with malignancy and receiving chemotherapy, 1 of which was associated with injury). The child in case 1 had possibly been suffocated (overlain) and certainly had features that suggested that there had been an ALTE. In case 14, the child had a disturbed conscious level, which may have been a result of head injury or an ALTE.
The child protection experts decided, on the basis of the information available at the time of presentation of the index incident, that by current standards a multiagency discussion would have been appropriate in 50% of the presentations.
For 12 years all children who attended the Royal Hospital for Sick Children AED and/or were admitted have had their diagnoses coded by using the ICD-10 when they left the hospital. ICD-10 code R04 was most helpful and included pulmonary hemorrhage outside the neonatal period, epistaxis, and hemoptysis. ICD-10 code K13.7 indicated ulcers and mucositis in the mouth that, after review of the notes, were not associated with bleeding. Hemoptysis and pulmonary hemorrhage seemed to be extremely rare and genuine events.
Whereas overall 1 in 400 children who attend the AED will present with an epistaxis, attendance for this problem is rare for children under the age of 2 years (1 in 6400 <1-year-olds and 1 in 6500 1- to 2-year-olds), and it is very likely that they will be admitted. In an average year in Lothian, there are ∼1.6 nosebleeds for a population of 2-year-olds (17000), which gives 0.94 nosebleeds per 10000 of this population.
This study shows that blood on the face from the nose or mouth in the first 2 years of life is rare and that spontaneous epistaxis is also rare.4 This study also confirms that this sign is associated with some serious underlying diagnoses. In the absence of physical injury, it would be appropriate to check the patient's platelet count and perform a clotting screen as serious illness (malignancy or chronic liver disease5) comes into the differential diagnosis.
Oronasal hemorrhage has been described in a number of case reports of imposed suffocation in which the perpetrator later confessed.1,2,6–11 It was part of the history of 11 of 38 parents of children with recurrent ALTEs who were seen by covert video surveillance to suffocate their infants. It proved to be a highly significant discriminator between this group and a control group found to have a natural cause for repeated ALTEs.2 This series, however, was highly selected, and a population denominator was impossible to define. Krous et al12 found facial blood present at autopsy in 7% of 406 cases of sudden and unexpected deaths in infancy, and in 14 of 28 cases the blood was present before or in the absence of cardiopulmonary resuscitation. The authors suggested that when this sign is found at autopsy, it is reasonable to question whether the diagnosis is truly sudden infant death syndrome.
The mechanism of oronasal hemorrhage is often unclear. Blunt trauma as part of accidental injury may be obvious and may also be apparent when there are other signs of physical abuse (eg, bleeding gums, pinch marks on the nose).13 When found in association with an ALTE, hemorrhagic pulmonary edema secondary to upper airway obstruction has to be considered. Case reports of confessed suffocation and strangulation suggest that this is a possible mechanism and usually produces frothy or pink blood-stained fluid8,9 and sometimes produces respiratory distress.9 Pathologists clearly recognize that the presence of either large amounts of blood in the lungs or intraalveolar siderophages in cases of sudden infant death syndrome may be a marker for imposed suffocation.11,14 For cases in which “blood and mucus” is described, it is less clear that the origin may be the lungs.7
No children described in this report suffered injury from definite abuse or from injury that was sufficiently worrying to trigger child protection proceedings at the time of injury, although by current standards a multiagency discussion would have been appropriate in half of the cases. The circumstances of the index event or hindsight available from subsequent events might raise child protection anxieties in 44% of them. In the first case, it was thought more likely that it was an overlying event rather than suffocatory abuse. In many of the other cases, the injuries were in a child at the toddling age, when injuries from falls are not infrequent. It is the repeated nature, varying stories, or other oddity that suggests that reconsideration may be warranted. The pediatrician has a dilemma. Should the concerns be voiced, leading to anger and affront in the parents, or should injuries be “accepted” as accidental and inevitable in the small child, leaving them at further risk? The law in the United Kingdom is unclear, although the lay press often takes a simplistic view of what is always a complex situation.
Do general practitioners refer such cases to ENT departments rather than pediatricians? All children who are referred for ENT consultation in Lothian are evaluated at the Royal Hospital for Sick Children and are entered into our database. Only 2 cases (6 and 9) of the 16 were seen in the ENT department; in neither case was a diagnosis made.
Two other issues need to be addressed to determine the completeness of these discussions and the final impact of the problem, neither of which can be answered by this study. First, do parents of a young child discovered with facial blood or a nose bleed always seek medical help? If there is a strong family history of nose bleeds in older children, will the parents manage it from their previous experience and a knowledge that ensuing harm is unlikely? Second, do primary care practitioners always refer a child of this age with facial blood for a hospital opinion? We have no data to answer either of these questions, but it is noteworthy that all infants who attended our AED during this period and with this problem were admitted for observation, which indicates that the AED staff took each individual event seriously.
Facial blood from a nose or mouth bleed or pulmonary hemorrhage is rare in the first 2 years of life. The problem should be taken seriously by medical staff, because there are important underlying conditions that may warrant urgent evaluation and management. A pediatrician with some experience in child protection should always be involved, and cases should not be referred directly to ENT surgeons.
We thank Tony Wright, for the 4D data analysis, and the Coding and Records staff at the Royal Hospital for Sick Children (Edinburgh).
- Accepted August 10, 2007.
- Address correspondence to Neil McIntosh, DSc(Med), Department of Child Life and Health, University of Edinburgh, 20 Sylvan Place, Edinburgh EH9 1UW, United Kingdom. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Truman TL, Ayoub CC. Considering suffocatory abuse and Munchausen by proxy in the evaluation of children experiencing apparent life-threatening events and sudden infant death syndrome. Child Maltreat.2002;7 :138– 148
- ↵Southall DP, Plunkett MC, Banks MW, Falkov AF, Samuels MP. Covert video recordings of life-threatening child abuse: lessons for child protection. Pediatrics.1997;100 :735– 760
- ↵Burton MJ, Dorée CJ. Interventions for recurrent idiopathic epistaxis (nosebleeds) in children. Cochrane Database Syst Rev.2004;(1) :CD004461
- ↵Makar AF, Squier PJ. Munchausen syndrome by proxy: father as a perpetrator. Pediatrics.1990;85 :370– 373
- ↵Rubin DM, McMillan CO, Helfaer MA, Christian CW. Pulmonary edema associated with child abuse: case reports and review of the literature. Pediatrics.2001;108 :769– 775
- Piatt JH Jr, Steinberg M. Isolated spinal cord injury as a presentation of child abuse. Pediatrics.1995;96 :780– 782
- Copyright © 2007 by the American Academy of Pediatrics