Statistics from Altmetric.com
It’s 2am and you are called to review a ‘well-looking child’ in the emergency department who has presented with a new non-blanching rash. He has been hot at home with some coryzal symptoms. Mum is worried, she thinks the rash has spread in the last hour!
What are you going to do?
In this article, we discuss the aetiology and initial assessment of non-blanching rashes in children.
Non-blanching rash (NBR) is a term for any rash in which the colour is unchanged with direct pressure. The presence of a NBR is of concern to both parents and clinicians as it is associated with a wide range of underlying diagnoses, some of which are life threatening. The term is usually used to refer to the presence of petechiae/purpura (figure 1), and in this form it is a relatively common presentation to the emergency department (ED), accounting for around 2% of all attendances.1 2
In this article, we discuss the aetiology and an initial assessment of NBR in children.
Any serious bacterial infection (SBI) can result in a NBR via disseminated intravascular coagulation (DIC). Some infections, however, feature a NBR as an early sign. The most common infections associated with a NBR as an earlier sign are as follows.
Enterovirus and adenovirus are the most common infectious causes of NBR in children.3
Of these infections, MD is arguably the one we worry about most in the UK but how commonly is MD responsible for a NBR in a child?
This is surprisingly difficult to answer because studies are difficult to compare due to their heterogeneity. For example, one study looking at all presentations of fever (>38°C) and NBR presenting to the ED found that only 1% of children had MD as the cause.2 In contrast, studies of hospitalised children with fever and NBR have reported rates as high as 23%.3
A mechanical cause is identified in almost a quarter of NBR in children,3 the most common being straining, coughing or vomiting. This causes raised pressure within the superior vena cava (SVC), with consequent pinpoint petechiae in the distribution of the SVC alone (above the nipple line).1
However, the early stages of serious bacterial infections may present with a localised NBR meaning that a mechanical cause is often arrived at through a process of exclusion.
Direct trauma can result in bruising that can appear identical to a true NBR. There is usually a clear history of trauma. In cases where a traumatic cause is likely, it is important to consider safeguarding. This is especially important when lesions are localised to the genital area, buttocks, are unusual or linear, or when the history is unclear.7
Henoch-Schonlein purpura (HSP) is the most common vasculitic cause in children, with other less common causes including atypical Kawasaki disease, polyarteritis nodosa and antineutrophil cytoplasmic antibody-related vasculitis.8
HSP typically presents with palpable purpura found in a gravity dependent distribution—classically on the legs and buttocks.9
The main haematological causes likely to present are thrombocytopenia, leukaemia and coagulopathy.
Idiopathic thrombocytopenia purpura (ITP) is the most common haematological cause and presents with the sudden development of a NBR. In ITP, a full blood count (FBC) should show isolated thrombocytopenia, and a blood film should be normal other than thrombocytopenia.10 11
Other rare causes of thrombocytopenia include:
infection (eg, Epstein-Barr virus12);
drug induced (eg, vaccination, heparin, non-steroidal anti-inflammatory drugs, ranitidine8 13–15);
thrombotic thrombocytopenic purpura;
bone marrow failure.
Children with undiagnosed haematological malignancies can present with a NBR, either as an isolated finding or in conjunction with other features such as weight loss, fatigue, pallor and general malaise.13 16 Clinical features such as lymphadenopathy, hepatomegaly, splenomegaly, jaundice and anaemia should be sought,13 and any child with an abnormal blood film or deficiencies in multiple cell lines should be discussed with the local haematology service.
Coagulopathy is a rare cause of a NBR in children (0.4% of cases).3 A family history of coagulation disorders or a long history of easy bruising and/or a NBR that remains unexplained may suggest an underlying coagulation disorder.
It is worth considering whether a well-child’s rash is in fact a normal variant. A study of infants attending routine health checks found that petechiae were commonly identified in well infants with over one-fourth having one or more petechiae.17
Assessing the child
No guideline or algorithm will ever perform perfectly due to the range of possible causes. The approach discussed here is designed to assist the thought process but is not a substitute for clinical reasoning or experience.
The steps involved are outlined in figure 2 and include:
An initial assessment of wellness to identify those requiring immediate treatment.
An attempt to make a positive diagnosis.
If no positive diagnosis can be made then consider if it is appropriate to discharge.
Is the child well?
Any child who appears unwell with a NBR should be presumed to have a SBI and be treated accordingly.1–5 In the context of NBR, presenting features of irritability, lethargy or a prolonged capillary refill time confer a significantly increased risk of SBI1–5—children with these features should be treated as per national guidance immediately.
Meningococcal disease (meningitis (bacterial) and meningococcal septicaemia in under 16 s: recognition, diagnosis and management).
Sepsis: recognition, diagnosis and early management.
The presence of fever is a key component of the history and examination, though it should not be relied on in isolation to make clinical decisions. While not all children with fever and NBR will have SBI, it is also important to note that not all children with SBI and NBR present with a fever—up to 20% of cases of MD have no fever at presentation to ED.1 18 While it may be reasonable to withhold antibiotics from children who appear well, it is important to note that children can initially appear well and deteriorate, mandating a period of active observation.
Can I make a positive diagnosis?
If the child appears well, it is still important to attempt to make a diagnosis. While undertaking this process, it is important to carefully monitor the child for signs of deterioration. As even well-appearing children may be harbouring an occult SBI.
When searching for a diagnosis, a number of tests can be considered (table 1). National Institute for Health and Care Excellence (NICE) meningitis (bacterial) and meningococcal septicaemia in under 16 s: recognition, diagnosis and management guidance includes the assessment of NBR in febrile children and advises.19
C-r eactive p rotein(CRP) ;
c oagulation screen;
b lood culture;
w hole-blood PCR for Neisseria meningitidis ;
b lood glucose;
b lood gas.
These investigations focus on the diagnosis of MD. While this is important, table 1 below outlines some additional investigations and anticipated results for other causes of a NBR:
Special considerations are required for children presenting with purpura. These children are at higher risk of MD and other SBI than those with petechiae alone, and some therefore advocate that all children with purpura should be treated for suspected MD/sepsis.1–5 This approach, while safe, leads to overtreatment of children with HSP.5 There are no unifying diagnostic criteria for HSP, but the presence of palpable purpura in a characteristic distribution in an otherwise well child suggests HSP as opposed to MD.5 However, it is known that even experienced paediatricians may misdiagnose MD as HSP, leading to treatment delays.5
A list of common causes and possible investigations are outlined below in table 1.
Can I rule out serious illness?
If a positive diagnosis cannot be made but the child otherwise appears well, then the more difficult question is ‘Can I rule out SBI and other serious causes?’
Proving a negative is always more difficult in medicine and this is where the real challenge in managing childhood NBR occurs, given that so few have a serious underlying cause. Where determining a cause is not possible, the challenge is deciding who is safe to be discharged and who should be treated3? If a decision is taken to discharge a child, it is important to provide clear advice to return if there is any: deterioration in the child’s health, spread of the or change of the rash.
The current best evidence for the management of this group comes from the Newcastle-Birmingham-Liverpool algorithm (NBL).5 This algorithm has been validated with a reported sensitivity of 100% and a specificity of 82% for the diagnosis of MD.5 In the NBL algorithm, a child can be discharged if the child remains well, has no purpura, no spread of the rash over 4–6 hours of observation and a CRP <6 and white cell count 5–15 10^9/L.5 This approach outperformed current NICE guidance in a comparative validation exercise, with NICE guidance displaying a sensitivity of 97% and specificity of 50%.5 The difference in the performance of the two algorithms was statistically significant (P<0.001).5 Both NICE and NBL algorithms are designed as ‘rule-out’ algorithms and as such both are highly sensitive but poorly specific. This means that very few cases of MD will be missed but that many children will receive unnecessary treatment.
NBRs are a common reason for children presenting to healthcare, often with non-specific findings. While SBI is rare, it is important to promptly identify and treat those at greatest risk. For well-appearing children, a structured approach can lead to a positive diagnosis in many, coupled with safe discharge decision-making.
Test your knowledge
A 3-year-old child presents with a petechial rash seen on the trunk. He is otherwise well apart from being a bit tired over the last few weeks. He is afebrile and examination is otherwise unremarkable.
Full blood count: haemoglobin 64 g/l; white cell count 33 10^9/L; platelets 23 10^9/L; neutrophils 7 10^9/L.
Which of the following is the best next step?
Intravenous ceftriaxone and send blood cultures
Reassure and discharge
Discuss with haematology
An 8-month-old girl presents with a purpuric rash over her extremities with a temperature of 39.6. She has a capillary refill time of 5 s and is lethargic. Heart rate is 160 and blood pressure is 83/56 mm Hg.
What is the most appropriate initial management?
Lumbar puncture and blood cultures
Contact the transport team
Fluid bolus and intravenous antibiotics
In an 8-year-old child with palpable purpura for 3 days on the legs and buttocks with normal observations, the most likely diagnosis is?
Antineutrophil cytoplasmic antibody-related vasculitis
A 6-month-old with three petechial spots is bought to accident and emergency department by his parents as they noted that the spots did not disappear with the ‘cold glass test’. They have been there for 1 day and have not spread. He had normal observations and appeared well. At 6 hours, his observations remained within the normal limits. His C-reactive protein was <6, white cell count 10 10^9/L and no further petechiae have appeared.
What is the next most appropriate step in management?
Contact social services
Discharge with safety netting advice
Discharge with oral antibiotics
Lumbar puncture and blood cultures
A 9-month-old child comes in with a 5-day history of cough and coryza. You diagnose bronchiolitis and they are medically fit to be discharged. As mum is changing the nappy, you notice a linear bruise on one buttock.
What would be the next appropriate action(s)?
Lumbar puncture and blood cultures
Top to toe examination and consider safeguarding background checks
Discharge home with safety netting advice
Topical emollient and discharge
Answers can be found after the references.
C Discuss with haematology
E Fluid bolus and intravenous antibiotics
B Henoch-Schonlein purpura
B Discharge with safety netting advice
C Top to toe examination and consider safeguarding background checks
Contributors TW conceived the idea and wrote the majority of the article. ED contributed to sections on haematological causes, vasculitic causes and mechanical causes. ML provided a comprehensive review of the article and expert analysis. All authors agreed the final version.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.