• 
• 

Three Part Question

Do [children with abdominal pain and HSP] [treated with steroids compared to children treated without steroids] show a more [rapid resolution to their symptoms]?

Clinical Scenario

Hannah is a 7 year old girl with Henoch-Schonlein purpura (HSP). She has a lot of abdominal pain which is not settling with simple analgesia. An ultrasound scan reveals that she does not have an intussusception. The SHO on-call tells you that her handbook of paediatrics says that such pain can be treated with steroids, but is there really any evidence to support this?

Search Strategy

Medline 1966–2004:
Best Bets
Cochrane
Medline: "Henoch Schonlein Purpura" AND "steroids" AND abdominal pain" 21 citations 2 relevant. "Henoch Schonlein Purpura" AND "abdominal pain" 169 citations . "Henoch schonlein purpura" AND "gastrointestinal" 169 citations 1 relevant. "Henoch Schonlein Purpura" AND ("steroids" OR "prednisolone" OR "hydrocortisone" OR "dexamethasone") AND "pain" 26 citations 5 relevant.
Best Bets: "Henoch Schonlein purpura"; match all/any words. none relevant. Steroids abdominal pain"; match all words. none relevant.
Cochrane: "henoch schonlein purpura" (MeSH) 11 hits. "steroids" and "abdominal pain". 43 hits

Search Outcome

6 relevant

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Leung, 2001	Case reports of 2 patients (both aged 5) with HSP and abdominal pain treated with intravenous hydrocortisone	Case series 4	Resolution of pain	Rapid and complete relief of abdominal pain within 10 minutes	2 case reports Rapid relief but pain relapse at later date. Little mention of conventional analgesia
Lin et al, 1998	27 children (6.7±0.5 y) with HSP and abdominal pain treated with corticosteroids	Retrospective study 4	Resolution of pain	Resolution of pain in 2.4±0.2 days	English abstract only. Limited data
Reinehr et al, 2000	101 children (mean 6 y) with HSP. 57 with severe pain or bleeding treated with steroids	Retrospective study	Resolution of pain	Steroid treated children: 77%—pain resolved in 24 h Non-steroid treated: persistent pain for 5 days (median), range 1–28 days	No side effects observed. 1 patient treated with steroids and 2 not treated developed intussusception
Gunasekaran et al, 2000	4 children with confirmed duodenojejunitis and clinical manifestations of HSP without the typical rash Patients followed up for 3 years	Case series	Resolution of pain	Marked improvement in pain in 48 h
			Repeat endoscopic examination at 8–12 weeks	Normal

Comment(s)

Henoch-Schonlein purpura is the most common vasculitic disease in childhood, most commonly affecting the skin, joints, gastrointestinal tract, and kidneys. Gastrointestinal involvement is said to occur in approximately 80% of patients, ranging from mild symptoms such as abdominal pain, nausea, and vomiting, to more severe manifestations such as gastrointestinal bleeding and intussusception. Some textbooks suggest that the abdominal pain of HSP may respond to steroids, with some suggesting that there is a benefit in their use and describing a regimen. No randomised controlled trials have ever been carried out to assess this problem and there have been no systematic reviews to date looking at the available data. The studies that are available include retrospective studies and case series. These studies show that children with HSP who are treated with steroids experience a quicker resolution of their pain than those not treated with steroids. This is seen within 24 hours of commencing treatment in the studies by Rosenblum and Reinehr et al. Although the groups were similar for some characteristics, randomisation and blinding was not carried out—thus there is little to ensure that patients were equal in terms of factors such as severity of illness. While steroids have been described in these studies as having a beneficial effect on abdominal pain, they are also known to have adverse effects, some of which have been noted in these studies—for instance, the masking of associated intra-abdominal pathology such as intussusceptions and bowel perforation. A randomised controlled trial seems the natural next step in order to answer this question. If we assume that a trial looking at the effect of steroids for severe abdominal pain will have a power of 80% at a 5% significance level and assume 15% complete resolution of pain at 24 hours in placebo treated children and 25% resolution of pain in children treated with steroids, we would need 247 children in each group to complete this trial. Larger effects would be easier to detect, but even assuming a doubling of pain relief using steroids we would still need over 100 subjects per arm. A large district general hospital serving a population of 100 000 children would only see 18 children a year with HSP, of whom only six might have severe abdominal pain. It is clear that this has affected why a prospective trial has not been carried out to date, as to do so would involve the detection of a small treatment effect, of an uncommon symptom (severe abdominal pain) in an uncommon condition. Ideally a large multicentre trial is needed, but an alternative approach may be a well designed large cohort study; one possibility may be to conduct it under the aegis of a body such as the British Paediatric Surveillance Unit.

Clinical Bottom Line

Case series and retrospective analyses show an improvement in pain when steroids are given to patients with HSP and abdominal pain. (Grade D) Further studies are needed to look at the magnitude of effects of steroids in alleviating abdominal pain in HSP and also to look at their possible adverse effects. (Grade D) Steroids should be used with caution to alleviate abdominal pain in HSP, particularly with regard to their effect in masking other intra-abdominal pathology. (Grade D)

References

1. Leung SP. Use of intravenous hydrocortisone in Henoch-Schonlein purpura. J Paediatr Child Health 2001;37:309–10.
2. Lin SJ, Chao HC, Huang JL. Gastrointestinal involvement as the initial manifestation in children with Henoch-Schonlein purpura—clinical analysis of 27 cases. Chung Hua Min Kuo Hsiao Erh Ko I Hsueh Hui Tsa Chih 1998;39:186–90.
3. Reinehr T, Burk G, Andler W. Does steroid treatment of abdominal pain prevent renal involvement in Henoch-Schonlein purpura? J Pediatr Gastroenterol Nutr 2000;31:323–4.
4. Gunasekaran TS, Berman J, Gonzalez M. Duodenojejunitis: is it idiopathic or is it Henoch-Schonlein purpura without the purpura? J Pediatr Gastroenterol Nutr 2000;30:22–8.