Sickle cell anaemia: febrile children treated as outpatients were more likely to be admitted in the next two weeks.

Clinical bottom line (level 4)

  1. Children with sickle cell disease may be treated as outpatients for febrile episodes if they are reasonably well. However, the chance of admission in the next two weeks was increased (NNF = 4 for 2 weeks) .
Wilimas et al: New England Journal of Medicine 1993; 329 (7): 472-476
Expires April 2002

The study

Unblinded concealed quasi-randomised trial without intention-to-treat
Setting: emergency room of teaching hospital, USA

86 patients (aged range 6 months to 12 years; median 24 months, 54% male) children with sickle cell anaemia with temperature >38 ° C

Excluded if
  • seriously ill appearance (blood pressure <70 mmHg if one year old or <70 mmHg+2 x child's age in years older; poor perfusion; temperature >40 ° C; white cell count >30 or <5; history of pneumococcal sepsis)
  • severe pain
  • dehydration
  • infiltration of lung segment or larger
  • penicillin or cephalosporin allergy


    • Control Group: (n = 42, 42 analysed): in patient care, based on 'plan of attending haematologist'; typically one additional dose of ceftriaxone
    Experimental Group: (n = 44, 42 analysed): outpatient care- followed up 20-30 hours later, or sooner if any problems; second dose of ceftriaxone 50 mg/kg.
    All children had a pneumococcal vaccine and penicillin prophylaxis until age five (and were then entered into the study). All had i.v. ceftriaxone 50 mg/kg soon after arrival in the emergency room. Patients were only given oral antibiotics if the source of infection was known.
    98% followed for 2 weeks

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)    		ARR
    (95% CI)    		NNH
    (95% CI)
    hospitalised 2 weeks 1
    (2.4%)    		 11
    (26%)    		 -1000%
    (-8000% to -49%)    		 -23.8%
    (-37.9% to -9.74%)    		 4
    (3 to 10)

    Outcome Control Group
    (SD)    		 Experimental Group
    (SD)    		 Mean Difference
    (95% CI)
    mean cost per episode ($) 2744
    ()     		1549
    ()     		1195
    ( to )

  • The majority of outpatients who were admitted were for reasons unrelated to their disease.
  • 70/233 (30%) of patients admitted were high risk- 7/86 (8%) had bacteraemia. No randomised patients had bacteraemia.
  • Mean hospital stay was three days (range ine to six days). High risk patients had a mean stay of four days (range one to eight); p<0.001.

    • Comments

    1. Quasi-randomisation- patients were allocated to groups depending on the number of patients already allocated to the group.
    2. Two patients in the outpatient group failed to return for follow-up.
    3. These results are from a centre with experience in sickle disease; those who are less familiar may be wise to admit children more often.

    Citation

    1. Wilimas JA, Flynn PM, Harris S, et al: A randomized study of outpatient treatment with ceftriaxone for selected febrile children with sickle cell disease. New England Journal of Medicine 1993; 329 (7): 472-476
    Contributor: Chris Ball and Clare Wotton, May 2000
    Reviewer: Sudha Chaudhari

    Clinical Question.
    Patient sickle cell anaemia
    Intervention or Exposure out patient care
    Comparison in patient care
    Outcome hospitalisation