Page 125 - 15Diarrhoeaandvomiting
P. 125
Diarrhoea and vomiting caused by gastroenteritis in children under 5 years
participants was 6 years (range 0–15 years). In total, 33/278 patients developed complete HUS
and 4/278 developed incomplete HUS. There were three fatalities.
Children who reportedly vomited (29/153) were statistically significantly more likely to develop
HUS than those who had not vomited (8/125) (RR 3.0; 95% CI 1.4 to 6.2). Although more children
who had bloody diarrhoea or fever developed HUS, these differences were not statistically
significant (RR 2.0; 95% CI 0.5 to 7.7, and RR 1.8; 95% CI 0.8 to 4.1, respectively).
To investigate early predictors, the risk of HUS was evaluated according to clinical outcomes
measured within the first 3 days of illness. Vomiting remained a statistically significant risk
factor in this time interval (RR 1.9; 95% CI 1.0 to 3.5) and the association was modified by age.
Vomiting in children younger than 5.5 years was strongly associated with HUS development
(RR 3.5; 95% CI 1.4 to 9.4), but the association was not evident in children older than 5.5 years
(RR 1.0; 95% CI 0.4 to 2.4).
The use of medications was also analysed. In total, 50 children received a range of antibiotics
in the first 3 days of illness. These children were more likely to live in a household with annual
income over $29,000 (RR 1.7; 95% CI 1.0 to 2.8). Eight of these children went on to develop HUS
compared with 28/218 who did not receive antibiotics (P = 0.56). In total, 31 children received
antimotility agents in the first 3 days of illness. Six went on to develop HUS, compared with
20/234 who received no antimotility treatments (P = 0.10). There was no statistically significant
difference in the development of HUS in children who received adsorbant and antimotility drugs
compared with those who did not (P = 0.26).
There were no statistically significant associations between HUS development and haematocrit,
platelet count, blood urea nitrogen, segmented neutrophils count or band forms at presentation.
However, children who had a white blood cell count of over 10 500 per microlitre were at
increased risk of developing HUS (RR 5.2; 95% CI 1.6 to 17.0; P < 0.01), and for those with a
white blood cell count of over 13 000 per microlitre this risk was larger (RR 7.2; 95% CI 2.8 to
18.5; P < 0.01).
Evidence summary
There was consistent evidence from two studies that a raised white blood cell count in children
with E. coli O157:H7 was a risk factor for the subsequent development of HUS. In one study,
vomiting in children younger than 5.5 years was strongly associated with the risk of developing
HUS. However, there was conflicting evidence on the effect of antimicrobials. One study
reported that antimicrobial treatment was an independent risk factor for HUS but the study lacked
precision for this finding. The second study did not find treatment with antimicrobials or with
antimotility agents (with or without adsorbant agents) was associated with increased risk of HUS.
7.9.2 Salmonella
One retrospective review 158 conducted in Malaysia sought to characterise the incidence, risk
factors and outcome of invasive non-typhoid salmonella gastroenteritis in children aged between
1 month and 14 years. [EL = 2+]. Participants were 131 children with positive stool cultures
for salmonella species but no second enteropathogen, seen in an outpatient department. Of
these, 67% of children were younger than 1 year. Demographic, clinical (diarrhoea, vomiting,
fever, hydration status), blood and stool outcome measures were recorded from case notes and
examined. Overall, 124 children were found to have non-invasive salmonellosis and seven had
invasive complications (five bacteraemia, two meningitis). Three risk factors were identified for
the development of invasive salmonellosis. In total, 45 (85%) of the 124 with non-invasive
disease were younger than 6 months compared with six of the seven with invasive disease
(P < 0.01). Only 53 of those in the non-invasive group had a temperature of over 38 °C,
compared with all seven of the invasive group (P < 0.003). Dehydration was found in five of the
seven with invasive complications, but in only 25 of the 124 with non-invasive disease. One
infant with bacteraemia died while awaiting a blood culture result. The authors suggested that
empirical antibiotic treatment should be given to infants younger than 6 months who are febrile
and dehydrated.
100