I have some concerns about the interpretation of data in the paper by
Aggarwal and colleagues on efficacy and safety of azithromycin for
uncomplicated typhoid fever [1]. While many children in the study sample
had typhoid fever, evidence is unconvincing that all of them had the
specific disease. The inclusion criteria included at least 4 days of fever
with clinical features suggestive of typhoid fever, such as abdominal pain
(present in 69%), diarrhea or constipation (61%), splenomegaly (27%) and
hepatomegaly (73%). While the sensitivity of such a spectrum would be high
to include typhoid fever, the specificity would be quite low, except for
splenomegaly. This suspicion is strongly supported by the very low blood
culture yield (17 children, 15.5%). Moreover, other diagnostic
possibilities were not excluded or apparently even considered.
If the diagnosis is not firm, the drug trial is on
slippery grounds. Azithromycin is well known to be safe. As for efficacy,
azithromycin is active against several pathogens and as pointed out above,
such agents could have been the cause of fever in some children. Its
efficacy evaluation against typhoid fever in this study is not valid since
the case definition was not stringent enough. Moreover, children seen with
fever on the fourth day, with only mild discomfort and no localizing signs
may well be left alone with symptomatic support and close monitoring;
their recovery cannot be attributed to the drug therapy. Many children
with uncomplicated fever have surprised pediatricians by their recovery
without antibiotic treatment, while their blood cultures have yielded
Salmonella typhi. The data on the 17 culture-proven cases, of whom 16
completed the study – their response and final outcome are essential to
consider the efficacy of the drug, but they are not specifically
mentioned. The report says that 5 of them "required add-on antibiotics"
[1], suggesting treatment failure in one-third.
Three questions arise. One, what are the indications to
start an antibiotic in children with fever of less than one week and
without specific diagnosis? Second, what are the criteria to diagnose
typhoid fever when blood culture result is negative and when blood culture
was not attempted? Third, what are the criteria to choose azithromycin in
typhoid fever in preference to other inexpensive oral drugs? These
questions have no answer in the paper. The IAP Task Force’s guideline on
treatment of enteric fever seems to have been ignored by the
investigators; it includes azithromycin as an ‘alternative’ agent for
treating uncomplicated typhoid fever [2]. However, the IAP Task Force did
not provide minimum diagnostic criteria – clinical and laboratory – for
typhoid fever, but apparently alludes that blood culture is essential [2].
In short, the present study cannot be taken as a
precedent to diagnose typhoid fever without defined criteria. The
rationale to choose azithromycin as the first drug to be given when
typhoid fever in ‘suspected’ or even ‘diagnosed’ needs further
clarification. Perhaps azithromycicn is justified when no cause for fever
is available but the child is ill, the cause is suspected to be infection
which could include a variety of agents including S typhi.
References
1. Aggarwal A, Ghosh A, Gomber S, Mitra M, Parikh AO.
Efficacy and safety of Azithromycin for uncomplicated typhoid fever: an
open label non-comparative study. Indian Pediatr. 2011;48: 553-6.
2. Writing Committee. IAP Task Force Report: Management of Enteric
Fever in Children. Indian Pediatr. 2006;43: 884-7.