Study did not prove benefit
Editor--Hossain et al conclude that vitamin A along with standard antibiotic reduces the severity of acute shigellosis and that vitamin A supplementation should be added to the standard treatment for acute shigellosis.[1] We disagree with those conclusions.
The benefit of vitamin A as adjuvant treatment for shigellosis can be judged only in a study in which patients receive adequate antimicrobial treatment.[2] In Hossain et al's study 64% of patients were infected with strains of shigella resistant to nalidixic acid, which was used to treat study patients. It is unclear how many patients were given an alternative effective antibiotic or when treatment was changed.
The inadequacy of antimicrobial treatment in this study is reflected in clinical cure rates of less than 50% in patients who did and did not receive vitamin A supplements. That is not acceptable as adequate antimicrobial treatment routinely achieves a cure rate of 65% or more.[3 4] Indeed, Hossain and colleagues calculated their sample size based on a 98% cure rate in the vitamin A group.[1]
Their definition of cure included the absence of mucus or blood in stools. These measures are not good indicators of overall clinical status and are subject to variation between observers. It is important to know which component of the definition of clinical cure (three or fewer formed stools, no visible stool blood or mucus, afebrile) used in the study accounted for the difference in cure rates between groups and if there were any differences in more objective outcome measures, such as total number of stools during the study or duration of fever.[3 5]
The authors contend that the differences in reported outcome between groups may have been because vitamin A supplementation stimulated phagocytosis and cell mediated killing of pathogens. The similar rate of clearance of shigella observed in the two groups seems to contradict this contention. The rapid divergence (by 24 hours) between groups in the proportion of patients with clinical cure suggests a more rapid response than could reasonably be expected to be achieved by vitamin A "enhanced repair of the micro-ulcers in the gut epithelium."
Vitamin A has an important role in public health in developing countries. The rationale for its use, however, must come from well conducted studies in which children receive effective primary treatment for their illness (in this case effective antibiotics), and for which there are clear, reproducible, and clinically relevant definitions of treatment outcome.
[1] Hossain S, Biswas R, Kabir I, Sarker S, Dibley M, Fuchs G, et al. Single-dose vitamin A treatment in acute shigellosis in Bangladeshi children: randomised double blind controlled trial. BMJ 1998;316:422-6.
[2] Salam MA, Bennish ML. Antimicrobial therapy of shigellosis. Rev Infect Dis 1991;13(suppl 4):S332-41.
[3] Salam MA, Bennish ML. Therapy for shigellosis. I. Randomised, double blind trial of nalidixic acid in childhood shigellosis. J Pediatr 1988;113:901-7.
[4] Salam MA, Dhar U, Khan WA, Bennish ML. Randomised comparison of ciprofloxacin suspension and pivmecillinam for childhood shigellosis. Lancet 1998:352:522-7.
[5] Salam MA, Seas C, Khan WA, Bennish ML. Treatment of shigellosis. IV. Cefixime ineffective in shigellosis in adults. Ann Intern Med 1995;123:505-8.
Authors' reply
Editor--Alternative effective antibiotics were given to 65% (54/83) of children in our study: 63% (26/41) in the placebo group and 67% (28/42) in the vitamin A group. All had Shigella dysenteriae type 1 strains. After the organism was identified by culture, the antibiotic was changed to pivmecillinam within 48 hours because of the known sensitivity pattern of the organism in our hospital.
All children with nalidixic acid sensitive shigella (S flexneri or S boydii) in the vitamin A group were cured by day 5 compared with 53% in the placebo group. For subjects with nalidixic acid resistant shigella, 17% (5/28) in the vitamin A group and none in the placebo group were cured by day 5.
Children were defined as clinically cured if they passed three or fewer formed stools in a day without any visible blood and mucus, were afebrile, and had no abdominal pain or tenderness on that day. We observed a difference in clinical cure between the groups after 48 hours, not rapid divergence by 24 hours. Stool outcomes were the main determinants of clinical cure (table). Salam et al's definitions of clinical cure as "no bloody mucoid stools, six or fewer total stools, and no fever"[1] and failure as "more than nine stools or more than two watery stools, or were afebrile" are more arbitrary than ours.[2] Although there is no universally accepted definition of clinical cure in shigellosis, we believe our definition is more relevant and practical within the context of clinical care. Research staff at the International Centre for Diarrhoeal Diseases Research, Bangladesh, are well acquainted with stool consistencies, blood, and mucus. To minimise interobserver variation, we standardised study nurses' assessment of outcome measures and doctors arbitrated on any uncertainty. Moreover, a double blind randomised trial provides little scope for such bias to determine the outcome.
Evaluation of the individual severity symptom of placebo and vitamin A group on study day 5
(*) According to definition
The cellular and tissue response to a large dose of oral vitamin A can be measured in hours and a few days.[3] Vitamin A might similarly be predicted to act rapidly on colonic immune and mucosal function and integrity. The rapid effect of vitamin A in diarrhoea related to measles also suggests fundamental and rapidly reversible host mechanisms that cope with established systemic infection,[4] although it is unclear whether the effect is due to adjuvant properties or nutritional activity, or both. Strong evidence exists on the effect of vitamin A on cell mediated immunity and on the synthesis and secretion of cytokines.[5] Excretion of viable organistas in the stool does not necessarily reflect events in intracellular colonic tissue and by itself does not refute the possibility of vitamin A enhanced immune activity against shigella.
[1] Salam MA, Dhar U, Khan WA, Bennish ML. Randomised comparison of ciprofloxacin suspension and pivmecillinam for childhood shigellosis. Lancet 1998:352:522-7.
[2] Bennish ML, Salam MA, Khan WA, Khan AM. Treatment of shigellosis, III. Comparison of one-or-two-dose ciprofloxacin with standard 5-day therapy. Ann Intern Med 1992;117:727-34.
[3] Sommer A, West K. Vitamin A deficiency: health, survival and vision. Oxford: Oxford University Press, 1996.
[4] Coutsoudis A, Broughton M, Coovadia HM. Vitamin A supplementation reduces measles morbidity in young African children: a randomized, placebo-controlled, double-blind trial. Am J Clin Nutr 1991;54:890-5.
[5] Semba RD. The role of vitamin A and related retinoids in immune function. Nutr Rev 1998;56:S38-48.
COPYRIGHT 1999 British Medical Association
COPYRIGHT 2000 Gale Group