Passive malaria surveillance in a low endemic area of India: validation of a clinical case definition.

BACKGROUND

In India, 2.55 million cases of malaria were reported during 1997; roughly one-third were due to Plasmodium falciparum. Malaria cases are identified by passive and active surveillance and all patients with fever are treated with chloroquine (10 mg/kg body weight). Since all fevers are not malaria, this results in overtreatment and has a bearing in terms of the parasites developing resistance. We aimed to test the validity of a clinical algorithm for passive malaria surveillance by primary care doctors (fever with pallor or splenomegaly) in a low endemic, Plasmodium vivax-predominant area of Ballabgarh block in Faridabad District, Haryana.

METHODS

Passive surveillance was carried out at the general and paediatric outpatient departments (OPDs) of Ballabgarh hospital. All persons with fever attending the OPD were examined for the presence of fever, pallor and splenomegaly by the treating doctor. A blood smear was prepared and examined in all these cases.

RESULTS

A total of 3119 slides for malaria were made at Ballabgarh hospital but clinical details in the requisition form were available for only 2616 patients who form the subjects of this analysis. A total of 59 malaria cases (30 P. vivax cases and 29 P. falciparum) were diagnosed. The presence of fever with pallor or splenomegaly had a sensitivity of 28.8% (95% CI: 18.1-42.3); specificity of 88.6% (95% CI: 87.3-89.8), positive predictive value of 5.5% (95% CI: 3.3-8.8) and negative predictive value of 98.2% (95% CI: 97.5-98.7).

CONCLUSION

The algorithm did not have sufficient sensitivity to detect malaria cases by passive surveillance.