Validation of malaria surveillance case reports: implications for studies of malaria risk.

STUDY OBJECTIVE The aim of the study was to investigate the quality of national malaria surveillance reports in the United Kingdom. DESIGN Persons with malaria reported to the Malaria Reference Laboratory (MRL) in 1987 were contacted by post to verify existing records with respect to key variables. The MRL data set was then analysed for inaccuracies. SETTING The study was confined to UK residents. PARTICIPANTS 602 persons with malaria in 1987 responded (53%). MEASUREMENTS AND MAIN RESULTS Review of case reports showed few missing data except for duration of residence in the UK, detailed chemoprophylactic regimens, and compliance. There were more missing surveillance data in reports of ethnic minority groups, principally in dates of travel (p = 0.008) and chemoprophylaxis use (p less than 0.0001). Patient recall in the survey was at variance with the surveillance reports in dates of travel and onset of infection, chemoprophylaxis use, and in compliance. Surveillance reports overestimated the number of days between leaving a malarious area and onset of symptoms (by 9 d for P falciparum and by 24 d for P vivax), and underestimated the delay between onset and diagnosis of P falciparum by 3 d. Over 50% of patients who had recalled the use of chloroquine, proguanil, pyrimethamine/dapsone, and pyrimethamine had not been recorded as having taken these drugs on the surveillance reports. Reported compliance also differed between the two data sets. CONCLUSIONS It is recommended that research units test the quality of their surveillance data before embarking on analytical studies used to generate health policy guidelines.