OBJECTIVE: To suggest changes in the sampling procedures for estimating the larval density of Aedes aegypti in the state of São Paulo, Brazil, by assessing the efficacy and level of complexity of the current procedures. METHODS: The "Superintendência de Controle de Endemias do Estado de São Paulo" -- SUCEN's (Superintendence for Endemic Control of the State of São Paulo) sampling procedures for estimating the density of Aedes aegypti larvae was assessed. The Breteau index with one-stage cluster sampling was applied. An average of 35 blocks, 14 buildings per block, of a total of 510 buildings per sample, were assessed. One hundred and eleven samples were obtained in 6 municipalities of São José do Rio Preto. RESULTS: The Breteau index estimates higher than 3 had coefficients of variation smaller than 30% in 71% of the samples. The design effect was 1.19 and 1.79 for indexes lower and higher than 5, respectively. The estimates were obtained in less than 3 days, requiring a daily average of 5.5 men per sampling. CONCLUSION: This sampling design revealed to be a simplified one: fast, economic and easily managed. The operational easiness was due to the use of the one-stage cluster sampling, when there was not need to prepare lists of addresses and then sort and locate the buildings. Though the design effect shows a slight loss of the estimates' accuracy with the use of one-stage cluster sampling, this was proven to be under acceptable range. Simplifying changes in the process of determining the sample size and assessing sampling errors are proposed.
Aedes; Larva; Sampling studies; Mosquito control; Health plans and programmes; Dengue; Evaluation; Dengue; Rapid assessment; Aedes aegypti; Simplified sampling design
