RESUMEN
BACKGROUND: Attempts have been made to link procurement of long-lasting insecticidal nets (LLIN) not only to the price but also the expected performance of the product. However, to date it has not been possible to identify a specific textile characteristic that predicts physical durability in the field. The recently developed resistance to damage (RD) score could provide such a metric. This study uses pooled data from durability monitoring to explore the usefulness of the RD methodology. METHODS: Data from standardized, 3-year, prospective LLIN durability monitoring for six LLIN brands in 10 locations and four countries involving 4672 campaign LLIN were linked to the RD scores of the respective LLIN brands. The RD score is a single quantitative metric based on a suite of standardized textile tests which in turn build on the mechanisms of damage to a mosquito net. Potential RD values range from 0 to 100 where 100 represents optimal resistance to expected day-to-day stress during reasonable net use. Survival analysis was set so that risk of failure only started when nets were first hung. Cox regression was applied to explore RD effects on physical survival adjusting for known net use environment variables. RESULTS: In a bivariate analysis RD scores showed a linear relationship with physical integrity suggesting that the proportion of LLIN with moderate damage decreased by 3%-points for each 10-point increase of the RD score (p = 0.02, R2 = 0.65). Full adjustment for net care and handling behaviours as well as other relevant determinants and the country of study showed that increasing RD score by 10 points resulted in a 36% reduction of risk of failure to survive in serviceable condition (p < 0.0001). LLINs with RD scores above 50 had an additional useful life of 7 months. CONCLUSIONS: This study provides proof of principle that the RD metric can predict physical durability of LLIN products in the field and could be used to assess new products and guide manufacturers in creating improved products. However, additional validation from other field data, particularly for next generation LLIN, will be required before the RD score can be included in procurement decisions for LLINs.