Risk and protective factors for canine visceral leishmaniasis in the Americas: a systematic review update with meta-analysis.

BACKGROUND

Identification of risk and protective factors for canine visceral leishmaniasis (CVL) is essential to understand the epidemiology of the disease and direct prevention and control strategies. Building on a previous systematic review, this study presents new findings relating to the associations between CVL and a range of variables.

METHODS

The systematic review included articles from the previous review (up to September 2011) and additional studies published thereafter regarding factors associated with CVL in the American continent. The inclusion criteria encompassed studies that analyzed associations between CVL and socioeconomic, environmental, household-level, or dog-level variables, regardless of the diagnostic method employed. Meta-analyses were conducted using random-effects models and subgroup analyses, while the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was used to classify levels of evidence.

RESULTS

Of the 111 studies included in the analysis, the vast majority were cross-sectional (87.4%), conducted in Brazil (95.5%), and employed serological diagnostic tests alone (77.5%). The variables most consistently associated with CVL, based on moderate levels of evidence, were short-haired dogs, dogs dwelling in the peridomicile or with free access to the street, and proximity of the domicile to green areas. Male dogs, large dogs, ectoparasite-infested animals, dogs in contact with horses, and living in homes with yards also had higher odds of infection, although the levels of evidence were low. Other variables associated with CVL but assigned with very low levels of evidence included the presence of chickens, cats, sand flies, and other dogs in the dog environment; previous cases of CVL in the domicile; substandard custodianship; and guardians with poor education and low income. The quality of the studies has improved since the last review, although methodological limitations were still present, in particular, the absence of control for confounders.

CONCLUSIONS

The analyses performed in this review strengthen the current knowledge of CVL and highlight the importance of further research to better understand some of the associated variables. Additional cohort and case-control studies are required, particularly those utilizing molecular diagnostics and adequate control for confounding factors. This review represents progress in understanding the determinants of CVL in the Americas and provides support for improving prevention and control strategies.