| Sumario: | Introduction: One of the leading causes of death in under-five children globally is diarrhoea. House flies, which commonly interact with filthy environment and food for human consumption, are known to carry diarrhoea causing bacteria. Despite this knowledge, interventions to control diarrhoea have ignored vector control options and mainly focused on treatment and other preventive measures like rotavirus immunisation, hand washing and general improved sanitation. This study was conducted to determine the impact of the Vestergaard Frandsen ZeroFly® Livestock Netting (VZFLN) on density of flies and bacteria contamination on meat in selected butcheries in Dagoretti sub- county, Nairobi, Kenya.
Methods: A controlled longitudinal study was conducted in 30 butcheries, randomly selected and allocated to intervention and control arms. A baseline survey to determine the species and density of flies and meat and butchery surface contamination by enteric bacteria was conducted in the 30 study butcheries prior to introduction of the VZFLN (treatment) in the intervention butcheries. Samples of flies, meat and swabs of the wooden log used for chopping meat were obtained from the butcheries on a weekly basis for 5 consecutive weeks. Interviews were conducted with butchery owners pre- and post-intervention to determine practices that promote meat contamination and willingness to use the VZFLN in the control of flies in butcheries.
Results: Baseline data identified seven species of flies in butcheries namely; Calliphora spp, Crysomya putoria, Crysomya chloropyga, Lucillia spp, Musca domestica and Sarcophaga spp. Of 682 flies collected at baseline, Musca domestica was dominant (89%) followed by Crysomya choloropyaga (4%) and Crysomya putoria (4%). All fly species except Calliphora spp were contaminated with one or more enteric bacteria species. E.coli (56%) was the predominant contaminant on all flies followed by Proteus spp (13%), Shigella spp (9%), Salmonella spp (3%) and mixed contaminations (18%). Meat (33%) and butchery surface swabs (27%) samples were contaminated by enteric bacteria. E.coli was also predominant on meat and butchery surfaces.
Post intervention results were calculated using a mixed effects Poisson regression in a Generalized Linear Model. These showed that the VZFLN accounted for a 57% difference (IRR 0.434, 95% CI 0.235-0.804, P = 0.008) in the mean number of flies caught in the intervention (mean=8, 95% CI, 5-12) and control (mean=19, 95% CI, 10- 28) groups in week two. At week five, the mean catch of flies in intervention butcheries was 6 (95% CI 2-10) and 19 flies (95% CI, 7-32) in the control group, representing a 67% difference (IRR 0.326, 95% CI, 0.126-0.845, P=0.021). The results were adjusted for baseline differences in fly numbers. Adjusting the results
for temperature, location, and baseline differences in fly numbers in the same model gave a 51% difference in the means of flies in the intervention and control butcheries (IRR=0.496, 0.254-0.969, p=0.04), mean=7 (95% CI, 3-12) in intervention and mean=15 (95% CI 11-18) in control butcheries.
Conclusion and recommendation: The VZFLN was effective at reducing fly density in butcheries and is therefore recommended for use. Trial studies however should be done in the dry season when fly numbers are relatively high.
|
|---|