This study shows that, there is a significantly higher prevalence of diarrhea among children residing in non model households compared to those residing in model households. This difference may be due to the fact that, out of three implementation components of HSEP (provision of community based health package, capacity building of potential families to be role model HHs, service delivery), health extension workers spend more time on capacity building part for model HHs and they gave extra training, support and follow up to those who were selected to be role models. This training, support and follow up may bring knowledge and skill development to the model and it made them practice health packages well compared to non model HHs . Findings from previous studies [8, 9] revealed that, model families have good utilization of maternal health services. Mothers utilizing those services are more likely to gain access to other preventive services which commonly target children under five year of age. In addition to that, national workshop on maternal and new born health recommend that, the package for model families should include essential indicators of maternal and newborn care to advance maternal and neonatal health at the community level. This may indicate that they consider model households to have advantages compared to non model households .
About 99(30%) of model HHs have three or more rooms compared to 65(20%) in non model HHs, which is a significant difference between the two groups. Model households are likely to improve their housing conditions. According to a study conducted in Keffa Sheka, south west Ethiopia , fewer number of rooms was a risk factor associated with under five diarrhea. This may be due to the fact that when there is overcrowding in the HH, the chances for contamination of water and food would be high.
Other important improvements in household condition are having a separate sleeping place for domestic animals and having separate kitchen. About 74(27%) and 37(15%) did not live with animals in the same house and 88(27%) and 51(16%) had a separate kitchen in model and non model HHs respectively showing significant difference that indicates a possible effect of model HH training. During the training period, health extension workers made follow up, provided supportive supervision and given health education might be attributed to the behavior changes being observed after training. A study conducted in Guinea Bissau indicated that, having domestic animals in the house is a risk factor for diarrhea in children . Living with animals in the same house increases unhygienic condition of the HH and probability of getting zoonotic diseases .
This study found that 26(8%) and 86(27%) have no latrine to use, and feces were seen around the house in 57(17%) and 91(28%) model and non model HHs respectively, again showing possible effect of the training towards lowering under five diarrhea. Availability of latrine reduces diarrhea  and unclean environment associated with under five diarrhea . Availability of latrine reduces fecal contamination in the domestic environment and, in turn, this prevents transmission of disease-causing organisms to human beings. This is also true for unclean environment, which can be good media for pathogens [15, 16].
In this study a statistical difference in practicing hand washing after toilet use was observed.
In 106(32%) model HHs and 145(45%) non model HHs care takers didn’t wash their hands after toilet visit. This difference may be due to model HH training. One study conducted in Pakistan indicated that hand washing after toilet visit significantly decreases incidence of diarrhea in under five children .
The study found good coverage of vitamin A supplementation, introduction of supplementary feeding in the age between 6-9 months and exclusive breast feeding in both model and non model HHs with no statistically significant difference between the two groups. This may be due to the expansion of health extension program [4–10]. Hence these programs are noted to play a great role in the prevention of under five diarrhea [18, 19].
This study has its limitations and strengths. Limitations include; the information on the prevalence may not reflect the actual situation that may be observed in the various seasons of the year which could be addressed by longitudinal study by other researchers. Moreover, the absence of clear demarcation between model and non model with reference to distance (closeness of model and non model) may have created information contamination as well as diarrheal disease transmission to the model HH members and vice versa. The effect of food hygiene was not assessed due to resource limitations. The strength of this study is that, data collectors were blinded regarding whether each household was model or non model inorder to reduce interviewer bias.