Use of an Extended Premise Condition Index for detection of priority areas for vector control actions

Highlights

• The Premise Condition index (PCI) could be used as good tool for vector control.

• Premises with higher PCI corresponded to higher risk of Aedes aegypti presence.

• The inclusion of the animals' presence increased the PCI discriminatory power.

Abstract

The Premise Condition Index (PCI), proposed by Tun-Lin and colleagues in 1995, is a score that considers the conditions of a premise as well its yards and degree of shading. They hypothesized that the higher its value the greater the probability of the premise having the presence of Aedes aegypti. This study aimed to evaluate if there is a correspondence between PCI and Ae. aegypti infestation in four areas of a large city in the State of São Paulo, Brazil, if the inclusion of new categories related to the presence of animals in premises would increase the probability of detecting predictive areas for vector control actions and, if so, to propose an expanded PCI. The positivity of the premises for the presence of Ae. aegypti was modeled considering a Bernoulli probability distribution, in a Bayesian context using the Integrated Nested Laplace Approximation. The study showed that, in general, the higher the value of the PCI of a premise, the more likely it is to have the presence of Ae. aegypti, and the inclusion of information on the animals' presence can increase the discriminatory power of PCI. These results support the proposition of an extended PCI that would consider, in addition to the conditions of the premise, the presence of animals to classify it regarding the risk of the presence of Ae. aegypti.

Introduction

Strategies to prevent dengue fever are based on vector control through integrated action management, active search for cases based on an adequate health information system, emergency preparation, capacity building and training of agents responsible for actions, and developing research in vector control. Vaccines are currently being developed for dengue control, but there are many challenges to be overcome, such as the immunological response in patients who have never had dengue fever (Vannice et al., 2016). Moreover, there is recent information from the National Health Surveillance Agency that considers the possibility of a vaccinated individual who has never had dengue to acquire a more severe form of the disease in case of contact with the virus through the bite of the transmitting mosquito (Rocha, 2017). Thus, it can be said that there are few alternatives currently to reduce or prevent the transmission of the virus (if not by vector control), and what is observed is that prevention programs have been insufficient and ineffective (Guzman et al., 2010). For example, Olliero et al. (Olliaro et al., 2018) in a systematic review did not find robust studies on the impact of fogging on transmission control, which shows the need for greater investment in vector control.

The lack of an impact evaluation of the measures adopted in Aedes aegypti control program may be one of the reasons for the difficulty in controlling the mosquito and reliable as well as practical indicators are needed (Barrera, 2016). An important limitation of larval indicators is that they depend on the visual location of the breeding sites, which do not reflect the true prevalence of Ae. aegypti when there are critical breeding sites, that is, those with difficult access that are not always visually identified (Barrera, 2016; Bermudi et al., 2017).

Considering these aspects, the question arises regarding potential Ae. aegypti breeding sites available on the premise. The possibility of finding premises that, due to certain characteristics, may offer more mosquito breeding sites could help the vector control program, reducing resources and the time spent eliminating them. Thus, Tun-Lin et al. (Tun-Lin et al., 1995a) presented an index known as the Premise Condition Index (PCI), which considers the conditions of the premise as well as its yard and degree of shading by assigning a standard score for this information and defines an index that may be related to the supply of containers. Tun-Lin et al. (Tun-Lin et al., 1995b) further demonstrated in another study that there is some degree of stability concerning positive households by classifying them as key premises, and these premises, or certain types of containers, contribute disproportionately to the maintenance of Ae. aegypti population. Chadee et al. (Chadee, 2004) also assessed the presence of the key premises in a study conducted in Trinidad and suggest that focusing efforts on these premises to control mosquito density may be more cost-effective.

Peres et al. (Peres et al., 2013) conducted a study in municipalities in the northern region of Rio de Janeiro, Brazil, and found as the number of eggs in oviposition traps increased, the PCI also increased, showing that the PCI can be used as an indicator of the presence of breeding sites and mosquitoes. Another study showing a positive relationship between the presence of adult larvae, pupae and mosquitoes and PCI values was performed in the municipality of Marília, State of São Paulo, Brazil (Andrighetti et al., 2009). Also, Nogueira et al. (Nogueira et al., 2005) in Botucatu, State of São Paulo, found the same association between the presence of Ae. aegypti eggs and the PCI.

Despite these studies, the use of the PCI in the activity routine of the Brazilian Vector Control Program has not yet been effective, because it still needs further studies for its validation and better adaptation to housing conditions and Brazilian customs and practices. One possibility would be the inclusion of components in the PCI that represent the widespread behavior in Brazilian society of keeping animals in their homes (Bortoloti and D'Agostino, 2012; Schoendorfer and Germano, 2001). In this regard, a study conducted in a medium-sized municipality in the state of São Paulo showed that the presence of dogs and chickens increased the chance of finding the vector in the premises in 51% and 103%, respectively (Favaro et al., 2013).

Once it is shown that increasing PCI values are directly related to the increased probability of Ae. aegypti identification on premises, this indicator has the potential to be used to identify areas of higher risk within municipalities for the presence of the vector, which could optimize its control. However, studies so far have not taken into account the spatial dependence of the studied phenomenon (Andrighetti et al., 2009; Nogueira et al., 2005; Peres et al., 2013), an important step for the validity of the indicator.

To better validate the PCI and evaluate the inclusion of other components in the PCI composition, this study aimed to evaluate if there is a correspondence between PCI and Ae. aegypti infestation in four areas of a large city in the State of São Paulo, if the inclusion of new categories regarding the presence of animals in premises in the PCI would increase the probability of detecting predictive areas for vector control actions and, if so, to propose an expanded PCI.