Frequency of typhoon occurrence accounts for the Poisson distribution of human leptospirosis cases across the different geographic regions in the Philippines

Derick Erl P. Sumalapao, Benjamin Kyle M. Del Rosario, Lara Beatrice L. Su?ga,Catherine C. Walthern, Nina G. Gloriani

1Department of Medical Microbiology, College of Public Health, University of the Philippines Manila, Manila, Philippines

2Department of Biology, College of Science, De La Salle University, Manila, Philippines

3Senior High School Department, De La Salle University, Manila, Philippines

Keywords：Human leptospirosis Typhoon Poisson distribution Philippines

ABSTRACT Objective: To study the distribution of human leptospirosis cases across the different geographic regions in the Philippines in 2015-2017 and to determine the relationship between the frequency of typhoon occurrence and human leptospirosis cases. Methods： Information on the frequency of leptospirosis cases and typhoon occurrence across the different regions in the Philippines from 2015 to 2017 was retrieved from the databases of the Department of Health and the Philippine Atmospheric, Geophysical, and Astronomical Services Administration,respectively. Descriptive measures on the prevalent cases and occurrence of typhoons across the different regions were summarized. Linear regression analysis was employed to establish the functional relationship between leptospirosis cases and typhoon occurrence. The distribution of human leptospirosis cases was assessed using the Poisson distribution. Results：The frequency of typhoon occurrence accounted for the signi ficant linear variation in the geographic distribution of human leptospirosis cases in the Philippines (P＜0.001). Moreover,the human leptospirosis cases obeyed a Poisson distribution (λ=6.89, P＜0.001). Conclusions：The Philippines has frequently experienced severe weather perturbations such as typhoons resulting in flooding and subsequently increasing the risk of transmitting bacterial infections including leptospirosis. Information obtained regarding the determinants and distribution of human leptospirosis will provide better understanding of the disease propagation for subsequent design of optimal disease prevention measures, appropriate resource allocation,effective control strategies, and necessary public health programs.

1. Introduction

The Philippines is a third world nation located in Southeast Asia with a land area of 298 170 km2[1] and houses over 96.6 million people[2]. Due to the archipelagic nature of this country, surrounding waterways and reservoirs serve as a signi ficant source of livelihood,which residents rely on. However, in recent years, it has become evident that this dependency on available waterways only increased the rate of pollution due to improper waste management practices.The state of the country also cultivated a prime breeding ground for several viral and bacterial diseases that plague the people of the local community subsequently posing certain health risks.As with several diseases, the living and sanitary conditions of the people largely determine their susceptibility and vulnerability to pathogens. The most notable of these is leptospirosis, which spreads through water contaminated with rat urine[3]. Concurrently, in a test conducted in Kanda, Sri Lanka, where leptospirosis is endemic,73.9% of rats tested were found positive for this disease[4]. This in turn was translated to its prevalence within humans exposed to infested waters which greatly increased the likelihood of contracting the infection, amounting to a further cause for concern. In addition,contaminated water merging with other sources of water, such as rivers, hastens the spread of this disease, as water travels between neighboring communities.

Moreover, the Philippines is often subject to severe weather perturbations such as typhoons which result in flooding with subsequent significant damage in affected areas and increasing risk of possible contraction of leptospirosis. Flooding, as further heightened by the prevalent pollution in the city, increases the risk of transmitting bacterial infections. In terms of flooding as a factor of disease propagation, the spread of leptospirosis rises dramatically during the rainy season in the Philippines with most seropositivity identified in areas involved in agriculture[5]. Moreover, the lack of sanitation in communities exacerbates the propagation of the disease making the residents in the slum areas highly vulnerable to this infection[6]. During heavy rain, water becomes a medium for the spread of waste. Since fecal matter, urine, and other pollutants compromise the safety of waterways, contaminated water can also result from exposed drainage systems, which may be accessed beneath roads. Rat population throughout certain areas also increases the risk of water contamination, as their urine becomes a vector for the propagation of leptospirosis[3]. Due to the relatively unsanitary circumstances present in the environment within the regions of the Philippines, it has become evident that the health of several Filipinos had been severely affected. Several individuals in various age groups are exposed to different pathogens, includingLeptospiraspp.,a health hazard which can subsequently affect a person′s overall quality of life, rendering the safety and health of those afflicted with waterborne diseases a significant issue. Hence, the present study identi fied the distribution of human leptospirosis cases across the different geographic regions in the Philippines in 2015-2017.Speci fically, a linear functional relationship between frequency of typhoon occurrence and reported human leptospirosis cases was established using the available information from the databases of the Philippine Atmospheric, Geophysical, and Astronomic Services Administration[7] and the Department of Health[8]. Information derived on the determinants and distribution of human leptospirosis cases across the different regions in the Philippines will provide better understanding of the disease propagation and identify highrisk populations of acquiring the infection eventually intended for the design of optimal disease prevention measures, appropriate resource allocation, effective control strategies, and necessary public health programs to reduce disease morbidity and mortality rates.

2. Materials and methods

Information on the frequency of leptospirosis cases and typhoon occurrence across the different regions in the Philippines from 2015 to 2017 was retrieved from the databases of the Department of Health[8] and the Philippine Atmospheric, Geophysical, and Astronomical Services Administration[7], respectively.

In order to measure the susceptibility of a region to acquiring leptospirosis, the proportion of cases to regional population per year was computed. Descriptive measures on the prevalent cases and occurrence of typhoons across the different regions were summarized. Moreover, zoonosis is prevalently underreported due to barriers to healthcare experienced by its primary afflicted demographic, those below the poverty line. Thus, these estimates may be used to identify which regions are likely to contract leptospirosis in the four weeks following a typhoon. Only cases reported in the following four weeks of a typhoon were considered since leptospires have an incubation period of 5 to 14 d, with a range of 2-30 d[9]. Linear regression analysis was employed to establish the functional relationship between leptospirosis cases and typhoon occurrence. The distribution of human leptospirosis cases was assessed using the Poisson distribution[10]. Statistical analyses were performed using Microsoft Excel? and STATA? software at 5%level of signi ficance. Spatial analysis on the distribution of cases was described using Quantum GIS 2.18 software (QGIS Development Team).

3. Results

Reported frequencies of leptospirosis cases in the Philippines from 2015 to 2017, retrieved from the databases of the Department of Health, were summarized (Table 1). For the year 2015, there was no available information regarding these human leptospirosis cases during the months of September and October. In 2016, there was a complete reporting of these cases. Only reported cases from January to September in 2017, in addition to the reported cases in the previous years, were included for analysis in this paper since data during the last quarter of 2017 were not available to the authors.Hence, the parameter estimates of the regression model and the Poisson distribution reported in this study are descriptions of the only information accessible to the authors.

Table 1 Distribution of human leptospirosis cases across the different regions in the Philippines, 2015-2017 (Frequencies of cases retrieved from the databases).

In 2015, Region Ⅵ (Western Visayas) and Region Ⅲ had 16 leptospirosis cases per 1 000 000 people, while Region Ⅰ (Ilocos Region) had 10 cases per 1 000 000 people. Region Ⅰ had 63 leptospirosis cases per 1 000 000 people in 2016, whereas RegionⅥ had 44 cases per 1 000 000 people. In 2017, Region Ⅵ remained to have 44 cases per 1 000 000 people, while Region Ⅰ had 32 cases per 1 000 000 people. However, the National Capital Region (NCR)consistently yielded the highest frequencies of leptospirosis cases notably in the incubation period following the onset of a typhoon,with a peak value of 34 leptospirosis cases per 1 000 000 people in 2016. With the high population density of the region, it rendered signi ficant reduction in the proportion of leptospirosis cases within the region. The distribution of these human leptospirosis cases in the Philippines from 2015 to 2017 was presented in Figure 1.

Moreover, information about the frequencies of typhoons that hit the different regions in the Philippines from 2015 to 2017 and the number of human leptospirosis cases associated with these typhoons were summarized (Table 2). In the present study, only cases reported in the following four weeks of a typhoon were considered since leptospires have an incubation period of 5 to 14 d, with a range of 2-30 d[9]. In 2015, the region with the greatest number of cases was Region Ⅲ, with 150 cases. On the other hand, Regions Ⅻ, ARMM,and CARAGA had no reported cases. Region Ⅱ had the highest frequency of typhoons, a total of nine, relatively comparable to Region Ⅲ. However, Region Ⅱ only had 15 cases of leptospirosis,4.7% of the total leptospirosis cases in 2015, despite of the given number of typhoons that hit the area. Moreover, NCR was hit by four typhoons in 2015, and this was associated with 18 cases. In regions that had experienced a similar typhoon frequency, varying leptospirosis case frequencies were recorded. For example, both Region Ⅹ and Region Ⅻ, with numerically comparable populations,were hit by same number of typhoons in 2015, but Region Ⅹ had six leptospirosis cases, while Region Ⅻ had none.

In 2016, Region Ⅰ had the greatest number of reported leptospirosis cases (a total of 150 cases) and had the highest frequency of typhoons (Table 2). Similarly, Region Ⅱ had the same typhoon frequency with only 20 cases. On the other hand,Region Ⅰ had a signi ficant increase in the number of cases (from 32 cases in 2015 to 150 cases in 2016) as there was an increase in typhoon occurrence. Regions Ⅸ, Ⅺ, Ⅻ, and ARMM reported no leptospirosis cases in 2016. These regions, except Region Ⅺ, were not hit by any typhoon. Region Ⅲ had relatively similar number of cases and typhoon frequencies in the years 2015 and 2016. On the other hand, there was a signi ficant increase in the number of leptospirosis cases in NCR despite of relatively the same number of typhoons. In 2017, NCR had the highest number of cases while Region Ⅻ and ARMM had none (Table 2). Region Ⅳ-B had only three leptospirosis cases even if it was hit by the greatest number of typhoons. Interestingly, Regions Ⅰ and Ⅲ had a signi ficant drop in the number of leptospirosis cases from 2016 to 2017.

Figure 1. Hotspot analysis of human leptospirosis cases in the Philippines, 2015-2017.

Results of the linear regression analysis between the occurrence of typhoon and the frequency of leptospirosis cases across the different regions in the Philippines revealed a moderate correlation(r=0.481 8) (Figure 2). Despite of the moderate correlation obtained,the regression model was identi fied to be statistically signi ficant(P=0.000 3). The linear regression equation,y=7.13x-3.13, is adequate in quantifying the contribution of frequency of typhoon occurrences (x) to the frequency of human leptospirosis cases (y)in a particular region. Moreover, human leptospirosis cases were determined to have a Poisson distribution (λ=6.89,R2=0.960 4,P＜0.000 1) using the method suggested by Hoaglin[10]. Regarding the typhoon occurrence across the different regions in the Philippines from 2015-2017, with a total of 1 509 reported human leptospirosis cases (Figure 3), the highest likelihood of 15.12% that a region will be hit by 6 typhoons, followed by 7 typhoons with 14.89%probability was identi fied.

Table 2 Frequency of leptospirosis cases during typhoon episodes and frequency of typhoon occurrence across the different geographic regions in the Philippines, 2015-2017(typhoon-related cases recorded following a typhoon).

Figure 2. Regression analysis of leptospirosis cases and typhoon occurrence(slope=7.13, intercept=--3.13, r=0.481 8, P=0.000 3).

Figure 3. Expected frequencies of human leptospirosis cases using the Poisson distribution (λ=6.89, R2=0.960 4, P＜0.000 1, n=1 509).

4. Discussion

The susceptibility of a particular geographic region in acquiring leptospirosis can be computed as the proportion of cases in a given regional population. In the Philippines, Region Ⅰ(Ilocos Region) and Region Ⅵ (Western Visayas) consistently had the highest proportions of leptospirosis cases. This may be attributed to both regions having topographies with large expanses of coastlines and multiple bodies of water, such as rivers and lakes[11]despite of experiencing varying typhoon cases. Moreover, the NCR consistently yielded the highest frequencies of leptospirosis cases notably in the incubation period following the onset of a typhoon, this is due to slums prevalently being erected within flood plains around Metro Manila which increase the susceptibility of residents to this waterborne disease[6]. However, with the high population density of the region, it rendered signi ficant reduction in the proportion of leptospirosis cases within the region. Another notable situation is seen in the Cordillera Administrative Region with relatively low frequency of cases but with frequent typhoon occurrence experienced by the region yearly. The topography of the region is composed of mountain ranges which lessen the chances of water from pooling and becoming a vector forLeptospiraspp. In fact, Cordillera Administrative Region is one of the regions with the highest proportion in an antithetical manner to that of NCR due to the relatively low population of the region. Regions with the lowest prevalence of leptospirosis over the past years were Region Ⅻ(SOCCSKSARGEN) and the ARMM. The topography of ARMM is known for its hilly geography, although there are lowland areas like plains[12]. This predominantly highland geography could be a reason for the diminished cases of leptospirosis. Similarly, RegionⅫ has highland territories interspersed with lowlands[13], with agriculture as its major industry[14]. It is likely that both regions are less susceptible to leptospirosis due to fewer typhoon occurrences,an average of one typhoon per year. Over the years, these computed percentages appeared to be relatively small, and with the absence of global estimates for leptospirosis morbidity and mortality rates,classi fied leptospirosis a neglected disease[15,16].

The present study determined the functional relationship between the occurrence of typhoon and the frequency of leptospirosis cases across the different regions in the Philippines in 2015-2017. Despite of the moderate correlation obtained, the regression model was identi fied to be statistically signi ficant and adequate in quantifying the contribution of frequency of typhoon occurrence to the frequency of human leptospirosis cases in a particular region. The obtained model can possibly generate an underestimate or an overestimate of the observed reported leptospirosis cases in some geographic regions. The moderate strength of correlation was exempli fied by a case in 2015, wherein Region Ⅲ had eight typhoons with 150 cases(predicted 54 cases) suggesting an underestimate. The excess cases were possibly due to other factors, such as the lowland topography of Region Ⅲ[11]. On the other hand, Region Ⅱ experienced nine typhoons in 2016 with 20 leptospirosis cases (predicted 61), an overestimate, possibly due to Region Ⅱ having several mountain ranges and areas of forestland[11].

For prediction purposes, the present study was able to identify the statistical distribution of the human leptospirosis cases as in fluenced by the frequency of typhoon occurrence. These human leptospirosis cases were determined to have a Poisson distribution. Over the years 2015-2017, a total of 1 509 human leptospirosis cases were reported. On the average, any geographic region in the Philippines is most likely to be hit, on the average, by 6-7 typhoons per year. This suggests further that the likelihood decreases when less than 6 or more than 8 typhoons are expected to hit a particular geographic area.Given that the frequency of typhoon occurrence accounts for the Poisson distribution of human leptospirosis cases in the Philippines,an average of 228 additional cases are expected to be reported in a given region if it is hit by 6 typhoons.

5. Conclusion

In this study, the human leptospirosis cases in the Philippines were determined to have a Poisson distribution and there is some functional relationship between the frequency of typhoon occurrence and these human leptospirosis cases. The resulting moderate correlation, possibly due to the variations in the topography,living and sanitary conditions of the different regions in the country, warrants further investigation by incorporating additional information involving the human host′s behaviour, practices, and even animal-environment contact patterns. A better understanding of leptospirosis including its determinants and distribution will provide additional information regarding disease propagation and identify high-risk populations of acquiring the infection for the design of optimal disease prevention measures, appropriate resource allocation, effective control strategies, and necessary public health programs.

Conflict of interest statement

The authors declare that there is no con flict of interest.

Acknowledgments

The authors are grateful to Mr. Christian E. Supsup for the GIS-generated image presented in this paper.