# Spatiotemporal patterns and entomological predictors of dengue transmission in Urban Surat, India (2016–2020): A surveillance-based risk modelling study

## TL;DR

This study analyzes dengue transmission patterns in Surat, India, identifying high-risk zones and factors influencing mosquito density to improve urban vector control.

## Contribution

The study provides a surveillance-based risk model for dengue transmission in Surat, integrating spatiotemporal and entomological data to guide targeted interventions.

## Key findings

- Dengue incidence declined from 11.1 to 2.2 cases per 100,000 population between 2016 and 2020.
- Vector indices remained high in the South and South-East zones, with 36% of cases in areas with HI ≥ 1%.
- A predictive model for vector density showed moderate accuracy (AUC = 0.688) with high specificity but low sensitivity.

## Abstract

Dengue fever is an increasing public health concern in urban India due to rapid urbanization, inadequate vector control, and underreporting of cases. Surat, a densely populated city in Gujarat, has remained a recurrent dengue hotspot, yet detailed spatiotemporal patterns and entomological determinants of transmission remain insufficiently explored. This study aimed to assess spatiotemporal patterns of dengue transmission in Surat and identify entomological and demographic predictors of elevated vector density to guide targeted urban interventions. A retrospective longitudinal analysis was conducted using five years of dengue surveillance data (January 2016–December 2020). A total of 1,658 laboratory-confirmed dengue cases reported to the sentinel surveillance system of the Surat Municipal Corporation were included. Monthly entomological surveillance assessed vector indices—House Index (HI), Container Index (CI), and Breteau Index (BI)—across administrative zones. Associations were examined using chi-square analysis, and binomial logistic regression identified predictors of moderate-to-high vector density (HI ≥ 1%) considering temporal, spatial, and demographic variables. Model performance was evaluated using receiver operating characteristic analysis. During the study period, reported dengue incidence declined from 11.1 to 2.2 cases per 100,000 population; however, elevated vector indices persisted, particularly in the South and South-East zones. Approximately 36% of cases occurred in households located in areas with HI ≥ 1%. Adolescents and young adults (median age 21.7 years) were disproportionately affected. Public sector facilities reported 2.6 times more cases than private providers (p < 0.001), suggesting substantial underreporting. Logistic regression identified year, month, zone, and age as significant predictors of elevated vector density (p < 0.001). The model showed moderate discrimination (AUC = 0.688), high specificity (87.3%), and low sensitivity (35.9%). Despite declining reported incidence, persistently high larval indices and post-monsoon peaks indicate ongoing transmission risk, emphasizing the need for zone-specific vector control and strengthened surveillance systems.

Dengue, transmitted by Aedes aegypti, is a major public health concern in tropical cities, but understanding how disease risk relates to mosquito abundance and urban dynamics remains challenging. In Surat, India, we analyzed five years (2016–2020) of laboratory-confirmed reported Dengue cases alongside mosquito surveillance data to identify patterns of transmission, high-risk areas, and population groups most affected.

We found that reported Dengue cases declined sharply, particularly in 2020—likely due to COVID-19 disruptions—yet mosquito densities remained high in many parts of the city, showing that vector abundance alone does not predict disease incidence. Cases peaked during the post-monsoon months, and younger adults, especially males, were disproportionately affected. Spatial analysis revealed persistent hotspots in the South and South-East zones, while private healthcare facilities reported substantially fewer cases, highlighting surveillance gaps.

A predictive model for moderate-to-high vector density showed moderate accuracy, suggesting that integrating environmental, climatic, and socioeconomic factors could improve risk prediction. These findings provide actionable evidence for targeted vector control, improved disease surveillance, and public health strategies to reduce Dengue transmission in rapidly growing urban settings.