| Title: |
Identifying clusters of leprosy patients in India: A comparison of methods |
| Authors: |
Taal, Anneke T.; Garg, Akshat; Lisam, Suchitra; Agarwal, Ashok; Barreto, Josafá G.; van Brakel, Wim H.; Richardus, Jan Hendrik; Blok, David J. |
| Contributors: |
Lau, Eric H. Y.; Dutch Postcode Lottery |
| Source: |
PLOS Neglected Tropical Diseases ; volume 16, issue 12, page e0010972 ; ISSN 1935-2735 |
| Publisher Information: |
Public Library of Science (PLoS) |
| Publication Year: |
2022 |
| Collection: |
PLOS Publications (via CrossRef) |
| Description: |
Background Preventive interventions with post-exposure prophylaxis (PEP) are needed in leprosy high-endemic areas to interrupt the transmission of Mycobacterium leprae . Program managers intend to use Geographic Information Systems (GIS) to target preventive interventions considering efficient use of public health resources. Statistical GIS analyses are commonly used to identify clusters of disease without accounting for the local context. Therefore, we propose a contextualized spatial approach that includes expert consultation to identify clusters and compare it with a standard statistical approach. Methodology/Principal findings We included all leprosy patients registered from 2014 to 2020 at the Health Centers in Fatehpur and Chandauli districts, Uttar Pradesh State, India (n = 3,855). Our contextualized spatial approach included expert consultation determining criteria and definition for the identification of clusters using Density Based Spatial Clustering Algorithm with Noise, followed by creating cluster maps considering natural boundaries and the local context. We compared this approach with the commonly used Anselin Local Moran’s I statistic to identify high-risk villages. In the contextualized approach, 374 clusters were identified in Chandauli and 512 in Fatehpur. In total, 75% and 57% of all cases were captured by the identified clusters in Chandauli and Fatehpur, respectively. If 100 individuals per case were targeted for PEP, 33% and 11% of the total cluster population would receive PEP, respectively. In the statistical approach, more clusters in Chandauli and fewer clusters in Fatehpur (508 and 193) and lower proportions of cases in clusters (66% and 43%) were identified, and lower proportions of population targeted for PEP was calculated compared to the contextualized approach (11% and 11%). Conclusion A contextualized spatial approach could identify clusters in high-endemic districts more precisely than a standard statistical approach. Therefore, it can be a useful alternative to detect ... |
| Document Type: |
article in journal/newspaper |
| Language: |
English |
| DOI: |
10.1371/journal.pntd.0010972 |
| Availability: |
https://doi.org/10.1371/journal.pntd.0010972; https://dx.plos.org/10.1371/journal.pntd.0010972 |
| Rights: |
http://creativecommons.org/licenses/by/4.0/ |
| Accession Number: |
edsbas.379F2852 |
| Database: |
BASE |