Scholarlink

Background <p>Nigeria contributes nearly 27% of the global malaria burden, with over 68&#xa0;million cases and 194,000 deaths reported in 2021. Despite large-scale interventions, malaria remains heterogenous in space and time, complicating control efforts. Household surveys such as the NMIS provide robust prevalence estimates but are conducted only every five years, limiting their utility for timely programme action. In contrast, routine health facility data from the DHIS2 offer high-frequency, geographically disaggregated case and testing data, but require careful adjustment for reporting quality and denominators. Spatio-temporal modelling provides a way to harness these routine data to generate stable, decision-relevant risk estimates for sub-national stratification.</p> Method <p>We applied a Bayesian hierarchical spatio-temporal Poisson model to DHIS2 under five malaria case data from 2014 to 2023 in Kano and Lagos States, using the number of children tested as an offset. Spatial structure was modelled via a BYM2 prior, temporal drift with a random walk, and space–time interaction as Gaussian exchangeable noise. Model adequacy was assessed with DIC, WAIC, and mean log- CPO. GRID3 facility geocoordinates were used to derive an “underserved” proxy (≥ 30% of LGA land area &gt; 5&#xa0;km from a facility) to explore associations between access and risk.</p> Result <p>The Poisson likelihood provided the best fit (Kano: DIC = 6109, WAIC = 5977, mean log-CPO = − 7.76; Lagos: DIC = 2833, WAIC = 2772, mean log-CPO = − 7.90). Results show that by 2023, 73% of Kano LGAs were classified as high-risk (Pr(RR &gt; 1) ≥ 0.8) compared with 55% in Lagos. Persistent hotspots (high risk in 10/10 years) included Kunchi, Bagwai, Tudun Wada, Takai, Kumbotso, and Ungogo in Kano, and Ibeju-Lekki, Ifako-Ijaye, and Oshodi-Isolo in Lagos. In Kano, some high-risk LGAs overlapped with underserved areas e.g., Tudun Wada RR = 1.20 (1.19, 1.21), suggesting compounded vulnerability. Conversely, in Lagos only one LGA (Epe) was underserved (&gt; 30% beyond 5&#xa0;km) but showed low risk (RR = 0.72, 95% CrI: 0.71–0.73). Non-linear effects indicated that increased proximity of settlements to facilities was associated with reduced risk, while greater settlement counts were linked to elevated risk, reflecting both transmission potential and stronger surveillance capture.</p> Conclusion <p>Routine DHIS2 data, though imperfect, can yield stable and actionable malaria risk maps when analysed with Bayesian spatio-temporal methods. Our results highlight persistent high-burden LGAs in Kano and peri-urban hotspots in Lagos, providing evidence directly relevant to Nigeria’s High Burden to High Impact strategy and LGA-level stratification. This approach complements intermittent surveys by enabling continuous sub-national surveillance, guiding where to intensify interventions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Spatio-temporal modelling of malaria risk in Kano and Lagos States using local health system data

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年9月12日

查看全文
  • Olamide Akeboi,
  • Ebuka Nwafia,
  • Idim Godwin,
  • Ganiyat Eshikhena,
  • Obi Charles,
  • Koko Aadum,
  • Dupsy Akoma,
  • Ifeoma Ezenyi,
  • Chukwu Okoronkwo,
  • Onyebuchi Okoro,
  • Ezra Gayawan,
  • Chijioke Kaduru

摘要

Background

Nigeria contributes nearly 27% of the global malaria burden, with over 68 million cases and 194,000 deaths reported in 2021. Despite large-scale interventions, malaria remains heterogenous in space and time, complicating control efforts. Household surveys such as the NMIS provide robust prevalence estimates but are conducted only every five years, limiting their utility for timely programme action. In contrast, routine health facility data from the DHIS2 offer high-frequency, geographically disaggregated case and testing data, but require careful adjustment for reporting quality and denominators. Spatio-temporal modelling provides a way to harness these routine data to generate stable, decision-relevant risk estimates for sub-national stratification.

Method

We applied a Bayesian hierarchical spatio-temporal Poisson model to DHIS2 under five malaria case data from 2014 to 2023 in Kano and Lagos States, using the number of children tested as an offset. Spatial structure was modelled via a BYM2 prior, temporal drift with a random walk, and space–time interaction as Gaussian exchangeable noise. Model adequacy was assessed with DIC, WAIC, and mean log- CPO. GRID3 facility geocoordinates were used to derive an “underserved” proxy (≥ 30% of LGA land area > 5 km from a facility) to explore associations between access and risk.

Result

The Poisson likelihood provided the best fit (Kano: DIC = 6109, WAIC = 5977, mean log-CPO = − 7.76; Lagos: DIC = 2833, WAIC = 2772, mean log-CPO = − 7.90). Results show that by 2023, 73% of Kano LGAs were classified as high-risk (Pr(RR > 1) ≥ 0.8) compared with 55% in Lagos. Persistent hotspots (high risk in 10/10 years) included Kunchi, Bagwai, Tudun Wada, Takai, Kumbotso, and Ungogo in Kano, and Ibeju-Lekki, Ifako-Ijaye, and Oshodi-Isolo in Lagos. In Kano, some high-risk LGAs overlapped with underserved areas e.g., Tudun Wada RR = 1.20 (1.19, 1.21), suggesting compounded vulnerability. Conversely, in Lagos only one LGA (Epe) was underserved (> 30% beyond 5 km) but showed low risk (RR = 0.72, 95% CrI: 0.71–0.73). Non-linear effects indicated that increased proximity of settlements to facilities was associated with reduced risk, while greater settlement counts were linked to elevated risk, reflecting both transmission potential and stronger surveillance capture.

Conclusion

Routine DHIS2 data, though imperfect, can yield stable and actionable malaria risk maps when analysed with Bayesian spatio-temporal methods. Our results highlight persistent high-burden LGAs in Kano and peri-urban hotspots in Lagos, providing evidence directly relevant to Nigeria’s High Burden to High Impact strategy and LGA-level stratification. This approach complements intermittent surveys by enabling continuous sub-national surveillance, guiding where to intensify interventions.