Association between the december 2020 mass long-lasting insecticidal net distribution campaign and reported malaria incidence in Kinshasa, Democratic Republic of the Congo: an interrupted time-series analysis of routine surveillance data.

@article{mbuku2026association,
  title = {{Association between the december 2020 mass long-lasting insecticidal net distribution campaign and reported malaria incidence in Kinshasa, Democratic Republic of the Congo: an interrupted time-series analysis of routine surveillance data.}},
  author = {Nanou Mbuku and Bien-Aimé Mandja and Georges Muntenge and Victoire Nsabatien and Philippe Lukanu and Sandra Wanet and Thethé Mpongo and Baudouin Matela and Josué Zanga and Jean-Paul Gonzalez and Christophe Luhata and Bobette Matulonga and Nono Mvuama},
  journal = {BMC public health},
  year = {2026},
  month = aug,
  doi = {10.1186/s12889-026-28892-y},
  url = {https://europepmc.org/article/MED/42968019},
  note = {CC BY-NC-ND},
}

TY  - JOUR
TI  - Association between the december 2020 mass long-lasting insecticidal net distribution campaign and reported malaria incidence in Kinshasa, Democratic Republic of the Congo: an interrupted time-series analysis of routine surveillance data.
AU  - Nanou Mbuku
AU  - Bien-Aimé Mandja
AU  - Georges Muntenge
AU  - Victoire Nsabatien
AU  - Philippe Lukanu
AU  - Sandra Wanet
AU  - Thethé Mpongo
AU  - Baudouin Matela
AU  - Josué Zanga
AU  - Jean-Paul Gonzalez
AU  - Christophe Luhata
AU  - Bobette Matulonga
AU  - Nono Mvuama
PY  - 2026
DA  - 2026/08/05
JO  - BMC public health
DO  - 10.1186/s12889-026-28892-y
UR  - https://europepmc.org/article/MED/42968019
AB  - Background  Long-lasting insecticidal nets (LLINs) remain a cornerstone of malaria vector control in sub-Saharan Africa. Although mass LLIN distribution campaigns substantially increase household ownership and access, their population-level effect on malaria incidence may vary over time. We assessed the association between the December 2020 mass LLIN distribution campaign and malaria incidence in Kinshasa, Democratic Republic of the Congo (DRC), from 2018 to 2023.Methods  We conducted an ecological interrupted time-series (ITS) study using monthly routine malaria surveillance data from Kinshasa between January 2018 and December 2023. Negative binomial regression models were fitted with population included as a logarithmic offset. The primary segmented ITS model included three explicit ITS terms: calendar time, a binary intervention indicator, and post-intervention time. Seasonality was controlled for using a binary seasonal indicator and Fourier harmonic terms. Sensitivity analyses were conducted using January 2021 and February as alternative intervention start dates and by applying heteroskedasticity and autocorrelation-consistent (HAC/Newey-West) standard errors. Residual temporal autocorrelation was assessed using autocorrelation function (ACF) and partial autocorrelation function (PACF) plots and formal Ljung-Box tests.Results  In a simple pre/post model, the post-distribution period was associated with higher malaria incidence than the pre-distribution period (incidence rate ratio [IRR] = 1.23, 95% confidence interval [CI]: 1.07-1.41; p = 0.003). After adjustment for harmonic seasonal terms, this association remained significant (IRR = 1.16, 95% CI: 1.03-1.32; p = 0.013). In the primary segmented ITS model, the post-intervention period was associated with a higher incidence level (IRR = 1.18, 95% CI: 1.07-1.31; p = 0.001), while post-intervention time was associated with a modest but significant monthly increase in incidence (IRR = 1.01, 95% CI: 1.01-1.02; p < 0.001). Sensitivity analyses using January 2021 and February 2021 as alternative intervention dates, as well as HAC/Newey-West standard errors, yielded highly consistent findings. Significant residual autocorrelation remained present across model specifications.Conclusions  This ecological ITS analysis did not demonstrate a sustained decline in reported malaria incidence following the December 2020 LLIN distribution campaign in Kinshasa. Findings were robust across alternative intervention dates and autocorrelation-consistent sensitivity analyses but should be interpreted cautiously because of residual autocorrelation, potential reporting changes, the absence of a control group, and the ecological nature of the study.
SN  - 1471-2458
ER  -