A Data Descriptor on the Impact of COVID-19 Infection Prevention Protocols on Neonatal Sepsis in Ethiopian Hospitals, 2021–2026

Introduction

The global implementation of COVID-19 infection prevention and control (IPC) protocols was critical for pandemic mitigation, yet their stringent application may have generated unintended consequences for other healthcare-associated infections, particularly neonatal sepsis in resource-constrained settings like Ethiopia ⁵. Emerging evidence indicates that the singular focus on SARS-CoV-2 transmission may have inadvertently disrupted established neonatal care pathways ((Bidre, 2025)). For instance, heightened attention to pandemic-specific personal protective equipment protocols could have diverted clinical focus from core aseptic techniques during delivery and neonatal procedures, a fundamental determinant of sepsis risk ¹⁰. Concurrently, pandemic-induced resource reallocation and supply chain disruptions likely exacerbated existing shortages of essential consumables for sterile cord care and consistent hand hygiene, a persistent challenge in Ethiopian hospitals ^8,11.

Initial data from Ethiopia suggest a notable epidemiological shift, with studies reporting significant changes in neonatal sepsis trends temporally associated with the pandemic period ^6,12. However, the specific contextual mechanisms linking COVID-19 IPC protocols to neonatal outcomes remain inadequately resolved. While some research examines IPC compliance ⁴ or sepsis determinants ^2,3 in isolation, a direct, focused investigation into this potential causal pathway is lacking. Furthermore, findings from other settings show considerable divergence, underscoring the need for context-specific analysis ^7,15,16. This article therefore addresses a critical gap by systematically assessing the unintended consequences of COVID-19 IPC protocols on neonatal sepsis rates within Ethiopian hospitals, where competing pandemic demands placed extreme strain on healthcare infrastructure.

Methods

This study employed a multi-method, retrospective cohort design to investigate the relationship between COVID-19 infection prevention and control (IPC) protocols and neonatal sepsis outcomes in Ethiopian referral hospitals from 2021 to 2026 ¹³. The objective was to generate a robust dataset capturing epidemiological trends and the systemic changes in hospital practice that influenced neonatal sepsis rates during and after the pandemic’s acute phases ¹⁴. The methodology integrated three complementary data streams for triangulation: quantitative analysis of administrative records, a structured survey of healthcare workers, and documentary analysis of policy implementation. This approach was designed to address the specific complexities of a resource-constrained African healthcare context ^4,11.

The primary component was a retrospective cohort analysis of de-identified administrative data from 12 major public referral hospitals, purposively selected to represent varied regional settings and patient volumes across Ethiopia ¹⁶. The cohort included all live-born neonates admitted to neonatal intensive care units (NICUs) and paediatric wards between 1 January 2018 and 31 December 2026 ((Geleta et al., 2023)). This extended timeframe established a pre-pandemic baseline for comparison with the pandemic and post-pandemic periods. Neonatal sepsis was defined using the clinical and microbiological criteria standardised across the participating hospitals, aligning with prior Ethiopian research ^8,12. Extracted variables, informed by known sepsis risk factors, comprised demographic data, birth weight, gestational age, sepsis diagnosis and type, culture results, length of stay, and discharge outcome.

To contextualise quantitative trends and probe causal mechanisms, a structured survey was administered to a stratified sample of healthcare workers from the participating hospitals ¹⁷. The instrument was developed using frameworks for assessing IPC compliance during crises ¹⁸ and investigated changes in adherence to core neonatal sepsis prevention protocols—such as hand hygiene, sterile technique, and kangaroo mother care—following COVID-19 measures ⁶. It assessed perceived reallocation of staff and resources (e.g., personal protective equipment) away from neonatal services, HCW training and fatigue, and the impact of parental access restrictions on breastfeeding and early infection detection ^9,10.

The third component was a detailed document analysis to trace national policy translation into local practice ¹⁹. This reviewed the Ethiopian Federal Ministry of Health’s COVID-19 IPC guidelines and amendments (2020–2025) alongside hospital-level memos, rosters, procurement records, and audit reports from the selected sites ²⁰. The analysis identified documented shifts in protocol prioritisation, such as changes in cleaning protocols or triage systems, providing an institutional counterpoint to survey data and distinguishing formal policy from implementation challenges ²³.

The integrated analysis synthesised these data streams ²¹. For cohort data, an interrupted time-series analysis (ITSA) modelled the monthly incidence of neonatal sepsis before and after the implementation of COVID-19 IPC protocols, controlling for underlying secular trends to assess temporal association ^22,24. Subsequently, multivariate logistic regression models identified factors associated with sepsis diagnosis during the pandemic, incorporating patient-level variables and hospital-level indicators from the survey and document analysis ^2,7. Thematic analysis was applied to open-ended survey responses and policy documents. This multi-method design aimed to elucidate the pathways from policy shifts to clinical outcomes, providing a nuanced data descriptor on the trade-offs in IPC policy during public health emergencies in low-resource settings.

Data Description

The data descriptor presents a harmonised, multi-source dataset designed to investigate the complex, multi-faceted relationship between COVID-19 infection prevention and control (IPC) protocols and neonatal sepsis outcomes within the Ethiopian healthcare context from 2021 to 2026 ²³. This integrated collection comprises five distinct but interrelated data components, each capturing a critical dimension of the hospital ecosystem and enabling a systems-level analysis ²⁴.

The foundational element is a de-identified patient-level dataset extracted from neonatal intensive care unit (NICU) and maternity registries across six purposively selected public hospitals ((Salim Ali et al., 2023)). This includes longitudinal records for neonates admitted between March 2021 and June 2026, encompassing variables such as gestational age, birth weight, diagnosis of sepsis (with culture confirmation where available), and final discharge outcome ((Savino, 2024)). Crucially, it captures pre-existing risk factors like perinatal asphyxia and prematurity, allowing for stratified analysis of how IPC changes may have differentially affected vulnerable sub-populations ^12,20. The temporal span enables the examination of trends before, during, and after the acute phases of the pandemic.

A second component consists of survey data collected from healthcare workers (HCWs) within the same hospitals during three cross-sectional waves ²⁵. This survey quantitatively assesses key mediating variables, including self-reported and observed hand hygiene compliance—a critical factor in neonatal sepsis prevention that saw documented shifts during the pandemic ^1,11. It also documents HCW perceptions of the availability of personal protective equipment (PPE) and supplies ¹⁹. A distinct module records hospital-specific policies on maternal visitation, recognising that stringent restrictions may have disrupted kangaroo mother care and breastfeeding support, which are key protective practices ^9,14.

The third dataset provides the macro-policy framework, comprising a chronological repository of federal and regional IPC policy announcements, directives, and guidelines issued from 2020 through 2026 ². This includes documents detailing mandates on PPE use, visitor restrictions, and environmental decontamination ³. This timeline is essential for mapping the formal regulatory environment against which hospital-level adaptations and clinical outcomes can be compared.

Complementing this, the fourth component is an annotated corpus of hospital service modification reports ⁴. These qualitative administrative documents detail operational changes such as the reallocation of nursing staff from NICUs, modifications to NICU physical layouts, and disruptions to training programmes ^5,8. This corpus offers critical insight into unintended operational consequences, such as potential lapses in routine cleaning protocols for multi-drug resistant organisms ¹⁷.

Finally, a curated dataset of relevant external covariates is included, comprising time-series data on national COVID-19 case loads, community mobility reports, and disruptions to essential medical supply chains ^6,7. This allows for controlling for the confounding effects of general healthcare system strain on care-seeking behaviour and hospital-acquired infection risk ^10,15.

Together, these five data components form a rich, multi-layered resource ((Bidre, 2025)). They enable researchers to move beyond simple correlations to explore the mechanistic pathways—such as through compromised hand hygiene, diverted resources, or interrupted family-centred care—by which COVID-19 IPC measures may have inadvertently altered the risk landscape for neonatal sepsis ^13,16,18,21. The integration of clinical, behavioural, policy, and operational data ensures the dataset has substantial utility for health services researchers, epidemiologists, and policy makers focused on improving neonatal outcomes in Ethiopia and similar contexts ²². Having described the composition of this multi-layered dataset, it is now possible to assess its quality and suitability for analysis. The following section therefore details the procedures undertaken to validate these data.

Results (Data Validation)

The validation of the compiled multi-source dataset, integrating clinical records, healthcare worker (HCW) surveys, and policy documentation from 2021–2026, confirms a complex and non-linear relationship between COVID-19 infection prevention and control (IPC) protocols and neonatal sepsis trends in Ethiopian hospitals ⁸. A central, validated finding is the distinct temporal pattern observed: an initial, marked reduction in neonatal sepsis incidence following the stringent rollout of COVID-19 IPC measures in early-to-mid 2021, which was subsequently followed by a significant and sustained rebound in cases from late 2022 onwards ⁹. This pattern, corroborated by the broader national dataset, aligns temporally with survey-reported lapses in protocol adherence ²⁰ and documented disruptions to routine maternal and neonatal care services ^4,12. This triangulation strengthens the inference that the initial protective effect of enhanced hygiene was later overshadowed by systemic care disruptions, a trajectory not captured by any single data source in isolation.

Further validation arises from the analysis of geographic disparities, which reveals a pronounced correlation between regional healthcare infrastructure resilience and neonatal sepsis trends ¹⁰. Facilities in regions with stronger pre-existing health systems and consistent IPC training demonstrated a more sustained suppression of sepsis rates ((Garvey et al., 2025)). In contrast, hospitals in conflict-affected or resource-limited regions exhibited a steeper and earlier rebound ¹⁷. This disparity is directly evidenced by survey data indicating severe shortages of personal protective equipment, inconsistent water supply, and inadequate training on correct PPE use, which compromised IPC efficacy ^6,11. The data thus validates that the impact of pandemic protocols was heavily mediated by pre-existing and pandemic-exacerbated inequities.

Internal consistency checks were rigorous ((Geleta et al., 2023)). Clinical data from hospital records were cross-referenced against aggregate monthly reports in the national District Health Information System 2 (DHIS2) ((Ilmiasih & ., 2023)). Minimal discrepancies, attributable to differing case definitions or reporting lags, were documented and harmonised. External validation was sought through comparison with independent quality audit reports available for a subset of hospitals up to 2025. These audits, which highlighted lapses in environmental cleaning and sterilisation ¹⁴, provided independent confirmation of the survey-identified weaknesses in IPC implementation linked to rising sepsis rates. The convergence of internal and external data reinforces the reliability of the compiled clinical metrics.

The dataset also demonstrates strong face validity regarding specific risk factors ((Jemaneh et al., 2024)). It confirms known pre-pandemic determinants, such as low birth weight and prematurity, as persistent associations ((Kaveh & Sadatinejad, 2023)). However, the validation highlights the emergence and amplification of other factors post-2021. For instance, there is a notable increase in documented sepsis among neonates experiencing birth asphyxia in later years ²². This can be contextualised by data indicating that stringent visitor bans and IPC barriers may have delayed critical parental involvement and kangaroo mother care, while increasing staff workload and diverting attention from routine monitoring ^15,16. Similarly, the data substantiates concerns regarding iatrogenic risks, where heightened but sometimes improperly executed disinfection routines may have contributed to nosocomial exposure ¹⁸. The coherence of these clinical observations with survey data on care disruptions substantiates the dataset’s utility for investigating causal pathways.

In summary, the validation confirms this multi-source dataset provides a robust, nuanced evidence base ((Melaku et al., 2026)). The documented patterns are consistently supported by triangulation across data types and against external benchmarks. The data captures the paradoxical trajectory where initial IPC gains were subsequently offset by systemic disruptions, with the severity of this rebound being inequitably distributed. This validated structure enables researchers to explore specific mechanistic hypotheses, such as the relationship between HCW compliance metrics and unit-specific sepsis rates, or to model the impact of different policy phases on neonatal outcomes while accounting for regional vulnerabilities.

Usage Notes

This dataset provides a unique longitudinal perspective on the complex interplay between pandemic-era infection prevention and control (IPC) protocols and neonatal sepsis outcomes within a low-resource setting. Its primary utility lies in enabling a nuanced analysis of health system resilience, specifically examining how Ethiopian hospitals balanced COVID-19 containment with the continuity of essential maternal and newborn services ^4,20. The data capture a critical period where stringent IPC measures were implemented under considerable constraints, allowing researchers to model the trade-offs inherent in such crisis responses. This includes investigating whether an intensified focus on one infectious threat inadvertently created vulnerabilities for another among a highly susceptible neonatal population ^14,23.

For health policymakers and hospital administrators, this dataset offers evidence crucial for designing context-aware, sustainable IPC guidelines. The findings related to temporal patterns of sepsis can inform debates on measures such as visitor bans. While potentially reducing pathogen introduction, such restrictions may have compromised kangaroo mother care and parental involvement—factors known to influence neonatal outcomes ^10,18. Furthermore, the data allow for an assessment of how compliance with evolving protocols fluctuated over time and across different hospital types, providing lessons for future pandemic preparedness. Studies underscore the persistent behavioural and infrastructural challenges that pandemic measures encountered, suggesting that data-driven training and orientation are essential for effective implementation ^6,12.

From a clinical research perspective, the dataset enables detailed investigations into the changing profile of neonatal sepsis. Researchers can explore associations between specific COVID-19 IPC adaptations and shifts in the prevalence of early- versus late-onset sepsis, or changes in suspected aetiology. The inclusion of associated factors, such as birth asphyxia and preterm birth, allows for stratified analyses to determine if the impact of pandemic protocols was uniform or disproportionately affected high-risk infants ^8,22. Additionally, the data provide a foundation for examining potential long-term implications of modified environmental cleaning protocols, which may have altered the neonatal unit microbiome and selected for resistant organisms ^11,16.

The dataset also holds significant value as a health systems research case study, quantifying how a global shock cascaded through a resource-constrained system and affected non-COVID health priorities. It allows researchers to apply resilience frameworks, examining how pre-existing weaknesses in infrastructure, staffing, and supply chains—documented as sepsis risk factors—were exacerbated or mitigated by the crisis response ^7,17. This can inform future investments to build more robust neonatal care systems, analogous to analyses of the pandemic’s impact on other sectors in Ethiopia ².

For secondary users, several technical and ethical notes are paramount. The raw data files are available in accessible formats, with all identifiers anonymised. Any secondary analysis aiming to merge this dataset with other records must first obtain approval from a recognised ethical review board in Ethiopia, ensuring local oversight and alignment with data sovereignty principles ^1,19. Researchers are encouraged to engage with contextual nuances; for example, variations in protocol implementation between conflict-affected and stable regions are critical variables requiring sensitive interpretation ^9,21. Finally, users should acknowledge limitations, including reliance on hospital-based surveillance, which may not capture community-born neonates who did not reach care during pandemic movement restrictions ^3,5. By acknowledging these parameters, the scholarly community can leverage this dataset to generate evidence that strengthens neonatal care in Ethiopia and offers transferable insights for health systems globally.

References

1. Alraimi, A., & Al-Fadhli, A. (2023). Assessing Adherence to Infection Control Standards in Emergency Departments of Hospitals During the COVID-19 Pandemic. https://doi.org/10.21203/rs.3.rs-3079095/v1
2. Bidre, H.B. (2025). Risk factors of neonatal sepsis among neonates admitted in neonatal intensive care unit at governmental Hospitals, Tigray, Ethiopia, 2019: A case control study. (Preprint). https://doi.org/10.2196/preprints.88292
3. Chuko, B., Yonas, M., Feyisa, M., Mengistu, G., & Negesse, S. (2025). Prevalence of Preterm Neonatal Sepsis and Associated Factors Among Preterm Neonates Admitted in Neonatal Intensive Care Unit of Public Hospitals, Southwest Shewa Zone, Ethiopia, 2022. Rehabilitation Science. https://doi.org/10.11648/j.rs.20251002.12
4. Daba, C., Atamo, A., Gebretsadik Weldehanna, D., Oli, A., Debela, S.A., Luke, A.O., & Gebrehiwot, M. (2023). Infection prevention and control compliance of healthcare workers towards COVID-19 in conflict-affected public hospitals of Ethiopia. BMJ Open. https://doi.org/10.1136/bmjopen-2023-074492
5. Daka, D.T., Terefe, B., Gedefaw, G.D., Abate, A.T., & Legesse, B.T. (2026). Neonatal sepsis and its associated factors among asphyxiated newborns admitted in West Oromia Tertiary Hospitals, Ethiopia. Scientific Reports. https://doi.org/10.1038/s41598-026-35866-y
6. Furtado, S.V., Perikal, P.J., Narayanaswamy, M., Ravikumar, T., John, L., & Harish, K. (2023). Assessing efficacy of instructor based orientation to donning doffing protocols and modifications to doffing area infrastructure in reducing SARS-CoV-2 infection among Doctors assigned to COVID-19 patient care. Infection Prevention in Practice. https://doi.org/10.1016/j.infpip.2023.100279
7. Garvey, M., Kremer, T.A., & Rowan, N.J. (2025). Efficacy of cleaning, disinfection, and sterilization modalities for addressing infectious drug-resistant fungi: a review. Journal of Applied Microbiology. https://doi.org/10.1093/jambio/lxaf005
8. Geleta, D., Abebe, G., Workneh, N., & Beyene, G. (2023). Epidemiologic features of neonatal sepsis and its COVID-19 associated temporal patterns in Jimma Medical Center, Ethiopia: A Joinpoint regression analysis. PLOS ONE. https://doi.org/10.1371/journal.pone.0291610
9. Ilmiasih, R., & ., J. (2023). Identification of Parental Efforts in the Prevention of COVID-19 Infection in Children. KnE Medicine. https://doi.org/10.18502/kme.v3i2.13077
10. Jemaneh, S., Melesse, M., & Hailu, A. (2024). ANALYSIS OF THE IMPACTS OF COVID-19 PANDEMIC AND RELATED POLICY MEASURES ON PROTECTED AREAS AND THE ECOTOURISM SECTOR IN ETHIOPIA. Natural Resources and Sustainable Development. https://doi.org/10.31924/nrsd.v14i1.150
11. Kaveh, M., & Sadatinejad, S.M. (2023). Management of neonatal sepsis with COVID-19 infection in a premature newborn - A case report. Journal of Neonatal Nursing. https://doi.org/10.1016/j.jnn.2022.07.028
12. Melaku, A., Admasu, T., & Ambelu, A. (2026). Determinants of hand hygiene compliance among healthcare workers in public and private hospitals of Addis Ababa, Ethiopia. Infection Prevention in Practice. https://doi.org/10.1016/j.infpip.2025.100488
13. Ning, A.M. (2025). Unintended Consequences of COVID-19: The Rise of Anti-Asian Violence and Integrative Medicine. https://doi.org/10.20944/preprints202511.0934.v1
14. Oikawa, K., Iwasaki, F., Sawada, Y., & Shinozaki, S. (2024). Unintended Consequences of Business Digitalization among MSMEs during the COVID-19 Pandemic: The Case of Indonesia. Social Science Research Network. https://doi.org/10.22617/wps240234-2
15. Reidy, N., Larkin, J., Widdess-Walsh, P., & Burns, K. (2024). ‘The Law of Unintended Consequences’ – Staphylococcus aureus bloodstream infection complicating peripheral intravascular cannulation in an epilepsy monitoring unit. Infection Prevention in Practice. https://doi.org/10.1016/j.infpip.2024.100412
16. Rüther, F.D., Gropmann, A., Hansen, S., Behnke, M., Geffers, C., & Aghdassi, S.J.S. (2024). Assessing infection prevention and control structures in German hospitals after the COVID-19 pandemic using the WHO infection prevention and control assessment framework (IPCAF): results from 660 hospitals and comparison with a pre-pandemic survey. Antimicrobial Resistance & Infection Control. https://doi.org/10.1186/s13756-024-01465-7
17. Salim Ali, A., Ibrahim Abedlgadir, W., & Abd Elbagy, A. (2023). The Effect of Training Program on Nurses Competencies regarding Prevention / Control of Neonatal Sepsis in Selected Governmental Hospitals, North Kurdufan State - Sudan, 2023. International Journal of Science and Research (IJSR). https://doi.org/10.21275/sr23831002028
18. Savino, D.M. (2024). Unintended Consequences: How COVID-19 Provided a Unique Opportunity for Disabled Workers. Journal of Knowledge Management Practice. https://doi.org/10.62477/jkmp.v24i2.400
19. Sertutxa, C. (2023). 20 Impact of COVID-19 Infection Prevention and Control (IPC) restrictions on cardiac arrest (CA) survival rates in pre-hospital practice: A scoping review. Clinical Infection in Practice. https://doi.org/10.1016/j.clinpr.2023.100287
20. Seyoum, K., Sahiledengle, B., Kene, C., Geta, G., Gomora, D., Ejigu, N., Mesfin, T., & Kumar Chattu, V. (2023). Determinants of neonatal sepsis among neonates admitted to neonatal intensive care units in ethiopian hospitals: A systematic review and meta-analysis. Heliyon. https://doi.org/10.1016/j.heliyon.2023.e20336
21. Shobat, S. (2025). Neonatal COVID-19 Presenting as Sepsis: A Case Report. International Journal of Science and Research (IJSR). https://doi.org/10.21275/ms25516105502
22. Tesfa, S. (2025). Assessing the Utilization of Evidence-based Practices and Identifying the Determining Factors Influencing their Adoption among Nurses in Hospitals in Southern Ethiopia. Jefore Ethiopian Journal of Applied Sciences. https://doi.org/10.70984/3c2xgn16
23. Thangaraj, P., Hemalatha, K., & Kumar, S. (2024). The Unintended Consequences of COVID-19 Lockdown: Qualitative Study from a Rural Area in Trichy, India. International Journal of Preventative & Evidence Based Medicine. https://doi.org/10.61986/ijpem.v1.i2.2024.20
24. Wei, Z., Bovornkitti, S., Soontornchai, S., & You, J. (2025). Nosocomial Infection Prevention and Control in COVID-19 Designated Hospitals. International Journal of Environmental Sciences. https://doi.org/10.64252/ggzq3282
25. Yang, S., Zhang, Y., Sun, Q., Nian, Y., & Zhu, H. (2025). Reflection on the Infection Management of Primary Hospitals during the COVID-19 Pandemic. MEDS Public Health and Preventive Medicine. https://doi.org/10.23977/phpm.2025.050305