Journal of Global Health最新文献

India released its 2021 official death registration data in May 2025, showing a dramatic undercount of reported COVID-19 mortality in 2021 - a year that included the lethal second wave from the Delta variant in April to June 2021. The civil registration system (CRS) documented 10.2 million deaths in 2021 - a 26% increase from 2020 - compared to only 335 004 reported COVID-19 deaths. We calculated excess deaths as the difference between observed deaths and expected deaths had pre-pandemic mortality trends continued. Our analysis reveals India experienced approximately 2.4 million excess deaths in 2021, representing a 7.2× undercount compared to reported COVID-19 deaths. This aligns with the estimates derived from several epidemiological models during and after the pandemic, indicating excess-to-reported COVID-19 death ratios ranging from 4.4× to 11.9×. State-level analyses revealed considerable variation in reporting fidelity, with excess-to-COVID-19 death ratios ranging from under 2× in Kerala and Goa to over 40 in Gujarat. Limited disaggregated data showed excess death rates were substantially higher in men than women (2.2 vs. 1.3 per 1000 population), and in urban than rural areas (2.3 vs. 1.4 per 1000 population). Since the CRS data are incomplete in terms of age-stratified deaths, the proportional allocation of deaths by age according to external sources suggests excess death rates were higher in the 65 and older age group than in the population under 65 (14.1 vs. 0.9 per 1000 population). The four-year data delay and systematic underreporting underscore urgent needs for modernising India's surveillance system during acute phases of the pandemic and restructuring the official vital registration systems through protocol standardisation, real-time linkages, and infrastructure investments. Robust mortality tracking strengthens crisis preparedness and broader public health response.

Anonymisation is intended to confer protection in research, yet in research where disclosure poses minimal threat, blanket anonymisation may disempower participants who explicitly seek recognition. Drawing on fieldwork examining task shifting and sharing in India, we present an example where the leadership of a non-profit questioned why they must stay anonymous, viewing research participation as a rare chance to share learnings, document impact, and build credibility; opportunities otherwise constrained by funding and capacity limitations. Looking through an epistemic injustice lens, we argue that mandatory anonymisation policies reflect global north institutional assumptions about protection rather than the preferences of participants, potentially perpetuating patterns where researchers advance their goals whilst organisations whose knowledge forms their evidence remain invisible. While anonymisation is critical for sensitive research where disclosure could do harm, we call for flexible ethical guidelines that empower participants to make informed choices regarding recognition if risks are minimal. We recommend tiered consent processes which allow participants to select their levels of identification, differentiate between organisational and individual visibility, and use context-dependent frameworks like the Global Ethical Research toolkit to move towards proportionate, context-dependent decisions that truly protect those we study.

Background: The World Health Organization recommends two passive immunisation strategies to prevent respiratory syncytial virus (RSV) disease in young infants. Both are being introduced in high-income settings, but their affordability and cost-effectiveness have not been evaluated in many low- and middle-income countries. Preliminary estimates of cost-effectiveness are needed to guide immunisation policy and planning in Nepal.

Methods: We estimated the potential health impact and cost-effectiveness of introducing a maternal vaccine (RSVpreF) or long-acting infant monoclonal antibody (mAb) (nirsevimab) over the period 2025-34 in Nepal. We compared both interventions to the status quo (no intervention) and to each other. Model inputs included health care cost estimates from a recent prospective cost-of-illness study in Kathmandu, as well as the latest efficacy data from clinical trials. The primary outcome measure was the incremental cost (2023 USD) per disability-adjusted life year (DALY) averted from a governmental health perspective. We conducted a range of deterministic analyses, including scenarios that incorporated a societal perspective and a seasonal approach. Additionally, we performed probabilistic uncertainty analyses to assess decision uncertainty and estimated the likelihood of cost-effectiveness for each intervention across a range of willingness-to-pay thresholds.

Results: Introducing a maternal vaccine (USD 5/dose, 81% coverage, 69% efficacy, 6 months protection) or long-acting infant mAb (USD 5/dose, 97% coverage, 77% efficacy, 5 months protection) could prevent >2300 deaths and >50 000 hospital admissions over ten years. The discounted immunisation programme costs were estimated to be USD 30 and USD 35 million, respectively. Compared to the status quo, the maternal vaccine and the long-acting infant mAb were estimated to cost USD 387 and USD 486 per DALY averted, respectively, which is around 0.3 times and 0.4 times the national gross domestic product (GDP) per capita. There was a 95% probability that the maternal vaccine would be cost-effective at USD 5 per dose, assuming a willingness-to-pay threshold of 0.5 times the national GDP per capita. With our base case assumptions, the maternal vaccine dominated the mAb (i.e. generated more health benefits at a lower cost). However, the results (and the rank order of interventions) were sensitive to the dose price, efficacy, duration of protection, and RSV disease burden estimates. Cost-effectiveness of the mAb improves with timely administration or when a seasonal approach is implemented.

Conclusions: New passive immunisation strategies have the potential to prevent a substantial number of RSV-related hospitalisations and deaths in Nepal. Cost-effectiveness and product choice will heavily depend on the price negotiated for each product.

Background: Vaccine hesitancy persists because definitive evidence regarding the underexplored safety of COVID-19 vaccines in pregnancy is still lacking, particularly concerning their effects on the maternal-foetal interface (MFI) and potential links to miscarriage. We aimed to verify whether COVID-19 vaccines modulate gene expression at the MFI, thereby influencing recurrent miscarriage.

Methods: We conducted an RNA sequencing analysis on decidual tissues from six pairs of early pregnancy participants, both vaccinated and unvaccinated. We extracted the data sets associated with COVID-19 placenta (GSE181238) and recurrent miscarriage (GSE22490) from the Gene Expression Omnibus database for further bioinformatic analysis, focussing on the expression, function, and distribution of core genes at the MFI.

Results: Compared to the control group, 879 differentially expressed genes (P < 0.05; fold changes >1.5; false discovery rate <0.05) were identified in the vaccinated group. Complement activation and cell adhesion pathways were up-regulated, while the graft-vs-host response was down-regulated. The vaccine down-regulated some genes overexpressed in recurrent miscarriage cases. Three significant genes - FOS, FOSB, and LY96 - associated with miscarriage were identified; these genes are up-regulated during infection but suppressed by the vaccine. Functional enrichment analysis revealed the vaccine's immune activity, similar to but weaker than COVID-19 infection, and it inhibited certain miscarriage-related pathways, such as the tumour necrosis factor signalling pathway. Gene set variation analysis suggested a positive influence of the vaccine on immune tolerance at MFI.

Conclusions: This study indicates that the COVID-19 vaccine may exert nonnegative effects on the maternal-foetal immune micro-environment and is unlikely to increase the risk of miscarriage.

The COVID-19 pandemic catalysed the development of digital health interventions across the globe, including low- and middle-income countries (LMICs) such as Malaysia. However, moving from pockets of innovation to sustainable implementation at scale remains a major challenge. This viewpoint presents insights from a digital health training programme using a multi-stakeholder engagement series convened by the National Institute for Health and Care Research Global Health Research Unit on Respiratory Health (RESPIRE) during the pandemic. Through co-designed workshops involving policymakers, healthcare providers, small and medium-sized enterprises, and academic researchers, participants examined systemic barriers to scaling digital health innovations in Malaysia, including issues with infrastructure, regulation, and workforce readiness. We used a home-based pulmonary rehabilitation initiative as a case study to explore these dynamics in practice. Broader lessons include the importance of ecosystem-building, capacity development, regulatory clarity, and inclusive design. Our findings offer transferable insights for strengthening digital health systems in LMICs.

Background: Frailty in older adults has become a major concern. It is influenced by biological, social, psychological, and environmental factors, with social frailty playing a particularly significant role. The relationship between social frailty and health outcomes can accelerate frailty transitions, offering new insights into strategies to improve health in older adults.

Methods: We obtained the data from the Fourth Sample Survey of the Aged Population in Urban and Rural China, with 2017 as the baseline and 2019 as the follow-up. We used the frailty index (FI) to assess physiological frailty and the social vulnerability index (SVI) to assess social frailty. We used logistic regression to analyse the impact of SVI on frailty transitions.

Results: The analysis included 9093 older individuals, with an average age of 71.2 (standard deviation = 7.0) years, comprising 4495 women and 4598 men. Of these, 39.3% were robust, 45.1% were prefrail, and 15.6% were frail. Both the FI and SVI increased with age, and women showed higher frailty levels. Over the two years, 56.2% maintained stable frailty status, 14.2% improved, and 29.6% deteriorated. Correlation analysis revealed a moderate relationship between FI and SVI (r = 0.337; P < 0.001). Logistic regression analysis indicated that higher social frailty was associated with an increased risk of worsening frailty in non-frail (odds ratio (OR) = 1.017; P < 0.05) and prefrail individuals (OR = 1.021; P < 0.05), but had no effect on those who were already frail.

Conclusions: Frailty deterioration with age is more common than improvement. Increased social frailty is a significant risk factor for the worsening of frailty, particularly in non-frail and prefrail individuals. Early identification and prevention of social frailty may help delay its progression.

Background: The impact of blood pressure fluctuations on the prognosis of stroke has been well documented, but little is known about the association between long-term systolic blood pressure (SBP) levels and the risks of cardiovascular outcomes in patients with ischemic stroke (IS).

Methods: In this retrospective cohort study, we included a total of 11 357 eligible IS patients hospitalised in Shenzhen, China between 1 July 2017 and 1 October 2023. One-year levels of SBP after IS patient discharge were identified using group-based trajectory models (GBTM). Propensity score-overlap weighted Cox regression models were used to assess the associations between SBP levels and the risks of recurrent stroke and major adverse cardiovascular events (MACE; including recurrent stroke, ischemic heart disease, and heart failure) within a 36-month follow-up period, respectively. Furthermore, we quantitatively assessed the benefits potentially gained from optimal SBP levels by calculating age-scale restricted mean survival times.

Results: Three one-year GBTM-derived SBP level patterns were identified: normal (n = 2120), high-normal (n = 7949), and uncontrolled SBP (n = 1288). During a median follow-up of 1.75 years, IS patients with normal and high-normal SBP were associated with lower risks of recurrent stroke or MACE, with weighted hazard ratios (95% confidence interval (CI)) ranging from 0.68 (95% CI = 0.54-0.86) to 0.89 (95% CI = 0.78-1.02), compared to those with uncontrolled SBP. Furthermore, IS patients aged 45 to 70 years with normal or high-normal SBP may derive greater health benefits, with the event-free survival time ranging from 7.12 to 0.27 years.

Conclusions: Maintaining sustained normal or high-normal SBP levels one year after discharge may be associated with a reduced risk of adverse cardiovascular events and potentially yields greater health benefits for IS patients.

Background: Chronic obstructive pulmonary disease (COPD) is a leading cause of mortality and morbidity globally, disproportionately affecting low- and middle-income countries (LMICs). Despite pulmonary rehabilitation (PR) being a key intervention, uptake and adherence remain low due to economic, geographical, and sociocultural barriers. We explored the lived experiences of individuals with COPD and their caregivers in Malaysia to identify contextually grounded self-management strategies.

Methods: We employed a qualitative photo-elicitation approach between January and December 2024. We purposively sampled adults with COPD and their caregivers based on age, gender, and ethnicity from a hospital-based outpatient PR centre in Selangor, Malaysia. We conducted semi-structured dyadic interviews at two time points. Participants documented their experiences through photographs, which guided the discussions. Lastly, we transcribed the interviews verbatim and thematically analysed them.

Results: Nine dyads (participant-caregiver pairs) completed two interviews. Participants were men with a mean age of 65.3 (standard deviation (SD) = 3), with GOLD stage 3-4. Caregivers were women, with a mean age of 56.4 (SD = 11). Six dyads identified themselves as Malay ethnicity. Four themes emerged: Navigating economic constraints in COPD self-management, where participants substituted costly devices with low-cost tools (e.g. loaded trolleys); Culturally embedded self-management: integrating practices like Qigong and reframing daily chores (e.g. folding laundry) as rehabilitation; Technology as a tool for home-based COPD care with participants adapting exercises from internet (e.g. Facebook) while caregivers expressed concerns over unverified content; and Family as partners in COPD management, where caregivers not only monitored symptoms but also exercised alongside participants, reporting mutual health benefits. These strategies were seen as essential for sustaining engagement in COPD care.

Conclusions: Photo-elicitation and dyadic interviews revealed how cultural traditions, digital adaptations, and reciprocal caregiving intersect in everyday life, shaping COPD self-management in low-resource settings. Interventions should build on these lived strategies, prioritising context-sensitive, low-cost, and inclusive care models for COPD in LMICs.

Background: Mental health interventions in low- and middle-income countries (LMICs) face significant sustainability challenges, often leading to 'programme drift' (protocol deviation ) and 'voltage drop' (reduced effectiveness). While implementation science frameworks emphasise fidelity, they often fail to explain how frontline providers in resource-constrained settings maintain services. Here, we investigate how adaptive mechanisms function as legitimate sustainability strategies within Lagos, Nigeria's Mental Health in Primary Care programme, which contends with chronic underfunding, high staff turnover, and community stigma.

Methods: We conducted a convergent mixed-methods study in six Lagos local government areas. Data were collected from 130 stakeholders (policymakers, managers, health workers, care recipients) through quantitative surveys and from a nested subsample of 70 participants through in-depth interviews and institutional ethnography. We analysed quantitative data using multiple regression and qualitative data using thematic analysis, systematically integrating the findings through triangulation to produce meta-inferences about sustainability dynamics.

Results: Systemic constraints, particularly underfunding (<2% of health budget) and high staff turnover (30% annually), drove programme drift and community stigma, deterring 40% of patients and contributing to voltage drop. However, this drift often manifested through constructive adaptive mechanisms, including informal peer mentoring networks and role flexibility, which maintained service continuity. Multiple regression (R² = 0.45) identified leadership (β = 0.42), infrastructure (β = -0.35), and stigma (β = -0.30) as significant predictors of sustainability. Mixed-methods integration revealed these adaptations were the primary mechanism through which effective leadership operated - a dynamic invisible to quantitative measures alone.

Conclusions: Adaptive mechanisms represent legitimate and necessary sustainability strategies in resource-constrained settings, not implementation failures. We propose 'functional fidelity' (maintaining core outcomes through flexible processes) and 'adaptive capacity' as crucial theoretical extensions for implementation science in LMICs. Sustainable mental health integration requires frameworks that recognise and support frontline innovation while ensuring quality safeguards are maintained, offering a more realistic pathway to closing the global mental health treatment gap.