Strengths (evidence-based)

• Wide geographical synthesis linking viral species to reservoir hosts and regional clinical syndromes (HPS vs HFRS) with an extensive reference list (n=127) Watson et al. 2013: Species–reservoir mapping
• Balanced discussion of documented human-to-human transmission limited to Andes virus with cluster epidemiology described (Argentina, Chile) Watson et al. 2013: Andes person-to-person clusters
• Integration of ecological drivers (mast cycles, ENSO, biodiversity/dilution effect) that provide mechanistic hypotheses linking environment → reservoir dynamics → human incidence Watson et al. 2013: Ecology drivers

Limitations and blindspots

• Heavy reliance on published case reports and national registries introduces surveillance and publication bias; the authors acknowledge underdiagnosis in Africa and parts of SE Asia where seroprevalence exists but clinical reports are sparse Watson et al. 2013: surveillance bias note
• Quantitative synthesis (meta-analysis) is absent — narrative review cannot provide pooled estimates or formally assess heterogeneity/confounding across studies; causal attributions (e.g., China's case decline) are plausible but not proven by controlled intervention studies in the review Watson et al. 2013: narrative limits
• Limited assessment of diagnostic test heterogeneity and cross-reactivity (serology vs molecular diagnostics) across settings — important because seroprevalence comparisons depend on assay specificity/sensitivity, which the review notes but does not systematically analyze.

Concrete, high-value improvements (for authors / future reviews)

1. Perform a systematic review + meta-analysis of seroprevalence and outbreak cluster data with stratification by diagnostic method (ELISA, IFA, RT-PCR) and quality scoring to estimate pooled prevalence and secondary-attack rates.
2. Standardize and report data collection templates and surveillance definitions (case definitions, laboratory confirmation algorithms) to reduce comparability bias across countries.
3. Incorporate reservoir monitoring data (rodent abundance/seroprevalence time series) alongside human surveillance into joint ecological–epidemiological models to test mechanistic hypotheses (mast/ENSO → reservoir → human cases) formally.
4. Assess potential role of domestic animals (pigs, pets) using molecular detection (RT-PCR) and viral isolation to move beyond serology and determine infectiousness/excretion potential.

Confidence-weighted takeaways

• High confidence: Rodent reservoirs drive most human hantavirus infections via inhalation of contaminated excreta; Andes virus is the notable exception with documented close-contact human-to-human clusters Watson et al. 2013: Andes transmission confidence
• Moderate confidence: Environmental drivers (mast, ENSO, land-use change, biodiversity loss) influence reservoir abundance and human risk but effects are regionally heterogeneous and require local modelling to predict outbreaks Watson et al. 2013: ecology evidence
• Lower confidence: Attribution of national incidence declines (e.g., China, 2000->2007) to specific interventions (rodent control, vaccination) is plausible but not proven causally within the review — controlled program-evaluation studies are lacking in the cited literature Watson et al. 2013: China decline caution

Immediate, actionable research/data steps

• Assemble a harmonized database of surveillance reports, serosurveys, rodent monitoring and molecular detections (format: geo-tagged time series) to enable quantitative eco-epidemiological analysis.
• Re-analyze Andes cluster datasets to estimate secondary attack rates by exposure type (sexual contact, sleeping in same room, caregiving) using standardized case definitions.
• Prioritize molecular testing (RT-PCR) of suspect human cases and matched local rodent trapping with sequencing to link human cases to reservoir populations and identify emergent strains.