Title : Estimating the role of heterogeneity and lockdown effectiveness in the first-wave dynamics of COVID-19 in Nigeria
Abstract:
Heterogeneity in susceptibility or connectivity can influence epidemic dynamics by altering transmission, reducing the effective reproduction number, and lowering the herd immunity threshold. During the first COVID-19 wave, non-pharmaceutical interventions, including lockdown measures, were widely implemented to reduce transmission before vaccines became available. Understanding their combined effects is important for evaluating epidemic spread and intervention effectiveness. We developed a modified Susceptible-Exposed-Infected-Recovered model with Gamma-distributed heterogeneity in susceptibility to estimate COVID-19 transmission dynamics and the impact of lockdowns using under-reported incidence data. The model accounted for individual variation in susceptibility, imperfect case reporting, and lockdown-related reductions in transmission. Parameters were estimated using a Bayesian Markov Chain Monte Carlo framework. The approach was validated with simulated data and subsequently applied to incompletely reported COVID-19 incidence data from twelve Nigerian states, including two with lockdown measures, obtained from the Nigerian Centre for Disease Control. Model performance was compared with a homogeneous model using information criteria. We estimated an average basic reproduction number of 1.833 (95% credible interval 1.817-1.848), a coefficient of variation in susceptibility of 3.662 (95% CrI: 3.300-4.055), and a reporting probability of 0.006 (95% CrI: 0.005-0.007). Information criteria indicated that the heterogeneous SEIR model was better supported than the homogeneous model and yielded substantially lower estimates of herd immunity, from 43.2% under homogeneity to 4.7%, and first-wave infection proportion, from over 70% to about 9%. The model also estimated that lockdown reduced transmission by only 0.3% on average. These findings demonstrate that heterogeneity plays a critical role in shaping COVID-19 transmission dynamics and should be incorporated into epidemic models. Ignoring heterogeneity may substantially overestimate infection burden, herd immunity thresholds, and intervention requirements. The framework further shows that lockdown effectiveness can be quantified despite heterogeneity and severe under-reporting, providing a practical approach for analysing epidemic interventions in data-limited settings.
Keywords: COVID-19; Heterogeneity; Lockdown; Mathematical Modelling; Non-Pharmaceutical Interventions; SEIR Model; Susceptibility

