Spatial dynamics of TB within a highly urbanised Asian metropolis using point patterns.

Singapore is a high-income country in a region with a high prevalence of tuberculosis. The Singapore Tuberculosis (TB) Elimination Program (STEP) was set up in 1997, and the better surveillance and clinical management practices initiated under STEP led to a decade-long decline in the incidence levels. However, incidence rates started to rise again since 2008. The reasons for this rise are unclear. This study involved a spatial analysis of the epidemiology of TB among Singapore residents. More than 30 000 cases reported during 1995-2011 and their residential addresses were analysed for spatial risk and spatial clustering, using spatial point pattern methodology. The principal factor responsible for the increasing resident TB incidence in Singapore is the changing age profile of the population. In particular the burgeoning population aged above 65 years accounts for the increase in reported cases. Singapore's population has one of the world's lowest fertility and mortality rates, and the elderly population is projected to grow substantially over the next few decades. Tuberculosis rates may therefore continue to rise even with static or improving case management and surveillance.

Time series analysis of demographic and temporal trends of tuberculosis in Singapore.

BACKGROUND: Singapore is an intermediate tuberculosis (TB) incidence country, with a recent rise in TB incidence from 2008, after a fall in incidence since 1998. This study identified population characteristics that were associated with the recent increase in TB cases, and built a predictive model of TB risk in Singapore. METHODS: Retrospective time series analysis was used to study TB notification data collected from 1995 to 2011 from the Singapore Tuberculosis Elimination Program (STEP) registry. A predictive model was developed based on the data collected from 1995 to 2010 and validated using the data collected in 2011. RESULTS: There was a significant difference in demographic characteristics between resident and non-resident TB cases. TB risk was higher in non-residents than in residents throughout the period. We found no significant association between demographic and macro-economic factors and annual incidence of TB with or without adjusting for the population-at-risk. Despite growing non-resident population, there was a significant decrease in the non-resident TB risk (p < 0.0001). However, there was no evidence of trend in the resident TB risk over this time period, though differences between different demographic groups were apparent with ethnic minorities experiencing higher incidence rates. CONCLUSION: The study found that despite an increasing size of non-resident population, TB risk among non-residents was decreasing at a rate of about 3% per year. There was an apparent seasonality in the TB reporting.