Spatial Panel Data Analysis on the Relationship between Provincial Economic Status and Enrolment in the Social Security Scheme amongst Migrant Workers in Thailand, 2015-2018

BACKGROUND

Thailand has a large flow of migrants from neighbouring countries; however, the relationship between economic status at the provincial level and the insured status of migrants is still vague. This study aimed to examine the association between provincial economy and the coverage of the Social Security Scheme (SSS) for migrants.

METHODS

Time-series data were analysed. The units of analysis were 77 provinces during 2015-2018. Data were obtained from the Social Security Office (SSO). Spatiotemporal regression (Spatial Durbin model (SDM)) was applied.

RESULTS

Migrant workers were mostly concentrated in Greater Bangkok, the capital city and areas surrounding it, but SSS coverage was less than 50%. However, the ratio of insured migrants to all migrants seemed to have positive relationship with the provincial economy in SDM. The ratio of insured migrants to all migrants was enlarged in all regions outside Greater Bangkok with statistical significance.

CONCLUSIONS

Low enforcement on employment law in some areas, particularly Greater Bangkok, can result in lesser SSS coverage. The provincial economic prosperity did not guarantee large SSS coverage. Interventions to ensure strict insurance enrolment are required.

Faster Detection of Poliomyelitis Outbreaks to Support Polio Eradication

As the global eradication of poliomyelitis approaches the final stages, prompt detection of new outbreaks is critical to enable a fast and effective outbreak response. Surveillance relies on reporting of acute flaccid paralysis (AFP) cases and laboratory confirmation through isolation of poliovirus from stool. However, delayed sample collection and testing can delay outbreak detection. We investigated whether weekly testing for clusters of AFP by location and time, using the Kulldorff scan statistic, could provide an early warning for outbreaks in 20 countries. A mixed-effects regression model was used to predict background rates of nonpolio AFP at the district level. In Tajikistan and Congo, testing for AFP clusters would have resulted in an outbreak warning 39 and 11 days, respectively, before official confirmation of large outbreaks. This method has relatively high specificity and could be integrated into the current polio information system to support rapid outbreak response activities.