METHODS: Secondary data for 9,850 participants were obtained from the Social Security Organisation Return To Work database. The dependent variable was the Return To Work programme outcome, successful return to employment (same employer or different employer) or unsuccessful return. Logistic regression analysis with weighted sum contrasts was performed to assess the odds ratios with 95% confidence interval (95% CI) for successful return to employment across the various subgroups of participants.

RESULTS: Overall, 65.5% of participants successfully returned to employment, either with their former employers or with new employers. Successful return to employment was found to be significantly higher than the overall proportion among those participants who had had commuting accidents, followed by those who had had workplace accidents. Successful return to employment was also associated with injuries of the upper and lower limbs, employers who were interested in hiring disabled workers, motivation to participate in the programme, an intervention period of 3 months or less, age 29 years or younger, and male participants.

METHOD: Seven volunteer post-call doctors were recruited to go through an EEG recording before and after their on-call rotation while at rest and subsequently while carrying out Stroop Test, putting their cognitive function at work.

RESULTS: The doctors have worked up to 33 hours in a row and have had sleep of an average of 1.5 hours. It is found that during task there is a statistically significant increase in theta (frontal and occipital regions) and beta (occipital region) band power while at task post-call. Alpha band power is increased in the frontal and reduced in other regions. Correlation with Stroop Test results indicated that those who have higher alpha, beta, and lower relative theta powers at the frontal region at post-call rest have higher percentage of correct congruent trials.

SUBJECTS: Patients who were admitted to the University of Malaya Medical Centre due to cardiac events.

METHODS: Eight different machine learning models were evaluated. The models included 3 different sets of features: full features; significant features from multiple logistic regression; and features selected from recursive feature extraction technique. The performance of the prediction models with each set of features was compared.

RESULTS: The AdaBoost model with the top 20 features obtained the highest performance score of 92.4% (area under the curve; AUC) compared with other prediction models.