Since 1998 the Synthetic Personality Inventory (SPI), a well-established aptitude test, has been used to select residents at St. Luke's International Hospital. The aims of our study were to analyze the association of the SPI with several characteristics of residents, e. g., intellectual capability, results of examinations in medicine, and evaluation during 2 years of residency, and to clarify the significance of the SPI for selecting residents. General mental ability measured with the SPI was strongly correlated with the intellectual capability of residents but was not correlated with results of examinations in medicine. High scores for general mental ability, for activity, and for tendency of extroversion were correlated with high evaluations of residents for 2 years, whereas a cooperative personality, reasoning ability, and practicality were related to an improvement in grades from the first to second years. These results demonstrate that the results of the SPI are correlated with several characteristics of residents.

BACKGROUND : Previous studies have suggested positive association between residents' workload and medical errors. However, few studies have investigated the possible associations between procedural errors, workload, and the individual characteristics of residents, including personality, mental state and job satisfaction.
OBJECTIVE : To explore possible associations of workload and individual characteristics of residents with their procedural error rates.
DESIGN : Prospective observational study based on a daily questionnaire.
PARTICIPANTS : Residents of postgraduate year 1 and 2.
MEASUREMENTS : Residents' workload (on-calls, work hours, sleep and napping hours), residents' physical and mental health state, personality inventory, and procedural error rate (defined as procedural error counts divided by overall procedural attempts).
RESULTS : On average, the residents (N=49) were responsible for 9.8 inpatients per day (range, 1.9-23.1), worked for 16.0 hours per day (range, 12.6-19.8), slept for 4.4 hours per day (range, 2.8-5.7), napped for 0.2 hours per day (range, 0-0.7), and experienced 1 overnight work shift every 7.2 days. The procedural error rate was 2.2 per 10 procedures (range, 0.4-5.0). Using a multivariable adjusted regression model, significant factors associated with lower error rates included : longer napping ; reflective personality ; better mental state ; higher job satisfaction ; and, less on-call frequency.
CONCLUSIONS : Procedural error of residents is positively associated with higher on-call frequency and inversely associated with napping, reflective personality, better mental state, and higher job satisfaction. For reducing procedural error among residents, improvement of modifiable factors, such as workload and mental health, is needed.