Validation of a one item screen for problem gambling

A valid 1 item screen (“In the past 12 months, have you ever had an issue with your gambling?”) is potentially important in both a primary care setting and in research applications where an extended screen is not practical or affordable. Prior research by Thomas, Jackson, Browning, and Piterman (2010) and Thomas, Piterman and Jackson (2008) found that the 1 item screen provided acceptable performance when compared to the 9-item Problem Gambling Severity Index (PGSI, Ferris & Wynne, 2001). The current study sought to replicate these findings in a different geographic area. Data were collected from a telephone survey of 1,292 respondents in Queensland. Results from the phone survey showed a very high proportion of false negatives (79%), indicating that approximately 4/5 of positively identified problem gamblers from the PGSI refused to admit to having any “issues with (their) gambling” in the last 12 months. Further investigation using Rasch analysis (Andrich, 2004) revealed that while the internal construct validity of the PGSI was good, the 1 item screen was ineffective as an indicator of those who are presently at risk for problem gambling. One possible explanation for the discrepancy with past results is that the present study employed the qualifier of “In the past 12 months” to match with the questions on the PGSI, whereas the prior research excluded this timeframe limit. Nevertheless, the current study casts doubt on the ability or willingness of people with severe gambling problems to self-diagnose their behavior as problematic.