Consistency of self-reported drug use events in a mixed methods study of injection drug users

Objectives: We illustrate the use of intraclass correlation coefficient (ICC) to assess the consistency of information provided during a mixed methods study of injection drug users living in Los Angeles and San Francisco. Background: Mixed methods research is gaining popularity in public health research. Little is known about the consistency of information provided during quantitative and qualitative portions of studies. Methods: Age of first use and first injection of heroin, methamphetamine, marijuana (use only), and cocaine/crack cocaine was provided during an in-person computer-assisted quantitative interview followed by an in-depth qualitative interview (n=62). Qualitative and quantitative data were used to create timelines detailing events in participants' lives, separate per participant (Timeline Maker Professional). Timelines were used to create a dataset containing nine items per participant containing age in years at which drug use events occurred from both quantitative and qualitative data sources. Items that the participant had not experienced or did not provide data for in both the quantitative and the qualitative interviews were excluded. Ages that were within 1 year of each other in the qualitative and quantitative interviews were considered to be in agreement and recoded to have the same age. Stata 11.2 was used for analyses. Results: ICCs ranged from .42-.95 (n=9-53). Data indicate age of first drug injection is reported more consistently than age of first drug use; age of first marijuana use was the least consistently reported item. Discussion: Consistency of self-reported data was adequate for most drug use events in this mixed method study.