How unusual are health survey participants? Differences in alcohol consequences
posted on: Sep 23, 2014
Exploration of the representativeness of health surveys has been carried out by researchers from the MRC/CSO Social and Public Health Sciences Unit at the University of Glasgow along with the MRC Biostatistics Unit in Cambridge, NHS Health Scotland, Information Services Division, NHS National Services Scotland and ScotCen Social Research in Edinburgh. National health surveys are vital for monitoring population health but participation by individuals who are atypical overall of the population can bias findings. In this study the researchers aimed to assess the extent of such bias in the Scottish Health Surveys conducted in 1995 to 2010, with a focus on alcohol-related outcomes. Follow-up health information on survey respondents was available via linkage to mortality and hospitalization records. When levels of alcohol-related harm (hospitalization or mortality) among respondents were examined, they were found to be around 30% lower than in the general population; deaths from any cause were 10% lower and the differences were greatest among people living in deprived areas. These insights are now being used to make thorough correct for bias in survey data.
The full article can be read here.