Sociodemographic Characteristics of Suicide Risk Question Response and Nonresponse, Monroe County Florida Behavior Risk Factor Surveillance Survey, 2016
Presentation
Overview
Overview
Description
Introduction. U.S. population-based health surveys are a primary method for obtaining critical behavioral health information to inform the prevention of disease and injury, including suicide risk. Unfortunately, selective nonresponse among completed population-based surveys adds to the difficulty of obtaining representative data. Moreover, suicide risk questions are even more subject to selective nonresponse due to the sensitive nature of the question items, potentially biasing results and leading to incomplete risk profiles. Those not responding to population-based survey questions in general have been found to represent a more vulnerable population of lower socio-economic status and worse health conditions. To date, no studies have explored associations between sociodemographic characteristics and population-based suicide risk question nonresponse. Therefore, this study explored associations between suicide risk question nonresponse and sociodemographic characteristics. Study findings identify the sociodemographic characteristics associated with suicide risk question nonresponse thereby providing guidance for suicide research and prevention efforts.
Methods. A secondary analysis of the Monroe County Florida 2016 BRFSS data was conducted to assess nonresponse to suicide risk questions among adult residents (n = 528). Univariate analyses using weights and strata were conducted to generate prevalence estimates, standard errors, and 95% confidence intervals of suicide risk question response and non-response. Next, three types of comparisons were done using Chi-square tests of significance. First, demographic characteristic differences were assessed between suicide risk question response and nonresponse. Second, demographic characteristic differences were assessed between suicide risk question nonresponse (n = 109) and no suicide risk (n = 370). Third, demographic characteristic differences were assessed between suicide risk question nonresponse (n = 109) and those at risk for suicide (n = 49). Finally, a multivariate logistic regression was conducted to assess associations among sociodemographic characteristics significantly related to suicide risk question nonresponse.
Results. 32% (n = 109, CI = 26.5-38.4) of the population did not respond to suicide risk questions. Persons not responding to suicide risk questions were significantly younger (χ2 [1, n = 528] = 20.6, p = .002), less likely to be non-Hispanic White(χ2 [2, n = 517] = 33.1, p = .001), more likely to be Hispanic (χ2 [2, n = 517] = 33.1, p = .001), less likely to be retired (χ2 [3, n = 524] = 17.5, p = .024), more likely to be renters (χ2 [1, n = 523] = 16.3, p = .008), and made less than $35,000 per year (χ2 [1, n = 419] = 17.1, p = .018) as compared to those responding to these questions. Hispanic ethnicity was significantly related to suicide risk question nonresponse (B = 0.794, OR = 2.2, p < .019) after controlling for all sociodemographic characteristics.
Conclusion. Sociodemographic characteristics of persons not responding to suicide risk questions were comparable to characteristics of individuals identified as at risk for suicide. These findings from a population based study in southeastern Florida provide evidence to support prior findings among online surveys of college students that suicide risk question nonresponse appears to represent a group at risk for suicide. Notably, Hispanics were more likely than non-Hispanics to avoid answering suicide risk questions, even when controlling for all other socioeconomic differences.
What the work adds to knowledge on the topic: Individuals not responding to suicide risk questions in population-based surveys may represent an at risk group. The findings of ethnic disparities in suicide risk question nonresponse suggests a need for more targeted or alternative research methods for assessing population-based suicide risk.
Learning Objective: Identify issues with suicide risk question non-response in population-based studies.
