Variables Associated with Emotional Symptom Severity in Primary Care Patients: The Usefulness of a Logistic Regression Equation to Help Clinical Assessment and Treatment Decisions

The aim of this study is to contribute to the evidence regarding variables related to emotional symptom severity and to use them to exemplify the potential usefulness of logistic regression for clinical assessment at primary care, where most of these disorders are treated. Cross-sectional data related to depression and anxiety symptoms, sociodemographic characteristics, quality of life (QoL), and emotion-regulation processes were collected from 1,704 primary care patients. Correlation and analysis of variance (ANOVA) tests were conducted to identify those variables associated with both depression and anxiety. Participants were then divided into severe and nonsevere emotional symptoms, and binomial logistic regression was used to identify the variables that contributed the most to classify the severity. The final adjusted model included psychological QoL (p < .001, odds ratio [OR] = .426, 95% CI [.318, .569]), negative metacognitions (p < .001, OR = 1.083, 95% CI [1.045, 1.122]), physical QoL (p < .001, OR = .870, 95% CI [.841, .900]), brooding rumination (p < .001, OR = 1.087, 95% CI [1.042, 1.133]), worry (p < .001, OR = 1.047, 95% CI [1.025, 1.070]), and employment status (p = .022, OR [.397, 2.039]) as independent variables, ρ2 = .326, area under the curve (AUC) = .857. Moreover, rumination and psychological QoL emerged as the best predictors to form a simplified equation to determine the emotional symptom severity (ρ2 = .259, AUC = .822). The use of statistical models like this could accelerate the assessment and treatment-decision process, depending less on the subjective point of view of clinicians and optimizing health care resources. (AU)

Variables Associated with Emotional Symptom Severity in Primary Care Patients: The Usefulness of a Logistic Regression Equation to Help Clinical Assessment and Treatment Decisions.

The aim of this study is to contribute to the evidence regarding variables related to emotional symptom severity and to use them to exemplify the potential usefulness of logistic regression for clinical assessment at primary care, where most of these disorders are treated. Cross-sectional data related to depression and anxiety symptoms, sociodemographic characteristics, quality of life (QoL), and emotion-regulation processes were collected from 1,704 primary care patients. Correlation and analysis of variance (ANOVA) tests were conducted to identify those variables associated with both depression and anxiety. Participants were then divided into severe and nonsevere emotional symptoms, and binomial logistic regression was used to identify the variables that contributed the most to classify the severity. The final adjusted model included psychological QoL (p < .001, odds ratio [OR] = .426, 95% CI [.318, .569]), negative metacognitions (p < .001, OR = 1.083, 95% CI [1.045, 1.122]), physical QoL (p < .001, OR = .870, 95% CI [.841, .900]), brooding rumination (p < .001, OR = 1.087, 95% CI [1.042, 1.133]), worry (p < .001, OR = 1.047, 95% CI [1.025, 1.070]), and employment status (p = .022, OR [.397, 2.039]) as independent variables, ρ2 = .326, area under the curve (AUC) = .857. Moreover, rumination and psychological QoL emerged as the best predictors to form a simplified equation to determine the emotional symptom severity (ρ2 = .259, AUC = .822). The use of statistical models like this could accelerate the assessment and treatment-decision process, depending less on the subjective point of view of clinicians and optimizing health care resources.