Variation in Psychometric Testing in General Practice - A Nationwide Cohort Study.

Objective: Most mental disorders are diagnosed and treated in general practice. Psychometric tests may help the general practitioner diagnose and treat mental disorders like dementia, anxiety, and depression. However, little is known about the use of psychometric tests in general practice and their impact on further treatment. We aimed to assess the use of psychometric tests in Danish general practice and to estimate whether variation in use is associated with the provided treatment and death by suicide in patients. Methods: This nationwide cohort study included registry data on all psychometric tests performed in Danish general practice in 2007-2018. We used Poisson regression models adjusted for sex, age, and calendar time to assess predictors of use. We used fully adjusted models to estimate the standardized utilization rates for all general practices. Results: A total of 2,768,893 psychometric tests were used in the study period. Considerable variations were observed among general practices. A positive association was seen between a general practitioner's propensity to use psychometric testing and talk therapy. Patients listed with a general practitioner with low use had an increased rate of redeemed prescriptions for anxiolytics [incidence rate ratio (95% confidence interval):1.39 (1.23;1.57)]. General practitioners with high use had an increased rate of prescriptions for antidementia drugs [1.25 (1.05;1.49)] and first-time antidepressants [1.09 (1.01;1.19)]. High test use was seen for females [1.58 (1.55; 1.62)] and patients with comorbid diseases. Low use was seen for populations with high income [0.49 (0.47; 0.51)] and high educational level [0.78 (0.75; 0.81)]. Conclusion: Psychometric tests were used mostly for women, individuals with a low socioeconomic status, and individuals with comorbid conditions. The use of psychometric tests depends on general practice and is associated with talk therapy, redemptions for anxiolytics, antidementia drugs, and antidepressants. No association was found between general practice rates and other treatment outcomes.

Concurrent validity of the anxiety symptom-scale compared to two well-being scales: results from the Lolland-Falster health study.

OBJECTIVE: To examine the concurrent validity of the Anxiety Symptom-scale against two well-being scales, the Cantril Ladder (CL) and World Health Organization Well-Being Index (WHO-5), to test the algorithm defining anxiety against these scales, and identify cut-off points for the Anxiety Symptom-scale sum score. SUBJECTS: 14,405 adult respondents completing all psychometric questions in the Lolland Falster Health Study. METHOD: Receiver operating characteristic analyses comparing Anxiety Symptom-scale WHO-5 and CL. RESULTS: 2.5% of respondents had an anxiety disorder (3% female and 2% male) according to the Anxiety Symptom-scale algorithm. The area under the curve (AUC) was 0.87 for CL and 0.90 for WHO-5 (using inverse scores), indicating high concordance with anxiety disorder as identified by the scale. A score solely ≥2 on item 10 is a relevant cut off to low wellbeing. Anxiety disorder covers a broad range on the scale's sum score, with 3 to 4 indicating low well-being in this population sample and a sensitivity of 0.85 - 0.99 against CL and WHO-5. CONCLUSION: The Anxiety Symptom-scale is a sensitive and valid instrument for the identification of patients in low well-being with symptoms of anxiety. A score ≥2 on the functional impact (Item 10) of all symptoms is a relevant indicator of anxiety associated with low well-being in this sample. A higher Anxiety Symptom-scale sum score is coherent with lower well-being, though without specific cut-off points. Further validation of the Anxiety Symptom-scale in a clinical setting is recommended.

Antidepressant treatment initiation and public sick-leave compensation in the following year: a register-based prospective cohort study in Denmark.

OBJECTIVES: Mental disorders have caused increasing sickness absence and related benefit claims in the OECD countries. This study investigates the association between antidepressant treatment initiation and public sick leave compensation (PLSC) in the following year in Denmark. METHODS: The study was designed as a register-based prospective cohort study. We included 39,401 adults (aged 18-65 years) with at least 12 consecutive months of full-time labour market attachment who had initiated first-time antidepressant monotherapy for depression or anxiety between 1 January 2015 and 31 December 2018. PLSC was estimated for the year following the incident prescription of various antidepressants for depression or anxiety disorders. RESULTS: The most frequently prescribed antidepressant medication was SSRIs (66.8%), with sertraline being the leading choice. Compared with sertraline, mirtazapine and mianserin were associated with the highest risks of PSLC in the year following initiation, with IRRs of 2.74 (95% CI: 2.63 to 2.86) and 5.79 (95% CI: 5.18 to 6.47), respectively. Compared with sertraline, citalopram (IRR 1.22, 95%CI 1.17-1.28), venlafaxine (IRR 1.34, 95%CI 1.23-1.45) and duloxetine (IRR 1.48, 95%CI 1.35-1.62) were all associated with increased PSLC. In contrast, paroxetine (IRR 0.85, 95% CI 0.74-0.98), fluoxetine (IRR 0.51, 95%CI 0.42-0.62) and vortioxetine (IRR 0.78, 95% CI 0.63-0.97) were all associated with a significantly lower risk of PSLC compared with sertraline. CONCLUSIONS: Antidepressant treatment initiation was associated with PLSC. The highest risk of PLSC was seen for antidepressants with sedative side effects. Some types of antidepressants were associated with a lower risk of PLSC in the year following treatment initiation.

Prevalence of anxiety disorders and association to socioeconomic position. Results from the Lolland Falster Health Study.

PURPOSE: The purpose of this study is to describe the prevalence of anxiety disorders in a general population and the association to socioeconomic position (SEP), which has not been described in a Danish context before. MATERIAL AND METHODS: We present data on anxiety symptoms from respondents in the rural-provincial Lolland-Falster population Health Study (LOFUS). Analyses of the questionnaire responses to the Anxiety Symptom Scale were done by descriptive statistics and logistic regression analyses adjusted for sex and age. RESULTS: 14,834 LOFUS respondents who completed the Anxiety Symptom Scale were included; According to the original algorithm 371 (2.5%) had an anxiety disorder. The adjusted odds ratios (aOR) for anxiety disorder were strongly associated to SEP. We found aOR for anxiety to be: 3.8 (confidence interval (CI 95%) 2.54 - 5.92) for respondents with no postsecondary education compared to those with 3+ years of postsecondary education; 11.9 (CI 8.89 - 16.01) for respondents temporarily out of a job compared to those working; 9.4 (CI 6.06 - 14.51) for those experiencing constant financial strain compared to those not experiencing financial strain. Relaxing the criteria for anxiety to item 10 > 1, the prevalence raised to 3.9%. The association was unchanged related to education; however, the aOR dropped to 9 and 8 respectively, for being temporally out of job, or in financial strain every month- when doing same comparisons. CONCLUSIONS: The 14-day prevalence of anxiety disorder seems low but strongly associated to SEP especially for individuals temporarily out of a job or experiencing financial strain.

Clinimetric properties of the Smoking Abstinence Expectancies Questionnaire.

Smoking abstinence expectancies are beliefs about negative and positive short-term psychological and physiological consequences of not smoking. The Smoking Abstinence Expectancies Questionnaire (SAEQ) is a widely used Patient-Reported Outcome Measure (PROM) to assess smoking abstinence expectancies. It has four subscales: negative mood, somatic symptoms, harmful consequences, positive consequences. Although studied from a psychometric perspective, the SAEQ needs further evaluation. Clinimetrics, and its Clinimetric Criteria for Patient-Reported Outcome Measures (CLIPROM), offers a robust method to evaluate the SAEQ. We verified construct validity and sensitivity of the Italian version of the SAEQ applying CLIPROM criteria. A total of 293 adult Italian smokers were consecutively enrolled at two smoking cessation clinics and assessed via the SAEQ. Item Response Theory models (i.e., combining Rasch and Mokken analyses) were used to test construct validity and sensitivity. The total score of the SAEQ was not found to be unidimensional but each SAEQ subscale score was. PSI (0.90) indicated that the total score of the SAEQ could reliably discriminate between respondents with different levels of the trait under assessment, whereas SAEQ subscales on negative mood and harmful consequences could reliably distinguish between different groups but not between different subjects (PSI ranging from 0.77 and 0.78). Overall, the total score of the SAEQ is a sensitive screening PROM and can be used at smoking cessation clinics to discriminate between subjects with different levels of smoking abstinence expectancies. SAEQ subscales should be used to detect severity and subjective burden of a wide range of expected effects of nicotine abstinence.

Construct validity of the Smoker Complaint Scale: A clinimetric analysis using Item Response Theory (IRT) models.

A number of rating scales evaluating symptoms of nicotine withdrawal have been developed over the years but insufficient attention has been devoted to the assessment of their clinimetric properties. Clinimetrics, the science of clinical measurements, is an innovative approach, particularly useful for assessing the validity of rating scales. This is the first study using clinimetric principles to test the construct validity of the Smoker Complaint Scale (SCS), a self-rating scale specifically developed to assess acute symptoms of nicotine withdrawal. Construct validity was evaluated via Item Response Theory (IRT) models (i.e., combining Rasch and Mokken analyses). A total of 366 subjects (mean age = 34.0, SD = 11.3 years) participated in the study. IRT analyses showed that SCS was a multidimensional measure of symptoms of nicotine withdrawal, including unidimensional subscales, particularly a four-item subscale (the SCS4), which was found to entail the clinimetric property of construct validity. IRT analyses also revealed that affective symptoms of nicotine withdrawal preceded cognitive ones. The SCS should be considered as an item bank, including a particularly valid subscale, the SCS4 that can be used as a screening or outcome measure to evaluate the severity of cognitive and affective symptoms of nicotine withdrawal. SCS4 is a sensitive clinimetric index which differentiates "ceiling symptoms" of nicotine withdrawal (e.g., "feeling slowed down") from "floor symptoms" of nicotine withdrawal (e.g., "feeling lightheaded") that emerge in the severe form of nicotine withdrawal.