Dataset Information

Machine learning algorithm-based estimation model for the severity of depression assessed using Montgomery-Asberg depression rating scale.

ABSTRACT:

Aim

Depressive disorder is often evaluated using established rating scales. However, consistent data collection with these scales requires trained professionals. In the present study, the "rater & estimation-system" reliability was assessed between consensus evaluation by trained psychiatrists and the estimation by 2 models of the AI-MADRS (Montgomery-Asberg Depression Rating Scale) estimation system, a machine learning algorithm-based model developed to assess the severity of depression.

Methods

During interviews with trained psychiatrists and the AI-MADRS estimation system, patients responded orally to machine-generated voice prompts from the AI-MADRS structured interview questions. The severity scores estimated from two models of the AI-MADRS estimation system, the max estimation model and the average estimation model, were compared with those by trained psychiatrists.

Results

A total of 51 evaluation interviews conducted on 30 patients were analyzed. Pearson's correlation coefficient with the scores evaluated by trained psychiatrists was 0.76 (95% confidence interval 0.62-0.86) for the max estimation model, and 0.86 (0.76-0.92) for the average estimation model. The ANOVA ICC rater & estimation-system reliability with the evaluation scores by trained psychiatrists was 0.51 (-0.09 to 0.79) for the max estimation model, and 0.75 (0.55-0.86) for the average estimation model.

Conclusion

The average estimation model of AI-MADRS demonstrated substantially acceptable rater & estimation-system reliability with trained psychiatrists. Accumulating a broader training dataset and the refinement of AI-MADRS interviews are expected to improve the performance of AI-MADRS. Our findings suggest that AI technologies can significantly modernize and potentially revolutionize the realm of depression assessments.

SUBMITTER: Shimamoto M

PROVIDER: S-EPMC10932776 | biostudies-literature | 2024 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning algorithm-based estimation model for the severity of depression assessed using Montgomery-Asberg depression rating scale.

Shimamoto Masanori M Ishizuka Kanako K Ohtani Kento K Inada Toshiya T Yamamoto Maeri M Tachibana Masako M Kimura Hiroki H Sakai Yusuke Y Kobayashi Kazuhiro K Ozaki Norio N Ikeda Masashi M

Neuropsychopharmacology reports 20231220 1

<h4>Aim</h4>Depressive disorder is often evaluated using established rating scales. However, consistent data collection with these scales requires trained professionals. In the present study, the "rater & estimation-system" reliability was assessed between consensus evaluation by trained psychiatrists and the estimation by 2 models of the AI-MADRS (Montgomery-Asberg Depression Rating Scale) estimation system, a machine learning algorithm-based model developed to assess the severity of depression ...[more]

PMID: 38115795

Similar Datasets

Project description:BACKGROUND:It has been claimed that efficacy estimates based on the Hamilton Depression Rating-Scale (HDRS) underestimate antidepressants true treatment effects due to the instrument's poor psychometric properties. The aim of this study is to compare efficacy estimates based on the HDRS with the gold standard procedure, the Montgomery-Asberg Depression Rating-Scale (MADRS). METHODS AND FINDINGS:We conducted a meta-analysis based on the comprehensive dataset of acute antidepressant trials provided by Cipriani et al. We included all placebo-controlled trials that reported continuous outcomes based on either the HDRS 17-item version or the MADRS. We computed standardised mean difference effect size estimates and raw score drug-placebo differences to evaluate thresholds for clinician-rated minimal improvements (clinical significance). We selected 109 trials (n = 32,399) that assessed the HDRS-17 and 28 trials (n = 11,705) that assessed the MADRS. The summary estimate (effect size) for the HDRS-17 was 0.27 (0.23 to 0.30) compared to 0.30 (0.22 to 0.38) for the MADRS. The effect size difference between HDRS-17 and MADRS was thus only 0.03 and not statistically significant according to both subgroup analysis (p = 0.47) and meta-regression (p = 0.44). Drug-placebo raw score difference was 2.07 (1.76 to 2.37) points on the HDRS-17 (threshold for minimal improvement: 7 points according to clinician-rating and 4 points according to patient-rating) and 2.99 (2.24 to 3.74) points on the MADRS (threshold for minimal improvement: 8 points according to clinician-rating and 5 points according to patient-rating). CONCLUSIONS:Overall there was no meaningful difference between the HDRS-17 and the MADRS. These findings suggest that previous meta-analyses that were mostly based on the HDRS did not underestimate the drugs' true treatment effect as assessed with MADRS, the preferred outcome rating scale. Moreover, the drug-placebo differences in raw scores suggest that treatment effects are indeed marginally small and with questionable importance for the average patient.

Project description:BackgroundThe Montgomery-Asberg Depression Rating Scale (MADRS) is commonly used to examine depressive symptoms in clinical settings, including facilities treating patients for alcohol addiction. No studies have examined the validity of the MADRS compared to an established clinical diagnostic tool of depression in this population. This study aimed to examine the following: (i) the validity of the MADRS compared to a clinical diagnosis of a depressive disorder (using the Structured Clinical Interview for DSM-IV-TR [SCID-IV-TR]) in patients seeking treatment for alcohol dependence (AD); (ii) whether the validity of the MADRS differs by type of SCID-IV-TR-based diagnosis of depression; and (iii) which items contribute to the optimal predictive model of the MADRS compared to a SCID-IV-TR diagnosis of a depressive disorder.MethodsIndividuals seeking treatment for AD and admitted to an inpatient unit were administered the MADRS at day 2 of their detoxification program. Clinical diagnoses of AD and depression were made via the SCID-IV-TR at the beginning of treatment.ResultsIn total, 803 participants were included in the study. The MADRS demonstrated low overall accuracy relative to the clinical diagnosis of depression with an area under the receiver operating characteristic curve of 0.68. The optimal threshold for balancing sensitivity and specificity identified by the Euclidean distance was >14. This cut-point demonstrated a sensitivity of 66%, a specificity of 60%, a positive predictive value of 50%, and a negative predictive value of 75%. The MADRS performed slightly better for major depressive disorders compared to alcohol-induced depression. Items related to lassitude, concentration, and appetite slightly decreased the accuracy of the MADRS.ConclusionsThe MADRS does not appear to be an appropriate substitute for a diagnostic tool among alcohol-dependent patients. The MADRS may, however, still be a useful screening tool assuming careful consideration of cut-points.

Project description:BackgroundDepression and depressive symptoms are highly prevalent in old persons but are potentially reversible. Full recovery is the main goal in the treatment of depressive episodes. Compared to clinical trials, observational studies of patients with depression in late life (DLL) show poorer prognoses in terms of response and remission. However, observational studies on the course of DLL are scarce. The aims of this study were to examine the course of DLL in terms of response, remission and symptom-specific changes as measured by the Montgomery and Asberg Depression Rating Scale (MADRS), and to explore which clinical variables were associated with the response and remission.MethodsThis is an observational, multicenter and prospective study of patients aged 60 years and older who were referred to treatment of depression in the department of old-age psychiatry at specialist health care services in Norway. The patients were evaluated with the MADRS at admission to and discharge from hospital. The mean, median, minimum and maximum values for days stayed in hospital were 68, 53, 16 and 301, respectively. Effect size (ES) was calculated to determine which MADRS symptoms changed most during the treatment. To assess the predictors for change in the MADRS score (continuous variable) and for remission and response (both dichotomous variables), regression models adjusting for cluster effects within center were estimated.ResultsOf 145 inpatients, 99 (68.3 %) had a response to treatment (50 % or more improvement of the MADRS score). Remission (MADRS score ≤9 at discharge) was experienced in 74 (51.0 %) of the patients. Of the individual MADRS items, "reported sadness" (ES =0.88) and "lassitude" (ES = 0.80) showed the greatest amount of improvement, and "concentration difficulties" (ES = 0.50) showed the least amount of improvement during treatment. Having a diagnosis of dementia was associated with a lower remission rate and less improvement in the MADRS score during the treatment. Poorer physical health was associated with a lower response rate. Having experienced previous episode(s) of depression was associated with a lower remission rate.ConclusionsRecurrent episodes of depression, poor somatic health and a diagnosis of dementia were found to be negative prognostic factors for the course of DLL. Clinicians should therefore pay close attention to these factors when evaluating treatment.Trial registrationClinicalTrials.gov NCT01952366.

Project description:ObjectiveThe objective of this study was to determine which symptoms measured by the Patient Health Questionnaire (PHQ-9) and Montgomery-Asberg Depression Rating Scale (MADRS) improve in those treated with esketamine nasal spray in combination with oral antidepressant (AD) compared with those treated with placebo plus AD for adult patients with treatment-resistant depression (TRD). These results complement the interpretation of PHQ-9 and MADRS total scores.MethodsThe TRANSFORM 2 study evaluated the efficacy and safety of esketamine nasal spray in combination with AD. This post-hoc analysis used PHQ-9 and MADRS data to evaluate symptom changes. The total scores change and proportions of individual item change scores on the PHQ-9 and MADRS were evaluated at days 15 and 28; analysis of variance was used to test differences on total scores. Generalized estimation equations of logistic regression models were used to estimate the likelihood of improvement on instrument items.ResultsThe mean total score reduction of the PHQ-9, indicating improvement, was greater in the esketamine plus AD arm compared with placebo plus AD at day 15 (- 1.8; p = 0.045) and day 28 (- 2.8; p = 0.006). Proportions of those who improved (≥ 1 point on a 4-point scale and ≥ 2 points on a 7-point scale for the PHQ-9 and MADRS, respectively) was greater in the esketamine plus AD group compared with the placebo plus AD group across all items. The odds of improving for those in the esketamine plus AD group compared with the placebo plus AD group were over two times greater on the PHQ-9 items: "Little interest/pleasure in things" (OR 2.252, 95% CI 1.165-4.355); "Feeling down, depressed, or hopeless" (OR 2.767, 95% CI 1.400-5.470); and "Feeling tired or having little energy" (OR 2.171, 95% CI 1.153-4.087). The mean reduction in total scores on the MADRS, indicating improvement, was numerically greater at day 15 (- 2.0; p = 0.189) and statistically significantly greater at day 28 (- 4.4; p = 0.017) in the esketamine plus AD arm compared with placebo plus AD. The odds of improving for those in the esketamine plus AD group compared with the placebo plus AD group were over two times greater on the MADRS items measuring "Apparent sadness" (OR 2.007, 95% CI 1.096-3.674); and "Inability to feel" (OR 2.099, 95% CI 1.180-3.735).ConclusionImprovement in mean total scores in those treated with esketamine plus AD compared with placebo plus AD are important results to confirm efficacy. The odds of improving in those treated with esketamine plus AD was at least two times greater than with placebo plus AD on three patient- and two clinician-reported individual symptoms of TRD. These findings provide patient-relevant quantification of the esketamine plus AD treatment benefit, adding understanding as to which symptoms are most improved with treatment.Trial registrationClinicalTrials.gov identifier: NCT02418585; first posted 16 April 2015.

Dataset Information

Machine learning algorithm-based estimation model for the severity of depression assessed using Montgomery-Asberg depression rating scale.

Aim

Methods

Results

Conclusion

Publications

Machine learning algorithm-based estimation model for the severity of depression assessed using Montgomery-Asberg depression rating scale.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets