Predicting onset of major depression in general practice attendees in Europe: extending the application of the predictD risk algorithm from 12 to 24 months

Development and validation of a prediction score to assess the risk of depression in primary care

BACKGROUND

Major depression is the most frequent psychiatric disorder and primary care is a crucial setting for its early recognition. This study aimed to develop and validate the DEP-HScore as a tool to predict depression risk in primary care and increase awareness and investigation of this condition among General Practitioners (GPs).

METHODS

The DEP-HScore was developed using data from the Italian Health Search Database (HSD). A cohort of 903,748 patients aged 18 years or older was selected and followed until the occurrence of depression, death or end of data availability (December 2019). Demographics, somatic signs/symptoms and psychiatric/medical comorbidities were entered in a multivariate Cox regression to predict the occurrence of depression. The coefficients formed the DEP-HScore for individual patients. Explained variance (pseudo-R2), discrimination (AUC) and calibration (slope estimating predicted-observed risk relationship) assessed the prediction accuracy.

RESULTS

The DEP-HScore explained 18.1 % of the variation in occurrence of depression and the discrimination value was equal to 67 %. With an event horizon of three months, the slope and intercept were not significantly different from the ideal calibration.

LIMITATIONS

The DEP-HScore has not been tested in other settings. Furthermore, the model was characterized by limited calibration performance when the risk of depression was estimated at the 1-year follow-up.

CONCLUSIONS

The DEP-HScore is reliable tool that could be implemented in primary care settings to evaluate the risk of depression, thus enabling prompt and suitable investigations to verify the presence of this condition.

Questionnaire-free machine-learning method to predict depressive symptoms among community-dwelling older adults

The 15-item Geriatric Depression Scale (GDS-15) is widely used to screen for depressive symptoms among older populations. This study aimed to develop and validate a questionnaire-free, machine-learning model as an alternative triage test for the GDS-15 among community-dwelling older adults. The best models were the random forest (RF) and deep-insight visible neural network by internal validation, but both performances were undifferentiated by external validation. The AUROC of the RF model was 0.619 (95% CI 0.610 to 0.627) for the external validation set with a non-local ethnic group. Our triage test can allow healthcare professionals to preliminarily screen for depressive symptoms in older adults without using a questionnaire. If the model shows positive results, then the GDS-15 can be used for follow-up measures. This preliminary screening will save a lot of time and energy for healthcare providers and older adults, especially those persons who are illiterate.

Predicting the presence of depressive symptoms in the HIV-HCV co-infected population in Canada using supervised machine learning

BACKGROUND

Depression is common in the human immunodeficiency virus (HIV)-hepatitis C virus (HCV) co-infected population. Demographic, behavioural, and clinical data collected in research settings may be of help in identifying those at risk for clinical depression. We aimed to predict the presence of depressive symptoms indicative of a risk of depression and identify important classification predictors using supervised machine learning.

METHODS

We used data from the Canadian Co-infection Cohort, a multicentre prospective cohort, and its associated sub-study on Food Security (FS). The Center for Epidemiologic Studies Depression Scale-10 (CES-D-10) was administered in the FS sub-study; participants were classified as being at risk for clinical depression if scores ≥ 10. We developed two random forest algorithms using the training data (80%) and tenfold cross validation to predict the CES-D-10 classes-1. Full algorithm with all candidate predictors (137 predictors) and 2. Reduced algorithm using a subset of predictors based on expert opinion (46 predictors). We evaluated the algorithm performances in the testing data using area under the receiver operating characteristic curves (AUC) and generated predictor importance plots.

RESULTS

We included 1,934 FS sub-study visits from 717 participants who were predominantly male (73%), white (76%), unemployed (73%), and high school educated (52%). At the first visit, median age was 49 years (IQR:43-54) and 53% reported presence of depressive symptoms with CES-D-10 scores ≥ 10. The full algorithm had an AUC of 0.82 (95% CI:0.78-0.86) and the reduced algorithm of 0.76 (95% CI:0.71-0.81). Employment, HIV clinical stage, revenue source, body mass index, and education were the five most important predictors.

CONCLUSION

We developed a prediction algorithm that could be instrumental in identifying individuals at risk for depression in the HIV-HCV co-infected population in research settings. Development of such machine learning algorithms using research data with rich predictor information can be useful for retrospective analyses of unanswered questions regarding impact of depressive symptoms on clinical and patient-centred outcomes among vulnerable populations.

Predicting clinically significant response to primary care treatment for depression from electronic health records of veterans

OBJECTIVE

To reduce delays in referral to specialty mental health care, we evaluated clinical prediction models estimating the likelihood of response to primary care treatment of depression in the VA healthcare system.

METHODS

We included patients with a primary care depression diagnosis between October 1, 2015 and December 31, 2017, an initial PHQ-9 score ≥ 10 within 30 days, a follow-up PHQ-9 score within 2-8 months, and no specialty mental health care within three months prior to depression diagnosis. We evaluated eight ordinary least squares regression models, each with a different procedure for selecting predictors of percentage change in PHQ-9 score from baseline to follow-up. Predictors included patient characteristics from electronic health records and neighborhood characteristics from US census data. We repeated each modeling procedure 1,000 times, using different training and validation sets of patients. We used R², RMSE, and MAE to evaluate model performance.

RESULTS

The final cohort included 3,464 patients. The two best performing models included multiple iterations of backwards stepwise variable selection with R² of 0.07, RMSE of 41.45, MAE of 33.30; and R² of 0.07, RMSE of 41.39, MAE of 33.28.

LIMITATIONS

Wide follow-up interval, possibility of misclassification error due to use of EHR data.

CONCLUSIONS

Model performance did not suggest its use as a guide in clinical decision-making. Future research should explore whether obtaining additional risk factor data from patients (e.g., duration of symptoms) or modeling PHQ-9 scores over a narrower time interval improves performance of clinical risk prediction tools for depression.

The impact of communicating personal mental ill-health risk: A randomized controlled non-inferiority trial

AIM

Risk algorithms predicting personal mental ill-health will form an important component of digital and personalized preventive interventions, yet it is unknown whether informing people of personal risk may cause unintended harm. This trial evaluated the comparative effect of communicating personal mental ill-health risk profiles on psychological distress.

METHODS

Australian participants using a mood-monitoring app were randomly allocated to receiving their current personal mental ill-health risk profile (n = 119), their achievable personal risk profile (n = 118) or to a control group (n = 118) in which no risk information was communicated, in a non-inferiority trial design. The primary outcome was psychological distress at four-weeks as assessed on the Kessler Psychological Distress Scale.

RESULTS

There was high attrition in the trial with 64% of data missing at follow up. Per-protocol (completer) analysis found that the lower bounds of the confidence intervals of the estimated mean change of the current risk (m = 0.19, 95% CI: -2.59- 2.98) and achievable risk (m = -0.09, 95% CI: -2.84 to 2.66) groups were within the non-inferiority margin of the control group's mean at follow up. Supplementary intention-to-treat analysis using Multivariate Imputation by Chained Equations (MICE) found that 98/100 imputed datasets of the current risk profile group, and all imputed datasets of the achievable risk profile group showed non-inferiority to the control group.

CONCLUSIONS

This study provides preliminary support that providing personal mental health risk profiles does not lead to unacceptable worsening of distress compared to no risk feedback, although this needs to be replicated in a fully powered RCT.

Learning From Clinical Consensus Diagnosis in India to Facilitate Automatic Classification of Dementia: Machine Learning Study

BACKGROUND

The Harmonized Diagnostic Assessment of Dementia for the Longitudinal Aging Study in India (LASI-DAD) is the first and only nationally representative study on late-life cognition and dementia in India (n=4096). LASI-DAD obtained clinical consensus diagnosis of dementia for a subsample of 2528 respondents.

OBJECTIVE

This study develops a machine learning model that uses data from the clinical consensus diagnosis in LASI-DAD to support the classification of dementia status.

METHODS

Clinicians were presented with the extensive data collected from LASI-DAD, including sociodemographic information and health history of respondents, results from the screening tests of cognitive status, and information obtained from informant interviews. Based on the Clinical Dementia Rating (CDR) and using an online platform, clinicians individually evaluated each case and then reached a consensus diagnosis. A 2-step procedure was implemented to train several candidate machine learning models, which were evaluated using a separate test set for predictive accuracy measurement, including the area under receiver operating curve (AUROC), accuracy, sensitivity, specificity, precision, F1 score, and kappa statistic. The ultimate model was selected based on overall agreement as measured by kappa. We further examined the overall accuracy and agreement with the final consensus diagnoses between the selected machine learning model and individual clinicians who participated in the clinical consensus diagnostic process. Finally, we applied the selected model to a subgroup of LASI-DAD participants for whom the clinical consensus diagnosis was not obtained to predict their dementia status.

RESULTS

Among the 2528 individuals who received clinical consensus diagnosis, 192 (6.7% after adjusting for sampling weight) were diagnosed with dementia. All candidate machine learning models achieved outstanding discriminative ability, as indicated by AUROC >.90, and had similar accuracy and specificity (both around 0.95). The support vector machine model outperformed other models with the highest sensitivity (0.81), F1 score (0.72), and kappa (.70, indicating substantial agreement) and the second highest precision (0.65). As a result, the support vector machine was selected as the ultimate model. Further examination revealed that overall accuracy and agreement were similar between the selected model and individual clinicians. Application of the prediction model on 1568 individuals without clinical consensus diagnosis classified 127 individuals as living with dementia. After applying sampling weight, we can estimate the prevalence of dementia in the population as 7.4%.

CONCLUSIONS

The selected machine learning model has outstanding discriminative ability and substantial agreement with a clinical consensus diagnosis of dementia. The model can serve as a computer model of the clinical knowledge and experience encoded in the clinical consensus diagnostic process and has many potential applications, including predicting missed dementia diagnoses and serving as a clinical decision support tool or virtual rater to assist diagnosis of dementia.

Risk of Depression, Anxiety, and Stress During the Second Wave of COVID-19 in Slovenia

The spread of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) causing coronavirus disease 2019 (COVID-19) pandemic has led to numerous negative consequences on the mental health of the population throughout the world. The main aim of our study was to compare the risk for depression, anxiety, and stress during the second wave of the pandemic in Slovenia. An additional goal was to analyze the association of depression, anxiety, and stress, with the most relevant subjective factors that define the quality of life. Furthermore, we aimed at determining whether health workers have a higher risk for depression following the course of the pandemic. The study was conducted on the general population, between July 2020 and January 2021 through an online survey. The data of 1,728 respondents in two samples of respondents (782 at baseline - first measurement point and 946 during the second measurement point) of the second wave were analyzed using zero-inflated negative binomial regression and Mann-Whitney U-test. The findings of this study show that the rise the second wave was associated with a higher risk for depression, anxiety and stress. The risk for all three was higher for younger participants. Women showed a higher risk for anxiety and stress. Finances, relationships, and housing dissatisfaction were relevant predictors for depression, anxiety and stress. Health workers in our sample showed a higher risk for stress, but not for depression or anxiety, than the general population. Our findings highlight the urgent need for coordinating and developing mental health services and tailored interventions to reduce the mental health burden, especially in the younger.

Prognosis and Improved Outcomes in Major Depression: A Review

(Reprinted from Transl Psychiatry. 2019 Apr 3; 9(1):127. Open access; is licensed under a Creative Commons Attribution 4.0 International License).

Developing algorithms to predict adult onset internalizing disorders: An ensemble learning approach

A growing literature is utilizing machine learning methods to develop psychopathology risk algorithms that can be used to inform preventive intervention. However, efforts to develop algorithms for internalizing disorder onset have been limited. The goal of this study was to utilize prospective survey data and ensemble machine learning to develop algorithms predicting adult onset internalizing disorders. The data were from Waves 1-2 of the National Epidemiological Survey on Alcohol and Related Conditions (n = 34,653). Outcomes were incident occurrence of DSM-IV generalized anxiety, panic, social phobia, depression, and mania between Waves 1-2. In total, 213 risk factors (features) were operationalized based on their presence/occurrence at the time of or before Wave 1. For each of the five internalizing disorder outcomes, super learning was used to generate a composite algorithm from several linear and non-linear classifiers (e.g., random forests, k-nearest neighbors). AUCs achieved by the cross-validated super learner ensembles were in the range of 0.76 (depression) to 0.83 (mania), and were higher than AUCs achieved by the individual algorithms. Individuals in the top 10% of super learner predicted risk accounted for 37.97% (depression) to 53.39% (social anxiety) of all incident cases. Thus, the algorithms achieved acceptable-to-excellent prediction accuracy with a high concentration of incident cases observed among individuals predicted to be highest risk. In parallel with the development of effective preventive interventions, further validation, expansion, and dissemination of algorithms predicting internalizing disorder onset/trajectory could be of great value.

Prognosis and improved outcomes in major depression: a review

Treatment outcomes for major depressive disorder (MDD) need to be improved. Presently, no clinically relevant tools have been established for stratifying subgroups or predicting outcomes. This literature review sought to investigate factors closely linked to outcome and summarize existing and novel strategies for improvement. The results show that early recognition and treatment are crucial, as duration of untreated depression correlates with worse outcomes. Early improvement is associated with response and remission, while comorbidities prolong course of illness. Potential biomarkers have been explored, including hippocampal volumes, neuronal activity of the anterior cingulate cortex, and levels of brain-derived neurotrophic factor (BDNF) and central and peripheral inflammatory markers (e.g., translocator protein (TSPO), interleukin-6 (IL-6), C-reactive protein (CRP), tumor necrosis factor alpha (TNFα)). However, their integration into routine clinical care has not yet been fully elucidated, and more research is needed in this regard. Genetic findings suggest that testing for CYP450 isoenzyme activity may improve treatment outcomes. Strategies such as managing risk factors, improving clinical trial methodology, and designing structured step-by-step treatments are also beneficial. Finally, drawing on existing guidelines, we outline a sequential treatment optimization paradigm for selecting first-, second-, and third-line treatments for acute and chronically ill patients. Well-established treatments such as electroconvulsive therapy (ECT) are clinically relevant for treatment-resistant populations, and novel transcranial stimulation methods such as theta-burst stimulation (TBS) and magnetic seizure therapy (MST) have shown promising results. Novel rapid-acting antidepressants, such as ketamine, may also constitute a paradigm shift in treatment optimization for MDD.

Development and validation of a clinical prediction tool to estimate the individual risk of depressive relapse or recurrence in individuals with recurrent depression

OBJECTIVES

Many studies examined predictors of depressive relapse/recurrence but no simple tool based on well-established risk factors is available that estimates the risk within an individual. We developed and validated such a prediction tool in remitted recurrently depressed individuals.

METHODS

The tool was developed using data (n = 235) from a pragmatic randomised controlled trial in remitted recurrently depressed participants and externally validated using data (n = 209) from a similar randomised controlled trial of remitted recurrently depressed participants using maintenance antidepressants. Cox regression was used with time to relapse/recurrence within 2 years as outcome and well-established risk factors as predictors. Performance measures and absolute risk scores were calculated, a practically applicable risk score was created, and the tool was externally validated.

RESULTS

The 2-year cumulative proportion relapse/recurrence was 46.2% in the validation dataset. The tool included number of previous depressive episodes, residual depressive symptoms, severity of the last depressive episode, and treatment. The C-statistic and calibration slope were 0.56 and 0.81 respectively. The tool stratified participants into relapse/recurrence risk classes of 37%, 55%, and 72%. The C-statistic and calibration slope in the external validation were 0.59 and 0.56 respectively, and Kaplan Meier curves showed that the tool could differentiate between risk classes.

CONCLUSIONS

This is the first study that developed a simple prediction tool based on well-established risk factors of depressive relapse/recurrence, estimating the individual risk. Since the overall performance of the model was poor, more studies are needed to enhance the performance before recommending implementation into clinical practice.

Development of a prognostic model for predicting depression severity in adult primary patients with depressive symptoms using the diamond longitudinal study

BACKGROUND

Depression trajectories among primary care patients are highly variable, making it difficult to identify patients that require intensive treatments or those that are likely to spontaneously remit. Currently, there are no easily implementable tools clinicians can use to stratify patients with depressive symptoms into different treatments according to their likely depression trajectory. We aimed to develop a prognostic tool to predict future depression severity among primary care patients with current depressive symptoms at three months.

METHODS

Patient-reported data from the diamond study, a prospective cohort of 593 primary care patients with depressive symptoms attending 30 Australian general practices. Participants responded affirmatively to at least one of the first two PHQ-9 items. Twenty predictors were pre-selected by expert consensus based on reliability, ease of administration, likely patient acceptability, and international applicability. Multivariable mixed effects linear regression was used to build the model.

RESULTS

The prognostic model included eight baseline predictors: sex, depressive symptoms, anxiety, history of depression, self-rated health, chronic physical illness, living alone, and perceived ability to manage on available income. Discrimination (c-statistic =0.74; 95% CI: 0.70-0.78) and calibration (agreement between predicted and observed symptom scores) were acceptable and comparable to other prognostic models in primary care.

LIMITATIONS

More complex model was not feasible because of modest sample size. Validation studies needed to confirm model performance in new primary care attendees.

CONCLUSION

A brief, easily administered algorithm predicting the severity of depressive symptoms has potential to assist clinicians to tailor treatment for adult primary care patients with current depressive symptoms.

Target-D: a stratified individually randomized controlled trial of the diamond clinical prediction tool to triage and target treatment for depressive symptoms in general practice: study protocol for a randomized controlled trial

BACKGROUND

Depression is a highly prevalent and costly disorder. Effective treatments are available but are not always delivered to the right person at the right time, with both under- and over-treatment a problem. Up to half the patients presenting to general practice report symptoms of depression, but general practitioners have no systematic way of efficiently identifying level of need and allocating treatment accordingly. Therefore, our team developed a new clinical prediction tool (CPT) to assist with this task. The CPT predicts depressive symptom severity in three months' time and based on these scores classifies individuals into three groups (minimal/mild, moderate, severe), then provides a matched treatment recommendation. This study aims to test whether using the CPT reduces depressive symptoms at three months compared with usual care.

METHODS

The Target-D study is an individually randomized controlled trial. Participants will be 1320 general practice patients with depressive symptoms who will be approached in the practice waiting room by a research assistant and invited to complete eligibility screening on an iPad. Eligible patients will provide informed consent and complete the CPT on a purpose-built website. A computer-generated allocation sequence stratified by practice and depressive symptom severity group, will randomly assign participants to intervention (treatment recommendation matched to predicted depressive symptom severity group) or comparison (usual care plus Target-D attention control) arms. Follow-up assessments will be completed online at three and 12 months. The primary outcome is depressive symptom severity at three months. Secondary outcomes include anxiety, mental health self-efficacy, quality of life, and cost-effectiveness. Intention-to-treat analyses will test for differences in outcome means between study arms overall and by depressive symptom severity group.

DISCUSSION

To our knowledge, this is the first depressive symptom stratification tool designed for primary care which takes a prognosis-based approach to provide a tailored treatment recommendation. If shown to be effective, this tool could be used to assist general practitioners to implement stepped mental-healthcare models and contribute to a more efficient and effective mental health system.

TRIAL REGISTRATION

Australian New Zealand Clinical Trials Registry (ANZCTR 12616000537459 ). Retrospectively registered on 27 April 2016. See Additional file 1 for trial registration data.

External validation of the international risk prediction algorithm for major depressive episode in the US general population: the PredictD-US study

BACKGROUND

Multivariable risk prediction algorithms are useful for making clinical decisions and for health planning. While prediction algorithms for new onset of major depression in the primary care attendees in Europe and elsewhere have been developed, the performance of these algorithms in different populations is not known. The objective of this study was to validate the PredictD algorithm for new onset of major depressive episode (MDE) in the US general population.

METHODS

Longitudinal study design was conducted with approximate 3-year follow-up data from a nationally representative sample of the US general population. A total of 29,621 individuals who participated in Wave 1 and 2 of the US National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) and who did not have an MDE in the past year at Wave 1 were included. The PredictD algorithm was directly applied to the selected participants. MDE was assessed by the Alcohol Use Disorder and Associated Disabilities Interview Schedule, based on the DSM-IV criteria.

RESULTS

Among the participants, 8 % developed an MDE over three years. The PredictD algorithm had acceptable discriminative power (C-statistics = 0.708, 95 % CI: 0.696, 0.720), but poor calibration (p < 0.001) with the NESARC data. In the European primary care attendees, the algorithm had a C-statistics of 0.790 (95 % CI: 0.767, 0.813) with a perfect calibration.

CONCLUSIONS

The PredictD algorithm has acceptable discrimination, but the calibration capacity was poor in the US general population despite of re-calibration. Therefore, based on the results, at current stage, the use of PredictD in the US general population for predicting individual risk of MDE is not encouraged. More independent validation research is needed.

Predicting Depression among Patients with Diabetes Using Longitudinal Data. A Multilevel Regression Model

INTRODUCTION

This article is part of the Focus Theme of Methods of Information in Medicine on "Big Data and Analytics in Healthcare".

BACKGROUND

Depression is a common and often undiagnosed condition for patients with diabetes. It is also a condition that significantly impacts healthcare outcomes, use, and cost as well as elevating suicide risk. Therefore, a model to predict depression among diabetes patients is a promising and valuable tool for providers to proactively assess depressive symptoms and identify those with depression.

OBJECTIVES

This study seeks to develop a generalized multilevel regression model, using a longitudinal data set from a recent large-scale clinical trial, to predict depression severity and presence of major depression among patients with diabetes.

METHODS

Severity of depression was measured by the Patient Health Questionnaire PHQ-9 score. Predictors were selected from 29 candidate factors to develop a 2-level Poisson regression model that can make population-average predictions for all patients and subject-specific predictions for individual patients with historical records. Newly obtained patient records can be incorporated with historical records to update the prediction model. Root-mean-square errors (RMSE) were used to evaluate predictive accuracy of PHQ-9 scores. The study also evaluated the classification ability of using the predicted PHQ-9 scores to classify patients as having major depression.

RESULTS

Two time-invariant and 10 time-varying predictors were selected for the model. Incorporating historical records and using them to update the model may improve both predictive accuracy of PHQ-9 scores and classification ability of the predicted scores. Subject-specific predictions (for individual patients with historical records) achieved RMSE about 4 and areas under the receiver operating characteristic (ROC) curve about 0.9 and are better than population-average predictions.

CONCLUSIONS

The study developed a generalized multilevel regression model to predict depression and demonstrated that using generalized multilevel regression based on longitudinal patient records can achieve high predictive ability.

Learning and behavioral deficits associated with the absence of the fragile X mental retardation protein: what a fly and mouse model can teach us

The Fragile X syndrome (FXS) is the most frequent form of inherited mental disability and is considered a monogenic cause of autism spectrum disorder. FXS is caused by a triplet expansion that inhibits the expression of the FMR1 gene. The gene product, the Fragile X Mental Retardation Protein (FMRP), regulates mRNA metabolism in brain and nonneuronal cells. During brain development, FMRP controls the expression of key molecules involved in receptor signaling, cytoskeleton remodeling, protein synthesis and, ultimately, spine morphology. Symptoms associated with FXS include neurodevelopmental delay, cognitive impairment, anxiety, hyperactivity, and autistic-like behavior. Twenty years ago the first Fmr1 KO mouse to study FXS was generated, and several years later other key models including the mutant Drosophila melanogaster, dFmr1, have further helped the understanding of the cellular and molecular causes behind this complex syndrome. Here, we review to which extent these biological models are affected by the absence of FMRP, pointing out the similarities with the observed human dysfunction. Additionally, we discuss several potential treatments under study in animal models that are able to partially revert some of the FXS abnormalities.

A prediction algorithm for first onset of major depression in the general population: development and validation

OBJECTIVE

Prediction algorithms are useful for making clinical decisions and for population health planning. However, such prediction algorithms for first onset of major depression do not exist. The objective of this study was to develop and validate a prediction algorithm for first onset of major depression in the general population.

METHODS

Longitudinal study design with approximate 3-year follow-up. The study was based on data from a nationally representative sample of the US general population. A total of 28 059 individuals who participated in Waves 1 and 2 of the US National Epidemiologic Survey on Alcohol and Related Conditions and who had not had major depression at Wave 1 were included. The prediction algorithm was developed using logistic regression modelling in 21 813 participants from three census regions. The algorithm was validated in participants from the 4th census region (n=6246). Major depression occurred since Wave 1 of the National Epidemiologic Survey on Alcohol and Related Conditions, assessed by the Alcohol Use Disorder and Associated Disabilities Interview Schedule-diagnostic and statistical manual for mental disorders IV.

RESULTS

A prediction algorithm containing 17 unique risk factors was developed. The algorithm had good discriminative power (C statistics=0.7538, 95% CI 0.7378 to 0.7699) and excellent calibration (F-adjusted test=1.00, p=0.448) with the weighted data. In the validation sample, the algorithm had a C statistic of 0.7259 and excellent calibration (Hosmer-Lemeshow χ(2)=3.41, p=0.906).

CONCLUSIONS

The developed prediction algorithm has good discrimination and calibration capacity. It can be used by clinicians, mental health policy-makers and service planners and the general public to predict future risk of having major depression. The application of the algorithm may lead to increased personalisation of treatment, better clinical decisions and more optimal mental health service planning.