Materiały źródłowe

• #1

  https://pmc.ncbi.nlm.nih.gov/articles/PMC4336916/

  Psychiatric treatment relies on a solid armamentarium of pharmacologic and nonpharmacologic treatment modalities that perform reasonably well for many patients but leave others in a state of chronic disability or troubled by problematic side effects. […] Thus, there is an urgent need for better tools that will provide a means for matching individual patients with the most effective treatments while minimizing the risk of adverse events. […] This review will consider the current state of the science in predicting treatment outcomes in psychiatry. […] Improved prediction of treatment response could have many benefits for patients and reduce the health care costs that accrue from the use of inappropriate or suboptimal treatments. […] However, prediction is a challenging endeavor, since many different variables can affect individual treatment outcomes.

• #2 Personalized Prognostic Prediction of Treatment Outcome for Depressed Patients in a Naturalistic Psychiatric Hospital Setting: A Comparison of Machine Learning Approaches

  https://pmc.ncbi.nlm.nih.gov/articles/PMC6924628/

  Knowledge of which patients are likely to exhibit a poor outcome may have important clinical implications regarding treatment recommendations (e.g., a more intensive, alternative or combination treatment) and can inform more careful symptom and treatment progress monitoring. […] The present study sought to predict treatment outcomes for depressed patients in an intensive psychiatric treatment setting (i.e., partial hospital program). […] The best performing model in the training sample, ENR, was then applied to the holdout sample, where the R2 was .38. […] The performance of ENR in the training sample and in the holdout sample indicate that this model could provide clinicians with reasonably accurate predictions of a patients depression status at the time of discharge on the basis of pre-treatment clinical and demographic characteristics.

• #3 Machine learning model to predict mental health crises from electronic health records | Nature Medicine

  https://www.nature.com/articles/s41591-022-01811-5

  The timely identification of patients who are at risk of a mental health crisis can lead to improved outcomes and to the mitigation of burdens and costs. […] A follow-up 6-month prospective study evaluated our algorithms use in clinical practice and observed predictions to be clinically valuable in terms of either managing caseloads or mitigating the risk of crisis in 64% of cases. […] Therefore, identifying patients at risk of experiencing a crisis before its occurrence is central to improving patient outcomes and managing caseloads. […] This research explores the feasibility of predicting any mental health crisis event, regardless of its cause or the underlying mental disorder, and we investigate whether such predictions can provide added value to clinical practice. […] We evaluated whether a tool predicting and presenting risk of mental health crisis provides added value for clinical practice in terms of managing caseloads and mitigating the risk of crisis.

• #4 The PSYchiatric clinical outcome prediction (PSYCOP) cohort: leveraging the potential of electronic health records in the treatment of mental disorders | Acta Neuropsychiatrica | Cambridge Core

  https://www.cambridge.org/core/journals/acta-neuropsychiatrica/article/psychiatric-clinical-outcome-prediction-psycop-cohort-leveraging-the-potential-of-electronic-health-records-in-the-treatment-of-mental-disorders/73CDCC5B36FF1347E6419EC7B80DEC48

  The quality of life and lifespan are greatly reduced among individuals with mental illness. To improve prognosis, the nascent field of precision psychiatry aims to provide personalised predictions for the course of illness and response to treatment. […] Implementation of prediction models in clinical psychiatry will likely improve treatment and, hopefully, increase the quality of life and lifespan of those with mental illness.

• #5

  https://pmc.ncbi.nlm.nih.gov/articles/PMC4336916/

  In recent years, considerable effort has gone into developing biomarkers of treatment response that might supplement clinical factors. […] Pharmacogenomics aims to reduce acute and longterm adverse events and optimize response rates by using genetic information to match medications to individual patients. […] In this review, the current state of treatment prediction research in psychiatry will be reviewed, with the aim of summarizing progress and highlighting areas where research findings have the best potential for near-term translation to clinical practice. […] Prediction of treatment outcomes is a specific case of the broader and very vibrant field of prediction science. […] The best predictions arise from statistical methods that fit precise data to valid models describing how the relevant variables contribute to the outcome.

• #6 Prognosis and improved outcomes in major depression: a review | Translational Psychiatry

  https://www.nature.com/articles/s41398-019-0460-3

  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. […] 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. […] Clear evidence of an inverse relationship between duration of episode and treatment outcome (either response or remission) underscores the importance of early intervention in MDD. In particular, replicable prospective and retrospective studies indicate that shorter duration of untreated disease both in terms of first and recurrent episodes is a prognostic factor indicating better treatment response and better long-term outcomes.

• #7 Prognosis and improved outcomes in major depression: a review | Translational Psychiatry

  https://www.nature.com/articles/s41398-019-0460-3

  Another important clinical variable is time to antidepressant response. For instance, one meta-analysis found that early improvement was positively linked to antidepressant treatment outcome in 15 of 16 studies. […] This suggests that TRD defined loosely here as non-response to at least two adequate antidepressant trials and chronic depression (roughly defined here as non-response to any treatment) may have similar response slopes in the earliest treatment stages. […] The influence of sociodemographic factors such as age, age of onset, gender, and number of previous episodes on treatment outcome has been investigated with mixed results. […] High stress levels significantly influence outcomes in MDD patients who are prone to vulnerable states, such as those with high levels of neuroticism.

• #8

  https://link.springer.com/article/10.1007/s10488-024-01401-2

  The primary objective of this study was to determine whether change shown early in treatment by patients with a personality disorder could predict their ultimate treatment outcome, as measured by two outcome measures, the SIPP-SF and the GAPD. The results showed that early changes on a specific domain of the SIPP were the strongest predictors of case-mix corrected final scores on that same domain, and the proportion of explained variance ranged from 10.9 to 33.4%. This indicates that improvements or declines in specific personality domains early in treatment are highly indicative of the final outcomes in those same domains. The second most significant predictor for the domains of Self-Control, Relational Functioning, and Identity Integration was the early change on the OQ-45 SD scale. […] For the GAPD as an outcome measure, early changes in the GAPD itself at both intermediate measurement points were the strongest predictors of case-mix corrected GAPD post-treatment scores. Early changes on the OQ-45 SD scale and the SIPP domain Social Attunement were also significant predictors, indicating that improvements in overall symptom distress and Social Attunement might be relevant for predicting outcomes on the GAPD. […] In sum, in the context of personality disorder treatments, early assessments during the initial 8 weeks of inpatient care can reveal valuable insights into treatment responsiveness.

• #9 Logo for Cambridge Core from Cambridge University Press. Click to return to homepage.

  https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/prediction-of-longerterm-outcome-of-treatmentresistant-depression-in-tertiary-care/79683F91C6D3398D0877FC16828070B8

  Systematic studies on the outcome of treatment-resistant depression are scarce. […] To describe the longer-term outcome and predictors of outcome in treatment-resistant depression. […] The majority of participants attained full remission (60.2%), most of whom (48.3% of total sample) showed sustained recovery (full remission for at least 6 months). […] A substantial minority had persistent subsyndromal depression (19.5%) or persistent depressive episode (20.3%). […] Diagnosis of bipolar treatment-resistant depression and poorer social support were associated with early relapse, whereas strong social support, higher educational status and milder level of treatment resistance measured with the Maudsley Staging Method were associated with achieving quicker remission. […] Although many patients with treatment-resistant depression experience persistent symptomatology even after intensive, specialist treatment, most can achieve remission.

• #10 Logo for Cambridge Core from Cambridge University Press. Click to return to homepage.

  https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/prediction-of-longerterm-outcome-of-treatmentresistant-depression-in-tertiary-care/79683F91C6D3398D0877FC16828070B8

  Three main factors were associated with achieving remission during follow-up: severity of treatment-resistant depression measured with the MSM, educational achievement and level of social support. […] The two factors independently associated with relapse were poor social support and receiving a diagnosis of bipolar depression. […] Use of MAOIs while an in-patient was independently associated with remission at point of discharge controlling for other treatments, particularly for unipolar treatment-resistant depression. […] The present study represents one of the very few such studies, and confirms the often poor long-term outlook and the need for new treatment approaches to improve outcomes. […] This is underlined by the association between increasing symptom severity and mortality.

• #11 Prognosis and improved outcomes in major depression: a review | Translational Psychiatry

  https://www.nature.com/articles/s41398-019-0460-3

  Psychiatric comorbidity has been shown to influence outcome in both treated and untreated patients. […] MDD and several physical diseases including cardiovascular disease and diabetes appear to have bidirectional effects on disease trajectory, yet pathophysiologic links are most likely complex and have to be elucidated. […] Structural markers of antidepressant treatment outcomes suggest that hippocampal volumes are related to response and remission. […] While these interesting results suggest that fMRI measures could ultimately help classify biological subtypes of depression, these methods are far from ready for clinical application and results will have to be reproduced. […] This paper also reviewed biological markers, where research has grown exponentially to encompass enormous projects with potentially tens of thousands of subjects enrolled in real world studies. […] Recent studies have led to important insights into neurobiological disease markers that could result in improved disease stratification and response prediction in the near future.

• #12 Association between life satisfaction, sleep disturbance and mental health problems among the migrant older adults with children: a conditional process analysis based on per capita bedroom area | BMC Geriatrics | Full Text

  https://bmcgeriatr.biomedcentral.com/articles/10.1186/s12877-025-05957-y

  Life satisfaction was negatively associated with mental health problems, and sleep disturbance could mediate their association. […] Furthermore, the direct effect of life satisfaction on mental health problems and the indirect effect of sleep disturbance in the relationship were moderated by PCBA, but it only moderated the direct effect and the second indirect link (sleep disturbance-mental health problems) of the mediating effect. […] Life satisfaction had negative effect on mental health problems. PCBA moderated the direct effect of life satisfaction on mental health problems and the mediating effect of sleep disturbance on mental health problems. […] This study illustrated that the association between life satisfaction and mental health problems was partially mediated by sleep disturbance, which confirmed the Hypothesis 2 and also explained how life satisfaction could indirectly affect mental health problems.

• #13

  https://il.linkedin.com/posts/amit-etkin_can-eeg-and-machine-learning-ml-predict-activity-7054461144001908736-ozLk

  Can EEG and machine learning (ML) predict outcome with psychotherapy? In this new paper in Nature Mental Health, we developed and validated EEG ML biomarkers for predicting outcome with two types of psychotherapy for PTSD. […] Not only could EEG ML predict treatment, but models trained on one therapy could predict the other. Not only could EEG ML predict responders, but it could also identify non-responders….people for whom neither therapy works. This work is an exciting advance in precision psychiatry — finding the right treatment for each patient — and further shows the power of EEG and ML.

• #14 Psychosis Prognosis Predictor: A Continuous and Uncertainty-Aware Prediction of Treatment Outcome in First-Episode Psychosis | medRxiv

  https://www.medrxiv.org/content/10.1101/2023.07.05.23292252v1.full-text

  Importance Presently, clinicians face challenges in accurately predicting the prognosis of patients with psychosis. […] Objective We propose and evaluate a multi-task recurrent neural network architecture specifically designed to overcome these challenges, enabling trustworthy prediction of psychosis prognosis. […] Conclusions and Relevance Our approach involved constructing prediction models utilizing a flexible neural network architecture tailored to clinical scenarios derived from a time series dataset. […] Therefore, we advocate for the inclusion of time series data in future individualized prediction modeling endeavors within psychiatric research, as it holds substantial promise in improving prognostic accuracy and informing personalized treatment approaches. […] In our study, we present a machine learning framework that predicts multiple outcomes based on longitudinal patient data while integrating prediction uncertainty to facilitate more reliable clinical decision-making.

• #15 Psychosis Prognosis Predictor: A Continuous and Uncertainty-Aware Prediction of Treatment Outcome in First-Episode Psychosis | medRxiv

  https://www.medrxiv.org/content/10.1101/2023.07.05.23292252v1.full-text

  The LSTM architecture is flexible; it is possible to build one model that can be used at different points in time, adding the additional data that is available at a specific moment. […] The multi-task model provides a way for clinicians to aim for recovery in several domains. […] Thirdly, we used the uncertainty of predictions to adjust the prediction accuracy and build a novel decision-making module for more individualized decision-making. […] Using this flexible multi-task architecture that incorporates the uncertainty of individual predictions, we took a leap forward toward improving patient care with the help of machine learning prediction models. […] Although the accuracy of our models is not above the 80% threshold suggested by the APA when all patients are incorporated, we still consider our models to be clinically relevant, as they provide a certainty measure for each prediction. […] This is why we consider this feature an important step toward reaching our goal of building an interactive tool for individual psychosis prognosis prediction.

• #16 Personalized Prognostic Prediction of Treatment Outcome for Depressed Patients in a Naturalistic Psychiatric Hospital Setting: A Comparison of Machine Learning Approaches

  https://pmc.ncbi.nlm.nih.gov/articles/PMC6924628/

  Research on predictors of treatment outcome in depression has largely derived from randomized clinical trials involving strict standardization of treatments, stringent patient exclusion criteria and careful selection and supervision of study clinicians. […] This study sought to predict depression outcomes for patients seeking treatment within an intensive psychiatric hospital setting, and while comparing the performance of a range of machine learning approaches. […] The best performing model in the training sample was elastic net regularization (ENR; MSE= 20.49, R2 =.28), which had comparable performance in the holdout sample (MSE= 11.26, R2 = .38). […] Informed by pre-treatment patient characteristics, such predictive models could be used to communicate prognosis to clinicians and to guide treatment planning. Identified predictors of poor prognosis may suggest important targets for intervention.

• #17 Personalized Prognostic Prediction of Treatment Outcome for Depressed Patients in a Naturalistic Psychiatric Hospital Setting: A Comparison of Machine Learning Approaches

  https://pmc.ncbi.nlm.nih.gov/articles/PMC6924628/

  Knowledge of which patients are likely to exhibit a poor outcome may have important clinical implications regarding treatment recommendations (e.g., a more intensive, alternative or combination treatment) and can inform more careful symptom and treatment progress monitoring. […] The present study sought to predict treatment outcomes for depressed patients in an intensive psychiatric treatment setting (i.e., partial hospital program). […] The best performing model in the training sample, ENR, was then applied to the holdout sample, where the R2 was .38. […] The performance of ENR in the training sample and in the holdout sample indicate that this model could provide clinicians with reasonably accurate predictions of a patients depression status at the time of discharge on the basis of pre-treatment clinical and demographic characteristics.

• #18 Machine learning model to predict mental health crises from electronic health records | Nature Medicine

  https://www.nature.com/articles/s41591-022-01811-5

  On average, the CMHTs disagreed with only 7% of the model predictions, with the model outputs found to be clinically useful in 64% of individual cases. […] Notably, the importance of the algorithm for identifying at-risk patients who would otherwise have been missed emerged from the semi-structured interviews conducted with the clinicians as part of the qualitative evaluation. […] We have demonstrated the feasibility of predicting mental health crises by applying machine learning techniques to longitudinally collected EHR data, obtaining an AUROC of 0.797 for the general model. […] We further expanded the subgroup analysis to assess the algorithms fairness. […] The rising demand for mental healthcare is increasingly prompting hospitals to actively work on identifying novel methods of anticipating demand and better deploying their limited resources to improve patient outcomes and decrease long-term costs.

• #19 Personalized Prognostic Prediction of Treatment Outcome for Depressed Patients in a Naturalistic Psychiatric Hospital Setting: A Comparison of Machine Learning Approaches

  https://pmc.ncbi.nlm.nih.gov/articles/PMC6924628/

  A prognosis calculator could be used to identify, and subsequently closely monitor symptom progress for, patients predicted to have a poor prognosis, which may improve outcomes. […] Ultimately, research is needed to test whether such patients do in fact have better outcomes in an inpatient setting than a partial hospital program. […] The present study demonstrates the use of machine learning in predicting treatment outcome in a naturalistic psychiatric hospital setting, with findings translated into a patient-specific prognosis calculator.

• #20 Using Early Change to Predict Outcome in Cognitive Behaviour Therapy: Exploring Timeframe, Calculation Method, and Differences of Disorder-Specific versus General Measures | PLOS One

  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0100614

  Early change can predict outcome of psychological treatment, especially in cognitive behavior therapy. […] This study aimed to investigate how well early change predicted outcome depending on the week it was measured, the calculation method (regression slope or simple subtraction), the type of measures used, and the target disorder. […] We conclude that a simple calculation method with a disorder-specific measure at week four seems to provide a good choice for predicting outcome in time-limited cognitive behavior therapy. […] There is a need to investigate clinically useful methods and instruments for early prediction of treatment outcome. […] Even though it would be desirable if patient characteristics could prospectively predict whether a treatment will be successful or not, prediction studies of CBT for panic disorder, social anxiety disorder, and depression have failed to identify stable pre-treatment patient characteristics that reliably predict treatment outcome with the exception of baseline symptom severity.

• #21 Using Early Change to Predict Outcome in Cognitive Behaviour Therapy: Exploring Timeframe, Calculation Method, and Differences of Disorder-Specific versus General Measures | PLOS One

  https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0100614

  The general aim of the present study was to investigate clinically relevant aspects of early improvement in CBT delivered via the Internet for panic disorder, social anxiety disorder and depression. […] The findings also demonstrate that if disorder specific symptoms are considered the primary outcome then disorder specific measures should be used as predictors rather than general measures. […] More research is needed on general measures to further investigate for which disorders they are most suitable for predicting and measuring outcome.

• #22 Hearing vocals to recognize schizophrenia: speech discriminant analysis with fusion of emotions and features based on deep learning | BMC Psychiatry | Full Text

  https://bmcpsychiatry.biomedcentral.com/articles/10.1186/s12888-025-06888-z

  The combination of speech analysis and machine learning has achieved some results in the medical field. […] The study demonstrated significant differences in speech patterns between SZ and mentally healthy individuals. […] The diagnostic accuracy of the log-Mel spectrogram analysis using a convolutional neural network ranged from 80.9% to 85.7% when considering a single emotional stimulus. […] Emotion fusion improved both sensitivity and specificity, resulting in an overall accuracy of 87.0%. […] By combining different emotional stimuli and individualized features, the diagnostic model achieved a remarkable individual-discriminative performance, with an accuracy of 91.7% and an AUC of 0.963. […] The inclusion of individual information in an automatic detection model can closely approximate the diagnostic approaches of professional physicians, thereby addressing the challenge of heterogeneity and facilitating rapid and comprehensive schizophrenia detection based on patients specific conditions.

• #23 Logo for Cambridge Core from Cambridge University Press. Click to return to homepage.

  https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/prediction-of-longerterm-outcome-of-treatmentresistant-depression-in-tertiary-care/79683F91C6D3398D0877FC16828070B8

  The choice of treatment and presence of good social support may affect remission rates, whereas those with low social support and a bipolar diathesis should be considered at higher risk of early relapse. […] Understanding factors associated with outcome is important in order to better understand the aetiology of the disorder, and to identify potential targets for therapeutic intervention. […] The main outcomes of interest were remission, recovery, relapse and persistence of depressive episode. […] Overall, 48.3% (n = 57) of the cohort met recovery criteria, and 11.9% (n = 14) had met remission criteria without ever meeting recovery criteria. […] In terms of unfavourable outcome, 39.8% had persistent symptomatology throughout the follow-up period, either because of persistent subthreshold/subsyndromal symptoms (19.5%) or persistent depressive episode (20.3%).

• #24

  https://link.springer.com/article/10.1007/s10488-024-01401-2

  A significant proportion of patients with a personality disorder do not benefit from treatment. Monitoring treatment progress can help adjust ineffective treatments. This study examined whether early changes in symptoms and personality dysfunction during the first phase of therapy could predict treatment outcomes. Early changes within a specific SIPP domain were the strongest predictors of post-treatment outcomes in that same domain. Early changes in symptoms significantly predicted outcomes in Self-Control, Relational Functioning, and Identity Integration, while the GAPD predicted outcomes in Self-Control and Social Attunement on the SIPP. Thus, when it comes to personality dysfunction, early changes in a specific domain or measure are the best predictors of outcomes in that same domain. In sum, considering these factors, monitoring early change can be useful in assessing progress in the treatment of patients with personality disorders.

• #25

  https://pmc.ncbi.nlm.nih.gov/articles/PMC4336916/

  In the best of worlds, predictions for a specific individual could be based on the right combination of clinical findings, biomarker measures, and genetic information. […] The major limitation of predictors based on diagnosis and clinical features is dependence on clinical assessments that may be inaccurate, imprecise, or unstable over time. […] New diagnostic systems such as the research domain criteria (RDoC) seek to address this problem by focussing on dimensions of observable behavior and neurobiological measures. […] Recent literature on response to a variety of psychotropic treatments has highlighted the importance of early symptomatic improvement. […] The recent focus on genetics has also revived interest in the family history as a predictive tool. […] Prediction of treatment outcomes remains a significant challenge for psychiatry, as it does for the rest of medicine. […] The best predictions will need to take into account clinical as well as biomarker data. […] Integrative approaches that like the family history can jointly model genetic and environmental influences stand the best chance for generating clinically relevant predictions.

Zobacz więcej