Medical Comorbidity

This thesis considers applications of machine learning techniques in hospital emergency readmission and comorbidity risk problems, using healthcare administrative data. The aim is to introduce generic and robust solution approaches that can be applied to different healthcare settings. Existing solution methods and techniques of predictive risk modelling of hospital emergency readmission and comorbidity risk modelling are reviewed. Several modelling approaches, including Logistic Regression, Bayes Point Machine, Random Forest and Deep Neural Network are considered. Firstly, a framework is proposed for pre-processing hospital administrative data, including data preparation, feature generation and feature selection. Then, the Ensemble Risk Modelling of Hospital Readmission (ERMER) is presented, which is a generative ensemble risk model of hospital readmission model. After that, the Temporal-Comorbidity Adjusted Risk of Emergency Readmission (T-CARER) is presented for identifying very s...

This thesis considers applications of machine learning techniques in hospital emergency readmission and comorbidity risk problems, using healthcare administrative data. The aim is to introduce generic and robust solution approaches that can be applied to different healthcare settings. Existing solution methods and techniques of predictive risk modelling of hospital emergency readmission and comorbidity risk modelling are reviewed. Several modelling approaches, including Logistic Regression, Bayes Point Machine, Random Forest and Deep Neural Network are considered. Firstly, a framework is proposed for pre-processing hospital administrative data, including data preparation, feature generation and feature selection. Then, the Ensemble Risk Modelling of Hospital Readmission (ERMER) is presented, which is a generative ensemble risk model of hospital readmission model. After that, the Temporal-Comorbidity Adjusted Risk of Emergency Readmission (T-CARER) is presented for identifying very sick comorbid patients. A Random Forest and a Deep Neural Network are used to model risks of temporal comorbidity, operations and complications of patients using the T-CARER. The computational results and benchmarking are presented using real data from Hospital Episode Statistics (HES) with several samples across a ten-year period. The models select features from a large pool of generated features, add temporal dimensions into the models and provide highly accurate and precise models of problems with complex structures. The performances of all the models have been evaluated across different timeframes, sub-populations and samples, as well as previous models.

Objectives: To examine: 1) whether patient characteristics predict perceived self-management support (usefulness of information and collaborative care planning) by primary care providers after six months, and 2) the association between perceived self-management support and patient activation at 6 months Methods: We conducted a secondary analysis among 120 primary care patients aged 40 and over with a chronic physical condition and comorbid depressive symptoms who participated in a randomized controlled trial of a coaching intervention for depression self-management. Activation was measured at baseline (T0) and 6 months (T1). Self-management support was captured at T1 for physical and mood problems. Results: The sample of analysis was 120 patients who completed all relevant measures. At T1, the perceived usefulness of information for mood self-management was independently associated with activation. More severe depressive symptoms at T0 predicted lower perceived usefulness of chronic condition self-management information at T1. Lower T0 mental health-related quality of life predicted lower perceived usefulness of mood self-management information at T1. Conclusions: Perceived informational support for mood self-management may contribute to increased activation. Patients with more severe mental health symptoms or impairment perceive that they receive less useful self-management information from their care team. Practice implications: Care teams should determine whether patients with mood problems need greater self-management support.

Introduction: Autism Spectrum Disorders (ASD) are behaviourally defined syndromes where the etiology and pathophysiology are not very well understood. If the child has been diagnosed with autism, parents already face special behavioural challenges. What makes ASD even more difficult, are the many other medical health issues that often arise in this population. The purpose of this article is to summarise the latest understanding of autism's commonly associated physical and mental health conditions. Methods: An analysis of relevant literature, sources from the internet and published literature, personal experience and observations of the author. Recent findings: Autism is a disorder of the whole body. It is often in co-morbidity with: epilepsy, gastrointestinal disorders, neuro-inflammation and immunological disorders, asthma, eczema, sleep disturbances, eating and feeding disorders, food allergies, attention deficit and hyperactivity disorder (ADHD), headaches, anxiety disorders, bipolar disorder, depression, schizophrenia, obsessive-compulsive disorder, Fragile X syndrome, intellectual disability, nonverbal learning disorder, motor clumsiness, Tourette syndrome, sensory problems, tuberous sclerosis, oxidative stress, acquired mitochondrial dysfunction and metabolic abnormalities. Many of the issues outlined here can overlap each other. Conclusions: Improved understanding of the underlying pathology of ASD and associated conditions, and the development of a common purpose across multiple treating sites, can improve the consistent and coordinated healthcare of children with autism. There is need for the development of improved strategies for delivering effective health education and healthcare to this large population. Improving the ability of these persons to lead relatively independent lives has a great economic impact.

— There is a dearth of research work and scarcity of availability of data on psychiatric morbidity among railway employees in India. Present hospital based observational descriptive study was carried out at medical OPD of Central Hospital of Northern Western Railway, Jaipur with the specific aims to study occupational attributes of common psychiatric disorders. Total 450 randomly selected individuals were screened for psychiatric morbidities by using Mini-international neuropsychiatric interview-PLUS. Psychiatric diagnosis confirmed by two psychiatrists separately using the ICD, 10th revision (diagnostic and research criteria). A self designed, pre-tested & semi-structured detailed Performa was used to elicit demographical and occupational attributes. Univariate analysis was carried out initially to identify risk factors. The data was analyzed in terms of descriptive statistics, Chi-square test and stepwise binary logistic regression by using SPSS and Primer Version '6'. Most common diagnosed psychiatric disorders were Neurotic, stress related and somatoform disorders (28%), followed by mood disorders (25.56%). Common Psychiatric Disorders (CPD) i.e. depression, anxiety and somatoform disorders were the most common diagnosed psychiatric morbidities (49.78%). Working in rotatory shifts, perceived problems related to shift duties and stress in the working environment, work experience more than 20 years and non-executive type job were observed as the occupational risk factors for common psychiatric disorders. The present observations suggest that there is scope for psychiatric intervention for the management of perceived stress at work place, problems related to shift duties and other occupational challenges.

Background: It is widely accepted that many factors including medical comorbidity can influence the development of depression
in women. There is no evidence, however, that specific medical diagnoses are independent markers for severity of
depression.
Objective: This study aims to investigate the iron status of Turkish female patients with depression, to find out if the iron
deficiency (ID) is a risk factor in depression. In addition, it also aims to investigate a possible relationship between reduced
iron capacity and intensity of the disease.
Method: A retrospective chart review comparison study was carried out to investigate the iron status of Turkish women
with a diagnosis of depression, aged between 18 to 43 years old, living in Germany. The iron parameters were measured in
the blood of 180 Turkish women with depression, and of 102 Turkish women without depression. The depressed patients
were further sub grouped as mild (n=28), moderate (n=74) and severe depression (n=78) by using Hamilton Depression
Rating Scale D (HRS). The measurements for red blood cells (RBC), haemoglobin, free iron and ferritin of the participants in
the Community Outpatient Clinic were recorded, and the data was statistically analysed.
Patients with a family history of a blood disorder except iron deficiency anaemia, patients with increased acute phase reactants,
and those who were pregnant, or suffering from malignancies, diabetes mellitus, and/or mental disorders except
depression were excluded from this study.
Results: We have found a significant negative correlation between ID and the severity of depression. Levels of reduce hemoglobin
in peripheral blood, along with iron and ferritin and RBC were found in several of the patients suffering from severe
depression.
Conclusion: These findings have been presented with consideration of their clinical implications of ID and depression with
a view to conducting further research