Exploring comorbid depression and physical health trajectories: A case-based computational modelling approach.
Creator
Castellani Brian; Griffiths Frances; Rajaram Rajeev; Gunn Jane
Publisher
Journal of evaluation in clinical practice
Date
2018
2018-12
Description
While comorbid depression/physical health is a major clinical concern, the conventional methods of medicine make it difficult to model the complexities of this relationship. Such challenges include cataloguing multiple trends, developing multiple complex aetiological explanations, and modelling the collective large-scale dynamics of these trends. Using a case-based complexity approach, this study engaged in a richly described case study to demonstrate the utility of computational modelling for primary care research. N = 259 people were subsampled from the Diamond database, one of the largest primary care depression cohort studies worldwide. A global measure of depressive symptoms (PHQ-9) and physical health (PCS-12) were assessed at 3, 6, 9, and 12 months and then annually for a total of 7 years. Eleven trajectories and 2 large-scale collective dynamics were identified, revealing that while depression is comorbid with poor physical health, chronic illness is often low dynamic and not always linked to depression. Also, some of the cases in the unhealthy and oscillator trends remain ill without much chance of improvement. Finally, childhood abuse, partner violence, and negative life events are greater amongst unhealthy trends. Computational modelling offers a major advance for health researchers to account for the diversity of primary care patients and for developing better prognostic models for team-based interdisciplinary care.
Exploring comorbid depression and physical health trajectories: A case-based computational modelling approach.
Creator
Castellani Brian; Griffiths Frances; Rajaram Rajeev; Gunn Jane
Publisher
Journal of evaluation in clinical practice
Date
2018
2018-12
Description
While comorbid depression/physical health is a major clinical concern, the conventional methods of medicine make it difficult to model the complexities of this relationship. Such challenges include cataloguing multiple trends, developing multiple complex aetiological explanations, and modelling the collective large-scale dynamics of these trends. Using a case-based complexity approach, this study engaged in a richly described case study to demonstrate the utility of computational modelling for primary care research. N = 259 people were subsampled from the Diamond database, one of the largest primary care depression cohort studies worldwide. A global measure of depressive symptoms (PHQ-9) and physical health (PCS-12) were assessed at 3, 6, 9, and 12 months and then annually for a total of 7 years. Eleven trajectories and 2 large-scale collective dynamics were identified, revealing that while depression is comorbid with poor physical health, chronic illness is often low dynamic and not always linked to depression. Also, some of the cases in the unhealthy and oscillator trends remain ill without much chance of improvement. Finally, childhood abuse, partner violence, and negative life events are greater amongst unhealthy trends. Computational modelling offers a major advance for health researchers to account for the diversity of primary care patients and for developing better prognostic models for team-based interdisciplinary care.
Patient Computer Use To Prompt Doctor Adherence To Diabetes Management Guidelines
Creator
Haller N A; Gil K M; Gardner W G; Whittier F C
Publisher
Journal of Evaluation in Clinical Practice
Date
2009
2009-12
Description
Rationale, Aims and Objectives Doctor compliance with diabetic care guidelines is low and may be improved with system-wide changes that include patient involvement. The objective of this study was to determine if patients in an internal medicine teaching clinic would use a touch-screen computer to receive personalized information regarding their need for diabetes care. Outcomes included determining if this intervention would improve resident doctor compliance with diabetic guidelines. Methods In this prospective study a computer was available for patients to use independently in one clinic, while another computer was placed in a second clinic with nursing support. Patients responding they were diabetic to the first screen received screens covering HbA1c, blood pressure, cholesterol, foot, eye examinations and compliance with having labs drawn. Non-diabetic patients received three general health screens. A response-based report was printed for patients to share with their doctor. Chart reviews were conducted to assess diabetic health care delivery. Results The computer was used voluntarily by 20.6% of patients in the primary clinic and by 100% of patients in the nurse-assisted clinic. A total of 104 patients from both clinics responded they were diabetic; over 50% did not know what HbA1c meant and a minority responded their HbA1c, blood pressure and cholesterol were at good levels. Significantly more HbA1c tests conducted within 6 months were documented in patients' charts. Discussion Patients used the computer effectively without direction in the primary clinic. In this initial study, implementation of the computer program increased the number of HbA1c tests ordered. Future studies incorporating refinements may increase both usage and efficacy of this intervention.
Subject
complications; computer; delivery; diabetes; General & Internal; Health Care Sciences & Services; intervention; involvement; knowledge; management guidelines; Medical Informatics; Medicine; mellitus; patient; performance; preventive services; primary care; quality; risk; standards