A tale of two studies: the importance of setting, subjects and context in two randomized, controlled trials of a web-based decision support for perimenopausal and postmenopausal health decisions

OBJECTIVE: Knowledge relevant to women's peri- and postmenopausal health decisions has been evolving rapidly. Web-based decision supports can be rapidly updated and have the potential to improve the quality of patients' decisions. We developed and tested a web-based decision support for peri- and postmenopausal health decisionmaking. METHODS: We recruited 409 women aged 45-75 for one randomized, controlled trial and 54 women with an upcoming clinic appointment for a subsequent trial. Women were randomized to use the web-based decision support versus a printed brochure (first trial) and usual care (second trial). Outcomes were changes in decisional satisfaction, decisional conflict, and knowledge, both within each trial and compared across the trials. RESULTS: Intervention subjects had greater increases in decisional satisfaction in the second trial and knowledge in both trials. A high dropout rate among women randomized to the website in the first trial effectively negated benefits in that trial, but not in the second. CONCLUSIONS: The utility of this web-based decision support in two trials depended on a number of factors that appear related to the urgency of making a decision. PRACTICE IMPLICATIONS: Decision aids should be targeted to patients actively trying to make a decision.     

Patients' responsiveness to a decision support tool for primary prevention of cardiovascular diseases in primary care

OBJECTIVE: Assessment of patients' responsiveness to a decision support tool for primary prevention of cardiovascular diseases (CVDs). The booklet focuses on barriers at patient level. METHODS: Process evaluation of an intervention in primary care. Patients at high or potentially high-cardiovascular risk were asked by their GP to prepare themselves for a second consultation in order to participate in decisions on risk management. OUTCOMES: Patients' actually having read the booklet and returning for the second consultation; comprehension and perceived relevance of the information; perceived reassurance. RESULTS: 17 GPs, in the intervention arm of a cluster RCT, issued 276 decision support tools during the first consultation and were instructed to ask them to return for a second consultation to discuss their CVD risk. Patients had a mean age of 54 years, 47% were male and 19% actually had a high cardiovascular risk. Data on 239 patients, a mixture of returnees and non-returnees, showed that they all read the booklet; comprehension was fair to good; 85% perceived the information as relevant; 68% of the patients felt reassured by the information. Satisfaction with the first consultation was higher in the non-returnees. CONCLUSIONS: Cardiovascular prevention spread over two consultations with use of a decision support tool for patients is not easily applicable for GPs. However, based on the findings of good patients' responsiveness, we recommend further development and implementation of decision support tools in primary care. PRACTICE IMPLICATIONS: Decision support for primary CV-prevention is welcomed by patients but needs further adjustment of both the GP and the organization of CV-prevention in primary care. Sharing information between professional and patient on a personal CV-risk management plan is difficult, more training is needed.     

From complex questionnaire and interviewing data to intelligent Bayesian network models for medical decision support

Objectives(1) To develop a rigorous and repeatable method for building effective Bayesian network (BN) models for medical decision support from complex, unstructured and incomplete patient questionnaires and interviews that inevitably contain examples of repetitive, redundant and contradictory responses; (2) To exploit expert knowledge in the BN development since further data acquisition is usually not possible; (3) To ensure the BN model can be used for interventional analysis; (4) To demonstrate why using data alone to learn the model structure and parameters is often unsatisfactory even when extensive data is available.MethodThe method is based on applying a range of recent BN developments targeted at helping experts build BNs given limited data. While most of the components of the method are based on established work, its novelty is that it provides a rigorous consolidated and generalised framework that addresses the whole life-cycle of BN model development. The method is based on two original and recent validated BN models in forensic psychiatry, known as DSVM-MSS and DSVM-P.ResultsWhen employed with the same datasets, the DSVM-MSS demonstrated competitive to superior predictive performance (AUC scores 0.708 and 0.797) against the state-of-the-art (AUC scores ranging from 0.527 to 0.705), and the DSVM-P demonstrated superior predictive performance (cross-validated AUC score of 0.78) against the state-of-the-art (AUC scores ranging from 0.665 to 0.717). More importantly, the resulting models go beyond improving predictive accuracy and into usefulness for risk management purposes through intervention, and enhanced decision support in terms of answering complex clinical questions that are based on unobserved evidence.ConclusionsThis development process is applicable to any application domain which involves large-scale decision analysis based on such complex information, rather than based on data with hard facts, and in conjunction with the incorporation of expert knowledge for decision support via intervention. The novelty extends to challenging the decision scientists to reason about building models based on what information is really required for inference, rather than based on what data is available and hence, forces decision scientists to use available data in a much smarter way.