Question

This is a short answer paragraph type question of practical nature- provide your answer in the box below.

Please read first this case and then answer the two sub-questions in the same answer box below:

Computerised guidelines offer benefits to help a large hospital group to deliver consistent high quality treatment. For example in the following case of managing chronic heart disease (CHD):

The hospital group is tackling heart disease in an aggressive and co-ordinated way. All of the group's 29 practices have developed validated CHD registers, and are now implementing guidelines for managing CHD and hypertension. This activity has been supported by dedicated CHD facilitators, the production of a detailed electronic handbook covering disease register development and treatment guidelines, and the use of computerized templates for decision support and data recording. :

a. (30p) Consider what specific measures need to be taken in the above situation for ensuring the quality/correctness of information in the electronic handbook containing computerized treatment guidelines.

b. (30p.) How would you provide safeguards for the accuracy of data and for the completeness of data in the recorded data about treatment of patients with Chronic Heard Disease( CHD) and hypertension .

Answer 1

a. Measures taken are

Cardiovascular magnetic resonance (MR) use because imageswere available in eight cases all including a three-dimensional (3D) whole-heart sequence (free-breathing and isotropic) that allowed assessment of anatomy and relationship between different morphological structures, 2D cine images across the implantation site to infer information on dynamics and overall material properties and 2D phase contrast images to assess flow. Computed tomography (CT) was performed in five cases—one patient had both CT and MR investigations—using three different systems in one of these cases , four-dimensional (4D) CT was acquired to capture information on ventricular function and implantation site dynamic.

The simulation methodologies included FE analyses which were performed with the commercial software Abaqus/Explicit (Dessault Systemes) and CFD analyses carried out with the commercial packages Fluent (Ansys) and VMTKLab (Orobix). PPVI and CoA stenting were modelled by virtually deploying devices within the patient-specific reconstructed surface anatomies. The 3D surfaces were meshed with shell elements with size ranging from 0.3 to 1.8 mm following sensitivity analysis and matching the dimensions and complexity of each model. Materials of the cardiovascular structures were modelled with linear elastic characteristics: the Young's modulus of the different patient-specific models varied between 42 and 700 kPa according to the distensibility of the structures assessed from imaging and Poisson's ratio was set to 0.45.

b. The pilot nature of this study indicates that there are margins for improvement, as already identified. First of all, boundary conditions for patient-specific simulations based uniquely on routinely acquired clinical data and non-invasive images as in this study need to be improved to account for the variability encountered in this patient population. Hence, the impact of this limitation will need to be investigated by means of statistical studies which can take into account the level of uncertainties. Such analysis will need to consider other sources of uncertainties including the choices of algorithms, physical properties and operator dependency. Second, the series here reported is small, and more cases and more conditions need to be simulated in order to provide stronger clinical evidence on the usefulness of prospective computational models. Third, the influence of computational tools on the actual decision-making process has not been evaluated yet. Further studies, purposely designed, will be required to evaluate the impact that simulations might have on the different phases of the cognitive process, including comprehension of the problem, review of possible solutions, definition of decision criteria and, finally, selection of the most satisfying solutions. Finally, a rigorous cost analysis has not been performed. While the general opinion is that computational analyses are inexpensive, a refined examination which quantifies direct and indirect costs should be carried out in order to assess more precisely the cost/benefit ratio.