Research On Computer-Aided Clinical Pathway Management System Based On Task Scheduling And Woekflow Technology

Clinical Pathways (CP) are structured, multi-displinary plans of care designed to implement clinical guidelines and protocols. They are designed to support clinical care, clinical and non-clinical resource management, clinical audit and also financial management. CP have been applied in many hospitals. However, most CP implementations are paper-based, in which all detailed guidance (treatments, interventions etc…) with a specific condition over a given period, including progress and outcomes details are printed and recorded in paper. There are some defects in paper-based CP which impair the effect of CP, e.g. scatted information in different papers, uneasiness in collecting data, less control over implementation, overload of doctors, not well coordination among nurses.In view of this, we develop a Computer-Aided CP Management System (CACPMS). The system integrates task scheduling algorithm and workflow engine technology, which greatly improve efficieny and intensify control over CP implementation. Firstly, we analyze the basic needs as a CP management system. The main functions are designing of CP rules and recording of CP implementation progress. Modules for CP designers, doctors, nurses, patients and task schedulers are set to co-work in the whole care process. Clinical data during the process are recorded in the system for data mining.Secondly, to improve efficiency in implementation, we develop a task scheduling model in the system by which key clinical tasks are scheduled with the goal of assuring healthcare quality, improving efficiency and balancing workload. Clinical Tasks are classified into general tasks with fixed executing time and scheduling tasks with unfixed executing time, whose parameters such as roles, time and order are set. A Genetic Algorithm (GA) is designed to solve the tasks scheduling model, in which five-sector seven-digit real numbers are used to denote scheduling tasks;greedy rules and general tasks inserting rules are used for decoding. Other strategies such as elite reservation and self-adaptor operators are applied to optimize efficiency of computing.Thirdly, to intensify control over implementation, we study the workflow driven mechanism in CP management system, modeling the implementation process of CP as a workflow and developing workflow engine based on Relational Database. Workflow Enginee builds clinical tasks into activity instances, distributes resources for them by task scheduling algorithm, guides flow of information according to router rules and make real-time monitoring as time proceeds.The Computer-Aided CP Management System finally developed in this study realizes following functions: 1) designing and maintaining of multi CP; 2) retrieving and recording clinical data over CP implementation; 3) patient information quering and feedback; 4) scheduling key clinical tasks with the goal of optimizing resource allocation; 5) workflow engine to guide flow and monitor tasks status.