Workflow decisions

Process management, the next marketing-mix variable, refers to how Care-Rite Pharmacy incorporated the new Pharmacy Check-up Service into its existing practice. Workflow decisions were made regarding patient flow during the service. It was decided that the service would be by appointment only to minimize disruption of the other aspects of the practice. Schedules for pharmacists were created so that one pharmacist was responsible for clinic services while another covered the dispensing area. 

So what makes one system succeed and another fail This critical question was addressed in a review of 70 studies of decision support systems, which examined 15 different features of those systems (Kawamoto etal, 2005). Just under 70% of the systems showed positive results. There were four critical determinants of success. The most powerful factor was that the system automatically provided decision support as part of the clinician workflow. Decision support did not have to be provided by a computer for instance, in one study, diabetes care recommendations were attached to relevant paper records by support staff. Other determinants of success were that the system provided actual recommendations of what should be done, rather than simply an assessment or presentation of options, that the decision support was provided at the time the decision was being made and that the support was provided by a computer. In summary, a system will succeed if it is easy to use and provides clear recommendations at the time the clinician needs them. 

An important aspect in all pipelines is the evolution of the decision-making. As more data become available, the structure determination paths can be scrutinized thoroughly in order to increase the efficiency of the overall workflow. 

In Section 8.2, the aim of analysis is emphasized especially for the API (active pharmaceutical ingredient) and the drug product. The workflows and the rationale at major decision points during synthetic processing steps where HPLC can be applied in process development are elaborated upon. For example, a fast method is needed to monitor reaction conversion of two components. However, a more complex method would be needed for stability-indicating purposes where multiple degradation products, synthetic by-products, and excipient peaks need to be resolved from the active pharmaceutical ingredient. 

Eventually the workflow should be automated as much as possible into a computer-based decision support system (DSS) that aids the process of decision making when generating and using nontesting data for regulatory purposes. The DSS may prove invaluable as a means of promoting the regulatory use of in silico methods. 

A. Kumar, and J. Wainer, Meta Workflows as a Control and Coordination Mechanism for Exception Handling in Workflow Systems, Decision Support Systems, 40(1), 89-103 (2003). 

Liu, D.-R., Shen, M. Business-to-business workflow interoperation based on process-views. Decision Support Systems 38(3), 399-419 (2004)... 

Expert systems allow scientists to access, manage, and apply data and specialized knowledge from various disciplines to their own research. Exper Systeiiis in Chemisfr esearch explains the general scientific basis and computational principles behind expert systems and demonstrates how they can improve the efficiency of scientific workflows and support decision-making processes. 

A workflow is a network of well-described computational tasks that together accomplish a specific goal. The network defines the sequence of tasks in addition to decision points and resultant alternatives. Each task is described in terms of its input, output, and software dependencies. 

Clearly, information integration work, supported by ontologies, will be a key criterion for success in this area, but even with substantial scientific evidence, a team will not succeed if there is not sufficient coordination of team members and of the information that they find and/or produce. The problem here is to abstract and formalize the workflow, including decision points and timelines for deliverables for each member of the matrix, as well as coordination and dissemination of the information produced by each member of the team. This complex task is usually left to the team leader to solve however, it is an area in which ontologies can provide a level of abstraction and coordination that will help bind the team together. Work in this area is focused primarily in the Semantic Web or Semantic Grid space, which we discuss separately in this chapter. 

Define tasks and workflows Query the manufacturing administrator about the specifics of their business processes, map individual tasks to users across organizations, applications, and systems, and assist in developing specialized workflow models and decision support models. Then automatically configure the ERP workflow/task models, decision support invocations, and report formats. 

---