TY - JOUR TI - Clinical decision support system as a risk assessment tool to aid in earlier diagnosis of pancreatic cancer DO - https://doi.org/doi:10.7282/T3CJ8G9W PY - 2015 AB - Background: Pancreatic cancer is the most aggressive and the most deadly type of cancer. It is the fourth leading cause of cancer-related death with a dismal prognosis because most diagnoses are made during the advanced stage of the disease. Although, several studies have researched methods for early diagnosis and treatment of pancreatic cancer, the use of clinical decision support systems as a precursor for early diagnosis has not been extensively studied. Clinical decision support systems are computer programs that could be used to aid clinicians or others in their decision-making at the point-of-care of their patients. It can also be designed to aid the general population in evaluating their risk of developing a particular illness or other facets of health care related issues. Objectives: To develop a clinical decision support system that can identify pancreatic cancer risk levels in individuals, and also provide recommendations and alerts tailored towards each individual’s situation in order that pancreatic cancer diagnoses are made earlier than later. Design: A multi-method approach using quasi-experimental and quantitative study. Methods: This study posed three hypotheses that will be tested using the developed clinical decision support system. First, extensive review of literature and existing data were conducted in addition to data analysis of the Nationwide Inpatient sample to gather clinical evidence and facts about pancreatic cancer risk factors. The knowledge gathered and probabilistic basis were used to define the variables and parameters and their respective weighted scores. Five weight groups of 100, 60, 30, 15, and 5 were created with “100” as maximum risk and “5’ as minimum risk. Fourteen common risk factors were used and within these risk factors, 87 parameters were defined and categorized into one of the five weight groups. The variables used include; demographic information, smoking history, history of cancer, family history of cancer, and other risk factors of pancreatic cancer. Using IF/THEN rules, logical order of steps were created. With the inference engine of the clinical decision support system, the system is able to compute the sum of the weighted scores and provide immediate feedback in the form of total risk factor score at the end of the test. The system also displays the scoring chart for the three risk levels; high-risk, moderate-risk and low-risk at the end of the test for users to identify which one of the three levels they belong. Results: Twelve case scenarios were used to test the validity and reliability of the system. Among the 12 cases, nine were diagnosed with pancreatic cancer, one was a healthy individual with no diagnosis of any sort and two were diagnosed with other health conditions. The results were as follows; two low risk patients, three moderate risk patients and seven high risk patients. The healthy individual turned out to be high risk. In some cases, recommendations and alerts were generated for patients to seek immediate medical attention, screen for pancreatic cancer or get a scan of the pancreas. Conclusion: The results show that it is possible to develop a system that can identify high risk individuals for pancreatic cancer. The impact the system will have on patient care and whether the system can reduce the number of misdiagnoses, delayed diagnoses, or lead to earlier diagnoses of pancreatic cancer is uncertain. Further studies will need to be conducted to expand the knowledge in using clinical decision support system for pancreatic cancer risk assessment, and also to identify the precise parameters of the three risk level-scores. KW - Biomedical Informatics KW - Pancreas--Cancer--Diagnosis KW - Cancer--Risk factors LA - eng ER -