Data-Driven Generation of Medical-Research Hypotheses in Cancer Patients

• Main Authors: Hsin-HsiungChang, 張信雄
• Other Authors: Jung-Hsien Chiang
• Format: Others
• Language: en_US
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/68002360762922984062

碩士 === 國立成功大學 === 醫學資訊研究所 === 104 === Taiwan Ministry of Health and Welfare data shows that catastrophic illness care accounts for about thirty percent of the total healthcare expenditure, wherein the cost of cancer care is the highest. Reducing the psychological burden of patients with catastrophic illnesses and the cost of national health insurance is pressing. Most studies do not explore the relationship between cancer and other catastrophic illnesses. Moreover, the most important part of medical research is the hypothesis. If we have a good hypothesis, we can design experiments and verify it. Therefore, this study attempts to identify the relationship between cancer and catastrophic illnesses. Then, we use those relationships as hypotheses in clinical medicine research. We hope this method can help physicians quickly and correctly find research hypotheses. We also want to identify the association between cancer and other catastrophic illnesses, to remind healthcare workers and cancer patients that cancer patients may suffer from other catastrophic illnesses, and to reduce national health insurance expenditures. This study was designed to capture data from the Taiwan National Health Insurance Database, using an association-rule method to identify the associations between cancer and other catastrophic illnesses, while consulting with a medical expert. We used these relationships as hypotheses in clinical medicine research, and finally verified the associations by cohort study. This experiment was divided into two parts. In the first part, we used the Apriori algorithm to find associations between cancer and other catastrophic illnesses. In the second part, we used these associations as medical-research hypotheses and designed cohort studies to verify them. The result showed that patients with renal cell cancer are more likely to suffer from dialysis than patients without renal cell cancer (Log-rank P 〈 .001). In this study, we proved that the association-rules method could help clinical physicians quickly and correctly obtain clinical-medicine hypotheses. Those patients with renal cell cancer and lung cancer are more likely to suffer from dialysis and long-term ventilator dependent respiratory failure. We reminded healthcare workers and cancer patients that cancer patients can easily suffer from other, subsequent catastrophic illnesses.