Time-Series Design to Explore the Variation between the Statistical Significant Difference and the Clinical Significant Difference---The Quality of Life of Ovarian Cancer Patients as Example

The assessment of quality of life for cancer patients is a popular issue in past decade. Various levels of quality of life have been reported by cancer patient during illness, however, limited reports discuss this issue in the Taiwanese ovarian cancer patients. In the analysis of long-term quality of life data, “statistical significant difference＂ and “clinical significant difference＂ mean two different concepts. Our previous data from oral cancer patients showed that more items reached the “clinical significant difference＂ than “statistical significant difference＂. Also, ovarian cancer is named as “silence killer＂ due to the difficulty of early detection. There is a valid and brief screening tool that has been established in western world. Therefore, the purposes of this prospective longitudinal follow-up study include: 1) to explore the changes of quality of life reported by the ovarian cancer women during the 3-year study period; 2). to identify the “statistical significant difference” and “clinical significant difference” on quality of life reported by ovarian cancer women during the 3-year research period; 3). to explore the variation between the “statistical significant difference” and the “clinical significant difference” on quality of life; and 4). to verify the usability of the Chinese version of “Ovarian Cancer Symptom Index, OCSI”. During the 08-01-2009~07-31-2012, a time-series design with questionnaires distribution will be conducted in a medical center located in the northern part of Taiwan. It includes two sections. The first section is to collect 300 samples from the gynecological OPD or words. Then, ask them to fill up questionnaires ten times during the 3-year research period. The second section is to invite 1000 subjects from the gynecological OPD to fill up questionnaires as OCSI. Its major focus is to verify the usage of OCSI Returned questionnaires will be managed by SPSS-PC statistical program. The basic information of each variable will be presented by mean, standard deviation, percentage, and frequency. The longitudinal data will be analyzed by: GEE; effect size, raw change, percentage change scores, and multiple regressions. For the data from screening test, methods as intra-rater reliability, inter-rater reliability, sensitivity, specificity, false negative, false positive, predictive value will be performed. The expected outcomes are: 1). to establish a quality of life data bank for ovarian cancer patients that included both the “statistical significant difference” and the “clinical significant difference”, and 2). to establish a valid and easy screening tool of ovarian cancer. All these outcomes can provide the health profession aa reference to promote their quality of care.

Project ID:PC9902-1654
External Project ID:NSC98-2314-B182-058-MY2