ABSTRACT
Purpose@#In Korea, hospice palliative care (HPC) provision for cancer patients has increased recently.However, whether end of life (EoL) care practices have improved along with the developmentof HPC is unclear. We intended to investigate the changes in EoL care practices and theirassociation with HPC referral. @*Materials and Methods@#Retrospective medical record review of adult cancer patients who died at National CancerCenter Korea from 1 January 2009 to 31 December 2014 was performed. Changes of EoLpractices including chemotherapy within 2 weeks from death, death in intensive care unit(ICU), documentation of “do not resuscitate (DNR)” within 7 days from death and referral toHPC from 2009 to 2014 were analyzed as well as the association between referral to HPCand other practices. @*Results@#A total of 2,377 cases were included in the analysis. Between 2009 and 2014, referral toHPC increased and DNR documentation within 7 days from death decreased significantly.Cases for chemotherapy within 2 weeks from death and death in ICU didn’t change overthe study period. Patients referred to HPC were less likely to receive chemotherapy within2 weeks from death, die in ICU and document DNR within 7 days from death. @*Conclusion@#During the study period, EoL practices among cancer patients partly changed toward lessaggressive in our institution. HPC referral was associated with less aggressive cancer careat the EoL. Policies to promote EoL discussion are necessary to improve the EoL practicesof cancer patients.
ABSTRACT
PURPOSE: This study aimed to validate an automated calculating system developed for determining the adenoma detection rate (ADR). METHODS: To calculate the automated ADR, the data linking processes were as follows: (1) matching the selected colonoscopy results with the pathological results, (2) matching the polyp number from colonoscopy with that from pathology and confirming the histopathological results of each colonic polyp, and (3) confirming the histopathological results, especially the adenoma status of each colonic polyp. To verify the accuracy of the automated ADR calculating system, we manually calculated the ADR for 3 months through medical record review. Accuracy was calculated by measuring the error rate for each value. The cause of error was analyzed by additional order and chart review. RESULTS: After excluding 318 cases, 2,543 patients (1,351 men and 1,192 women; median age, 57.9 years) who underwent colonoscopy were included in this study. When the automated calculating system was used, polyps were found in 1,336 cases (52.6%) and adenomas were found in 1,003 cases (39.4%). When the manual calculating system was used, polyps were found in 1,327 cases (52.2%) and adenomas were found in 1,003 cases (39.4%). The accuracies of the polyp detection rate and ADR according to the automated calculating system were 99.3% and 100%, respectively. CONCLUSION: We developed a system to automatically calculate the ADR by extracting hospital electronic medical record results and verified that it provided satisfactory results. It may help to improve colonoscopy quality.
Subject(s)
Female , Humans , Male , Adenoma , Colon , Colonic Polyps , Colonoscopy , Colorectal Neoplasms , Electronic Health Records , Medical Records , Pathology , Polyps , Quality ImprovementABSTRACT
OBJECTIVES: New methods for obtaining appropriate information for users have been attempted with the development of information technology and the Internet. Among such methods, the demand for systems and services that can improve patient satisfaction has increased in hospital care environments. METHODS: In this paper, we proposed the Hospital Exam Reservation System (HERS), which uses the data mining method. First, we focused on carrying clinical exam data and finding the optimal schedule for generating rules using the multi-examination pattern-mining algorithm. Then, HERS was applied by a rule master and recommending system with an exam log. Finally, HERS was designed as a user-friendly interface. RESULTS: HERS has been applied at the National Cancer Center in Korea since June 2014. As the number of scheduled exams increased, the time required to schedule more than a single condition decreased (from 398.67% to 168.67% and from 448.49% to 188.49%; p < 0.0001). As the number of tests increased, the difference between HERS and non-HERS increased (from 0.18 days to 0.81 days). CONCLUSIONS: It was possible to expand the efficiency of HERS studies using mining technology in not only exam reservations, but also the medical environment. The proposed system based on doctor prescription removes exams that were not executed in order to improve recommendation accuracy. In addition, we expect HERS to become an effective system in various medical environments.