Implementing quality indicators using health insurance claims data linked to the hospital-based cancer registry.

Abstract

94 Background: To establish systematic monitoring of cancer care quality, we measured the quality of cancer care in several facilities through chart reviews by tumor registrars. However, this method required both extensive effort of and skills in registrars. To explore less-labor–intensive methods of measuring care quality, we assessed quality measurement using health insurance claims data linked to the Hospital Based Cancer Registry (HBCR). Methods: We previously developed 206 quality indicators (QIs) to assess cancer care processes in collaboration with clinical experts. Ten of these (stomach cancer, 1; colorectal cancer, 1; lung cancer, 2; breast cancer, 3; liver cancer, 1; and supportive care, 2) could be used for analyzing HBCR health insurance claims data. Patients treated at 7 designated cancer hospitals in Japan in 2010 were included. Their characteristics and tumor stages were obtained from HBCR, and processes of care administered to the patients in 2010–2011 were obtained from health insurance claims data. We calculated a score for each QI based on the proportion of patients receiving care among those eligible for QI. Results: Data of 4,785 patients were analyzed (stomach cancer, 1,181; patients with colorectal cancer, 1,077; lung cancer, 1,091; breast cancer, 1,184; and liver cancer, 252). Quality scores of essential laboratory tests were high; 91% patients underwent the HER2 test for invasive breast cancer and 95% underwent the liver function test using indocyanine green clearance before liver cancer surgery. However, indicator scores for adjuvant chemotherapy were relatively lower at only 59% for stomach cancer patients, 57% for colorectal cancer patients, and 56% for lung cancer patients receiving adjuvant chemotherapy. The supportive care scores had even more scope for improvement as only 43% patients received antiemetics for highly emetic chemotherapy and 66% patients received laxatives along with narcotics. Conclusions: These QIs can be implemented for health insurance claims data linked to HBCR and used to identify the potential target area for improvement. In future, such electronic systems will enable rapid cycles of quality measurement and feedback.