Creating the foundation data for building a population grouping methodology – lessons learned at the Canadian Institute for Health Information (CIHI)

摘要: Background In April 2013, CIHI initiated a project to develop population grouping methodology that stratifies based on past clinical information and produces risk measures (i.e., costs for the prospective year). The includes all individuals in at given moment, including those who are not using health system. foundation data this include historical administrative utilization linkable individual level. Ideally, such multiple years of cover sectors full population. For example, if two used profiles, predictive indicators aim estimate need one year future, then three consecutive needed. It is also important assess stability over time determine optimal review period classification; as result, there additional project. One goal was be useful majority Canadian provinces. A province will benefit from any where personlevel available apply (vs. develop) methodology. Clinical only needed final methodology, so they consistently high quality across This rationale influenced choice incorporate into does have pan-Canadian coverage patient-level sectors. Some partial or no within some provinces (e.g., emergency department, home care, long-term drug prescriptions). provinces, but access patientlevel physician billing). Additionally, while may sufficient present, can limited. These issues included data.