Assessing the impacts of the Quebec primary care enrolment policies on patient-physician affiliation

Patients’ relationships with, and affiliation to, primary care physicians can influence patients’ care experience, continuity of care and health outcomes. Many Canadian provinces are actively trying to increase the number of patients who have a regular medical doctor through patient-physician enrolment policies. In Quebec, two enrolment policies were introduced, one for the chronically ill and elderly in 2003 and one for the general population in 2009. While the association between having a regular primary care provider and better health outcomes is well established, we know less about the impact of enrolment policies on attributes of the patient-physician relationship.In this thesis, the primary goal was to evaluate the impact of primary care enrolment policies on measurable attributes of the patient-physician relationship. I focused on measuring the impacts of Quebec enrolment policies on aspects of patient-physician relationships that could be plausibly affected by enrolment and measured using health administrative data (HAD). This included the proportion of visits with the physician seen most that year and whether patients self-reported having a regular medical doctor. The first is directly available in HAD while the latter is not. In manuscript 1, I explore predictive modeling methods that can be used to predict patients’ report of having a RMD. The Canadian Community Health Survey has respondents’ answers to whether they have a RMD, but HAD does not. With linked data, I built predictive models for responses using data only available in the HAD. I used Random forests, a machine learning technique, in addition to conventional statistical models. In manuscript 2, I compare the prediction performance of the different methods identified in manuscript 1 for predicting whether patients have a RMD in HAD. This includes comparisons to the usual provider continuity index (UPC), the conventional measure of patient affiliation in HAD. Once I established that HAD could be used to accurately predict self-reporting having a RMD, I identified which predictors in the models were most important. By identifying the most important predictors, I was able to create a simple index that is highly predictive and can easily be applied by other health services researchers. This new measure is the Reporting a Regular Medical Doctor Index (RRMD). In manuscript 3, I evaluate the impacts of the Quebec enrolment policies on patient affiliation to a primary care provider, using both UPC and the RRMD index. I use a difference-in-difference analysis to evaluate the 2003 vulnerable enrolment policy and an interrupted-time-series to evaluate the 2009 general enrolment policy. For both policy evaluations, I found no evidence that the enrolment policies impacted patient-physician affiliation.The findings presented in this work offers valuable evidence on the effect of enrolment policies on patient-physician affiliation that can be used to inform future interventions aimed at increasing patient-provider affiliation. The new RRMD measure that predicts whether a person reports having a RMD using only HAD will be useful for both researchers and government institutions to evaluate the impacts of policy interventions on this health systems indicator

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.