Mark Harrison

Background: Hospital readmissions significantly impact patient outcomes and healthcare costs globally. While prediction models are widely used to identify patients at high risk of readmissions, their discriminative ability often falls short of achieving adequate accuracy. Integrating patient-reported outcome measures (PROMs) may enhance these models by considering patient perspectives on their own health conditions.Objective: To investigate the contribution of PROMs to the performance of prediction models for hospital readmissions by (1) systematically reviewing the literature on readmission prediction models, (2) assessing the incremental value of incorporating PROMs on the performance of risk prediction models, and (3) exploring extra predictive value of PROMs across various clinical and geographic subgroups.Methods: A systematic review of readmission risk prediction models that included PROMs was conducted through searching electronic databases. A retrospective population-based cohort study was performed using data from the British Columbia Acute Inpatient Survey linked to administrative healthcare databases. The study cohort comprised patients discharged from acute care facilities between September 2016 and March 2017 who had completed PROMs. Subgroups were stratified by the presence of ambulatory care sensitive conditions and distance from major hospitals relative to patients’ residences. Prediction models were constructed using both statistical methods and machine learning techniques. Model performance was compared with and without PROMs data for the overall cohort and subgroups.Results: The systematic review highlighted the potential of PROMs in predicting hospital readmission, although the effectiveness varied depending on how PROMs data were collected and used in the prediction models. The analysis included 9,148 patients discharged from BC’s acute care hospitals, with PROMs completed between 26- and 60-days following discharge. While global summary of PROMs were influential predictors, the inclusion of PROM data led to only modest improvement in model performance. Subgroup analyses indicated the need for adjusting risk thresholds to accommodate specific clinical and demographic characteristics.Conclusion: Incorporating PROMs into prediction models for unplanned hospital readmission or death within one year modestly enhanced model performance. However, the marginal gains in predictive accuracy must be balanced against the benefits to patients and the costs associated with collecting these data.

View record

The uptake of biosimilars has been slower than hoped in many jurisdictions across the globe. A concern has been raised that switch from originator to biosimilar would be associated with increased health care resource utilization (HCRU) such as increased physician visits and hospitalizations. However, the signals for these concerns have been based on small samples and potentially biased methods. In this study, administrative data were analyzed from British Columbia (BC), Canada where in the past 5 years, policies have led to patients with inflammatory bowel disease (IBD), inflammatory joint disease (IJD), and inflammatory skin disease (ISD) being switched from an originator to a biosimilar to maintain insurance coverage.A literature review was conducted using Embase, Medline, and Web of Science to update existing systematic reviews. HCRU and cost from current literature were summarized. An empirical analysis was undertaken using population-based, administrative databases (Population Data BC) from 2015 to 2021 including individuals with IBD, IJD and ISD based on ICD-9 and -10 codes. The cohort included those using originator TNF-α blockers (adalimumab, etanercept, or infliximab) before the mandatory switch policy. Descriptive analysis of HCRU components (physician visits, hospital days, emergency visits, and other medications) was conducted. An interrupted time series was used to assess the pre- and post-switch trend by treatment and cohort on total HCRU related costs.The literature review showed no pattern in increase HCRU and cost following a switch policy. The analysis on BC population included 6326, 6205, and 2474 patients in the IBD, IJD, and ISD cohorts, respectively. Based on descriptive statistics and interrupted time series analysis, no important changes in physician visits, hospital admission, or emergency visits following the policy occurred for all three cohorts. Overall, no important changes in HCRU costs due to the switch were observed, but a historical trend in decreasing hospitalizations in IBD cohort was identified.This study suggests concerns about biosimilar switch policies creating increases in other HCRU is unfounded in IBD, IJD, and ISD patients. Policy makers in other jurisdictions can feel reassured that mandatory biosimilar switching polices should not lead to increases in the cost of other healthcare resources.

View record

BACKGROUND Despite representing less than 2% of prescriptions, biologics accounted for nearly three of every ten dollars spent on prescribed medicines in Canada in 2018. Similar to generics for small molecule drugs, biologic biosimilars are one potential way in which payers can reduce drug spending. In 2019, the government of British Columbia became the first in North America to mandate switching from reference biologics to biosimilars. While a number of other provinces have followed, the impact of these policies remains unclear. Therefore, this thesis examined the current state of biosimilars use in Canada.METHODS This dissertation focused on uptake of and spending on infliximab, etanercept, and insulin glargine using two primary data sources: (1) data from the IQVIA Canadian Drugstore and Hospital Purchases Audit representing all Canadian provinces except Newfoundland and Labrador, and (2) British Columbia health administrative data from Population Data BC. Interrupted time series analysis was used to quantify the results of two eras of biosimilars policies in British Columbia, including mandatory biosimilars use for new starters and subsequently mandated switching for all users, among individuals with inflammatory arthritis and psoriasis. RESULTS We found that prior to 2019 uptake of biosimilar infliximab, etanercept, and insulin glargine was low across Canada. Coinciding with the introduction of mandatory switching policies, there was a large increase in utilization thereafter in a number of provinces. In British Columbia, we determined that the introduction of mandatory switching among individuals with inflammatory arthritis and diabetes mellitus resulted in an increase in biosimilars utilization beyond what would have occurred if payers maintained new start policies only. CONCLUSION Government-mandated switching policies have the ability to greatly increase use of biosimilars even in the context of a multi-payer system. Although the ability to ascertain savings is limited due to the proprietary nature of drug pricing in Canada, the enhanced use of biosimilars will likely create a more favorable environment for price-based competition among pharmaceutical manufacturers. Future work should continue to examine the impact of mandatory switching on patients and prescribers as well as on the market for biologic drugs on a longer time horizon. 

View record

BackgroundA 2010 workforce survey revealed British Columbia was facing a shortage of rheumatologists and a consequent crisis of access to rheumatology care. Rheumatic diseases are chronic, and early intervention is crucial to prevent progression and mitigate systemic damage. Recognizing nurses may be able to perform aspects of rheumatology care and thereby “free up” rheumatologist time, the Ministry of Health introduced billing code G31060 to facilitate nurse-supported consultations for the “complex” rheumatology cases most in need of attention. The objective of this thesis is to evaluate the impact of introducing this new billing code and model of care on access to rheumatology care for the population of BC living with rheumatic disease.MethodsI conducted an interrupted time series analysis with a comparator using administrative health data on outpatient visits from Population Data BC. Patients with rheumatic diseases were identified using International Classification of Diseases codes and classified as those who received the intervention (i.e. nurse-supported rheumatologist care) or ‘status quo’ (i.e. rheumatologist care alone). Access was defined as 1) number of unique patients treated per month 2) number of service units billed per month. In sensitivity analyses I explored the impact of more restrictive definitions of intervention which required more “consistent” (at least once in every year) and “high-intensity” (at least 30 per year) billing of G31060.ResultsThe primary cohort included 128,726 patients with rheumatic disease, seen by 29 intervention and 17 comparator rheumatologists. No statistically significant effect change in level or trend of unique patients (pβ6=0.682 & pβ7=0.231) or service units (pβ6=0.744 & pβ7=0.419) attributable to the introduction of G31060 was detected in the primary analysis. Sensitivity analyses revealed statistically significant, increases in patients seen for rheumatologists billing “consistently” (62%) and with “high intensity” (168%) in April 2015 as compared to ‘status quo’. ConclusionThe introduction of G31060 does not appear to impact the number of service units billed per month, nor does it necessarily increase the number of patients seen. However, consistent and high-intensity users of G31060 appear to increase the number of unique patients seen per month.  

View record

Introduction: Currently, there are ongoing clinical trials for preventative treatments that aim to prevent and minimize the progression to RA in high risk individuals. However, preferences that drive people’s decision making in the context of a preventative treatment are unknown. With these clinical trials reporting their results within the next 2 years, this thesis aims to understand preferences of those who are at high risk of RA around preventative treatment and guide how these preferences can be best implemented in preventative treatment programs.Objectives:1) To identify important attributes for uptake of a preventative treatment program for those who are at high risk of RA, 2) To identify the value that is placed on these attributes through a discrete choice experiment (DCE), and 3) To predict the potential uptake of a preventative treatment options in those who are at high risk of RAMethods To determine the attributes that were important for the uptake of a preventative treatment program for those who are at high risk of RA, individuals with RA, first-degree relatives and rheumatologists were interviewed. These interviews were analyzed through a Framework Method. A DCE provided insight into whether those who are at high risk of RA would be willing to take preventative treatment. Results:The qualitative Framework analysis of patient, first-degree relative, and rheumatologist focus groups yielded five different attributes to be included in a DCE. Including the five treatment related attributes in the DCE demonstrated that first-degree relatives and RA patients preferred preventative treatments that had high risk reduction of RA, were orally administered, minor reversible side effects, moderate certainty in estimates, and were preferred by the health care provider. Predicted uptake of preventative treatments ranged from 51% to 92%, with oral methotrexate having the highest and infusion rituximab having the lowest.Conclusion:This thesis provides understanding around preferences, and subsequent trade-offs that an at risk individual might make when considering preventative treatment for RA. Through these trade-offs, the most important attributes of a preventative treatment program have been identified, and the likely uptake of potential preventative treatments have been estimated.

View record