Implementation of the COVID-19 antiviral therapy Nirmatrelvir/Ritonavir across Canada in 2022

Abstract

Background: The COVID-19 antiviral Nirmatrelvir/Ritonavir (Paxlovid^TM, N/R) was approved for use in Canada in January 2022, with the Government of Canada assuming a procurement role and provinces, territories, and federal departments implementing usage within their respective healthcare systems. The objective of this analysis is to describe how N/R was implemented across various jurisdictions in the first six months after it was available for use and identify promising implementation practices.

Methods: Fourteen semi-structured discussions in small group settings were conducted with jurisdictional representatives involved in the implementation of N/R. A descriptive analysis of the eligibility criteria and service delivery model was conducted. A thematic analysis using the Consolidated Framework for Implementation Research and cluster analysis of the codes were then undertaken on NVivo 12 to identify key themes.

Results: Overall, the eligibility criteria were similar across jurisdictions, and three types of service delivery models were identified. Ten main themes emerged as facilitators and eight as challenges to the implementation. Partnership, collaboration, communication and flexibility were among the facilitators identified, while the complexity of the intervention (e.g., drug-drug interactions), perceived evidence gaps in effectiveness by prescribers, and resource limitations were identified as key implementation challenges.

Conclusion: While there were jurisdictional variations in the implementation of N/R, communication and collaboration, and the availability of rapid testing for COVID-19 emerged as key facilitators. Drug-drug interactions, resource pressures and limited evidence were some of the key challenges. Overall, these facilitators and challenges were similar across jurisdictions and may help inform future therapeutic implementation plans for pandemic preparedness.

Introduction

Nirmatrelvir/Ritonavir (Paxlovid^TM, N/R) is an oral antiviral for treatment of SARS-CoV-2 in adults with mild to moderate symptoms at high-risk of progressing to severe disease or death. On January 17, 2022, Health Canada authorized its use following the interim results of the Phase 2/3 double-blind placebo-controlled EPIC-HR (Evaluation of Protease Inhibition for COVID-19 in High-Risk Patients) trial demonstrating reduction in COVID-19-related hospitalization and all-cause mortality ^{Footnote 1Footnote 2}. Nirmatrelvir/Ritonavir was the first oral antiviral approved to treat COVID-19 in Canada. The Government of Canada assumed the role of procurement, with jurisdictions responsible for delivering the treatment to their populations. Given the unprecedented context of the fast-changing COVID-19 pandemic and the novel role of procurement, the Public Health Agency of Canada (PHAC) developed an evaluation framework in conjunction with jurisdictional stakeholders. One of the components of the framework aimed to address questions on best practices in light of the limited experience with this new therapeutic option, and to inform implementation, which this study focuses on. The evaluation of the implementation aimed to describe how the rollout of N/R took place across Canada and to identify the facilitators and challenges associated with its implementation.

Evaluation objectives

The overarching aim of this evaluation was to answer the following questions:

1. How has N/R been administered across Canada?
2. What are the most promising strategies to deliver therapeutics in outpatient settings?

Methods

Adopting a qualitative methodology, this study employed semi-structured group discussions as the primary method of data collection. Informed by the Donabedian framework ^{Footnote 3}, the Consolidated Framework for Implementation Research (CFIR) ^{Footnote 4}, the Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) framework ^{Footnote 5}, and behavioural science theory models ^{Footnote 6Footnote 7}, a structured questionnaire guided the discussions, covering key topics such as eligibility, service delivery, communication, training, and procurement experiences.

Participants were identified and recruited from various federal, provincial, and territorial (FPT) working groups using a snowball sampling technique, ensuring a diverse representation that included managers and healthcare professionals in COVID-19 therapeutics planning. Ultimately, of all jurisdictions invited, 12 provinces and territories and two federal departments participated.

One-hour interviews were conducted, and each session was recorded, transcribed, and summarized. These summaries, validated by participants, served as the foundation for descriptive and thematic analyses guided by the CFIR. The approach primarily followed a deductive method aligning with CFIR domains and constructs, while remaining open to inductive additions to the coding scheme as necessary. All factors of the CFIR were included in the initial deductive coding scheme. Each coded passage was further assigned a sentiment indicating the direction of the influencing factor (i.e., barrier or facilitator). Passages detailing the participating jurisdictions' eligibility criteria and service delivery models were not coded but summarized.

The CFIR is a theoretical framework developed to systematically explore the intricate factors influencing the successful implementation of innovations in various organizational contexts. The CFIR's five domains examine critical elements, including intervention characteristics, outer and inner organizational settings, individual traits, and the implementation process. Within these domains, there are 39 CFIR constructs and subconstructs, representing the evidence-based factors most likely to impact the implementation of interventions. The CFIR was used to code the data and organize emerging themes post-data collection. For the analysis, the jurisdiction responsible for the implementation was the reference unit. Nirmatrelvir/Ritonavir was coded as the innovation; the provincial, territorial, and federal healthcare organizations were coded as the inner setting; and the federal government and any other external institution as the outer setting.

The data analysis process involved initial coding through NVivo 12 ^{Footnote 8} using the CFIR. This was followed by a cluster analysis to uncover patterns among the coded items based on their co-occurrence (Pearson coefficient) and, finally, matrix coding queries for constructs that were coded as positive (facilitator) or negative (challenge). Thematic analysis, inspired by Guest and Mclelan ^{Footnote 9}, was conducted to identify overarching themes when numerous salient themes emerged from the dataset. The findings were synthesized through a multi-step process, beginning with the identification of overarching themes through cluster analysis. Sentiments analysis was used to distinguish between facilitators and barriers, and major and minor themes were developed based on recognition of patterns within the data. The analysis included triangulation, member checking, and inter-observer reliability to bolster the credibility and transferability of the findings. Iterative rounds of analysis and refinement ensured a nuanced understanding of the complex implementation process.

Ethics approval

The PHAC policy on research activities was followed, however, a consultation with the Research Ethics Board was not required since the implementation evaluation is within PHAC's standard practices of assessing its programs. Consultations took place with the Privacy Management Division to ensure any personal information that might be disclosed during the evaluation was handled as per federal regulations and departmental policies.

Results

Two main phases to the rollout of N/R were identified during the interviews: the first, at the beginning of the rollout, when the supply of N/R was limited; and the second characterized by increased and stable supply. These phases directly impacted the decision-making processes and access of N/R for Canadians.

Eligibility criteria

At the start of the rollout, jurisdictions based their eligibility criteria for treatment with N/R on PHAC and the Canadian Agency for Drugs and Technology in Health (CADTH) guidance ^{Footnote 10}, the EPIC-HR trial results ^{Footnote 2}, the product monograph ^{Footnote 11}, and advice from expert advisory committees. Given the limited supply context and short lead time from drug authorization to implementation, prioritization of drug usage focused on individuals at highest risk of severe outcomes, including older, under- or unvaccinated individuals, Indigenous peoples, and those with immunosuppression, specific risk factors, or comorbidities (e.g., BMI ≥30, diabetes, lung and cardiovascular diseases). There was inter-jurisdictional variability within these criteria in regard to age thresholds for eligibility in conjunction with vaccination status and comorbidities, the number of comorbidities required for eligibility (one to three), and the comorbidities included (e.g., smoking, hypertension, chronic kidney disease).

As supply increased, most jurisdictions expanded their eligibility criteria, with a couple transitioning from defined criteria to guidance for prescribers. Changes included lowering age thresholds, inclusion of vaccinated populations, refinement of the definition of vaccination status (e.g., booster doses), and lists of comorbidities. Factors driving these changes included increased supply, new research findings (e.g., EPIC-SR trial results ^{Footnote 12}), evidence of a reduction in the effectiveness of neutralizing monoclonal antibodies (e.g., sotrovimab) against the new variants of concerns, and evolving epidemiology studies within the jurisdictions.

Service delivery models

The service delivery models varied between jurisdictions due to differences in health system infrastructure, resulting in three main models of service delivery: centralized, decentralized, and mixed, which combines both (see Figure 1). Initially, most jurisdictions relied on a centralized model, enabling more oversight on distribution, which was influenced by limited supply, logistical considerations, and limited evidence to support expanded usage. However, as supply increased with time, many transitioned to a decentralized or mixed model.

Figure 1: Descriptive text

This is a bar graph showing the distribution of service delivery models at the start of the implementation when compared to at the time of the evaluation. At the start of the implementation, there were two mixed, eight centralized, and three decentralized service delivery models. At the time of the evaluation there were five mixed, two centralized, and six decentralized service delivery models.

Authorized prescribers varied based on jurisdictional regulations, with physicians, nurse practitioners, and pharmacists authorized to prescribe N/R. Several jurisdictions amended regulations to expand the pool of designated healthcare providers authorized to prescribe and dispense the antiviral. At the time of the data collection, five jurisdictions permitted prescriptions by pharmacists, while one allowed community nurses to prescribe in an expanded role (the numbers have since increased). The expansion of access points and authorized prescribers was noted to have increased prescribing and access to N/R by some jurisdictions.

Thematic analysis

Based on the analysis of code relationships and their frequency of co-occurrence, seven essential themes emerged, defining the key drivers of N/R implementation in Canada (Table 1). The examination of barriers and facilitators to N/R implementation in Canada uncovered several key factors common across jurisdictions. These are summarized in Table 2, organized by CFIR domain.

Adaptability to the rapidly evolving context and collaboration among stakeholders emerged as crucial facilitators, evident across individual (characteristics and roles), implementation process, and inner and outer setting domains of the CFIR. The availability of guidance documents, provided by CADTH ^{Footnote 10}, PHAC and expert advisory committees, at the onset of the implementation coupled with its adaptation to jurisdictional context (e.g., prioritization of equity-seeking groups and high-risk populations) for the creation of screening and reporting tools enabled a streamlined process, facilitating efficient patient prioritization, assessment, and, in some cases, follow-up. Multi-jurisdictional and interdisciplinary collaboration facilitated the development of improved processes, aided by pre-existing systems for information dissemination, access to testing and screening tools, and timely drug dispensing. Comprehensive training, continued learning, and guidance materials for prescribers contributed significantly to the intervention's success, disseminated through various pre-existing and new channels, such as healthcare professional associations, academic institutions, and even through informal messages on a commonly used internal messaging application. The availability and admissibility of alternate testing modalities (i.e., rapid antigen and rapid molecular point-of-care tests) were instrumental in streamlining the rollout in the later stages, as well as ensuring access to N/R in remote and rural communities within the five-day eligibility window. Policies expanding the authority of pharmacists and community nurses to prescribe and administer N/R were reported as increasing uptake. Lastly, equity considerations informed the implementation of components/activities aiming to prioritize high-risk populations (e.g., long-term care residents, Indigenous populations, and racialized and marginalized individuals).

Elements contributing to barriers in N/R implementation in Canada were mainly observed across the individual (characteristic), implementation process, innovation, and inner and outer setting domains of the CFIR. The perceived lack of benefits and complexity of prescribing the drug, attributed to limited evidence and numerous drug-drug interactions, influenced attitudes towards its adoption. The short lead time and global supply constraint amidst a pandemic posed challenges for the health services infrastructure, particularly in establishing necessary administration processes, IT and work infrastructures, for timely identification of eligible and best candidate patients, as well as timely delivery of the therapy. Setting up systems for identifying positive cases through PCR testing was challenging initially, but improved with widespread use of rapid antigen tests. As it relates to procurement, some jurisdictions expressed the need for enabling centralized storage capacity and federal distribution of therapeutics, alongside expanding the FPT pandemic coordination processes to include clinical discussions. The global supply constraint also influenced communication strategies and service delivery models initially, until the Canadian supply stabilized. Infrastructural limitations and insufficient human resources, including healthcare providers, further impacted the N/R rollout in Canada. A patient trajectory to access N/R in Canada along with possible strategies to streamline the process was created base on the findings from the thematic analysis (see Figure 2).

Figure 2: Descriptive text

Figure showing the patient trajectory to access Nirmatrelvir/Ritonavir (N/R) and their dimensions including possible strategies. 1) Procurement/supply: global and domestic supply, allocations, distribution systems, and cost (cost was not assessed), possible strategies are central storage and longer term planning; 2) COVID-19 symptoms appear (not assessed): health seeking behaviour, knowledge attitudes and behaviours (KAB) regarding COVID-19, possible strategies are increased communication, community engagement, KAB studies, and strategies based in documented perceptions; 3) COVID-19 testing: risk perception and health seeking behaviour (not assessed) and testing criteria and testing accessibility, possible strategies are facilitated testing and ongoing availability; 4) Assessment for N/R and prescription: Eligibility criteria (supply, evidence, demand), 5-day window, drug-drug interactions, service delivery model, health system resources, and accessibility (accessibility was partially explored), possible strategies are adaptive trials, consolidate training, effective tools, resource stabilization, and streamlined service delivery models; 5) N/R dispensing: service delivery model, health system resources, and accessibility (accessibility was partially explored), possible strategies include data on accessibility gaps, KAB studies, adherence support strategies, and measures to increase accessibility; and 6) N/R administration (not assessed): Risk perception and side effects. Accessibility dimensions that were partially explored and their possible strategies in parenthesis include: approachability (e.g. language, complexity, and travel time), acceptability (e.g. cultural safety), availability and accommodation (e.g., wait times, 24h/7 day services, affordability (e.g., no cost), and appropriateness.

Discussion

This qualitative study sought to examine the implementation of N/R within the first six months of its approval in Canada during the COVID-19 pandemic. Its aim was to identify key facilitators and challenges encountered during this implementation process and derive lessons applicable to future outpatient therapeutic implementations, thereby informing pandemic preparedness readiness in the Canadian context. While there was jurisdictional variability in eligibility criteria and service delivery models, owing to differences in health system infrastructure, the findings of this evaluation have highlighted the importance of communication, partnership and collaboration, and flexibility and adaptability of policies and implementation processes to ensure equitable access to COVID-19 treatments. In effect, jurisdictions shared knowledge and resources, from guidance documents to testing and screening mechanisms, with one another to facilitate the rollout of N/R. The context in which N/R was implemented, short lead time to implementation amidst a global shortage, along with elements directly related to the therapeutic (e.g., complexity of the therapeutics due to numerous drug-drug interactions and limited evidence, and short window of eligibility), as well as infrastructural and workforce limitations, posed challenges to its implementation.

Through our analysis, 10 themes were delineated related to facilitators and eight themes related to challenges (see Table 2). While the factors affecting uptake and barriers vary by jurisdiction based on healthcare system structure, population characteristics, and data analysis capacity, synthesizing insights from participants and considering challenges and successful approaches, strategies have been proposed to address some of the barriers uncovered in this study (Figure 2). These strategies span across the N/R supply chain from procurement through dispensing or administration to patients. Healthcare system capacity notably affected the N/R rollout, with primary care capacity being a critical factor raised by evaluation participants. While some challenges are systemic and complex, actionable strategies could include consolidating provider training, streamlining patient pathways, and ensuring continued access to rapid testing for patients eligible for treatment. Addressing cost barriers by providing free testing and treatments is crucial, especially since COVID-19 impacts are not uniformly distributed across Canadian populations ^{Footnote 13}. Although variations in implementation are to be expected, given the wide breadth of the population, with each jurisdiction facing some unique challenges. Furthermore, lessons learned from the implementation of N/R can help guide future therapeutics implementation efforts and inform ongoing federal planning to bolster Canada's readiness for public health threats as Canada updates its pandemic preparedness plan ^{Footnote 14}. These can inform logistical planning of therapeutics implementation (e.g., testing, storage, and surveillance of distribution and dispensing), as well as the facilitation of FPT and Indigenous collaboration during the decision-making processes to ensure equitable access and distribution of therapeutics in future health emergency situations.

Limitations

There was a high participation level, with almost all jurisdictions participating in the evaluation. The participants in the discussion session were diverse and included individuals that contributed to different capacities (i.e., healthcare professionals and managers). As such, experiences from across the country were captured, giving a broad understanding of the implementation processes of N/R. There are, however, potential limitations to the evaluation. the patient perspectives and perspectives of other more local/regional components of the healthcare system were not explored, and any information obtained was from the participants' experiences. The potential dynamics between participants cannot be ruled as discussion sessions were held in groups for each jurisdiction. As well, although it was communicated that the results of the evaluation would be anonymized and the acquired data handled to ensure participant privacy, the potential for desirability bias remains, as the discussion sessions were led by PHAC employees. There is also the potential for interviewees' recall bias, as the evaluation covered the experiences in the first six months of the rollout. Furthermore, while the CFIR offers great insight into the implementation processes, the complexity and scope of its 39 constructs may pose a limitation to its generalizability and consistency of its application; these limitations have been mitigated by defining the scope of the settings to ease with reproducibility.

Conclusion

Patients' health-seeking behaviours have been found to be influenced by their knowledge, attitudes, and beliefs ^{Footnote 6Footnote 7} as well as accessibility to testing and treatment services. While this evaluation did not assess knowledge, attitudes and behaviours for patients and providers, future work in this area would further inform therapeutic implementation efforts and help understand and address some of the challenges identified in this study.

The evaluation assessed how N/R was administered in Canada in the first six months of the implementation, identifying pressure points and considerations for the initial stages of a therapeutic rollout. While jurisdictions have since modified their N/R programs, the lessons learned remain valuable for future therapeutic rollout in the context of a public health threat.

Authors' statement

• AW — Project management, conceptualization, methodology, leading discussion sessions, data analysis, interpretation, writing–original draft, writing–review & editing
• CT — Conceptualization, methodology, leading discussion sessions, data analysis, interpretation, writing–original draft, writing–review & editing
• NS — Supervision, conceptualization, methodology, leading discussion sessions, data interpretation, writing–review & editing

Competing interests

NS had some previous interactions with some of the participants of this study as part of other Public Health Agency of Canada activities.

ORCID numbers

• Aklile Workneh — 0009-0009-7098-6622
• Camilia Thieba — 0000-0002-7444-630X

Acknowledgements

We would like to thank all the representatives from the participating provinces, territories, and federal departments as their contributions and collaboration made this project possible.

Funding

This work was supported by the Public Health Agency of Canada. No additional funding was provided to the Public Health Agency of Canada to conduct this evaluation, nor was funding provided from the Public Health Agency of Canada to participating jurisdictions. All evaluation work was conducted using existing resources.

• Previous Page
• Table of contents
• Next page