FormalPara Key Points for Decision Makers
Table 1
Full size table

1 Introduction

Canada is among the highest spenders on health care in the Organisation for Economic Co-operation and Development (OECD) [1, 2]. In Canada, public insurance covers only 43% of prescription drug cost, while the remainder is paid by private payers (35%) or patients (out of pocket). In the public sector, provincial/territorial programs and federal direct drug subsidy programs are the main payers. In the private sector, payers include private health insurance and either households or individuals paying out of pocket [1, 2].

1.1 Drug Pricing and Reimbursement in Canada

In Canada, to obtain public reimbursement, cost-effectiveness reports are prepared by manufacturers in accordance with the Canadian Agency for Drugs and Technologies in Health (CADTH) economic evaluation guidelines [3] (submitted to the Common Drug Review (CDR), the pan-Canadian Oncology Drug Review (pCODR) [4]) and/or those of the Institut national d’excellence en santé et en services sociaux (INESSS) in Québec [5]. Drug submissions that receive a positive listing recommendation are submitted to the pan-Canadian Pharmaceutical Alliance (pCPA) for price negotiations (for the provincial price negotiation including Quebec) [6]. All provinces participating in the pCPA process require the submission of a budget impact analysis (BIA) that applies to their jurisdiction. The federally operated drug programs [i.e., the Non-Insured Health Benefits Program (NIHBP)] and private payers also require BIA reports for new drug submissions.

For patented drugs, since 1988, the Patented Medicine Prices Review Board (PMPRB) regulates and defines the ceiling price, i.e., the non-excessive price. Figure 1 illustrates the PMPRB as part of Canada’s pharmaceutical regulatory and reimbursement system. In 2005, the PMPRB initiated the development of the Canadian BIA guidelines on behalf of the National Prescription Drug Utilization Information System, and these were published in 2007 [7, 8]. The guidelines were initially developed to provide a standard for BIA accompanying new drug submissions to public drug plans. This was supplemented with an interactive Excel-based template to address provincial differences in drug regulations (e.g., drug prices, markups, professional fees, co-payments, discounts and cost analyses). However, over the last few years, some federal, provincial and territorial (F/T/P) drug plans have updated their specific BIA requirements [9,10,11,12]. The present study will provide a most up-to-date list of recommendations which could be applied towards a revision of the Canadian BIA guidelines and with the perspective of meeting the BIA requirements for new drug submissions to either public or private drug plans in Canada. We used the results of the present study and our previously published literature review [13] to create a comprehensive list of recommendations relating to three key elements of designing a BIA (i.e., analytical model structure, input and data sources, and reporting format). The two systematic reviews were complimentary with regard to drafting a consultation proposal for obtaining the Canadian stakeholders’ feedback on the BIA recommendations. A standard abstraction form was developed for extracting data from BIA guidelines by Foroutan et al. [13] and was applied for abstracting data from the Canadian public and private plans’ BIA guidelines in the current study. The first literature review [13] focused on national and transnational BIA guidelines including those of Australia [14], CanadaFootnote 1 [7], the United Kingdom (UK) [15], Belgium [16], Ireland [17], France [18], Poland [19], Brazil [20] and the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) [21], whereas in the present study, we performed a comparative review of the Canadian public, private and PMPRB 2007 BIA guidelines.

Fig. 1
figure 1

Drug approval and reimbursement process in Canada. CADTH Canadian Agency for Drugs and Technologies in Health, CDR Common Drug Review, INESSS Institut national d’excellence en santé et en services sociaux, pCODR pan-Canadian Oncology Drug Review, pCPA pan-Canadian Pharmaceutical Alliance, PMPRB Patented Medicine Prices Review Board, R&D research and development

Full size image

2 Methods

The websites of the CADTH (i.e., CDR and pCODR), pCPA, Canadian federal, provincial and territorial (F/P/T) drug plans (public drug plans which participate in pCPA negotiations) and five private payers (chosen from a list of private payers for drug submissions in a consulting company database) were manually searched (May 2018) for BIA guidelines, BIA guidance documents or (Excel-based) templates for new drug submissions. The first author also contacted private drug plans to get their BIA templates for new drug submissions. A total of 14 F/P/T public drug benefit programs (including Quebec), pCODR (CADTH) and five private payers’ websites were searched (Tables 1 and 2). Where guidelines were available in both English and French, we used the English version for our review. When a BIA guideline was updated, we only included the latest version of the BIA guideline in order to avoid duplication in data abstraction. In the present study, we used our standard Excel-based data abstraction form—which was developed in our previous peer-reviewed systematic literature review of national and transnational BIA guidelines [13]—for data abstraction. Then we highlighted similarities and differences between recommendations related to the BIA key elements provided by the Canadian PMPRB and F/P/T BIA guidelines or templates.

Table 1 Summary of the searched F/P/T drug plans and their BIA guidelines or templates published on their websites
Full size table
Table 2 List of Canadian private health insurance companies included in this review
Full size table

3 Results

The BIA guidelines for Canadian publicly funded drug programs were from Alberta [11], Ontario [9], Manitoba [10], Quebec [12], British ColumbiaFootnote 2 and the CADTH (pCODR) [22] (see Table 1 and Appendix 1). Only one BIA checklist for new drug submissions was found for private payers (Table 2). Some private payers (e.g., Green Shield and TELUS Health Benefits and Payment Solutions) in Canada use the Academy of Managed Care Pharmacy (AMCP) document (United States) [23] as their template for BIA (which is adopted in the ISPOR BIA guidelines).

Table 3 provides a summary of the recommendations relating to the BIA key elements extracted from the reviewed BIA guidelines, highlighting the similarities and differences amongst the recommendations in the PMPRB and Canadian F/T/P and private BIA guidelines/templates. A comprehensive list of recommendations was extracted from eight reviewed guidelines. The following sections provide a synthesis of the key similarities and differences among the eight guidelines.

Table 3 Summary of abstracted data of the PMPRB and Canadian F/P/T BIA guidelines
Full size table

3.1 Analytical Model Structure

3.1.1 Perspective

The perspective of the health care budget holders in Canada (public and private drug plans) should be adopted in conducting a BIA [7, 9,10,11,12, 24]. In Manitoba, BIA should be reported for the Manitoba Health, Seniors and Active Living Pharmacare program and additional information may be requested for drugs that significantly impact other Manitoba Provincial Government-sponsored drug programs [10]. In Alberta, an Alberta Health-sponsored drug programs’ perspective should be taken for all BIAs [11] (e.g., not the entire health care system). Inclusion of the patients’ perspective in addition to the primary payer’s perspective (in the case of co-payment) is recommended in the PMPRB BIA guidelines as a supplement to the base-case analysis [7, 24] as per the drug plan’s requirement (i.e., Alberta and Quebec).

3.1.2 Time Horizon

The recommended time horizon in the PMPRB and F/P/T and private BIA guidelines is 3 years [7, 9,10,11,12].

3.1.3 Modeling

The Ontario, British Columbia and 2007 PMPRB BIA guidelines [7, 9, 24] provided an Excel-based template for reporting BIA results. Only the PMPRB 2007 BIA guidelines [7, 24] recommended testing the face, internal and external validity of the structured BIA model.

3.1.4 Target Population

The target population is defined as “all drug plan beneficiaries who are expected to be diagnosed and treated for the conditions of interest and are eligible to use the new drug” [9,10,11,12]. For the target population estimation, there are two broad approaches: top-down or epidemiologic and bottom-up or claims-based (market-share) analyses. An epidemiologic approach is usually preferred if the submission indicates a superior therapeutic conclusion in clinical studies, whereas a market-share approach might be preferred if the submission indicates a non-inferior therapeutic conclusion.

The private payer whose BIA checklist was included in the review did not explicitly mention anything about the analysis for target population assessment. Private payers who use the US AMCP template adopt ISPOR BIA guidelines which apply only a top-down population assessment approach. In Canada, public plans accept both approaches. In Alberta, in the absence of epidemiologic data (e.g., disease prevalence), claims data could be applied with justification for calculating prevalence in this manner and necessary assumptions and sources appropriately cited [11]. In the epidemiologic analysis, disease severity shifts, incidence, and prevalence are required, and it is usually inevitable to use data from different sources [7, 24]. In Manitoba, the prevalence of the disease state and/or indication for which the medication is intended for the total Manitoba population and for the population covered by the Manitoba Pharmacare program should be provided [10]. A list of all new and currently covered indications for the proposed new medication and information on shifts in the target population or market expansion are recommended in the Alberta, Ontario and Manitoba BIA guidelines [9,10,11].

3.1.5 Comparators

According to the PMPRB 2007 BIA guidelines, the reference scenario is the current market share distribution of all comparators without the new drug, whereas the new drug scenario is the forecasted market share of the same comparators with the inclusion of the new drug [7]. One should note that comparator mix does not necessarily always match the comparator mix in the utilization (real world) of the different comparators that private payers see in their databases, and that could be due to differences in the public versus private formularies. Choice of comparators is important, and is sometimes consistent with the economic evaluations and sometimes not.

3.1.6 Costing

Clear costing methods description is required in all provincial guidelines, and the inclusion of cost items is directly related to the chosen perspective [9,10,11,12, 24]. The costs associated with changes in outcomes, disease complications, adverse drug reactions and resource utilization (e.g., hospitalization and emergency room admission) are excluded from BIA [9,10,11,12]. Manitoba requires reporting any significant impact on health care services (e.g., laboratory testing and diagnostic testing) if applicable [10]. The BIA should clearly state which unit of analysis is adopted in measuring the outcomes [9,10,11]. According to Ontario and Manitoba BIA requirements, in the case of medications where the recommended duration of use is less than 30 days (e.g., antibiotics), this should be specified and the cost calculated accordingly [9, 10]. All provincial guidelines require reporting on the total and incremental impact on the budget in BIAs. For drugs which require reconstitution or dose preparation, information should be provided on the method of dose preparation, dose stability and specifics around potential drug wastage [22]. INESSS and pCODR require the cost of supplies to the manufacturer and the payer, and any cost of companion diagnostic tests or medical devices should be reported [12, 22].

Some of the BIA guidelines recommend that the cost of the treatment should be adjusted to consider the markups [25], discounts, inventory charges, business-related costs to the pharmacy covered by the drug plans, dispensing fees and/or patient co-payments, as requested by drug plans [7, 10, 12, 24]. Regulations for covering markups (caps) are different across Canadian provinces. For instance, in Ontario, there is a provincial 8% markup cap (6% for high-cost drugs) [25]; however, markupsFootnote 3 [25] and dispensing fees should be excluded in BIA reports [9]. In contrast, in Alberta, effective May 2018, the Manufacturer List Price (MLP) is the price published in the Alberta Drug Benefit List plus the wholesaler (3%) and pharmacy (7% up to $100) allowable upcharges and the dispensing fee ($12.50), and if applicable, the least cost alternative (LCA) price for relevant drug comparators is recommended [11]. Also, in Manitoba, the total drug cost/patient/month should be based on the actual acquisition cost (AAC) of the medication, which includes the whole cost borne by the pharmacy; therefore, the AAC may include a wholesaler markup if applicable [10]. The PMPRB BIA guidelines [7, 24] and those for Ontario [9] provide Excel-based templates for reporting BIAs. Other provinces request submission of BIA results with accompanying methods of calculation (spreadsheets).

Only in the PMPRB BIA guidelines and those for British Columbia [7, 24] is the use of inflation rates permitted if there is justification for these to be included (e.g., confirmed information on pricing policy, implementation of an approved new policy rule shortly or price changes after patent expiration).

Inflation and discount rates are not applied; however, in the Canadian guidelines, they are permitted in certain circumstances if there is justification for them being included (e.g., confirmed information on pricing policy, implementation of an approved new policy rule in the near future or price changes after patent expiration).

3.1.7 Modeling and Model Validity

All submitted models should be transparent, simple and include confidential prices at the same time. Excel-based electronic models would be preferred if they have the ability to express results in either contract or fiscal years. In addition, it is recommended that the incident and prevalent patients and patients coming off the excess (if there is one) be shown in the model separately (e.g., in the case of biologics). Face, internal and external validity are recommended to be checked and documented. The model validity and transparency could be assessed using recommendations provided by ISPOR and the Society for Medical Decision Making task force report [26]. The detailed process of the validation is not required in Canadian BIA. The programming code should be documented, annotated, and undergo quality assurance and control methods for software engineering. The program created by the developer of the budget impact model to perform the analysis (source code) should be made available for review (on the condition that property rights are respected).

3.1.8 Handling Uncertainty

One-way (univariate) deterministic sensitivity analysis or scenario analysis (multivariate) is acceptable for the most critical variables such as prices, population, and market shares. Alberta and Manitoba recommend one-way and/or multi-way sensitivity analyses for direct prescription costs and incremental prescription costs (savings), and an explanation of the methods used to calculate the sensitivity analyses must be included as well as the assumptions used in calculating the values [10, 11]. Sensitivity analysis of drug dosage and duration [9] and cost of supplies for manufacturers [12] are also recommended. Scenario analysis is recommended in Quebec [12].

3.2 Input and Data Sources

Based on our analysis, regarding clinical safety and efficacy and market data (e.g., degree of implementation in the market), it is acceptable to use data from other jurisdictions in the case of a lack of real-world information for a specific disease/new drug (e.g., rare disease). Epidemiologic data should be captured from Canadian statistics as much as possible. Cost transfers from other jurisdictions are not accepted. In most of the reviewed guidelines, including the PMPRB BIA guidelines [7, 24], epidemiologic data (e.g., disease prevalence and incidence) has to be obtained from national and provincial statistics and registries. In Manitoba, data should ideally be Manitoba specific and not simply an extrapolation of Canadian national data or data from other provinces to the Manitoba population. If Manitoba specific data are not available, a justification for why this is so must be provided [10].

The best sources for the claims-based and market research information are the payer database and the manufacturer’s marketing department [7, 8]. In the PMPRB, Alberta, and Manitoba BIA guidelines data from foreign markets are accepted if local data are not available [7, 10, 11, 24]. The BIA reports from manufacturers with clear supporting data could also be helpful in the PMPRB, Alberta, and Manitoba guidelines [7, 10, 11, 24]. Consensus expert opinion is an option when market intelligence for forecasting the new drug market share is not available [7, 8, 24].

3.3 Presenting Results

There are a few specific requirements for reporting the results mentioned in the reviewed guidelines. In general, total and incremental impact on the primary payer’s budget should be presented [9, 10]. Results should be both aggregated and disaggregated in each year of the time horizon [7, 10, 12, 24]. A table of assumptions, inputs, and outputs, a schematic representation of any uncertainty analyses (e.g., tornado diagram), appendices, references [11] and netFootnote 4 and grossFootnote 5 impact [12] should be included.

4 Discussion

We conducted a comparative review amongst the Canadian BIA guidelines available from the PMPRB, F/T/P jurisdictions and private drug plans. A comprehensive list of recommendations was abstracted from seven reviewed guidelines. Table 3 summarizes the similarities and differences between the 2007 PMPRB, F/P/T and private payers BIA requirements.

The similarities amongst the different BIA guidelines include a time horizon of 3 years, terminologies used for defining current and new scenarios, the epidemiologic data requirements for the proposed indications, the real-world market analysis information, the market share or market capture estimates for the new drug and comparators, and the direct drug costs to be used in the BIAs. Limited information was available regarding private payers’ BIA requirements.

There are differences among provincial BIA guidelines in terms of the nature and number of the recommendations provided for analysts to conduct comprehensive BIA reports based on the provincial drug plan’s requirements (e.g., the AAC including or not including markups and dispensing fees, a supplementary patients’ perspective, cost of supplies, cost of health care utilization, a scenario analysis, and provision of Excel-based templates for cost analysis). To address these differences that were evident also in 2007 [7], the PMPRB BIA guideline was supplemented by an interactive Excel-based template and provided general rules for conducting a BIA that could help policymakers with formulary and reimbursement decisions [7, 8]. The most likely explanation for differences amongst the provinces is that provinces have different inclusion criteria for the coverage eligibility (target population), different access to the new drug, different pricing regulations for generic and brand drugs, and different allowable markups (upcharges), dispensing (professional) fees, and patient co-payments [25].

There are also differences among the updated provincial and 2007 PMPRB guidelines, for which the most suitable explanation is the passage of time and the development of more sophisticated BIA guidelines in which the later guidelines capture the following: inclusion of the patients’ perspective as supplementary to the base-case analysis, in the case of co-payment [11]; a clear description of the unit of analysis in the population and cost analyses [9,10,11]; the degree of implementation of the new intervention (i.e., substitution, combination and expansion) [9,10,11,12]; any significant impacts on health care services (e.g., laboratory testing and diagnostic testing) [10]; cost of supplies [12]; the LCA price for relevant drug comparators [11]; a scenario analysis [12]; the gross and net impact on the budget [12]; and detailed recommendations regarding inclusion of graphics and figures of the analytical framework, a schematic representation of uncertainty analyses, a table of assumptions, tables of inputs and outputs, appendices and references [10, 11].

There are literature reviews as part of the Canadian [7], Belgian [16] and French [18, 27] BIA guidelines. Our review is an update to the literature review published by Marshall et al. in 2008 [7]. They made a side-by-side comparison between the PMPRB 2007 BIA guidelines [8] and the Canadian provincial, other national and transnational BIA guidelines of the time, i.e., those of Alberta (2006), Manitoba (2003), Ontario (2006), Poland (2004), Australia (2002) and ISPOR (2007). In their review, they highlighted differences in the BIA costing approach, perspective, time horizon, opportunity cost, definition of target population and methods for performing sensitivity analysis between the PMPRB 2007 BIA guidelines and the others [7]. When comparing the results obtained in this review with BIA guidelines around the world (Australia [14], UK [15], Belgium [16], Ireland [17], France [18], Poland [19], Brazil [20] and ISPOR [21]), we identified that a considerable number of recommendations related to BIA key elements, including the open (dynamic) population, subgroup analysis in the target population assessment, catch-up effect (for chronic conditions and treatment switch), off-label indications in the eligible population assessment, opportunity costs, cost of clinical outcomes and disease complications, indirect costs, capital costs, staff training costs, applicable tax, patient adherence, total and incremental budget impact for the different health care payers, complicated modeling methods, probabilistic sensitivity analysis (PSA) and scenario analysis for handling the uncertainty, were not included or discussed differently in the 2007 PMPRB guidelines. Most of the Canadian provincial recommendations were in line with other reviewed guidelines, except for the following recommendations: (1) LCA price for relevant drug comparators is only mentioned in Alberta [11], (2) sensitivity analysis for drug dosage and duration in Ontario [9], (3) reporting gross impact on the budget is recommended in Quebec [12], and (4) for drugs which require reconstitution or dose preparation, the method of dose preparation, dose stability and specifics around potential drug wastage [22].

There were two important limitations to the current study: (1) only six (out of 15) F/P/T drug plans had published their template for BIA on their website and (2) there was limited information for private payers, which were found online. Thus we had to contact some private payers to get the required information.

The output from the present Canadian study has been used in conducting a qualitative research projectFootnote 6 in order to obtain stakeholders’ feedback and opinion on the relevance and applicability of the recommendations that were not included or were discussed differently in the PMPRB 2007 BIA guidelines. The next step will be developing a proposal for updating the guidelines.

5 Conclusion

The present study has provided a review of the current BIA guideline environment in Canada. It also identified where the PMPRB 2007 BIA guidelines might not have kept pace with the evolution of BIA guidelines over the past decade. The study has provided a foundation for updating those guidelines, which will occur after conducting a qualitative study to obtain Canadian stakeholders’ feedback and opinions. Future work will also need to address the diversity of BIA needs across the different provincial and territorial programs.