We will have limited operations from 15:00 Tuesday 24 December 2024 (AEDT) until Thursday 2 January 2025. Find out how to contact us during the holiday period.
Background
There is growing concern amongst some key stakeholders over the ready availability of codeine-containing medicines in over-the-counter (OTC) formulations and whether this may be contributing to a reported increase in codeine abuse and, consequently, negative health impacts and public health costs. A number of submissions to this effect were made to the July 2015 meeting of the Advisory Committee on Medicines Scheduling and, on 1 October 2015, the Delegate made an Interim Decision to restrict the availability of codeine-containing products to prescription only level. The Interim Decision was widely contested, resulting in deferral of the final decision to allow for thorough consideration of stakeholder input and a range of regulatory options.
KPMG was engaged by the Therapeutic Goods Administration (TGA) on 2 August 2016 to assist in the preparation of a Regulation Impact Statement (RIS) assessing the regulatory options (as identified by the TGA) through the provision of economic modelling and financial quantification services; specifically, the provision of input for the Impact Analysis and Regulatory Costing of the RIS. In the following four weeks, KPMG identified and documented potential regulatory, economic and social consequences of each of the codeine scheduling options.
At the request of the TGA, KPMG was asked to consolidate the results of its regulatory and economic modelling into a standalone report suitable for public release.
Purpose of this document
This document sets out KPMG's approach to undertaking both the regulatory and economic modelling, including summaries of consultation activities and explanations of the workings of the two models. Both an Impact Analysis and the Regulatory Costing, for inclusion in the RIS, are provided with this document (annexes C and D respectively). Annex E provides additional detail on the economic model to supplement that provided in the main body of the report.
Noting that this report may be read separately to the RIS produced by the TGA, some additional commentary is provided in the Executive Summary to provide more detailed explanations of the process undertaken, including its limitations. This additional commentary also provides some additional context for the inputs and outputs to the models that are likely to be of greatest interest to external stakeholders.
Overall context for this document
The modelling detailed in this report was conducted in accordance with the guidance provided by the Office of Best Practice Regulation (OBPR) for undertaking a cost-benefits analysis. It is worth noting that the policy driver for this study was not to reduce the regulatory burden on business but rather to achieve a public health outcome. As such, an understanding of the evidence base for the proposed rescheduling of codeine, as detailed in the TGA RIS, provides essential context for viewing the results of the regulatory and economic modelling. This report will not comment on the merits of this policy driver.
Throughout the report, reference is made to the base year of the report being 2017. Readers should not infer that this necessarily means that the proposed regulatory change will be in effect from 1 January 2017 or even 1 July 2017. Rather, 2017 was selected as the base year purely for the purposes of the regulatory and economic modelling.
Data limitations
The models, in particular the economic model, were developed in a 'data-poor' environment over an initial four-week period. Validated data was limited to essentially historical IMS[1] for the FY 12/13, which was extrapolated forward to provide the base year volume data. A key Input factor was provided by the Department of Human Services in relation to the public cost of additional GP and specialist pain consultations. Other inputs, such as the discount rate and the default wage rate were sourced from OBPR guidance. A limited review of the relevant academic literature was undertaken to inform the development of the economic model. The theoretical construct for this approach was that if conservative inputs for projected benefits were used, and the model still returned a net economic benefit across a range of sensitivity analysis for a particular scenario, then that scenario, of the scenarios examined, would provide the greatest overall benefit.
In addition to quantitative data sources, as detailed above, KPMG also sought inputs from a range of qualitative sources. In addition to desktop reviews of submissions made to the TGA with respect to this matter, KPMG conducted consultation activities within the four-week timeframe, engaging with representatives from:
- TGA;
- broader Department of Health;
- Office of Best Practice Regulation (OBPR);
- five pharmaceutical companies involved in the production and sale of low dose codeine medicines; and
- three peak bodies.
The models
For this project KPMG developed a regulatory costing model and an economic and social impacts model (health economic model), each informed by a range of sources, including the aforementioned industry and peak body consultations, as well as guidance from the TGA and the OBPR.
KPMG has sought to be as transparent as possible in relation to the assumptions made, and the inputs and outputs from the regulatory and economic models. Given its relative simplicity, a high level of transparency has been achieved for the regulatory model. However, the economic model is more complex; the population is segmented into five groups based on codeine usage characteristics and assumptions do not operate collectively across all groups but rather in combination, depending on the pathway of each group through the modelled time period (ten years). Hence transparency is achieved by detailing the model design and structure and comparing it against alternative potential structures (see Annex E).
Key findings
In consultation with TGA, KPMG and the TGA determined that Options 1- 6 (see Table ES1) (pertaining to distinct regulatory options), while separated in the modelling for the purposes of clarity, would, in practice, result in the following scenarios (logical grouping of regulatory options) due to the natural alignment and dependency of certain options.
| Scenario 1: (Option 1) |
No change to the status quo. |
|---|---|
| Scenario 2: (Options 2 and 5) |
Schedule 2 and Schedule 3 entries for codeine (including, but not limited to, cough and cold medicine preparations and codeine containing analgesics) be amended to reduce the pack size to not more than 3 days' supply and include a label warning that codeine can cause addition. Summary
|
| Scenario 3: (Options 3 and 5) |
The current Schedule 2 entries for codeine in cough and cold medicine preparations be up-scheduled to Schedule 3, and then all Schedule 3 entries (i.e. those currently Schedule 3 and those previously Schedule 2) for codeine (including, but not limited to, cough and cold medicine preparations and codeine containing analgesics), be amended to reduce the pack size to not more than 3 days' supply, and include a label warning that codeine can cause addiction. Summary
|
| Scenario 4: (Options 4 and 6) |
Schedule 2 and Schedule 3 entries for codeine (including, but not limited to, cough and cold medicine preparations and codeine containing analgesics) be up-scheduled to Schedule 4. Summary
|
Source: TGA and KPMG workshop 4 August 2016
As discussed in further detail throughout the report, only the implementation of Scenario 4 (Options 4 and 6) results in a net benefit to society (as detailed in Table ES2). OBPR specifies different methods for summarising ten years of projected regulatory and economic costs. Regulatory costs projected over ten years are summed (without discounting) and then summarised and reported as an average annual impact (total divided by ten). In contrast, economic costs for a cost-benefit analysis have their impact over the life of the proposed regulation (set at 10 years in this case) monetised and discounted at 7% per annum to obtain present value of the ten years of costs. Noting these differences in calculation methods, Tables ES2 sums the results of the regulatory costs (the undiscounted ten years summed) and economic modelling (ten years discounted to obtain a predicted net benefit). Subsequent tables in the Executive Summary provide additional detail on each of the component costs and benefits.
Option 5 has an increase in economic costs; the reduction in pack size means that consumers who are currently using packs of more than 3 days' supply (38% of packs sold in the 12 months to September 2013) will purchase more packs to maintain their use of the medicines, and smaller packs are at a higher cost per tablet. The economic benefits are driven by gains in quality of life, deaths prevented and also net financial savings to consumers, for example those who substitute OTC codeine with paracetamol and/or ibuprofen, post regulatory change. These benefits are only realised for Options 4 and 6, hence the absence of an economic benefit estimate for Options 2, 3 and 5 (Scenarios 2 and 3).
The health economic model sought to project the changes to a range of costs to the Commonwealth and individual consumers, deaths prevented and changes in quality of life as a consequence of the specified scenarios. There were a number of key assumptions regarding both baseline parameters (e.g. the proportion of consumers who are dependent on low dose codeine medicines) and projection parameters (the proportion of dependent users whose health outcomes will improve)) that have limited supporting evidence. By using the most appropriate model structure, conservative assumptions and sensitivity analysis, the following key results were demonstrated:
- only Scenario 4 will result in an expected net benefit;
- the drivers of the net benefit are improved quality of life following better treatment pathways, deaths prevented, and net financial savings to consumers; and
- even under sensitivity analyses that materially reduce benefits and increase costs, a positive net benefit, albeit small, remains the projected result of Scenario 4.
| Element | Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| Regulatory costs (total over ten years (2017-26) not discounted) | ($0.50) | ($1.30) | ($101.40) | ($1.30) | ($102.40) | ($22.10) |
| Economic costs (PV 2017-26 at 7% discount $M) | ($20.70) | ($409.87) | ($14.49) | ($409.87) | ($56.03) | ($209.87) |
| Economic benefits (PV 2017-26 at 7% discount $M) | 0 | 0 | 0 | 0 | $243.95 | $5,353.17 |
| Net benefit (option basis) ($M) | ($21.20) | ($411.17) | ($115.89) | ($411.17) | $85.52 | $5,121.20 |
| Net benefit (scenario basis) ($M) | ($432.37) | ($527.06) | $5,206.72 | |||
Note: the regulatory costs are undiscounted and the economic costs discounted, as is consistent with OBPR guidelines.
Source: KPMG
An extensive sensitivity analysis was undertaken to provide the necessary degree of confidence in the robustness of the model's key result: that the net benefit of Scenario 4 is positive whereas scenarios 2 and 3 are negative. The summary contains a discussion of the drivers and key inputs for each broad outcome; specifically (1) health gains and death prevented; (2) the net savings and costs to consumers; (3) additional MBS activity and costs; and (4) additional PBS scripts and their cost to PBS. This summary also includes a comparison of the outputs and results of the economic model with some relevant indicators of context: (1) additional MBS expenditure from Scenario 4 as a proportion of total national MBS expenditure; (2) the results of separate public domain analyses that investigated additional MBS cost and the share of all consumers who would continue to use low dose codeine if they required a prescription; and (3) the cost effectiveness of pharmaceuticals newly approved for listing on the PBS using a cost per Quality of life year (QALY) metric.
Additional commentary on key aspects of the findings
Regulatory compliance costs to pharmaceutical industry
In accordance with OBPR guidance, regulatory costings were annualised over 10 years. However, projected regulatory costs will likely be fully realised over the assumed implementation period (12 - 24 months) for a decision by the delegate that impacts on the labelling, pack-size or up-scheduling of low-dose codeine medicines. Regulatory costings were informed by insights provided by the industry consultations and assumed that any of the proposed options would result in some product portfolio rationalisation by current sponsors of codeine phosphate products listed on the Australian Register of Therapeutic Goods (ARTG). Annex D provides the detailed assumptions and inputs used to undertake the regulatory costing. Tables ES3 shows the projected industry compliance costs (excluding impacts on consumers, pharmacists and doctors). Scenario 4 will produce the least projected increase in compliance costs for industry (less than 20% of the figure for the other scenarios), primarily resulting from the projected significant rationalisation of product portfolios if OTC codeine based medicines were up-scheduled to S4.
| Element | Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| Regulatory compliance costs (industry) | ($0.49) | ($1.27) | ($0.4) | ($1.27) | ($0.09) | ($0.2) |
| Net cost (scenario basis) | ($1.76) | ($1.67) | ($0.29) | |||
Additional costs to doctors and pharmacists (regulatory model)
Data was not available on the current capacity of GPs to take on additional consultations or for specialist pain clinics (which is assumed to pose a greater limitation given anecdotal evidence of current wait times for appointments at such clinics). Some key clinical stakeholders noted that the initial increase in the number of people who are additionally treated and experience a health benefit is likely to decline in a few years until the system recalibrates. That is, patients who are currently using low dose codeine medicines and who, following up-scheduling will pursue therapeutic pathways that improve health outcomes, are part of a cohort. They will receive additional treatment and care for the following year, but this additional treatment, compared to what they would otherwise have received, will reduce each year. Stakeholders who raised this issue indicated that this additional activity would start to reduce after one to two years and maybe be absorbed into standard care after three or four years. Hence, the economic model captures this factor by assuming, in the base case, a 30% annual reduction on previous year's treatment and health gains. This factor was incorporated into the model by using an average (over ten years) number of additional GP appointments (200,000 additional appointments per year for Option 4 and 51,000 additional appointments per year for Option 6). Unlike the regulatory compliance costs for sponsors, detailed above, the additional regulatory compliance costs for doctors and pharmacists are likely to occur over the entire ten-year period used for the regulatory modelling. These costs are detailed in Tables ES4.
| Element | Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| Regulatory compliance costs (doctors and pharmacists) | ($0) | ($0) | ($6.91) | ($0) | ($2.52) | ($0.22) |
| Average net cost (scenario basis) per year for the period 2017-26 | ($0) | ($6.91) | ($2.74) | |||
Additional costs to consumers (regulatory model)
As noted above the cohort of consumers changing their behaviour from the status quo is likely to reduce over time. Tables ES5 details the projected costs to consumers (in relation to increased time spent), as per OBPR guidance. Scenario 4 will results in a significantly greater time impost (almost 450% greater than Scenario 3) on consumers. However, this results should not be considered in isolation but rather in conjunction with the costs and benefits as determined by the economic model.
| Element | Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| Regulatory compliance costs (doctors and pharmacists) | ($0) | ($0) | ($3.2) | ($0) | ($7.71) | ($1.97) |
| Average net cost (scenario basis) per year for the period 2017-26 | ($0) | ($3.2) | ($9.68) | |||
Additional deaths prevented and health gains (economic model)
A total of five (2017) and 50 (2017-26) prevented deaths are projected by the economic model. The number of deaths prevented per year was assumed to be five; equivalent to 0.445 deaths prevented per 100,000 current customers. This is a conservative assumption. An Australian study[2] found that over 2000-2013 (14 years) 1,437 deaths (accidental and intentional) were assessed by the authors as having 'codeine toxicity' as a 'contributory cause of death'. Of these codeine related deaths, 83.7% were 'attributed to multiple drug toxicity'. In 572 (41.1%) of these cases, the study's authors found that 'information about whether the codeine consumed before death was prescribed or obtained over the counter' had been recorded. Of these cases, in 229 'OTC codeine products were recorded'. In the remaining 343 cases, prescribed codeine products were recorded, most commonly Panadeine Forte. Hence, on average 16 (229/14) codeine related deaths per year had OTC codeine products recorded, and possibly up to 20 to take into account the deaths for which this information was missing (conservatively assuming that 10% of all the unidentified sources were OTC codeine) . The authors found that the rate of codeine related deaths had increased between 2000 and 2009, from 3.5 to 8.7 deaths per million population. They also noted that a potential driver of this increase 'may have been the introduction in Australia of OTC products containing larger amounts of codeine, including codeine combined with ibuprofen'. Given the underreporting of type of codeine product and the ongoing increase in codeine related deaths, then, for the purpose of this model, it was assumed that up to 30 codeine deaths in 2017 could occur and also have had OTC codeine as the codeine product. The critical assumption is however the number of deaths that could be prevented if OTC low dose codeine were rescheduled to S4. A conservative assumption, 5 rather than 16 or 30 deaths prevented was used to reflect the additional information also reported in the study: (1) 83.7% of all of the codeine related deaths were attributed to multiple drug toxicity; (2) the combination of mental health and chronic pain issues were found in 25.7% of cases; (3) 53.6% of cases had a history of mental health issues; and (4) 36.1% had a history of substance use problems. Hence, it is reasonable to assume that up scheduling from S3 to S4 could prevent some of these death and that 5 deaths (which represents between 15% and 30% of these deaths) is a conservative assumption.
In the economic model each prevented death is valued at $4.2 million, to ensure consistency with OBPR guidelines.[3]
A gain in quality of life years (QALY) of 9,208 in 2017 is the result of patients receiving treatment they would otherwise not have accessed that leads to more effective therapy compared to low dose codeine combination medicines. This gain is offset by the loss experienced by a small proportion of clients for whom low dose codeine combination medicines are the most effective option available to them and who do not attend a GP to obtain a prescription under Scenario 4. This loss is projected to be 134 QALYs, based on the assumptions about the rate at which this situation occurs (assumed to be small) and the annual QALY loss per person. This situation was identified by some stakeholders. Systematic reviews indicate[4] that, on average, current acute users and chronic users of low dose codeine combination medicines who change to OTC only paracetamol and/or ibuprofen will experience no change in pain relief and hence do not experience a QALY loss. This assumption was used in the economic model to ensure consistency with the clinical evidence. (See Table ES15 for details of the number of patients who participate in additional therapy and the average gains.)
The monetary valuation of a QALY ($182,000) was used to ensure consistency with OBPR guidelines.[5] This resulted in a gain of $1,651.5 million (in 2017) and a gain (present value) of $4,399.5 million over 2017-26. The ten year result is not simply ten times the first year result, for two reasons. First, the majority of the QALY gains occur in the first two years as customers pursue different options as a consequence of attending a GP consultation. Stakeholders stated that this initial activity would reduce over time. The model assumes the reduction in the number of patients participating in additional therapy occurs at a rate of 30% per year. (The QALY gained is assumed conservatively to occur only in the year that treatment occurs). Secondly, the QALY gained were discounted using a rate of 7% to ensure consistency with OBPR guidelines.
There are other potential sources of QALY gains and losses that are noted in the literature and by stakeholders but which were not quantified as part of the economic model. The reasoning for this is that the data on both the added gains and losses is limited but this might not impact the net QALY gain because it is plausible that these aims and losses will net each other out. To illustrate further, the sources of QALY gain relate to the reduction in risk of dependence and the reduction in risk of adverse events relating to overdose or long-term use of codeine, paracetamol and/or ibuprofen. Moreover, some stakeholders also identified that some patients might have an increased risk of adverse events as a result of changes to higher dose prescription codeine, for example. However, neither of these effects were quantified, and to the extent that the former outweighs the latter this results in an underestimate of the net QALYs gains under Scenario 4.
It was assumed that Scenario 4 would not result in an improvement in the quality of life for current customers whose use of codeine medicines is non-therapeutic.
Table ES6 presents the deaths prevented and QALYs gained in the first year, and Table ES7 the present value of these outcomes over 2017-26.
| Health outcomes 2017 ($M) |
Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| Deaths prevented | ||||||
| Estimated deaths prevented (number) | 0 | 0 | 0 | 0 | 0 | 5 |
| Monetised deaths prevented (using OBPR valuation of $4,200,000) ($M) | 0 | 0 | 0 | 0 | 0 | $21.0 |
| Improved quality of life (ex. Deaths prevented) | ||||||
| Patients with QALY benefits (number) | 0 | 0 | 0 | 0 | 0 | 221,034 |
| Above as % of all current customers | 0 | 0 | 0 | 0 | 0 | 19% |
| QALY gains from improved treatment (number) | 0 | 0 | 0 | 0 | 0 | 9,208 |
| QALY loss for some consumers no longer using Low dose codeine medicines (number) | 0 | 0 | 0 | 0 | 0 | (134) |
| Monetised QALY gain (using OBPR valuation of $182,000) ($M) | 0 | 0 | 0 | 0 | 0 | $1,651.5 |
| Monetised value of health gains and deaths prevented | ||||||
| Total monetised value of health gains and deaths prevented ($M) | 0 | 0 | 0 | 0 | 0 | $1,672.5 |
| Monetised health outcomes and deaths 2017-26 Present value (7% as per OBPR guidelines) (all dollar amounts are in italics) |
Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| Monetised present value of Deaths prevented | 0 | 0 | 0 | 0 | 0 | $147.5 |
| Monetised present value of net QALYs gained | 0 | 0 | 0 | 0 | 0 | $4,399.5 |
| Total monetised present value | $4,547.0 | |||||
Projected additional MBS consultations and costs (economic model)
The projected additional MBS consultations in 2017 arise for one of two reasons:
- First, under Option 6 some patients are expected to attend consultations to gain the benefits of alternative treatment options, including pain clinic consultations with specialists and pain management at general practices provided by GPs. The potential for some current customers to attend a medical practitioner to improve their current therapy was identified by all stakeholders and this increase in consultations for treatment purposes is consistent with best practice. The number of consultations for treatment for a given patient will depend upon whether they participate in more or less intensive pain management; current chronic dependent users were assumed to be the most likely to participate in more intensive pain management, including the care provided by pain clinics.
- The second reason is that some customers will attend GPs to obtain prescriptions for their ongoing use of low dose codeine medicines under Scenario 4. For both treatment and prescription only consultations, some will be accommodated in consultations that would otherwise have occurred. In the case of treatment consultations, the share that would otherwise have occurred depends on whether the patient requires a more or less intensive clinical pathway, with consultations related to less intensive pathways (primarily changed pharmacotherapy) morel likely to occur as part of consultations that would otherwise have occurred. For low dose codeine prescription consultations, 25% of consultations were assumed to be additional for both S2 (Option 4) and S3 (Option 6).
The additional low dose codeine prescription GP consultations under Option 6 (46,979) are substantially less than under Option 4 (185,184) (See table ES8), despite the estimated packs sold for current S2 products currently (3.7 million) being substantially lower than for S3 (20.8 million packs)[6]. The reason is the difference in the number of packs per year that are expected to be purchased under Scenario 4. First, the economic model assumes that up to five repeat prescriptions will be available as is currently the case for prescription codeine products on private prescription.[7] This allows for up to 6 scripts per visit to a GP. Hence, current acute users of S3 medicines, who are assumed to purchase five (20 tablet) packs a year currently (see Table ES14 for the basis of this assumption) will only require one consultation a year, and in 25% of cases are assumed to be additional consultations. In contrast, current S2 users who continue to use the product are assumed in Scenario 4 to require only one pack a year and hence one consultation is required per person per year, 25% of which are assumed to be additional.
The additional GP consultations were translated to an additional cost to the MBS using an average cost to MBS per consultation calculated from data provided by the Department of Human Services. This data provided the total cost to MBS for a given volume of consultations, each year for the next five years starting in 2017. The average cost per consultation was derived from this data for each year. This average cost was applied to the number of consultations projected by the model to determine the cost to the MBS in the base case and all of its sensitive analyses. This average cost increased each year. On advice from the data custodians, the values for the final year were kept constant and continued for each year up to 10 years. This average cost was calculated using assumptions made by the data custodians including: the mix of consultations; safety net effects and bulkbilling rates. This rate was $40.43 per average consultation in 2018 for example. An analogous method was used to derive the average cost per specialist consultation (pain clinic related costs), for example $89.89 in 2018.
Table ES8 details the additional MBS consultations and costs in the first year (2017), and Table ES9 the present value of these costs over 2017-26. The additional MBS costs in 2017 are estimated at $59.4 million, of which 83% relate to the costs of improved treatment for pain relief that would otherwise not have occurred. The remainder ($5.6 million) relate to the costs of additional consultations to obtain prescriptions for low dose codeine medications. The present value of additional costs to the MBS over the ten-year period 2017-26 are $185.2 million, of which 61% relate to treatment costs. This difference in the proportion that relate to treatment costs over the ten-year period is the result of the need for treatment for patients with chronic therapeutic use (dependent and non-dependent) decreasing as the use of low dose codeine medications reduces due to the requirement for patients to obtain a prescription from a GP.
| MBS Consultations and Costs 2017 (all dollar amounts are in italics) |
Scenario 2 | Scenario 3 | Scenario 4 | ||||
|---|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | Scenario 4 total | |
| General Practice Consultations | |||||||
| Treatment related | |||||||
| Consultations (number) | 0 | 0 | 0 | 0 | 0 | 1,062,891 | 1,062,891 |
| Cost to MBS of additional consultations ($M) | 0 | 0 | 0 | 0 | 0 | ($42.9) | ($42.9) |
| LDC prescription related | |||||||
| Number of consultations | 0 | 0 | 0 | 0 | 185,184 | 46,979 | 232,162 |
| Cost to MBS of additional consultations ($M) | 0 | 0 | 0 | 0 | ($7.5) | ($1.9) | ($5.6) |
| Specialist consultations (All treatment related) | |||||||
| Number of consultations | 0 | 0 | 0 | 0 | 0 | 80,333 | 80,333 |
| Cost to MBS of additional consultations ($M) | 0 | 0 | 0 | 0 | 0 | ($7.2) | ($7.2) |
| Total additional MBS | |||||||
| Number of consultations | 0 | 0 | 0 | 0 | 185,184 | 1,190,203 | 1,375,387 |
| % that are treatment related consultations | 0 | 0 | 0 | 0 | 0% | 96% | 83% |
| Cost to MBS of additional consultations ($M) | 0 | 0 | 0 | 0 | ($7.5) | ($52.0) | ($59.4) |
| % that are cost of treatment related consultation | 0 | 0 | 0 | 0 | 0% | 96% | 84% |
| MBS Costs 2017-26 Present value (7% as per OBPR guidelines) (all dollar amounts are in italics) |
Scenario 2 | Scenario 3 | Scenario 4 | ||||
|---|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | Scenario 4 total | |
| GP consultation costs | 0 | 0 | 0 | 0 | ($56.03) | ($129.16) | ($185.19) |
| Comprising: | |||||||
| treatment related | 0 | 0 | 0 | 0 | 0 | ($16.56) | ($16.56) |
| Low dose codeine scripts related | 0 | 0 | 0 | 0 | ($56.03) | ($112.59) | ($168.63) |
| Specialist consultation costs | 0 | 0 | 0 | 0 | 0 | ($19.27) | ($19.27) |
| Total Additional MBS costs | 0 | 0 | 0 | 0 | ($56.03) | ($148.43) | ($204.46) |
Additional prescription medicines and additional costs to PBS (economic model)
The additional prescription medicines dispensed are primarily the result of patients who have chronic conditions and current chronic use of low dose codeine medicines and, under GP or specialist pain management, are prescribed pharmacotherapy with less risk of adverse events/and or improved pain management outcomes compared to low dose codeine combination medicines and OTC options. Stakeholders provided examples of these medications; the appropriateness of the regimen (medicine(s), dose, frequency number of repeats, pack size and duration of therapy) depending on the situation for the individual patient as assessed by the GP or specialist: tramadol, high dose codeine,[8] pregabalin, and gabapentin.
The model assumes that 223,149 (20%) of all current consumers are 'chronic users' (see tables E13 and E14 below) and that 174,056 (78%) of these users will have some kind of additional pharmacotherapy in Scenario 4. Also, some patients with acute conditions, such as migraines, are expected to be prescribed alternative analgesics such as anti-migraine preparations where they are more effective compared to low dose codeine medicines and/or have less side effects. The model assumed that 46,979 (5% of acute consumers) would participate in pain management that would lead to additional health benefits compared to low dose codeine and that pharmacotherapy would be one part of this management.
A share of these medicines would attract a PBS subsidy, however, there was little data available to inform an estimate of either the share of these scripts that would attract a subsidy, or the size of that subsidy if it occurred. Some of these alternative analgesic medicines will be below the general concessional co-payment ($38.30 at July 2016) and around $10 above the concessional co-payment ($6.20 at July 2016), for example, tramadol and high dose codeine combination medicines. Others will be above the general co-payment, for example, pregabalin. PBS will only incur a cost if the medicine is above the threshold for the relevant patient group (concessional or general beneficiary). In the absence of data and evidence of the likely uptake of these medicines the following simplifying assumption was made: 50% of all prescription medicines would attract a PBS subsidy and the average subsidy would be $10.
Clinical stakeholders indicated that these additional scripts would reduce over time as chronic pain for this group became more appropriately managed and new cases that would otherwise have used low dose codeine went directly to their GP and received the appropriate pharmacotherapy. The model accommodated this advice.
Additional costs to PBS for these additional medications that are expected to improve pain relief and or reduce the risk of side effects relative to low dose codeine medications is analogous to a new drug with additional effect and additional cost compared to usual care approved through the PBAC process because it is cost effective. Hence, given these medications are already approved for PBS listing it is reasonable to assume that this additional cost represents a cost effective increase in PBS expenditure. The issue of assessment of cost effectiveness is discussed in more detail in the section 'Results in Context' at the end of this Executive Summary.
Low dose combination medicines are assumed not to be listed on the PBS under options 4 and 6. This assumption is consistent with stakeholder perspectives. Hence patients prescribed and dispensed this medicine will not attract a PBS subsidy.
Table ES10 presents the additional prescription medicines dispensed (4.6 million) and costs to the PBS ($23.1 million) in the first year (2017), and Table ES11 the present value of the additional costs to the PBS ($61.4 million discounted at 7%) over 2017-26. Only some of the 4.6 million additional prescriptions in 2017 that are expected to occur under Scenario 4 will incur cost to the PBS.
| Prescription medicines and PBS Costs 2017 (all dollar amounts are in italics) |
Scenario 2 | Scenario 3 | Scenario 4 | ||||
|---|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | ||
| Total additional prescription medicines (Inc. under and over co-payment PBS scripts and private scripts) | 0 | 0 | 0 | 0 | 0 | 4,612,359 | |
| Additional Cost to PBS (this estimate excludes under co-payment and private scripts) | 0 | 0 | 0 | 0 | 0 | ($23.1) | |
| PBS Costs 2017-26 Present value (7% as per OBPR guidelines) (all dollar amounts are in italics) |
Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| PBS costs | 0 | 0 | 0 | 0 | 0 | ($61.4) |
Additional costs and benefits to consumers (economic model)
The calculations of additional cost to consumers due to changes in pack size for low dose codeine (S2 and S3, all scenarios) were calculated by starting with the IMS wholesale data provided by the Department of Health describing current volume, its distribution across pack size, and the total wholesale expenditure by pack size. The cost per tablet was calculated and then the total number of smaller packs that represent an equivalent total tablet volume to the total tablets with current pack sizes was estimated. Smaller pack sizes have a higher cost per tablet, hence the wholesale cost are expected to be higher under all scenarios for the same volume. The cost to consumers was calculated by assuming a 44% mark-up on wholesale price, based on a review of the published priced compared to the prices by pack size derived from IMS data. For S3 medicines, this change in pack size would result in an increase of 23% in the number of packs sold in a year to maintain volume of tablets (from 20,844,727 to 25,750,705) and a 33% increase in the average expenditure per year per consumer from $138 to $185. The smaller sales of S2 items (an estimated 3.7 million packs in 2016) results in a smaller increase in out-of-pocket costs due to the change in pack size.
Another source of additional cost to consumers are the co-payments for additional GPs and specialists consultations, which were assumed to be at an average of $10 and $75 respectively; these are weighted average co-payments and take into account that some of these consultations will be bulkbilled. The weighted value of $10 corresponds to 20% of consultations requiring a co-payment at $50 per consultation and the remaining requiring no co-payment ($50 x 20% =$10). The cost per prescribed medicine was assumed to be $15 and takes into account the factors discussed in the PBS section, above. Additional OTC medications (paracetamol and/or ibuprofen) were assumed to be purchased at an average cost per pack of $3.[9]
The additional financial savings due to reduced use of low dose combination medicines are the source of financial savings to consumers in Scenario 4. In a conventional regulatory situation, an additional cost impost on an item that leads to a reduction in its use and the replacement with a lower cost substitute would indicate a reduction in consumer surplus. In the case of a health economic model, where the loss of interest is health (QALYs), and the lower cost alternative is demonstrated to be no more effective and no greater risk of adverse events compared to the higher cost alternative, there is no loss in health, only a reduction in expenditure. Hence, the reduction in out-of-pocket-cost to consumers who have substituted lower cost OTCs that are equally effective and/or with less side effects comes, on average, without a loss to the consumer. From table E13 it can be seen that chronic users could be spending between $600 and $700 on low dose codeine combination medications currently per year. The additional savings are calculated relative to Scenario 1, with no change in pack size.
Table ES12 presents the additional benefits (deaths prevented, QALYs gained and net financial savings) and costs (net out of pocket costs) in the first year (2017), and Table ES13 the present value of these outcomes over 2017-26. This table is presented to provide the overall perspective of the gain and loss to consumers. The basis of the estimates of deaths prevented and net QALYs gained are presented in a previous section. These two tables also include the net financial costs (in Scenarios 2 and 3) and net financial savings to consumers under Scenario 4.
| Consumer costs and benefits 2017 | Scenario 2 | Scenario 3 | Scenario 4 | ||||
|---|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | Scenario 4 total | |
| Deaths prevented and QALYs gained | |||||||
| Deaths prevented (number) | 0 | 0 | 0 | 0 | 0 | 5 | 5 |
| net QALYs gained (number) | 0 | 0 | 0 | 0 | 0 | 9,074 | 9,074 |
| Financial savings to consumers | |||||||
| Additional cost of co-payments for MBS and PBS, OTC medications less the savings from less purchases of low dose codeine ($M) | 0 | 0 | 0 | 0 | $32.5 | $40.2 | $72.73 |
| Comprising: | |||||||
| Savings from reduction in Low dose Codeine ($M) | 0 | 0 | 0 | 0 | $34.4 | $132.8 | $167.19 |
| Additional costs of medicines (OTC and prescription) and co-payment ($M) | 0 | 0 | 0 | 0 | ($1.9) | ($92.6) | ($94.47) |
| Net additional out of pocket costs to consumers | |||||||
| Additional cost due to change in pack size less savings due to less purchases of low dose codeine in options 2 and 3 ($M) | ($2.8) | ($54.6) | ($1.9) | ($54.6) | 0 | 0 | $0.00 |
| Consumer costs and benefits 2017-26 Present value of QALYs gained, and financial costs or savings (7%) | Scenario 2 | Scenario 3 | Scenario 4 | |||
|---|---|---|---|---|---|---|
| Option 2 | Option 5 | Option 3 | Option 5 | Option 4 | Option 6 | |
| Deaths prevented and QALYs gained | ||||||
| Deaths prevented (number) | 0 | 0 | 0 | 0 | 0 | 50 |
| net QALYs gained (number) | 0 | 0 | 0 | 0 | 0 | 24,173 |
| Net financial savings or cost | ||||||
| Net financial savings or costs ($M) | ($20.7) | ($409.9) | ($14.5) | ($409.9) | $244.0 | $806.2 |
Other key model assumptions (Economic model)
The first important element of the model design for Option 6 was that it used five different patient groups, rather than an average patient, to determine: (1) current use; (2) response by consumers under Option 6; (3) the resource use for consumers; and (4) the health impact. All stakeholders identified that there were distinct groups amongst current users: (1) therapeutic and non-therapeutic users; (2) chronic and acute users: and (3) dependent and non-dependent chronic therapeutic users. Stakeholders also identified that these consumers would respond differently, and, importantly for the net benefit calculations, only a small minority of acute users would have a health gain from important treatment options whereas chronic users, would potentially benefit from improved treatment options (less adverse events and/or more health effects). Critically for the economic modelling, there was little data available to distribute consumers across these types, nor was there data on the total number of consumers.
There was data on the total volume of sales for the 12 months to September 2013, by pack size, and this was assumed to have reduced by 10% to approximate 2017 use, based on advice from stakeholders that the market had reduced since that time. This represented a total of 20.8 million packs, as presented in table ES14 as the total of the row labelled 'number of packs'.
Some stakeholders provided possible daily and annual use patterns for these types of consumers; there was no single agreed figure and the assumptions presented in Table ES14 for daily use, and days per year of use are broadly consistent with these expert opinions, but do not reflect any single opinion. An important consideration is that both duration of use and daily use were indicated as drivers of potential adverse events. Chronic dependent users were assumed to exceed the maximum recommended daily dose, whereas chronic non-dependent were assumed to not exceed the dose but to have a duration of use that was beyond the recommended use and increased the risk of dependency.
To obtain the result of packs per year and pack size for each group, it was assumed that acute users purchase smaller packs than chronic users. The actual distribution of pack size in the IMS was maintained in the model and the packs reported in the IMS data were essentially allocated from smallest (first) to largest (last) across acute therapeutic users and last, dependent non therapeutic users. This approach resulted in the estimates of packs per person and average pack size presented in Table ES14, with acute user packs smaller than those for chronic users.
The number of people in each group was estimated by 'back solving' given the modelled input 'constraints' of daily and annual individual use, proportion in each group, and overall volume. Hence the number of people in each group in table ES14 below is the only possible 'solution', that is consistent with inputs regarding total volume, use profile and proportion in each group. Solution is referring to both: (1) total number of customers in a year; and (2) distribution of customer numbers across customer groups. This approach ensured that for the sensitivity analysis, if any of the inputs of user profile or share in each group changed, then the total volume of sales (a fixed and relatively certain input) was maintained, and the 'slack' was taken up by changing the number of customers in total and/or their distribution across groups.
| S3 use only | Therapeutic use | Non therapeutic use | ||||
|---|---|---|---|---|---|---|
| Acute | Chronic dependent | Chronic not dependent | Chronic | Acute | Total | |
| Annual use of individual consumers | ||||||
| Tablets per day when taken (8 is max) | 8 | 12 | 8 | 20 | 6 | |
| Days per year taken | 12 | 250 | 250 | 365 | 20 | |
| Average tablets per pack | 20 | 40 | 28 | 40 | 20 | 28 |
| Average packs per year | 5 | 75 | 71 | 183 | 6 | 18 |
| Expenditure per year per person | $25 | $701 | $591 | $1,706 | $30 | $129 |
| Total baseline use by type of consumer | ||||||
| Number of consumers | ||||||
| Number of consumers in each group | 939,572 | 44,630 | 178,519 | 1,174 | 10,570 | 1,174,465 |
| As % of all users | 80.0% | 3.8% | 15.2% | 0.1% | 0.9% | 100% |
| As % of Australian population over 12 | 4.73% | 0.22% | 0.90% | 0.01% | 0.05% | 5.91% |
| Packs | ||||||
| Number of packs | 4,624,637 | 3,347,226 | 12,595,103 | 214,340 | 63,421 | 20,844,727 |
| As % of all packs sold | 22.2% | 16.1% | 60.4% | 1.0% | 0.3% | 100% |
| Total baseline expenditure | ||||||
| Total expenditure ($M) | $23.1 | $31.3 | $105.5 | $2.0 | $0.3 | $151.7 |
A number of important characteristics of the current profile are suggested by the results presented in ES14
- Around 6% of the Australian population over the age of 12 (19,874,413 at June 2015) purchase at least one pack of low dose codeine per year. Given that there will be an estimated 20.8 million packs sold in 2017 and that many consumers purchase more than one pack, this result of 6 per cent of the population purchasing at least one pack has face validity.
- 80 per cent of all consumers are 'acute', however, as a group they purchase 22.2% of all packs.
- 19% of all consumers are chronic therapeutic users and they purchase 76.5% of all packs sold.
A second key element of the model structure was that the use of resources (consultations and medicines) and the health benefits (QALY gains) would vary across consumers, and that chronic consumers were more likely to participate in therapeutic options that would require medical consultations and prescription medicines. In summary, six potential pathways were identified, and were defined by increasing use of pharmacotherapy and GP and specialist consultations and increasing average health benefits. The first two pathways, use of OTC only and ongoing use of low dose codeine prescriptions, had no health effect. The remaining four had an average health effect for patients. The participation in increasingly intensive treatment pathways is higher for chronic therapeutic users compared to acute therapeutic users, and then higher for dependent compared to non-dependent therapeutic users.
Table ES15 presents the components that constitute the estimated QALY gains and reflects the allocation of the population across treatment pathways (as described above). Only 5% of acute users were assumed to participate in improved therapy that leads to QALY gains, but they represent 21% of all participants in such treatment. An example of acute users in this group is people with migraines. (Migraines were an acute condition treated with low dose codeine combination products and potentially more effectively treated with prescription medicines that was most frequently cited by stakeholders.) In contrast, 90% of dependent and 75% of non-dependent therapeutic users are projected to participate in improved therapy with an average QALY gain per year therapy of 0.05 and 0.045 respectively. These higher gains per person come at additional cost to the MBS, PBS and the consumer, as discussed. However, from Table E15 it can be seen that chronic users could currently be spending between $600 and $700 on low dose codeine combination medications. Hence these costs of alternative treatments will be offset by the reduced expenditure on low dose codeine medicines.
| Participation in treatment that could potentially lead to a gain in QALYs Only gains in first year are presented - proportion participating in treatment is reduced by 30% a year each year |
Therapeutic use | Non therapeutic use | Total | |||
|---|---|---|---|---|---|---|
| Acute | Chronic dependent | Chronic not dependent | Chronic | Acute | ||
| % in group or participate | 5% | 90% | 75% | 0% | 0% | |
| Number who participate | 46,979 | 40,167 | 133,889 | 0 | 0 | 221,034 |
| % of all participants who are in group | 21% | 18% | 61% | 0% | 0% | 100% |
| Number who attend pain clinics | 0 | 4,463 | 8,926 | 0 | 0 | 13,389 |
| Average QALY gain from therapy per participant | 0.025 | 0.05 | 0.045 | 0 | 0 | |
| Total QALY gain from improved therapy | 1,174 | 2,008 | 6,025 | 0 | 0 | 9,208 |
| % of all QALY gains from improved therapy | 13% | 22% | 65% | 0% | 0% | 100% |
Sensitivity analysis (economic model)
The result of a positive net benefit for Scenario 4 (see table ES2) is robust to a wide range of sensitivity analyses, including the following set of cost maximising and benefit minimising assumptions:
- no deaths prevented;
- the costs of gaining any quality of life improvements through improved treatment for chronic pain, for example, are increased by 80%; and
- the quality of life gain from treatment is reduced by 80%.
The variables that this result of a positive net benefit were most sensitive to were the average QALY gain resulting from additional treatment received, and the number of repeat scripts. The result was only moderately sensitive to the discount rate, number of deaths prevented, and the co-payment for GP and specialist consultations. The net benefit for Scenario 4, however, remained positive across all sensitivity analysis conducted.
Results in context
The tables above report results either for 2017, or for the present value over the period 2017-26, at a discount rate of 7%, to ensure consistency with OBPR guidelines. The following graphic presents the results on an undiscounted basis for each year over this period. It demonstrates that the rate of reduction in treatment related costs and also in the benefits of treatment (QALYs) decline over time, as indicated by stakeholders. It also demonstrates the ongoing use of low dose codeine by some consumers.
Figure ES1: Annual additional MBS treatment consultation costs, MBS low dose codeine medicine prescription consultation costs PBS costs and QALYs - Scenario 4
The additional costs to MBS for Scenario 4, in the first year, represent 0.5% (for treatment related) and 0.1% (for low dose codeine prescription related) of total MBS expenditure ($9.8 billion) in the professional attendances group in 2015-16.[10] The difference in treatment related consultations and low dose codeine prescription related consultations are discussion in more detail in the section above (Projected additional MBS consultations and costs (economic model)). In summary, the former relate to improved treatment and management of pain which occurs as a consequence of change in patient behaviour, triggered by Scenario 4. The latter relate to the additional costs to the MBS of patients continuing to use low dose codeine medicines without any other changes in therapy, care or health related outcomes.
One way of summarising and assessing the health and financial consequences of Scenario 4 is as a net benefit, where the QALYs and deaths are summarised as monetised values and compared to the costs. This approach is consistent with OPBR guidelines and presented in this report. Another approach, which is used by the Pharmaceutical Benefits Advisory Committee (PBAC) and the Medical Services Advisory Committee (MSAC), is cost effectiveness analysis: a comparison of the additional costs with the additional effects (QALYs) expressed as an incremental cost effectiveness ratio (ICER). A formal cost effectiveness analysis was outside the scope of this project. It is possible to use the results generated by the economic model to make a broad estimate of the ICER of Scenario 4 compared to Scenario 1 (current practice). Using the results report in previous tables, the additional costs to MBS and PBS from Scenario 4 relative to current practice were compared to the additional QALYs, for both the first year and the present value over ten years and estimates $9,092 per QALY and $11,097 per QALY respectively are reported. (See Table ES16). Although the decision threshold for PBAC is not publicly reported, it is reasonable to conclude that at around $10,000 per QALY that these preliminary results indicate that Scenario 4 is projected to achieve additional QALYs at a cost per QALY that would be considered to be 'cost effective'.
| Cost and effect (QALYs) - all sourced form previous tables | 2017 | PV of 2017-26 |
|---|---|---|
| Additional MBS low dose codeine prescription consultation costs ($M) | $9.36 | $72.60 |
| Additional MBS treatment consultation costs ($M) | $50.08 | $134.22 |
| Additional PBS costs ($M) | $23.06 | $61.43 |
| Total MBS and PBS costs ($M) | $82.50 | $268.25 |
| Additional QALYs (ex. Deaths) (QALYs) | 9,074 | 24,173 |
| ICER = additional cost/additional QALYS ($) | $9,092 | $11,097 |
The Macquarie University Centre for the Health Economy (MUCHE) reported the results of a study on the value of OTC medicines, the gain in value should some medicines be rescheduled from S4 to S3, and the loss in value should some of them not be available OTC.[11] The report stated that their survey found that, in the case of analgesics/pain relievers, when asked what they would do if the medication for their condition became unavailable over the counter: 63% of respondents said they would see their doctor, 24% said they would do nothing, 15% said they would use a home remedy and 7% said they would 'supplement'.
More options were explored in the model presented in this report. The 48% of all current users who do not go to their doctor to obtain the now prescription only product, could instead use OTC paracetamol and/or ibuprofen medicines. Another 12% were assumed to visit their GP who would advise them to use OTC analgesics and/or non-pharmaceutical pain management options. It was not possible to determine from the MUCHE report whether these patients would go to their doctor only for low dose codeine medicines or whether they would undertake alternative therapeutic options that could have less risks or are more effective. Of those who do go to their doctor, some will continue to use low dose codeine medications and others will use different pharmacotherapy options as part of a pain management strategy.
It was not possible to verify the sample frame used in the study to determine the representativeness of the sample with respect to the current Australian consumers. For example, it was not possible to determine whether the sample frame was pharmacy customers. Selection bias (and reduced representativeness) would occur in this case because customers who are more frequent purchasers would attend the pharmacy more often and hence more likely to be invited to participate in such as study. Hence frequent purchasers of OTC products would be over represented in this study and while 63% of such customers might go to their doctor, it is possible that only 20% of non-frequent purchasers would go to the doctor. Hence the result of the survey is not representative of current consumers. In addition to not being able to verify the sample frame to assess selection bias, it was also not possible to verify exactly what questions had been asked of these respondents and whether this study referred specifically to S3 products or OTC medicines that are also available at a supermarket.
A report prepared for the Pharmacy Guild by Cadence[12] projected that there would be an additional 8.7 million GP visits as a consequence of up-scheduling low dose codeine combination medicines at a cost of $316.44 million as a result of patients attending a doctor for these scripts. The Cadence model assumes that 53% of all current use will continue if low dose codeine medicines if they were up-scheduled and the additional GP costs correspond to one consultation per prescription. The main reason for the difference in the Cadence results and the results presented for this analysis is that the Cadence model assumes that there will be one visit for every script whereas the model presented in this report uses the assumption, provided by the TGA and clinical stakeholders, that that patients can have up to five repeat scripts (as is currently the case for prescription codeine products on private prescription), that is, a ratio of up to six scripts per GP visit, reducing the figure of $316.44 million to as low as $52 million. Further, the Cadence model assumes that patients will either continue to use low dose codeine or presumably use other OTC medicines. The model presented in this report provides other options for patients. Furthermore, this Cadence report uses the results of the MUCHE study (discussed above) to inform their estimate of 53% of usage continuing. The limitations of the representativeness of the MUCHE study are discussed in the previous paragraph. The modelling presented in this report moves away from the average consumer approach in order to accommodate the variations reported by all stakeholders across patients and their potential responses and resource implications, and the variation in opportunities for improvements in QALYs.
The public domain reports on the MedsASSIST initiative were reviewed, and the most recent public data that was found was from an August 16th publication.[13] Referencing the IMS data, the two million transactions recorded in MedsASSIST from March to July 2016 represent around 10% of the total sales in 2015. This lower than expected coverage could be a consequence of the 65% coverage of pharmacies generally, the rate of uptake over these five months or it could be the result of patients selecting to go to pharmacies that did not use MedsASSIST; it is not possible to determine these drivers this from available data. It is also not possible to determine from this public data whether the 43,000 instances of non-supply over this period were for 1,000, 20,000 or 40,000 people, nor how many individual customers had purchased these items. Therefore, this data was not used to inform the inputs in this model.
Conclusion
Only the implementation of Scenario 4 (Options 4 and 6 - up-scheduling S2 and S3 codeine-containing medicines to S4) results in a net benefit to society. The drivers of the net benefit are improved quality of life following better treatment pathways (as patients pursue different options as a consequence of attending a GP consultation), deaths prevented, and net financial savings to consumers. The result of a positive net benefit for Scenario 4 is robust to a wide range of sensitivity analyses.