Netflix and pill: is there a role for volume-delinked subscription-style payments beyond antimicrobials?’

KEYWORDS:

• Subscription
• pricing
• reimbursement
• antibiotics
• digital health
• infectious diseases
• cell and gene therapy

1. Volume-delinked subscription payments

Volume-delinked subscription-style payments, also known as ‘Netflix-models,’ are a novel pricing and reimbursement model in pharmaceutical markets. Under this type of agreement, manufacturers receive a fixed (recurring) payment for the supply of a new therapy, independent of the quantity sold. This is similar to paying a fixed fee for access to, say, a streaming service, with the fee being independent of the number of programmes watched. their differs from the standard pharmaceutical payment model, whereby a set price is received upon units sold, and revenue price times quantity of units sold.

2. Use for antibiotics

The ‘Netflix-model’ has prominently been discussed in the context of hepatitis C drugs [1] and recently new antibiotics, where it has been proposed to overcome the market failures that led to a drought in the pipeline for new antibiotics. Due to antimicrobial resistance (AMR), antibiotics are under stewardship guidelines which restrict usage. Consequently, quantities sold are low and revenue expectations are so low that investors walk away. Compensating for low revenue expectations by increasing the price per unit sold, does not work either. Unlike most other medicines, antibiotics are necessarily studied in non-inferiority trials and cannot objectively demonstrate to be worth more than another existing (often low-cost generic) antibiotic to justify a higher value-based price. Furthermore, there is no precedent for paying for the benefits of new antibiotics over and above health gain to the individual [2–4], a problem shared with some vaccination programmes [5]. Even if their full value could be demonstrated, the enormous health and economic burden of AMR on today’s healthcare systems [6] and societies might show a novel antibiotic’s value so high that it creates budget-impact challenges in the short term; another issue for which subscription-style payment models have been proposed in other areas, e.g. gene therapies [7].

The traditional model of reimbursement failed to stimulate sufficient investment in research and development and undermines successful product launches in this area [8]. In response, a ‘Netflix-type model,’ was successfully piloted in England [9,10]. This approach is transformative in overcoming market failure for new antibiotics, but can it also work for other effective therapies that struggle to achieve timely patient access at a value-based price?

3. Benefits of a ‘Netflix-model’ in different contexts

To explore whether there are other useful applications of a ‘Netflix-model,’ two key questions are: 1) What are the key features of the ‘Netflix-model’? and 2) What problems can these solve?

3.1. The payment is volume-delinked, i.e. it is unrelated to the quantity of units sold.

As in the example of antibiotics, this feature allows to make products available to patients in cases where health systems are willing to pay for the therapy to be available as an option but routine use is rationed (e.g. stewardship principles).

In other contexts, this feature could remove the incentive for payers to ration use for cost-containment purposes. Consider infectious diseases e.g. where it may be beneficial at a societal level for more patients to be vaccinated to stymie spread of infection, than a cost-based rationing would allow. To that end, subscription models were trialled in Louisiana [11] and Australia [12] for direct-acting antivirals (DAAs) for the treatment of hepatitis C.

In the case of high-cost treatments (such as DAAs), the volume-delinked fee also smooths the financial ‘shock’ to the healthcare payer of each additional treatment, thereby mitigating affordability concerns. This works best for treatments that have low marginal costs of 1) production relative to their value as reflected in the price under per-unit pricing [12,13], otherwise the ‘shock’ transfers to the manufacturer; and 2) therapy delivery, otherwise an element of ‘shock’ remains for the healthcare system.

At the extreme, when the marginal cost of production is (close to) zero, such as with digital technologies, delinked agreements are efficient by reducing the administrative burden of per-unit payment.

3.2. The payment takes the form of a subscription, i.e. it is paid at regular intervals over a prolonged time period.

Subscriptions spread a large total payment over time providing benefits similar to annuitisation of an upfront sum, previously discussed in relation to payment challenges for cell and gene therapies [7,14].

Spreading payments over time can also increase the cost-effectiveness of a technology from the perspective of the healthcare payer, as economic evaluations discount future costs compared to today’s costs. However, manufacturers also discount future revenues and may require compensation for future payment instalments (i.e. in the form of interest). If well-designed, however, the total benefits of a ‘Netflix-model’ compared to the status quo should outweigh this.

Manufacturers and healthcare payers may both favour the certainty offered by a subscription model. Payers know what they are paying for how long to have access as needed to a valuable therapy, and manufacturers receive a guaranteed revenue. Depending on the context, it is possible that either party could be financially better off under standard per-unit pricing, but certainty has a value of its own, as does finding an agreement between stakeholders that allows for patient access. Both parties would benefit from a ‘Netflix-model’ if the fixed payment level is set higher than the revenues expected under per-unit pricing, but lower than unit price * volume (assuming the volume increases when the incentive to ration is removed).

Patients benefit from this model when those who can benefit from treatment get access. Evidence from Australia on DAAs suggests that ~100,000 extra individuals with hepatitis C were treated than would have been under per-unit pricing. If these people had been treated under per-unit pricing, this would have increased costs to the payer from $1bn, to $7.5bn [12]. Because the model was optional, the participation of payers and manufacturers suggests it was favourable to them [12].

Based on the above, we can identify features of a technology for which a ‘Netflix-model’ might be beneficial (Table 1). We mapped a selection of health technologies, some of which have been trialled under a ‘Netflix-model,’ against these features. Consider digital technologies, there are contexts under which their widespread usage would be beneficial for population health (e.g. contact tracing or symptom tracking) and have a low marginal cost of production and delivery, making a ‘Netflix-model’ attractive. For cell and gene therapies, the main benefits to payers and manufacturers arise from the certainty of the payment, and the ability to spread payment over time. However, a ‘Netflix-model’ for cell and gene therapies may be hindered by their higher marginal cost of production for the manufacturer and higher costs of delivery to the healthcare system, both of which may be perceived as risks.

Table 1. Features of health technologies for which a ‘Netflix-model’ may be useful.

Feature of the technology	Antibiotics	Treatments for infectious diseases	Digital therapies	Cell and gene therapies
Benefits from rationed usage (stewardship)	Y	N	N	N
Positive externalities of no rationing (increased usage) on population health	N	Y	<depends on the therapy and indication>	N
High cost technology	Y	Y	<depends on the therapy and indication>	Y
Low marginal cost of production	Y	Y	Y	N
Low delivery cost to the healthcare service	Y	Y	Y	N

Key: Y= feature is present for this technology; N = feature is not present.

4. Remaining challenges

Designing and implementing a ‘Netflix-model’ is not easy. A subscription amount must be set and this would ideally be value-based, but is necessarily based on limited evidence. For antibiotics, determining a value-based subscription amount that at least equalled a positive expected Net Present Value (to entice investment) and appropriately reflected the (additional) value of the specific product above that minimum, was particularly challenging [9,10,15] but may be less daunting for other technologies with fewer externalities to consider.

The duration of the agreement also requires careful thought. It needs to be long enough to guarantee availability of the product to the healthcare system and accrue sufficient revenue to the manufacturer, but not so long to deter competition [16]. Contracts must not prevent other potentially more (cost-) effective products from entering the market or disincentivise innovation. A term of 3–5 years has been suggested to balance these priorities [17].

Considering practical implementation, we suggest that a ‘Netflix-model’ should be seen as a complementary (optional) route to reimbursement, as suggested for similar types of payment models [18]. There may be different considerations for implementation in multi-payer systems, which may require all payers to opt into the model to realize the full effects (such as disease eradication), increase bargaining power, and avoid patient churn [13], as opposed to single-payer systems. There, the decision to implement may be simpler, but single-payer systems are often larger and less agile than smaller private payers. There may also be legislative or administrative hurdles to practical implementation, such as maximum values for contracts, which need to be considered at the local level.

5. What is the future for the Netflix-model in healthcare?

Society is moving toward subscriptions for many goods and services. Entertainment, cloud storage, and communication services are all examples of volume-delinked subscriptions, whilst volume-based subscriptions are gaining traction for various household products and leisure items such as bikes.

Healthcare seems to follow this trend. To help stakeholders identify areas where volume-delinked subscriptions could be useful, we have outlined a set of health technology features that are amenable to such a model.

We call on researchers, manufacturers and policymakers alike to question whether the status quo is the best option, or just the ‘easy’ option. With alternative payment models, such as the Netflix model and similar proposals gaining traction, alternatives may well be available to help solve some important incentive, affordability and/or reimbursement problems, increasing economic surplus and patient welfare along the way.

Declaration of interest

LS and GH have received grants and consulting fees from various pharmaceutical companies to OHE, outside the submitted work.

Reviewer disclosures

A reviewer on this manuscript has disclosed that they have received consulting fees from OHE. Peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.

Acknowledgments

None stated.

Additional information

Funding

This paper was not funded.