We have arrived at a special moment in health care. Innovative, life-changing gene therapies are here that will cure or ease debilitating diseases. Yet these expensive treatments are entering a market structure that was not built to price them.

Congress will likely need to play a part in developing a new paradigm for financing such treatments. As a senator and a physician, I have been following this issue closely, and see several possible paths forward. There are multiple issues to address.

First, the current model places the responsibility for purchasing treatment on a patient’s health insurer at the time of the treatment. Gene therapies can carry multimillion-dollar price tags. Although they may actually save money in the long run compared to the health care costs associated with the diseases they treat, those savings might not accrue to the initial payer because patients often change health insurers. This model would incentivize an insurer not to cover the therapy, or to discourage access by delays in approving its use, requiring prohibitive cost sharing, or other means.

Second, individuals who have commercial insurance pay more for therapies than those covered by Medicaid. There will be some genetic diseases, such as sickle cell disease, in which the majority of patients are covered by Medicaid. There are two sides to this. If Medicaid covers the majority of the individuals with these diseases, it imposes a tremendous burden on taxpayers. The flip side is that Medicaid pays less for a treatment than commercial insurance, so pharmaceutical companies charge those who are commercially insured more. This can lead to serious problems with affordability for the privately insured. If a family with two children with a genetic disease are on a small group policy and the gene therapy is priced at $2 million, that would be a $4 million charge in one year spread over three to 50 people.

Regarding the impact upon taxpayers, a 2009 paper by Teresa L. Kauf and colleagues estimated that individuals with sickle cell disease consume almost $1 million worth of health services by age 45. With approximately 70,000 sickle cell disease patients nationwide, a gene therapy priced at $1 million to reflect the avoided cost of future health services would cost $70 billion. Since there would be a moral imperative to treat individuals as soon as possible, this would be incurred over a one- to two-year period. With an estimated 3,000 individuals with sickle cell disease in my home state of Louisiana, the expense would be $3 billion over one to two years. And this is for just a single genetic disease. In short, gene therapy may not be affordable for federal and state government budgets.

This leads to the third issue. Therapies given at different times of a patient’s life have differing values. To stick with the example of sickle cell disease, imagine if gene therapy for it is given to a child in utero or at birth. This could save $1 million in medical costs by the age of 45. But if the gene therapy is given to a 45-year-old with sickle cell disease, there may be negligible or no savings because the damage the disease does to the brain, heart, bones, and other parts of the body has already occurred and will require additional costly treatment. This means we need some judgment as to the relative worth of a product given at different ages and/or relative to the presence of preexisting irreversible damage to organs.

A fourth issue is that many conditions that may be cured by forthcoming therapies are for rare diseases with small or ultra-small populations The entirety of the expense for research and development must be recovered on delivering treatment to these few individuals. What is a reasonable price for such treatments? It is unlikely that a competitor treatment will be developed and marketed for such a small population, so competition cannot be relied on to lower prices. Is it whatever the pharmaceutical company decides the market will bear? Presumably, the launch price will be negotiated by payers and manufacturers, but it must be adequate to reward and encourage innovation.

This brings me to consider outcomes-based arrangements. Some treatments do not work for every patient, will work only partially, or are not needed for everyone with a potentially harmful genetic profile. As an example, take alpha-1 antitrypsin deficiency, an inherited disorder that can cause lung and liver disease. Individuals who inherit only one gene for the disease (heterozygotes) may never develop symptoms of the disease and if they do the symptoms may appear late in life. Among individuals who inherit two genes for the disease (homozygotes), symptoms appear earlier and are more severe. So a curative therapy for alpha-1 antitrypsin deficiency might have one price for homozygous patients and a lower price for heterozygotes based upon both the likelihood and the severity of the disease.

A final point to consider is that a particular gene therapy may not completely heal or cure an illness but rather may only ameliorate it. To illustrate this point: if a gene therapy is only partially effective, perhaps because it is given after irreversible damage has already occurred, an individual may survive but for the rest of his or her life require continued and potentially expensive medical support. In this case, the lifetime cost of care may actually be increased by gene therapy as it improves survival but the patient still requires intensive medical care.

With these considerations in mind, I describe five payment approaches that have been proposed. There are certainly others.

1. As Mark R. Trusheim, Peter B. Bach, and I described in an article in the Journal of the American Medical Association, the subscription or Netflix model establishes a single set of predetermined payments and as many affected patients are treated as possible for those payments. The state of Louisiana is currently using this model to treat people with hepatitis C.
2. A variant of the Netflix model involves pledging or setting aside either a given amount of money to pay for a curative therapy for a particular disease or the amount to be spent per patient. This could be an international fund, as some nations have a greater prevalence of some diseases than others. An innovative company can then decide what amount of resource devotion is worth that financial incentive.
3. A pay-over-time model would require the insurer currently covering a patient to pay a portion of the cost instead of the entire cost. One disadvantage of this model is that insurance companies would have an incentive to avoid insuring someone who had received gene therapy and for which there would be an obligation to pay. An advantage, however, is that it is a form of value-based reimbursement. I’ll use as an example the hypothetical gene therapy for sickle cell disease priced at $1 million. As mentioned, if given to a child in utero, it saves $1 million worth of medical expenses by age 45. The insurer’s annual payment would be $1 million divided by 45, or $22,222. If given in utero or at birth and the patient dies at age 10, then the insurers would have paid $22,222 for 10 years with a total payment of $222,220. If the gene treatment is given to someone age 45 and the patient dies within the year, the payment would only be for one year at $22,222.
4. Medicaid or some other government program could pay for gene therapies and commercial payers would not have to cover them. Instead, it would be fully backstopped by public funds. Society as a whole pays, as opposed to just the insurance pool to which the patient belongs at the time the therapy is given. This leaves open how to establish a price. Should taxpayers be willing to pay any price without a proportionate relationship to individual and social good?
5. Create or encourage a private-payer collaboration for a curative therapy fund. Curative therapies would be carved out of the usual capitated payments and/or premiums for all public and private health plans. A patient needing gene therapy would receive the treatment and the cost would be paid out of this separate account to which all insurance plans contributed and which would be dedicated to covering gene therapy. For each new therapy developed, the pricing would have to reflect the amount of money available or the income into the fund would have to increase. Since payment is carved out of premiums, all insurers/users of the gene treatment would get the same price.

Life-changing gene therapies are coming. We must give thought now on how to determine the price of these innovative, new-age treatments and how to finance them to ensure that we realize their full, beneficial potential while also ensuring that society can pay for them.

I do not believe that any single one of these proposed models can do that in all cases, and at least for now we need to explore them all. The tragedy would be if the benefits of innovation were denied because of an inability to afford the innovation. It is an issue which must be addressed.

Bill Cassidy, @SenBillCassidy M.D., has served Louisiana as a Republican senator since 2015.