Intended for healthcare professionals


Healing an ailing pharmaceutical system: prescription for reform for United States and Canada

BMJ 2018; 361 doi: (Published 17 May 2018) Cite this as: BMJ 2018;361:k1039
  1. Adam Gaffney, instructor in medicine1,
  2. Joel Lexchin, professor emeritus2,
  3. US,
  4. Canadian Pharmaceutical Policy Reform Working Group
  1. 1Department of Medicine, Cambridge Hospital/Harvard Medical School, Cambridge, MA 02139, USA
  2. 2School of Health Policy and Management, York University, Toronto, Ontario, Canada
  1. Correspondence to: A Gaffney agaffney{at}

Our pharmaceutical systems are broken, and only fundamental reform can ensure universal access to safer, more innovative, and more affordable drugs

Key messages

  • The US and Canadian pharmaceutical systems are dysfunctional

  • Costs are exorbitant, commercial goals distort drug development, misleading promotion fosters misuse, and medications are too often unaffordable for patients

  • We propose reforms that would provide universal drug coverage without fees at the point of use while reducing prices through negotiations with drug firms and, when needed, compulsory generic manufacture

  • Innovation would be enhanced through patent reform and by establishing new public agencies to fund drug development and clinical trials

  • Drug safety, efficacy, and prescribing quality would be improved by raising standards for approval and safety monitoring

  • While the proposals face formidable political obstacles, a popular mandate exists for pharmaceutical reform in both nations

Drugs are among medicine’s most powerful tools. Yet the pharmaceutical systems of the United States and Canada are mired in dysfunction. The industry’s pricing practices—charging whatever the market will bear, especially in the US—strain budgets and put vital medicines out of reach for many patients.1234 Despite some notable advances, the industry’s overall rate of real innovation remains incommensurate with our vast drug spending; many new drugs are marketed each year but few represent substantial clinical improvements.567 And commercial imperatives distort drug trials,8 research priorities, and drug regulation.910

While many recognize the need for change, proposed remedies vary3111213 and would fall short of achieving the fundamental reform that these deficiencies call for. The advocacy organization Physicians for a National Health Program therefore encouraged a working group of US and Canadian doctors, scholars, and advocates (the US/Canadian Pharmaceutical Policy Reform Working Group) to come together to craft a wide ranging reform proposal for both nations. Although political circumstances, including the influence of the pharmaceutical lobby, make full implementation of these reforms unlikely at present, shifting political winds may bring a more favorable policy climate. Hence, the working group aimed to craft an ambitious proposal for pharmaceutical reform to set an agenda for the future, including insurance coverage, pricing, drug development, clinical testing, regulatory approval, postmarketing monitoring, and promotion.

Although some of our recommendations (box 1) could be implemented within the existing US healthcare financing framework, full implementation would require a universal single payer system. Canada already has a single payer system but it would still require reforms because the system fully covers hospital and doctors’ services but not drugs out of hospital.1415

Box 1

Summary of proposed pharmaceutical reforms

1. Access to prescription drugs

  • Each nation would establish a formulary of all medically necessary prescription drugs

  • If agents with equivalent efficacy and safety were available, only the least expensive would be included

  • All residents would have full coverage for all formulary medications without copayments, co-insurance, or deductibles

  • When clinically necessary (eg, allergies), non-formulary alternatives would be covered

2. Drug prices

  • Government would negotiate with drug firms to lower prices

  • “Compulsory licensing” would allow generic manufacturers to produce essential patented medications if the patent holder refused to offer a reasonable price

  • Government would commission public production of essential drugs when price negotiation fails and no reasonably priced generic is available

  • New public divisions of the NIH and CIHR would develop non-patented drugs and make them available for low cost generic manufacture.

3. Preclinical drug development

  • Preclude patents for trivial modifications of existing agents, and restrict market exclusivity for me-too drugs unless they are shown to be more effective or convenient or have fewer side effects than others in the same class.

  • Repeal provisions of the Bayh-Dole Act in the US that allow private firms to obtain exclusive licenses for drugs developed through publicly funded research

  • Establish public drug innovation divisions in the US and Canada that would fund and oversee the early stages of drug development

4. Clinical testing

  • Require higher standards for clinical trials used in drug approval applications

  • Increase the transparency and public availability of (anonymized) clinical trial data

  • Publicly fund the majority of clinical trials through new “Clinical Trials Divisions” of the NIH and CIHR.

5 Drug approval reform

  • Full public funding of the drug regulatory agencies, ending their reliance on industry user fees

  • Less frequent use of expedited reviews

  • Restrict membership on regulatory advisory committees to experts without financial ties to drug companies

6. Postmarketing surveillance

  • Enforce requirements to promptly perform postmarketing studies

  • Increase funding and authority for regulatory agencies’ postmarketing monitoring programs

7. Promotion

  • Ensure that regulatory agencies have adequate resources to review promotional materials

  • Stiffen sanctions for misleading drug promotion

  • Eliminate tax deductions for expenditures for direct-to-consumer advertising and other marketing and, in some cases, exclude advertised drugs from the formulary

  • Promote academic detailing in lieu of industry detailing

  • Reduce the role of industry funding in continuing medical education and guideline development


Our proposal rests on six principles:

  • • Medical needs, not financial means, should determine access to medications

  • • Drugs must be affordable to society

  • • Drug development should be geared toward real innovation that maximizes population health

  • • The human right to health16 must take precedence over intellectual property rights (patents)

  • • The safety and effectiveness of medications must be independently and rigorously evaluated

  • • Comprehensive and unbiased information on drugs should be available to prescribers and patients.17

Access to prescription drugs

The right to essential medications is often compromised in both the US and Canada (fig 1). High out-of-pocket costs leave millions unable to fill prescriptions141518 and drive many into bankruptcy.1920 In the US an estimated 28 million people remain uninsured for healthcare,21 while 3.5 million in Canada lack drug coverage.14

Fig 1
Fig 1

Proportion of adults aged over 55 with cost related non-adherence to medication in 11 high income nations15

Cost sharing (copayments, deductibles, and co-insurance) also impedes access on both sides of the border.1522 It reduces needed and unnecessary care to similar degrees23; is a factor in reduced adherence2425; and, for some conditions, exacerbates racial disparities in health,26 raises non-drug healthcare spending, and worsens outcomes.2426 Notably, Wales, Northern Ireland, and Scotland have been able to provide universal drug coverage without cost sharing while using other cost control mechanisms to hold drug spending well below US or Canadian levels.2728

To improve access and population health, we propose universal,29 first dollar coverage (full insurance with no cost sharing) of all medically necessary drugs, echoing Archie Cochrane’s famous invocation that “all effective treatments must be free.”30 Each nation should establish a national formulary of covered drugs, which should include all medications shown to improve the length or quality of life—or the safest, most effective, and least expensive option when equivalent agents are available. A national technology assessment office would provide data on comparative effectiveness to guide formulary decisions. When clinically appropriate—eg, for allergies or other unique circumstances—off-formulary drugs should also be covered.

Drug prices

Spending on outpatient drugs is higher in the US ($1026 (£742; €833) per capita annually) and Canada ($713) than in other nations in the Organization for Economic Cooperation and Development (averaging $515, and as low as $240 in Denmark).27 High prices (especially in the US) rather than high use explain these differences. For example, in 2014 a daily 50 unit dose of insulin glargine cost $186.38 a month in the US (after applicable discounts) versus $63.65 in the UK and $46.60 in France.31

Despite claims to the contrary, research and development costs cannot justify these high prices.32 For instance, the total research and development expenditures of 10 firms that recently introduced new cancer drugs amounted to $9bn, while those drugs generated $67bn in revenues.33 Drug firms continue to sharply increase US prices decades after recouping development costs,1343536 and their mean profits are consistently threefold higher than the average of other Fortune 500 firms—23% v 7% in 2016.37

Several steps could reduce drug prices while ensuring that no uniquely effective medications are withheld. Each nation’s regulatory agency would continue to approve drugs without regard to price. Once approved, however, a public agency would negotiate with manufacturers over prices, guided (in part) by comparative effectiveness data. Experience internationally, and in the US, indicates that such negotiations can lower prices38—probably by about 50% for branded drugs in the US.27394041

While negotiations and a national formulary could reduce prices for many medications, when patented drugs lack competitors firms could still demand unreasonable prices, forcing nations to exclude the drug or strain their budgets.4243 Hence, additional options to assure reasonable pricing are necessary (fig 2). For instance, if price negotiations over branded drugs failed, governments would issue a compulsory license to allow generic manufacturing, a mechanism already sanctioned under international trade law,44 US patent law,45 and the Bayh-Dole Act.46 Indeed, in 2001, both the Bush (US)4447 and Chretien (Canada) administrations,48 facing fears of anthrax bioterrorism, threatened to break the patent on ciprofloxacin, causing Bayer to lower the price.

Fig 2
Fig 2

Pathways of drug development and pricing

In some circumstances, however, even compulsory licensing might not give reasonable prices; the cost of some generic drugs has soared after sole generic manufacturers cornered the market.135 We thus advocate creating public manufacturing capacity to produce drugs when no reasonably priced option is available. This capacity could also augment production during public health emergencies or drug shortages.49

Finally, drugs developed through public funding by public entities would remain unpatented and available for generic manufacture worldwide at greatly reduced cost.

Preclinical drug development

The patent protection and market exclusivity that prop up drug prices are typically portrayed as critical to encourage innovation. This portrayal is misleading for two reasons.

Firstly, despite achieving some important advances, the drug industry’s record on innovation is derisory relative to its vast revenues and profits.50 Most new drugs offer little new besides higher cost,2675152535455 while firms often extend market exclusivity through trivial modifications and secondary patenting.5657

Secondly, it is far from clear that patents are the most important stimulus to therapeutic advance. Throughout history, curiosity and the intrinsic rewards of discovery, rather than financial incentives, have often driven scientific breakthroughs. Even today, most basic research underlying later drug innovation is carried out in non-profit or public institutions and funded by the National Institutes of Health (NIH) and the Canadian Institutes of Health Research (CIHR). Before the 1980 Bayh-Dole Act, the fruits of publicly funded research remained in the public domain in the US. Since 1980, however, publicly funded researchers have been allowed to patent their discoveries and sell them to drug firms,58 as occurred with the hepatitis C drug sofosbuvir. Although Bayh-Dole permits government to break the patents of such drugs, this provision has never been used.46

Thus, we propose the repeal of Bayh-Dole to keep drugs developed with public funding in the public domain. Meanwhile, for drugs developed fully by the private sector, the patent system should be reformed to encourage innovative drugs, not more look alike, “me-too” agents.

In the US, the criteria for issuing drug patents have been stretched far beyond the original requirement that a patentable discovery had to be useful, novel, and non-obvious.5759 As others have argued,36061 patent reforms could both lower prices and advance innovation. Minor variations or combinations of existing agents, drug isomers,3 and tweaks to drug delivery devices that don’t add important functionality should not be patentable. Some countries have already mandated similar restrictions.62

Because the reforms we advocate risk reducing incentives for industry to develop marketable products from important new discoveries, we propose creating institutes for prescription drug development within the NIH and CIHR. The new institutes would have two divisions: for drug innovation and for clinical trials (fig 2). The drug innovation division would focus on the development of non-patentable agents to the point of clinical trials. This “public track” would—alongside private research—fund the development of novel pharmaceuticals. We propose public funding equal to about half of current preclinical private sector investment. All novel molecules developed by the division would remain in the public domain. This approach is a form of “delinkage” of drug development and pricing that others have proposed.6364

The drug innovation divisions might do some drug development themselves but would mostly fund efforts by academic or other non-commercial investigators. Priority would be given to potential drugs with the most clinical value, focusing on diseases that are neglected, commercially unprofitable, lacking effective treatments, or important for public health. The new, unpatented agents could be produced as generics by companies anywhere—a major advance for global health.

Clinical testing

Industry sponsored clinical trials have sometimes used unsound methods and reported incomplete findings, calling into question the interpretability, and sometimes the veracity, of their conclusions on safety and efficacy.8 For instance, trials have compared new agents with placebos rather than the best existing therapies, underdosed comparator drugs, or relied on surrogate endpoints65 that may not predict outcomes. Some commercially funded researchers have also selectively published (and republished) positive results 866 or concealed negative findings,67 while firms have ended trials prematurely for purely commercial reasons.68

Meanwhile, corporate ownership of trial data can obscure safety problems and impede further research.69 Although requiring preregistration of trials has been an important step forward, transparency problems persist.70

Drug regulatory agencies must therefore raise evidentiary standards. Trials should, whenever possible, compare new agents to existing therapies and use a superiority design to discourage investment in unneeded me-too drugs. When new agents mimic existing ones, they should generally be tested in patients who do not respond to (or tolerate) existing products. And with infrequent exceptions, trials should assess hard clinical (rather than surrogate) outcomes.71 Anonymized patient level data from all trials (including older trials), should be made publicly available70 (whether or not a drug gains approval) to facilitate accountability and further research.

Finally, because of concerns regarding the objectivity of industry funded trials and the need to test unpatented and unprofitable therapies, the clinical trials divisions within the new NIH and CIHR institutes would also fund and oversee trials (fig 2).6972 The divisions would select promising molecules developed by non-profit laboratories, academic investigators, and drug companies for clinical trials, which would mainly be designed and conducted by non-commercial investigators. They might also fund trials assessing new indications for existing agents or non-drug therapies.

Publicly funded trials would offer important benefits: minimizing commercial conflicts of interest; redirecting research from “me-too” drugs toward real innovations, and facilitating the development of unprofitable but essential treatments.6972 Although firms could still fund trials of their products,72 because clinical trials are costly and would be subject to enhanced regulatory scrutiny (based on past evidence of companies manipulating results), publicly funded trials would be likely to predominate in the long term.

Drug approval reform

Canadian and US regulatory agencies too often allow unsafe drugs to reach the market73747576 and inadequately monitor them after approval.777879 Both agencies’ independence has been eroded by their reliance, starting in the 1990s, on funding from fees paid by drug companies. In the US, the FDA’s receipt of these funds is explicitly linked to its shortening of review times.7375

Meanwhile, an increasing proportion of new drugs qualify for programs that further reduce review times. By 2014, 69% of drugs submitted to the FDA gained “expedited review” through various designations or pathways.80 The comparable figure for Canada for 1997-2012 was 26%.81 Although intended to accelerate the availability of innovative agents, these programs have been exploited to speed the marketing of many “me-too” drugs.780 Some of these expedited review pathways have weaker standards of evidence. The recently enacted 21st Century Cures Act in the US creates even more such pathways, and mandates that the FDA evaluate the potential use of “real world evidence”— ie, not from clinical trials—for approving new indications for drugs.8283

Such evidentiary changes may increase the risk that unsafe drugs will enter the market.7684 And most,7374758586 but not all,84 studies suggest that shorter review times are deleterious.

We propose several reforms to the drug approval process: Firstly, industry funding of drug regulatory agencies should be ended; governments should fully fund agency budgets. Secondly, expedited review should be reserved for drugs likely to offer genuine clinical advances. For instance, “first in class” drugs should not automatically qualify for expedited approval since many are not superior to existing products.7 Thirdly, requirements that trials use hard clinical endpoints and active comparators should be waived only in exceptional circumstances. Fourthly, while experts who receive commercial funding may appropriately offer testimony before advisory panels evaluating drugs, such experts should not be allowed to participate in the panels’ voting or decision making.87 Finally, drugs should be required to demonstrate superiority—whether in efficacy, safety, or convenience of dosing or administration—over any existing agents to be eligible for market exclusivity.

Postmarketing surveillance

As regulatory agencies have approved more drugs based on surrogate endpoints and smaller or fewer clinical trials, they have often mandated postmarketing studies to confirm benefits or exclude serious risks.77 However, this approach has serious shortcomings. Though large postmarketing studies are critical to assuring safety (especially for rare side effects), they should not be an excuse for weakening preapproval safety requirements. And while big data approaches to pharmacosurveillance (eg, the FDA Sentinel System) hold promise, their results to date are modest and cannot substitute for clinical trials.88

Unfortunately, enforcement of mandated postmarketing studies is currently lax. The FDA has failed to fully use its authority to penalize firms that don’t complete such studies,7778 while Health Canada has allowed firms to continue marketing drugs for years without completing required trials.79

We propose several reforms to upgrade postmarketing safety efforts. Funding for such efforts within the FDA and Health Canada should be increased to a level on par with spending for review of new drug applications, and safety offices should have equal position in these agencies’ hierarchies to offices tasked with drug approval. Safety monitoring offices should be empowered to independently order safety warnings and remove unsafe drugs from the market, and agencies should use their legal authority more aggressively to pursue drug companies that fail to complete required postmarketing studies on time. Finally, information about delays must be made publicly available.

Some of these reforms could be accomplished without legislation: since 2007, for instance, the FDA has had authority to penalize companies that failed to conduct timely postmarketing studies. Yet it has not exercised that power in any meaningful way.78 Recent legislation allows Health Canada to levy substantial fines in case of company non-compliance.89


Drug promotion—including industry “detailing” of physicians’ offices—consumes billions of dollars annually, more than total expenditure on medical student education in the US909192; expenditures for sales and marketing exceed those for research and development.13 In addition to diverting funds that might be better used to develop lifesaving medications, such promotion is often misleading or inaccurate.939495 This is especially true for direct-to-consumer advertising (DTC)—now widespread in the US96 and, in attenuated form, Canada.97 Advertising that mentions the brand name of a prescription-only medicine along with its indication is banned in all other developed nations except New Zealand.

Promotional spending dwarfs the tiny budgets of the FDA and Health Canada components that regulate marketing. The FDA is overwhelmed by the sheer volume of materials to review,9899 and Health Canada has delegated most of the regulatory oversight of promotion to third parties.97

We propose a major expansion of promotional review. Regulatory agencies need more (and more predictable) resources to carry out rigorous assessments of all promotional materials.99 They should not have to rely on funding contingent on meeting deadlines to complete reviews, which can foster a lenient approach, and money should come only from government to avoid conflicts of interest.99

Improved monitoring should be coupled with stiffer sanctions for misleading or off-label promotion. In the past, even massive fines haven’t deterred industry violations100 because, as one expert noted, “When you’re selling $1 billion a year or more of a drug, it’s very tempting for a company to just ignore the traffic ticket and keep speeding.”101 Hence, authorities should be empowered to suspend firms’ right to promote their products or, in extreme cases, pursue criminal complaints against drug executives.

While we also favor prohibiting direct-to-consumer advertising and industry detailing, constitutional challenges based on “commercial speech” rights may preclude such bans in the US.102 However, other tools are clearly constitutional, such as eliminating tax deductions for promotional activities; additionally, when alternative treatments are available, drugs promoted in these fashions might be excluded from the formulary. Industry detailing could also be countered by not-for-profit “academic detailing”103 to optimize physician prescribing practices.104

Finally, industry funding can bias continuing medical education (CME)105 and clinical guidelines.106 Licensing authorities should not accept industry funded CME for mandated credits. CME could, instead, be undertaken and coordinated by a body similar to the Australian NPS MedicineWise (, while clinical guideline development should, at a minimum, follow the recommendations outlined by the Institute of Medicine.107

Economics of a national pharmaceutical program

Although our proposal would have large economic and budgetary implications, a detailed examination of those effects is beyond the scope of this article. Others have estimated that a national pharmaceutical program for Canada could save $7.3bn of the $22bn currently spent annually on prescription drugs in that nation, although that estimate did not contemplate the new investments in drug research, development, and regulation that we advocate.108 For the US, we believe that savings on drug prices through the mechanisms detailed above could fully offset the added costs of universal, first dollar drug coverage and new public investments that we recommend.

Achieving change

Jonas Salk, inventor of the polio vaccine, eschewed patenting, declaring: “Could you patent the sun?” Today, in contrast, profiteering too often reigns, to the detriment of population health.

Our proposal calls for a fundamental reorientation of drug policy: it would make drugs more affordable for patients and society, promote innovation, strengthen efforts to assure the safety and effectiveness of medications, and upgrade the evidence available to prescribers and the public. Because drugs developed through the proposed new public pathways would remain in the public domain, they could be produced generically throughout the world, benefiting many nations.

The reforms we advocate face formidable political opposition, especially from drug firms, with those in the Fortune 500 in the US alone making total profits of $67.7bn in 2016.37 However, most Americans—both Democrats and Republicans—now favor government action to lower drug prices,109 and 91% of Canadians support a universal pharmaceutical benefit.110 These are unmistakable popular mandates for change. The trail from sentiment to policy will doubtless be arduous. Yet history is replete with examples of sweeping reforms—often enabled by unpredictable shifts in political circumstances—that overcame entrenched interests. We aim with this proposal to provide a blueprint for reform that anticipates—and may kindle—transformative changes in our nations’ pharmaceutical systems.


  • Contributors and sources: This proposal was drafted by a writing committee comprising Adam Gaffney (cochair), Joel Lexchin (cochair), Marcia Angell, Michael Carome, David U Himmelstein, Gordon D Schiff, Sidney M Wolfe, and Steffie Woolhandler. Other members of the US/Canadian Pharmaceutical Policy Reform Working Group were: Brook Baker, Monika Dutt, Marc-André Gagnon, Gordon Guyatt, Ritika Goel, Brian Hutchison, Richard Klasa, Michael C Klein, Danielle Martin, Barbara Mintzes, Karen S Palmer, Danyaal Raza, and Robert F Woollard.

  • Competing interests: We have read and understood BMJ policy on declaration of interests and declare the following interests: JL has received payments from non-profits for consulting on projects that investigated indication based prescribing and which drugs should be distributed free of charge by general practitioners. He received payment from a for-profit for being on a panel that discussed expanding drug insurance in Canada. This proposal has been endorsed by Physicians for a National Health Program and Canadian Doctors for Medicare; authors and working group members are active in both organizations. Physicians for a National Health Program is not-for-profit organization that advocates for a single-payer healthcare system for the United States. Canadian Doctors for Medicare is a not-for-profit organization that advocates on behalf of Canada’s public single-payer system.

  • Provenance and peer review: Commissioned; externally peer reviewed.


View Abstract