Ethical Justifications for Access to Unapproved Medical Interventions: An Argument for (Limited) Patient Obligations

Abstract

Many health care systems include programs that allow patients in exceptional circumstances to access medical interventions of as yet unproven benefit. In this article we consider the ethical justifications for—and demands on—these special access programs (SAPs). SAPs have a compassionate basis: They give patients with limited options the opportunity to try interventions that are not yet approved by standard regulatory processes. But while they signal that health care systems can and will respond to individual suffering, SAPs have several disadvantages, including the potential to undermine regulatory and knowledge-generation structures that constitute significant public goods. The “balance” between these considerations depends in part on how broadly SAPs are used, but also on whether SAPs can be made to contribute to the generation of knowledge about the effects of health interventions. We argue that patients should usually be required to contribute outcome data while using SAPs.

Keywords::

• health care delivery
• research ethics
• experimental therapies
• regulatory issues
• end-of-life issues
• patient rights
• special access

ACKNOWLEGMENTS

We thank Dr. Eva Pilowsky for helpful insights into the Australian regulatory system and, together with Prof. Alison Avenell and Prof. Shaun Treweek, for comments on an earlier draft of the article.

Notes

1. These schemes are distinct from the phenomenon of off-label use of interventions, which we do not seek here to investigate. We note, however, that off-label use raises some overlapping issues with SAPs, as it also involves providing an intervention for an indication for which evidence is lacking. See, for example, Largent, Miller, and Pearson (2009), Ghinea, Lipworth, Kerridge, and Day (2012), and Isbister et al. (2008).

2. Many jurisdictions having a lower evidentiary standard for devices compared to drugs. Pharmaceutical agents typically go through four phases of testing: Phase I trials assess a drug's safety, usually among healthy volunteers; Phase II trials assess efficacy in a small cohort of affected patients; Phase III trials test efficacy in a larger cohort (often as a randomized controlled trial [RCT]); and Phase IV trials occur postapproval and examine long-term safety and effectiveness.

3. Although we note the growing criticisms of industry funded research and its potential to skew the evidence base. See, for example, Lundh et al. (2012). For further discussion of rights claims in relation to the Abigail Alliance case, see Leonard (2009) and Robertson (2006).

4. In the United States, a series of legal cases in the 1970s dealt with this issue. See Supreme Court of United States (1979).

5. Complete information was not readily available for all countries. Note that in some locations (including the United Kingdom) regulations have recently been under review.

6. For more detail on these and related programs see Edwards (2013), Institute of Medicine (1991), and Schuklenk and Lowry (2009).

7. We lack the space here to comment in detail upon the effects of the differing rules across international jurisdictions. In addition, data on the prevalence and kinds of use of SAPs are unobtainable in many instances, hindering development of further detailed analysis of this kind.

8. DiMasi et al. (2003) calculated that new experimental drugs had 40% probability of moving from initial IND application to phase I trials, 75% probability of progressing from Phase I to Phase II trials, 48% probability of moving from Phase II to Phase III trials, 64% probability of moving from Phase III trials to an application for approval, and a 90% chance, once submitted for approval, of receiving it.

9. Of course, these are possible outcomes of any intervention—approval by a regulatory body does not ensure that an intervention will do more good than harm in individual cases.

10. Bevacizumab (Avastin), for example, received accelerated approval by the FDA in 2008 for breast cancer, despite the lack of evidence that it extended life. When this approval was revoked in 2011, people who had used it previously protested vigorously (Carpenter, Kesselheim, and Joffe 2011).

11. To our knowledge this liability has not yet been legally tested in any country.

12. This could have a further implication for equity, as burdens of research participation may come to fall disproportionately on the disadvantaged if experimental interventions can be accessed without cost via trials compared with self-funding the same intervention via an SAP (Schuklenk and Lowry 2009, 16). Even where manufacturers are only able to charge for costs for interventions accessed via SAP, these costs may be higher than some patients can readily afford (Edwards 2006).

13. For these reasons we do not consider making ineligibility for trial participation a condition for SAP use a preferable option. However, we consider our below proposals for collection of data from SAP usages to be another way to address these issues.

14. With regard to patients, Harris (2005) and Schaefer, Emanual, and Wertheimer (2009) use similar considerations to draw stronger conclusions than we do here.

15. For example, registry-derived data identified the long-term dangers of metal-on-metal hip implants in Australia (TGA 2012), and in the United States informed the FDA's approval of left ventricular heart implants (FDA 2010).

Additional information

Funding

The research for this article was funded by a Macquarie University Research Development grant. The funder had no control over the findings.