A life-cycle approach to monitoring benefits and harms of medicines

Regulatory agencies worldwide are giving greater recognition to postmarketing surveillance, as part of a life-cycle approach to maintaining safety of medicines.1,2 This approach involves monitoring the potential benefits and harms of each medicine, and managing risks, throughout development, preregistration testing, registration and postmarket use in the community.

The current grounds for initial and continued registration of medicines and devices in Australia with the Therapeutic Goods Administration (TGA) are “acceptable” safety, efficacy and quality. Risk-management plans — outlining how safety will be monitored and how risks will be minimised in the postmarketing period — have been a requirement of registration since 2009, reflecting the understanding that important evidence about drug safety continues to be acquired once a medicine is in use.3 In broad terms, postmarketing safety of medicines requires:

• ongoing acquisition of data to characterise benefits and harms

• quantitative integration of these data with the existing evidence

• management of identified risks

• management of uncertainty.

Acceptable safety can only be assured when appropriate data are available, and the types of evidence required differ according to the stage of the medicine life-cycle. When a medicine is first registered and introduced to the market, evidence regarding its efficacy, and particularly its safety, is often limited. Despite rigorous regulatory requirements, preregistration trials are constrained in terms of size and follow-up, and involve highly selected patient populations.4,5

Randomised controlled trials (RCTs) provide robust data on the efficacy of medicines (ie, the intended effects of treatment) because they avoid the confounding associated with selection to treatment.6 To avoid such confounding, it is essential to analyse RCT data on efficacy by intention to treat. However, largely because not all trial participants continue with their assigned regimens, these findings are generally biased towards finding no significant effect. This conservative and appropriate approach means that any observed differences in efficacy between trial arms are likely to be real. However, for adverse events, although analysis by intention to treat yields results that are unconfounded by treatment selection, estimates are also likely to be similarly diluted. Hence, in terms of quantifying adverse events, such estimates from RCTs are neither conservative nor precautionary.

Spontaneous reporting of adverse events provides important data on potential safety signals. Large-scale observational studies are suitable for investigation of medicine exposure in relation to generally unintended and unpredictable outcomes,6 which cover many adverse events, and enable detection of less common adverse events.

Ultimately, an assessment of safety depends on quantitatively weighing up the benefits and harms of a specific medicine. Even if the relative efficacy and safety are considered generalisable, overall absolute risk reductions and harms depend on the attributes of the actual clinical population that uses the product once it is marketed.7

Australia’s medicines safety system could be strengthened in a number of ways. First, communication of risk and engagement with consumers could be improved. The life-cycle approach to maintaining safety of medicines acknowledges that there are the higher levels of uncertainty about risks and benefits of medicines at the time of registration, and that evidence improves gradually. Consumers are central to the safety system, so it is vital that they are central to the life-cycle approach. Consumers need to be informed about the risks and uncertainties associated with specific medicines so that they can make informed decisions about medicine use. In recognition of this, a review of the TGA’s transparency recommended investigating an early postmarketing risk-communication scheme for therapeutic goods. An option proposed was the inclusion of an “early postmarketing” symbol alongside the names of medicines with a risk-management plan, to indicate to consumers and health professionals that these medicines are subject to enhanced postmarketing safety monitoring.8

Second, more pragmatic large-scale Phase IV RCTs are needed. These would enable greater generalisability of RCT results to the populations ultimately using medications.4

Third, there could be greater availability of and support for the establishment and analysis of large-scale datasets, to independently quantify medicine use and adverse events. Much postmarketing surveillance has been devolved to industry, on a drug-by-drug basis, which is relatively inefficient and lacks independence. Recent global initiatives are establishing postmarketing surveillance systems based on the electronic health data of millions of people to prospectively monitor the safety of medical products.9 In Australia, there is massive potential to contribute to this worldwide movement, building on unique data sources and existing high-quality Australian pharmacoepidemiological work. The establishment of integrating authorities to link our extensive health databases, along with recent postmarketing surveillance budget initiatives, are important developments, but greater support is required.

Maintaining the safety of medicines requires genuine and complex engagement of multiple stakeholders, with consumers central to the process. If appropriately supported and communicated, the life-cycle approach, combined with greater availability of high-quality data, has the potential to significantly improve safety of medicines in the coming decades.