From a drug safety perspective, one key issue in drug development is the (very) low probability of observing (very) rare adverse events in preapproval clinical trials, even in large therapeutic confirmatory trials. Such side effects are probabilistically much more likely to surface once the drug is widely used, and unfortunately some of these side effects may be extremely serious. The “rule of threes” is instructive here (Strom 2005). The number of individuals participating in a clinical trial that would be needed to be 95 % confident that a single case of an identified adverse event of interest would be seen is approximately three times the reciprocal of the frequency of the event in the general population. That is, for an event that occurs in 1/1,000 individuals, a sample size of 3,000 subjects would provide 95 % confidence of observing at least one event. For adverse events that are considerably more rare (e.g., rhabdomyolysis, torsades), much larger sample sizes would be needed (e.g., 30,000 and 300,000 for events with frequencies of 1/10,000 and 1/100,000, respectively). Trials of this magnitude are infeasible from both cost and time perspectives. Cobert (2007, p. 11–12) commented on this phenomenon as follows:

Should the [adverse drug reaction] be dramatic and rapidly discovered, such as torsades de pointes, aplastic anemia, or rhabdomyolysis, there will be a torrent of recriminations about why this was not discovered earlier during the [preapproval] clinical testing. The correct response is that the testing of only 5,000 to 10,000 patients could not pick up such a rare event. This response is usually lost in the clamor.

Therapeutic use clinical trials and postmarketing surveillance therefore play critical roles in cardiovascular drug safety and indeed in many other domains of drug safety.

A second issue concerns the fact that preapproval clinical trials typically employ relatively homogeneous participant samples. For example, potential participants who have other illnesses or medical conditions, including renal and hepatic impairment, are typically excluded, as are those taking (at least certain) concomitant medications. Additionally, the age range of participants can be fairly limited. A third issue concerns the length of time that a patient may take the new drug: for chronic diseases this is likely to be (very) much longer than the treatment period in preapproval clinical trials. The long-term safety of a drug that is suitable for chronic administration is therefore not fully known at the time that the drug is approved. Fourthly, and of particular salience for certain drug classes, its propensity for abuse is not known, nor is the likelihood that patients will develop a dependency on the drug (some of the discussions in the following chapter touch upon this issue).

Another important point concerns how drugs are actually taken by patients. Howren (2013) observed as follows:

Adherence is a term used to describe the extent to which an individual’s behavior coincides with health-related instructions or recommendations given by a health care provider in the context of a specific disease or disorder. The term has been used extensively in psychology and medicine in reference to acute, chronic, and preventive treatment regimens (e.g., a course of prescribed medication, wound self-care), preventive health screenings, dietary restrictions, exercise recommendations, and other health behaviors.

While the negative effects of patient nonadherence have been known for decades (Howran 2013) and authoritative sources such as the World Health Organization and the American Heart Association put average nonadherence among those with chronic diseases around 50–75 %, the literature still reveals a predominance of discussions of, and further research into, the problem rather than offering immediate action plans (Turner 2013). This topic is addressed in the following chapter in Sect. 15.​6.

Therapeutic use trials are conducted once a drug is on the market. They may be optional studies or studies required by a regulatory agency as a condition of approving the drug for marketing. In the former case, the biopharmaceutical company sponsoring a trial may wish to know more about the drug’s performance in individuals who were not well represented in preapproval trials, e.g., those with compromised liver function and/or taking several concomitant medications. Other sponsors, such as an institute within the National Institutes of Health, may wish to explore a drug’s place in future treatment guidelines, and hence conduct trials comparing its safety and/or effectiveness with other treatment options. Therapeutic use trials can be experimental or nonexperimental, nomenclature that is explained in the respective following sections.

While various definitions can be found in the literature, Man and Andrews (2002, p. xvii) regarded pharmacovigilance to be “the study of the safety of marketed drugs under the practical conditions of clinical usage in large populations.” Shakir and Layton (2002) provided a more encompassing definition that also included preapproval monitoring, detection, and evaluation of drug safety hazards. Stephens (2004, p. 2) provided a comprehensive list of the aims of pharmacovigilance:

Strom (2005, p. 3) defined pharmacoepidemiology as “the study of the use of and the effects of drugs in large numbers of people.” Dimensions of interest include safety, effectiveness, utilization, and cost. Going back to randomized clinical trials for a moment to allow a subsequent comment to be put into context, Matthews (1999, p. i) made the following comment at the turn of the millennium:

Over the last two to three decades, randomised concurrently controlled clinical trials have become established as the method which investigators must use to asses new treatments if their claims are to find widespread acceptance. The methodology underpinning these trials is firmly based in statistical theory, and the success of randomised clinical trials perhaps constitutes the greatest achievement of statistics in the second half of the twentieth century.

At the time that quote was published, it is likely fair to say that the statistical methodologies available to the discipline of pharmacoepidemiology were less well developed than for randomized clinical trials. In a paper published 8 years later entitled “In defense of pharmacoepidemiology–embracing the yin and yang of drug research,” Avorn (2007) commented as follows:

We forget how difficult it was to establish the rules of the road for conducting randomized trials. In terms of design theory and public policy, drug-epidemiology is now where randomized trials were in the 1950s. We have much to learn about methods, transparency, and protecting the public’s interest. But that work can be done, and we often have no other way of gathering vital information.

In the last decade, tremendous advances have been made: see Consiglio et al. (2013), Hennessy and Strom (2015), and Hennessy et al. (2016).

The terms used in the previous two sections, pharmacovigilance and pharmacoepidemiology, would probably both be suitable to reflect the discussions in this section. The term postmarketing surveillance has been chosen since it makes clear that our focus here is squarely in the post-approval domain.

Postmarketing surveillance involves surveillance for occurrences that have been identified in advance as potential safety concerns and also for unanticipated events. Regulatory agencies may approve a drug based on their belief at the time of approval that the benefit–risk balance is favorable, but also make it clear to the drug’s sponsor that their assessment of the benefit–risk balance’s favorability could change if certain adverse events suggested in preapproval trials materialize to a concerning degree during postmarketing surveillance.

In the previous chapter, we discussed the large cardiovascular safety outcome trials that are currently required for the prospective exclusion of unacceptable cardiovascular risk associated with new antidiabetic drugs for type 2 diabetes. There is currently considerable interest in alternative ways of exonerating a drug (see Sager et al. 2015): however, these methodologies can also be applicable in other therapeutic categories.

Evidence-based medicine was defined by Sackett and colleagues (1996) as follows:

Evidence-based medicine is the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients. The practice of evidence based medicine means integrating individual clinical expertise with the best available external clinical evidence from systematic research. By individual clinical expertise we mean the proficiency and judgment that individual clinicians acquire through clinical experience and clinical practice. Increased expertise is reflected in many ways, but especially in more effective and efficient diagnosis and in the more thoughtful identification and compassionate use of individual patients’ predicaments, rights, and preferences in making clinical decisions about their care.

There are two components to evidence-based medicine and two related sets of responsibilities. The first component is clinical research. Clinical research is a scientific endeavor that provides evidence concerning potential therapeutic interventions. Once clinical trials have been conducted, results are published in appropriate peer-reviewed journals. Everyone involved in clinical research has the responsibility to provide the best possible evidence in this manner (a topic discussed in the following chapter with particular attention to safety results).

The second component of evidence-based medicine is clinical practice. Clinicians have the responsibility of providing the best possible care to each of their individual patients. One part of being able to provide this optimum care is remaining aware of pertinent evidence that is published in clinical communications (no small task given the enormous amount of publications per year). It is also incumbent on clinicians to be able to decide for themselves if the evidence presented in a clinical communication is solid evidence and if the conclusion conveyed by the publication is justified based on the results presented. As Katz (2001, p. xvi) commented:

Part of the burden for the responsible cultivation of higher standards and better outcomes in medicine falls, naturally, to researchers and those who screen and publish their findings. But application is ultimately the responsibility of the clinician, who is obligated to consider not only the pertinence of particular evidence to his or her practice, but also the adequacy and reliability of the evidence itself.

Therefore, an appreciation of study design, experimental methodology, statistical analysis, and clinical interpretation is vital for clinicians who must decide whether the evidence presented in journal publications is adequate and reliable and therefore constitutes an appropriate basis for clinical care.