Post-marketing Observational Studies

Providing insight into real-life clinical practice regarding safety, efficacy and patients’ drug compliance.

---

by John Sampalis, PhD

Once the data from Phase I and II clinical trials have produced sufficient evidence to indicate a drug is safe and will produce the beneficial results that it was intended to, Phase III trials are conducted in order to further establish its effectiveness and to compare it to existing alternatives. The main purpose of Phase IV trials is to provide new or further confirmatory evidence regarding the safety and efficacy of a drug by expanding the number and spectrum of patients tested.

The results of Phase IV studies often provide evidence regarding new indications for a drug or identify populations with idiosyncratic response profiles. Therefore, even in Phase IV studies there is some degree of control of the conditions under which the drug is administered, the population of patients that is included, the duration and methods of followup, as well as the outcome measures used. Although less controlled than earlier trials, the Phase IV trial continues to provide data on efficacy rather than effectiveness. In this sense, Phase IV studies provide results that are more representative of “real-life” settings compared to other controlled trials.

In order to obtain evidence of benefits and safety of an intervention in real life, the real-life situation must be studied. This would require that the intervention is applied by clinicians in a non-controlled manner to patients that would, through the clinician’s judgment, be appropriate candidates to receive the intervention. The frequency and duration of followup should not adhere to a specific protocol but should follow the clinician’s routine practice. Furthermore, patient characteristics — including attributes affecting compliance, co-morbid conditions and concomitant medication use — should not be controlled for the purposes of the investigation. It is only under these conditions that data representing real-life effectiveness can be generated.

From a societal point of view, the effectiveness and safety of an intervention represents only one part of the equation. In order to estimate the benefit/risk ratio of a particular treatment, it is necessary to know the following: first, what are the expected benefits of the treatment in the target population; second, what are the expected potential harmful effects of the treatment in the target population; and third, what will be the overall benefit/risk ratio if the treatment was approved for use and reimbursement within the particular health-care system. For example, the high cost of a particular drug may limit its access by a significant proportion of the population, thus rendering it ineffective to society. As a consequence, a less-expensive and probably less-efficacious treatment may be approved and used.

The Post-marketing Observational Study (PMOS) is a non-interventional, observational, epidemiological study aimed at addressing these issues. Specifically, the PMOS provides data describing the safety and effectiveness of a treatment as it is applied in a real-life setting for the population and indication for which it was developed. The design of a PMOS differs significantly from that of Phase I to IV studies. The most important distinguishing characteristic is that there is no control by the investigator with respect to patients, duration and frequency of followup, and the method by which patients are managed in general. This truly reflects real-life circumstances. In addition, the treatment or prescription in the PMOS is not provided by a sponsor but is reimbursed either by the regular third-party payers or the patient.

Essentially, the patient will receive a prescription for the medication or will be referred for treatment as per usual practice. At that point in time, the patient is entered into the study cohort and will be followed prospectively without any intervention for the specified period of time, which is usually longer than the followup periods used in Phase I to IV studies. The data recorded for a PMOS are similar to those recorded for Phase IV trials but are somewhat less detailed. Information regarding patient demographics and characteristics, including co-morbidity and concomitant medication use, will be recorded at baseline, defined as the time period just prior to initiation of treatment. Information on other factors that might affect study outcomes, such as compliance with treatment, should be recorded at baseline or during the followup period. PMOS outcome measures are usually well-accepted clinical indicators of treatment efficacy but not standardized outcomes as those used in controlled studies.

In recent years, there has been increased concern regarding the safety of approved medications. The review process within regulatory organizations, such as the U.S. Food and Drug Administration or Health Canada, is under scrutiny. There is a perception that a significant gap exists in the post-marketing surveillance of approved drugs. Results obtained from clinical trials or studies using administrative databases have had significant impact on popular belief and have led to panic, generating a sensationalistic media frenzy. Unfortunately, these results have been interpreted as indisputable evidence and have caused the public to lose trust in two of the most important health-care stakeholders — the pharmaceutical industry and the government regulatory agencies. A properly designed PMOS may have prevented the current situation by either demonstrating that the safety concerns were not as significant as currently perceived, or by identifying the signal and sounding the alarm earlier, before the frenzy started.

The PMOS is a term used to describe a modern application of the prospective cohort epidemiological study in the pharmaceutical industry. This term, however, has been severely abused in recent years. Within this title, pure marketing initiatives such as seeding trials have been concealed. Although there have been attempts to disguise marketing campaigns as PMOS, there are fundamental and critical differences. By definition, the PMOS is an observational study. As such, no intervention with respect to the patients treated or the method by which treatment is allocated and administered is permitted. As a result, it would be inappropriate for the investigator to predetermine the selection process of study participants. While PMOS is aimed at providing data regarding real-life effectiveness, seeding trials and marketing initiatives are aimed at converting market shares. Of course, positive results demonstrating safety and efficacy would result in improved sales. However, this is not the objective of the PMOS — although the combination of good science and potential benefits to the sponsor would not be inappropriate.

If properly designed and executed, the PMOS will not only provide valuable information to regulatory agencies and patients regarding safety and effectiveness, but could also assist pharmaceutical companies in growing their profits. The challenge will be to understand that these benefits will not be accrued in the short term, but in the long term, probably in the next two or three fiscal years. The PMOS should be perceived as the process by which evolution of the drug’s utility to society is under continuous evaluation. The PMOS generates data that are translated to scientific information and then used as evidence to better understand a drug’s real-life effectiveness, providing accurate risk-management signals. Evidence generated in a PMOS should be used to convince health-care providers and patients that the specific drug is safe and effective. In addition, the data can be used to identify patient populations for which the effectiveness may be altered or that are at high risk for adverse reactions. Using this information to further guide health-care providers will improve the overall benefit/risk ratio of the drug from the societal perspective.

Within the framework of understanding how a particular treatment may be effective from a larger societal point of view, one important factor is only partially understood and often ignored: the process by which a clinician chooses one treatment over another. Medical decision-making involves consideration of available evidence, the patient’s individual characteristics, and expected results of intervention. The golden Hippocratic rule of “first, do not harm,” in today’s times has become “first, do less harm than good.” Clinicians are often challenged in their decisions by the delicate balance between risk and benefit of a treatment, taking into account the expected and accepted risk tolerance of the specific patient. Clinicians are expected to translate data into information, information into evidence, and evidence into clinical practice.

At this point in time, we have very little empirical evidence that will allow us to model the real multifactor process by which a treatment decision is made. Competing companies assume these decisions can be influenced by conventional marketing tools. It is interesting to observe how little focus has been placed on changing the mindset and beliefs of the prescribers, and how much is invested in temporarily changing their next individual prescription. We seem to neglect the fact that changing a singular action is only a short-term solution, whereas converting the scientific beliefs and biases of the prescribers has a lasting effect. The PMOS is the single most valuable tool for allowing us to study and understand what influences physicians and patients to prefer one course of treatment over another. Proper use of this information would allow marketing departments to develop programs that could convey information that would effectively change the beliefs and practices of physicians and inspire patient confidence. Inevitably, this would lead to increased market shares through methods that remain ethically sound and utilize scientifically valid methods. Furthermore, the execution of the studies should be entrusted to experts in research with proper training and credentials that would validate the study and provide credibility to its results.

In conclusion, we must accept the existing and increasing need for PMOS. It is critical to understand that these studies should be conducted within scientific aegis involving rigorous epidemiological methods to answer questions that are important to all health-care stakeholders — including patients, government and industry. With a properly designed PMOS, all parties will benefit.

John S. Sampalis, PhD, president of JSS Medical Research (Westmount, QC), is a clinical epidemiologist with undergraduate training in microbiology, immunology and neuroscience and graduate training in clinical epidemiology and biostatistics. Sampalis is also a tenured associate professor of surgery, and epidemiology and biostatistics at the faculty of medicine at McGill University (Montreal, QC).