by Patricia Blount, MD

Prospective Observational Studies Evaluating Standards of Care: Ethical Considerations for the Physician-Investigator

Prospective observational studies evaluating standards of care (SOC) aim to collect data about a treatment an individual receives in the course of clinical care by a physician. However, because the data are collected prospectively, the conduct of such a study may influence physician decision-making and subsequently the treatment a patient receives. As prospective observational studies gain popularity in the field of comparative effectiveness research (CER), it will be incumbent upon the physician-investigator to consider the ethical implications of deviating from the usual practice of medicine for purposes of the research. Measures can be taken to protect the patient-physician relationship when conducting a prospective observational study evaluating SOC.

CER is defined as the conduct and synthesis of research comparing the benefits and harms of various interventions and strategies for preventing, diagnosing, treating, and monitoring health conditions in real world settings.[1] CER is of increasing interest to health care providers, patients and third party payers.[2] Clinical treatments and procedures under investigation in comparative effectiveness research are generally referred to as SOC because they have been accepted by medical experts as appropriate treatments or procedures for a given type of disease or condition.[3],[4] There are two main types of studies utilized to evaluate SOC in CER: observational studies and interventional studies.

Observational studies evaluating SOC involve the collection of data generated in the course of treatment of a patient by a physician. The study design does not influence the physician’s decision-making process or the treatment the individual receives. In observational studies, IRBs consider risks of the research to be solely those associated with data collection activities. Conversely, in interventional studies (e.g. randomized clinical trials) evaluating SOC treatments, it is the protocol, not the patient’s physician, that determines the prescribed or assigned treatment the individual receives. In this case, IRBs also consider risks of the interventions to be risks of the research.[4]

Prospective observational studies are gaining interest in CER because, unlike retrospective studies, the study design takes place before treatment of the patient occurs. This permits the advance specification of data collection activities to meet research endpoints that cannot be met in a retrospective study due to missing data.[2],[3].[5] The CER literature is generally concerned with the rigor of the prospective observational study design.[2],[3].[5] It does not address the ethical implications of how the prospective design may influence physician decision-making and subsequently the treatment a patient receives.

By nature of its prospective design, physician-investigators conducting prospective observational studies are at risk for unconsciously shifting focus from recommending a treatment as a physician for benefit of the individual patient, to recommending a treatment as an investigator for purposes of the research. The risk of a change in physician treatment recommendations may be particularly high and more subtle in prospective observational studies involving SOC because, strictly speaking, the treatment is recommended by the physician, not assigned by the protocol. Furthermore, the physician-investigator may be accustomed, in the course of clinical practice, to routinely recommending all SOC treatments being compared in the study. For this reason, the physician-investigator may view all studies involving SOC as minimal risk data collection studies because the subject could have received the treatment as a patient in his or her practice. The physician-investigator may not recognize he/she is crossing the line between recommending a treatment to a patient in the course of clinical care (after giving careful consideration of what is best for that individual patient) and recommending a treatment to a patient to meet research-related goals. If this line is crossed, it can be argued that research-related risks also include those associated with the SOC treatment because, although not assigned to a subject by the protocol, the research has influenced the physician’s decision-making process and subsequently the treatment the patient receives.

According to the Office of Human Research Protection’s (OHRP) recent draft guidance on research comparing SOC, “….if an observational study compares two standards of care that are chosen by individuals with a given medical condition in consultation with their clinical care providers, and the research plan does not assign subjects to any specific standard of care, then the risks of the standards of care are not risks of research.  In this case, the risks of the standards of care are the same risks individuals would have been exposed to without participating in the research study.”[4]

In contrast, the same draft guidance indicates that if the protocol assigns individuals to an SOC treatment, the risks of the treatment that some individuals would be exposed to by participating in the study are different from the risks of the treatment those same individuals would have been exposed to in the course of their clinical care. Therefore, risks associated with SOC treatments delivered to individuals for purposes of the research and not in the course of their clinical care are risks of the research. All reasonably foreseeable risks associated with the research should be disclosed to subjects. This is especially important when one treatment may be associated with greater risk, but is potentially more effective, and the alternative has lower risk but is potentially less effective.[4],[6]

IRB members regularly review protocols submitted as prospective observational studies involving SOC or medical treatments delivered in a physician’s own practice. These protocols frequently read more like interventional studies. In such cases, IRB members may be unable to determine whether the SOC treatment is delivered in the course of clinical care or prescribed primarily to answer a research question. The IRB is therefore unable to determine absolute research-related risks. This is especially true when the protocol includes extensive eligibility criteria and when consenting and data collection activities begin prior to implementation of the SOC treatment. IRB members often express concern that the physician-investigator may not recognize whether he/she is unintentionally recommending a particular treatment for purposes of the research. If this occurs, associated risks of the treatment, as well as alternative treatments, may not be fully discussed with subjects because the treatment is delivered in the course of clinical care of a patient by a physician.

The ethical conduct of a prospective observational study requires placing a fence around the practice of medicine such that the patient-physician relationship remains unaffected by the investigator’s understandable desire to meet study goals. Ideally, the physician-investigator does not recruit subjects from his or her medical practice due to this inherent conflict of interest and potential for undue influence over patient decision-making. However, investigator-initiated research designed to answer questions about the physician’s own clinical practices leaves little room for alternative designs. The recommendations below are steps that can be taken to manage the conflict of interest between acting in the capacity of both the patient’s physician and the study investigator. These steps can be implemented regardless of whether the study is CER evaluating two or more SOC or investigator-initiated research which aims to answer questions about one particular treatment or procedure:

  1. Use best practices in clinical research protocol writing, which includes utilizing a set of protocol content guidelines for observational studies. Write a detailed protocol which clearly separates SOC treatments that occur as a matter of clinical care from research-related activities.
  2. If the physician-investigator is recruiting from his or her patient population, include measures in the protocol to prevent an unconscious change in the practice of medicine to meet study goals, and indicate these measures will be reviewed with clinic/study staff. Examples might include:
    • Ask colleagues in the same practice, who are not involved with the study, to identify patients to whom they have recommended the treatment of interest and refer those patients to study staff for consideration of the study. In this case, physician colleagues of the investigator recommend and deliver the treatment and the investigator conducts the research-related, data collection activities.
    • Recruit potential subjects from the subset of patients in the investigator’s practice who have already undergone the SOC treatment. In this case, the medical record is the source of baseline data and patient outcomes are collected prospectively.
    • If collection of study-specific data must occur before the patient undergoes the treatment of interest, instruct study staff to review patient records for study eligibility and recruit patients for the study only after there is clear evidence the patient has decided to undergo the treatment:
      • The patient has received a prescription for a drug.
      • The patient is scheduled to undergo a treatment.
      • The patient has signed a clinical consent for a treatment.
      • The patient arrives at the clinic to undergo a pre-scheduled treatment.
    • Educate study staff regarding the importance of avoiding undue influence of patients to undergo a particular treatment to meet research goals. Remind staff the physician-investigator is a physician first, who patients trust to deliver medical care for their individual benefit.

If the practice of medicine is conserved through the ethical conduct of prospective observational research, the risks associated with the research are only those related to data collection activities about the medical treatment delivered to a patient. However, if the practice of medicine is altered for research purposes, the research-related risks may include the risks of the SOC treatments. The best foundation for ethical conduct of prospective observational research yielding meaningful data about real world healthcare delivery is a well-written protocol that includes a clear description of subject protection measures.

Download: Best Practices in Clinical Research Protocol Writing

This whitepaper addresses some of the most frequently observed errors and omissions in protocol content encountered by IRB members during review of studies.

Download the Whitepaper

[1] Comparative Effectiveness Research-Implications of the Federal Coordinating Council’s Report. www.nejm.org/doi/pdf/10.1056/NEJMp0905631

[2] Grace principles: Good Research for Comparative Effectiveness. https://www.pharmacoepi.org/pub/1c29f69f-2354-d714-5100-1ef2b0e9abd9

[3] Effective Health Care Program: Developing a protocol for Observational Comparative Effectiveness Research-A User’s Guide. http://effectivehealthcare.ahrq.gov/ehc/products/440/1166/User-Guide-Observational-CER-130113.pdf

[4] Draft Guidance on Disclosing Reasonably Foreseeable Risks in Research Evaluating Standards of Care. http://www.hhs.gov/ohrp/newsroom/rfc/comstdofcare.html

[5] Prospective Observational Studies to Assess Comparative Effectiveness: The ISPOR Good Research Practices Task Force Report. http://dx.doi.org/10.1016/j.jval.2011.12.010

[6] Draft Guidance on Disclosing Reasonably Foreseeable Risks in Research Evaluating Standards of Care and the SUPPORT Trial. http://www.quorumreview.com/blog/2014/12/05/qf-post-fda-draft-guidance-support-trial/