Part 2: The Case for Better Health History Records – How Much History is Needed and How Should it be Reviewed?

Part 1 of this medical histories series focused on the regulations relevant to clinical trial subject eligibility, and the possible issues which can arise when ineligible subjects are enrolled. As noted, the outcomes vary depending on whether you are a clinical site, a CRO, or a study sponsor, but all are detrimental to the research in some form.

How then do you ensure you have adequate information to support eligibility?

Most studies that I have reviewed in the last two years or more have enrolled subjects based solely on verbal medical histories and completed general medical history forms, without support from medical records.

It is not clear when this change to subject-only reporting of medical histories occurred, but it appears to have happened slowly over the last decade, and it is certainly becoming more the norm in the pharmaceutical industry. Although the subject-provided history may seem like enough, it most definitely is not—for a few reasons.

Verbal history is not enough

First and foremost, people are notoriously bad personal historians when it comes to medically-related issues, even given good memory and mental health status. Indeed, at more than 50 percent of the clinical sites that I have personally audited in the last year in which someone was found to be ineligible, the review of written medical records and the subsequent resolution of conflicting information has shown that verbal histories were incorrect.

Even more subjects were found ineligible when sites followed my advice and requested historical medical records to support subject-provided histories, then re-reviewed eligibility status.

Second, verbal histories are, of course, very dependent on the memory and understanding of the person supplying it. Age, stress level, complicated medical histories, and familiarity with medical terminology can all play a part in reducing recall. Subjects may have unrecognized memory disorders; may have failed multiple treatment regimens, which could have lasted months or years; could have multiple medical conditions; or may have a disease with a complex etiology.

Third, financial motives can never be dismissed, particularly for studies in which there is recruitment of “healthy volunteers.”

Finally, subjects may just not know everything that has been documented by their physician(s) as long-term or potential health issues. A patient’s doctor, for instance, may be worried about their high blood pressure, which has been above normal for months or years. If a patient has not been medicated for it yet, they may not consider themselves to have hypertension, although they technically do.

Tips for gathering the proper medical records

To ensure that subjects meet eligibility criteria and do not face unknown risks by participating in a clinical trial, all necessary written medical records should be gathered to support each inclusion and exclusion criteria. As such, the study team should know the potential subject’s primary care physician’s contact information, along with that of as many specialists, hospitals, diagnostic centers, or clinics where the person has been examined and/or treated as is possible to collect.

Requested records should cover at least a six (6) month period, dependent on inclusion and exclusion criteria requirements (i.e., the subject may be required to have had a condition for at least a year). For all potential subjects, each set of records used to substantiate eligibility should include, at minimum, a summary of the most recent history and physical, a summary of the patient’s historical condition(s), and the most recent clinical laboratory and diagnostic test results. If such summaries are not available, records covering the missing information (fuller histories, recent clinical lab results, etc.) should be requested.

The period covered should be proportionately lengthened, however, if the disease under study has a longer etiology and/or prior treatments are required to have been received or not received (i.e., radiation therapy, specific immunotherapy, etc.). If the individual subject has a condition which was diagnosed/treated for longer than 6 months and which could affect their eligibility or safety, the included period should be extended for this situation as well.

If the volume of records is not prohibitive, it is a best practice to gather complete medical histories related to all inclusion and exclusion criteria, as well as all comorbidities.

Concomitant medications and other risks

In addition to reviewing the collected records for inclusion and exclusion criteria, and comorbidities, notes should be taken of concomitant medications, as they can provide clues to medical conditions that were neither verbally reported by the potential study subject nor explicitly listed in the medical records received.

To ensure accuracy, the list in the medical record(s) should also be compared to any list of medications that the study volunteer provided. Notably, if the subject is not adherent to prescribed therapies, it can be an indicator of potential subject compliance issues during the study.

The requested records should also be reviewed for any conditions which may complicate the study therapy, or are determined by the study investigator to potentially pose additional risks to the study subject, even though they were not listed in the exclusion criteria. These potential complicating factors can be comorbidities, historically adverse events while on specific treatments, or even participation in other clinical trials too recently.

Importantly, as all studies are reviewed by regulatory health authorities, like the FDA, and investigational review boards/ethics committees to ensure that risks and potential benefits are balanced, this last aspect of the medical history review should not be discounted. Certainly, protocol developers cannot foresee all potential risks that a patient population may face when there is a treatment under study. Since conditions outside the inclusion and exclusion criteria may be important to maintaining the approved risk-benefit balance when looking at an individual patient, the onus is on the investigator, as a physician, to ensure that his/her patient population is adequately screened and protected from undue risk.

Preventing risk to your patient can only be done with a full understanding and thoughtful review of each individual’s medical history in as complete a form as possible. Always keep in mind that failing to prevent the participation of a subject who is ineligible or poses other additional risks while on the clinical study not only affects clinical site subject management, but it also increases the workload related to pharmacovigilance and safety reporting teams, and can confound study results during statistical analysis.

Resolving conflicts in medical records

Even when you have all of the medical records, and are diligently reviewing them, histories may not always be clear cut, and may not match what the patient has reported. So, what is to be done when medical records conflict?

The simple answer is that the review proceeds just as it would for any new patient to a primary care practice. Historical and current treatments should be used to provide support for a diagnosis, and current physical examinations and clinical testing should be used in support as well. In addition, historical diagnostic studies should be requested if this has not already happened, and/or new diagnostic tests may need to be ordered.

All records should be reviewed to determine sources. If one record only relied on the patient’s own reporting (i.e., verbal or completed intake form), while another relied on exams and tests, the personal reporting should be ruled out.

As additional support, the general/primary care physician or even specialists can be asked to provide written input, resolution and/or historical perspective via a letter. And, when there is a lack of diagnostic evidence (i.e., no documented history of an exclusionary condition, no treatment in use), it is then that a verbal or written confirmation by the patient or their legally-authorized representative is most appropriately utilized as documented support.

In all cases where a conflict has had to be resolved, the records reviewed, the methods used for the review, and the final decision on diagnosis(es) needs to be documented by the investigator in a note to file, memorandum, medical record, or source worksheet.

Of course, all of this work depends upon getting the medical records that the study team needs. Given the multiplicity of electronic document formats, the cost of attaining records, and many other factors, procurement can be a daunting task. In Part 3 of this blog, we will discuss potential barriers to obtaining records.

by Barbara Rusin

About the author

Barbara A. Rusin is Regulatory Compliance Manager at MMS, providing technical oversight and direction to a global regulatory team, spread across four continents. Prior to MMS, Barbara was a Bioresearch Monitoring Investigator and Biologics Specialist with the Food and Drug Administration’s (FDA) Detroit District office.

Read Part 3 of the series