Multinational Chart Reviews for Addressing Real-World Evidence Needs

Multinational Chart Reviews for Addressing Real-World Evidence Needs

Traditionally, chart review studies have been employed to evaluate burden of illness and patterns of care, but, increasingly, this methodology is being applied to meet marketing authorization and risk management data requirements including data that reflect patterns of off label medication use. 

Drawing on experience across a broad range of chart review studies, we will be offering an interactive webinar, Multinational Retrospective Chart Reviews, to present this methodology and some of its applications, and highlight key design and operational challenges associated with this data gathering approach.

I hope you’ll join us!  In the meantime, here are some notes that we hope you will find helpful as you are contemplating your evidence gathering needs and methods you will need to employ.    

Multinational chart reviews are one way to effectively capture real-world data. A balance between good science and pragmatic methodology, the successful chart review study can result in tailored fit for purpose patient level datasets. There has been a marked increase in the conduct of these studies in Europe, largely because databases to inform multinational healthcare analytics are not routinely available. Moreover, in the absence of electronic medical records (EMRs), not widely available outside the United States, chart review studies offer an alternative for the capture and analysis of real-world patient level data.

 Real-world evidence needs addressed by chart reviews include but are not limited to: 

  1. Patient characteristics (demographics, medical history, concomitant illness, etc.)
  2. Patterns of care, such as visits by type of healthcare provider, medications, diagnostics, procedures, accident and emergency visits, hospitalizations
  3. Effectiveness (clinical outcomes)
  4. Patterns of medication use including off-label prescribing
  5. Unmet clinical need
  6. Safety (documented adverse events (AE), serious adverse events (SAE), etc.) 

A common application of chart review studies is in support of burden of illness studies including the use of data to populate health economic analyses. There is also an increased interest in employing chart reviews to achieve drug safety research objectives.  For example, chart review studies can evaluate long term outcomes in patients who have had a previously documented SAE, inform patterns of on and off-label medication use, and to verify physician adherence to Risk Evaluation and Mitigation Strategy (REMS) or Risk Management Plan (RMP) requirements. 

Strengths and Limitations

Relative to other observational study designs, the retrospective chart review is, by definition, void of the Hawthorne Effect – hence a purely naturalistic data source. True patterns of care can be evaluated that have not been influenced by a prospective study protocol, and the knowledge of physicians and patients that they are being observed. In addition, unlike many sources of administrative healthcare or claims data, clinical outcomes and results of diagnostic tests can also be accessed.

There are also some important limitations, however. Since patients are not assigned to treatment through any randomization process, the evaluation of treatment effects is made difficult by selection and channeling biases. In addition, study data are only as complete as the quality and completeness of the medical chart – which is highly variable from physician to physician. To enhance quality, however, strategies can be employed to ensure that the data that are available are correctly abstracted and entered into the study database; electronic data capture with sophisticated data checks and validations can be integrated into the EDC.  Regardless of these main limitations, however, retrospective chart reviews can be an effective means to gather the right evidence for the right purpose.

Operational Lessons Learned

Here at UBC, we’ve learned some valuable lessons through hard work and experience.  For example, to achieve a more streamlined and efficient study start-up, we recommend the following actions:  

  • Engage a regulatory expert with experience in observational, non-interventional studies to ensure a country-level understanding of such topics as the nuances of data and privacy protection laws, and ethics submission requirements specific to chart review methodologies
  • Query sites at the outset  as to when ethics committee meetings occur, what types of submissions are required, and if translations are necessary
  • Set up a study risk management plan with strategies to proactively mitigate these risks.
  • Maintain site engagement and motivation. Establishing a single point of contact, usually a clinical site specialist or a clinical research associate who speaks the native language, for each site is helpful in building relationships and optimizing open and transparent communication.
  • Review and test case report forms (CRFs) before finalizing and implementing them. Ensuring these forms are not overly complex or time consuming and are designed to gather the data needed is crucial, so engage a study nurse and the Sponsor safety team in the design and testing. 

Multinational, retrospective chart reviews are proving to be an effective methodology for the capture of tailored, patient-level data. These studies can be used to address a myriad of research objectives, and can be implemented anywhere medical charts exist. While manually extracting healthcare data from the medical charts of patients may seem to be “going back to basics”, these studies should not be approached in a “cookie cutter” manner. Although there are key design and operational principles that must be addressed in any rigorous protocol, each study must be carefully conceptualized in relation to the specific research aims and objectives at hand.