Overseas Studies

 A payor notes,

We routinely review and consider studies performed overseas. Elements such as study design and quality are universally important and are always a part of any review.  Additionally, for studies performed overseas, we evaluate whether the conclusions drawn would be applicable to US beneficiaries.  

For example, if physicians in another country follow cancer patients with an imaging schedule that is very different than that used in the US for a particular indication, study conclusions regarding lead times for MRD tests may be very different (and possibly not applicable) to patients in the US.  

Moreover, if a study population is primarily or exclusively composed of one racial or ethnic group, that might also impact the applicability of results to broader populations.  In this case, it is still possible that broader applicability across populations may have already been demonstrated in pre-existing clinical validity (CV) studies for the same analytes performed using a different test.

In short, our standards are the same, regardless of whether a study is performed in the US or overseas. Applicability to the intended-use population is a key feature no matter where a study is performed.

Very similar text is found in a June 2023 proposed evidence handbook for CMS NCDs.

https://www.cms.gov/medicare-coverage-database/view/medicare-coverage-document.aspx?mcdid=34&docTypeId=1&sortBy=title&bc=16

CMS routinely considers studies that are performed in whole or in part outside of the United States (OUS).  Whether outcomes from OUS studies may be generalized to the Medicare beneficiary population depends on multiple factors, but an important consideration is whether the study outcome depends on the care delivery context.  To the extent that health systems and practice standards differ between countries, an OUS study designed to show positive results may not be generalizable to the Medicare beneficiary population.  For example, an OUS study that aims to demonstrate that an intervention reduces hospitalizations may not be generalizable to the US if there are substantial differences in the types of (and coverage provided by) health insurance, hospital bed availability, and practice patterns between the US and the study country. Studies that include outcomes that may be sensitive to care delivery context (whether across different sites in the US or multi-country studies) should be appropriately designed and analyzed, potentially incorporating clustering or stratification into their statistical analysis plan.