Regulatory Open Forum

"One of the areas that…we believe we are poised to move forward on, is active surveillance here in the U.S., where we take large datasets, electronic datasets, and apply analytical software to more quickly look for safety issues."

My first reaction was maybe this was more like active surveillance of the results of some else's passive surveillance than actual active surveillance. This promptly led me to what is active surveillance, anyway? I realized I was very clear on passive surveillance, but when it came to active surveillance, not so much. I googled, and soon found out why. Turns out there are a lot of definitions and descriptions of passive surveillance out there, but when it comes to active surveillance, not so much.

The other twist is that there is surveillance as it applies to public health, and there is surveillance as it applies to medicine. Public health tends to have given much consideration to passive surveillance, where medicine has focused on active surveillance. Moreover, medicine views active surveillance as an alternative to active treatment, rather than as an alternative to passive surveillance.

The authorities on passive surveillance are, not surprisingly, the CDC and WHO. Their descriptions of passive health surveillance are pretty much like MAUDE, if it were done well. Possibly more like MedSun, but I haven't looked at it.

The National Cancer Institute's definition of active surveillance is somewhat interesting to apply to postmarket surveillance of medical devices:

A treatment plan that involves closely watching a patient's condition (in Shuren's statement, not watching actual patients, but large datasets) but not giving any treatment (recall or market withdrawal) unless there are changes in test (software analysis) results that show the condition is getting worse. Active surveillance may be used to avoid or delay the need for treatments (recall or market withdrawal) which can cause side effects or other problems (patients not experiencing side effects may suffer health consequences after being deprived of treatment that has been working well for them). During active surveillance, certain exams and tests are done on a regular schedule. Shuren didn't say anything about a schedule, but I wouldn't be surprised.

So I think for surveillance to be truly active, you have to actively ask about what you want to know, not just actively look at datasets. The problem with my model is that it's a certainty, the more you ask, the more you will hear. Most manufacturers don't relish the prospect of hearing about more and more adverse events associated with their devices. The model that Shuren describes…eh, don't know nothing about those datasets, but I guess it's better than just sitting there, waiting to get hit over the head.

"One of the areas that…we believe we are poised to move forward on, is active surveillance here in the U.S., where we take large datasets, electronic datasets, and apply analytical software to more quickly look for safety issues."

My first reaction was maybe this was more like active surveillance of the results of some else's passive surveillance than actual active surveillance. This promptly led me to what is active surveillance, anyway? I realized I was very clear on passive surveillance, but when it came to active surveillance, not so much. I googled, and soon found out why. Turns out there are a lot of definitions and descriptions of passive surveillance out there, but when it comes to active surveillance, not so much.

The other twist is that there is surveillance as it applies to public health, and there is surveillance as it applies to medicine. Public health tends to have given much consideration to passive surveillance, where medicine has focused on active surveillance. Moreover, medicine views active surveillance as an alternative to active treatment, rather than as an alternative to passive surveillance.

The authorities on passive surveillance are, not surprisingly, the CDC and WHO. Their descriptions of passive health surveillance are pretty much like MAUDE, if it were done well. Possibly more like MedSun, but I haven't looked at it.

The National Cancer Institute's definition of active surveillance is somewhat interesting to apply to postmarket surveillance of medical devices:

A treatment plan that involves closely watching a patient's condition (in Shuren's statement, not watching actual patients, but large datasets) but not giving any treatment (recall or market withdrawal) unless there are changes in test (software analysis) results that show the condition is getting worse. Active surveillance may be used to avoid or delay the need for treatments (recall or market withdrawal) which can cause side effects or other problems (patients not experiencing side effects may suffer health consequences after being deprived of treatment that has been working well for them). During active surveillance, certain exams and tests are done on a regular schedule. Shuren didn't say anything about a schedule, but I wouldn't be surprised.

So I think for surveillance to be truly active, you have to actively ask about what you want to know, not just actively look at datasets. The problem with my model is that it's a certainty, the more you ask, the more you will hear. Most manufacturers don't relish the prospect of hearing about more and more adverse events associated with their devices. The model that Shuren describes…eh, don't know nothing about those datasets, but I guess it's better than just sitting there, waiting to get hit over the head.