If you're using a TCP monitor it's just testing to see whether it can open a socket on the pool member so doesn't verify whether the underlying process/service is working and responding to users (really just useful for verifying that the VM is on and that its network stack is operational).
I don't know for sure if this was in place as far back as 6.0.2, which has been end of support for a while, but in all of the supported releases, if you're using TCP 8443 for an HTTPS service, you may be able to change the load balancer to utilize an HTTPS health check which will actually generate GET requests and verify that it's receiving 200 OK responses from the server.