If we have multiple surveys done on same diseases and in same area and time. Can we combine these datasets even though the sampling methods are not similar? What analysis can be done on the combined datasets?
The answer can be yes or no depending on many relevant parameters. These parameters include (though not an exhaustive list by any measure): the purpose of the study and potential primary use of data which will dictate the required rigor, the nature uses of the findings will determine the level to which it is crucial for the methodology to match, and the availability of alternative sources of data. The degree of variation in instrumentation and sampling designs are independent decision drivers in their own right.
They could be pooled if the actual variables are the same. But a clustered design for the analyses would be warranted. Also if the two populations are really different I am not sure what you would gain from pooling instead of using the two surveys seperately. You might gain some statistical power but you will have a hard to define population to generalise your results to if the samples included are totally different.