Jay Bhattacharya dismissed the test-negative design as 'crap.' The biostatistics literature tells a different story.