The first thing you need to do when looking into biocompatibility for a medical device, coated or not, is think about your application and categorize it on this chart:
For most hydrophilic coatings, the application requires a Limited exposure (<24 hrs) in a Circulating Blood environment. Therefore the chart reveals that you will need to do cytotoxicity, sensitization, irritation, systemic toxicity, in some limited circumstances genotoxicity, and finally, haemocompatibility. If your application is different, look it up on the chart and see which tests you need.
First, make sure that your devices are tested exactly as they are delivered clinically, i.e. coated and sterilized. You do not want to find out that your coating is fine but when sterilized it suddenly becomes cytotoxic, for example.
The other thing you really should do is look at each bicompatibility test and understand how it works. Do not just send out your samples with a checklist and rely on the results without thinking.
For example, the cytotoxicity test can be complicated, and that's one of the simplest ones. There are different variations on the MEM Elution test. In the first phase, the test article is submerged in a fluid to extract substances into it. This extraction can be done at different temperatures and for different lengths of time. Obviously a longer extraction time at a higher temperature is the toughest test. If you think you can pass the toughest test, go for it. If you can't, then you just need to go for the test that is acceptable to the Regulatory body you work with. In the second phase of the MEM Elution test, the extracted solution is exposed to cells, usually L929 fibroblasts. This extraction can be done for different lengths of time too, from 24 hrs to 72 hrs. The longer the exposure, the tougher the test. Extractions can also be done with polar and non-polar solvents, and the tests can be run by weight or by surface area which has a big effect on the results. Usually for hydrophilic coatings, you want to run the test by weight.
There are also vendor differences for these tests. For example, they use different amounts of serum in their media, and some add antibiotics to the solution and some do not. Ask your testing vendor to see the protocols. I am not able to conclude whether or not these protocol differences influence results but I can definitely tell you that different vendors give different results. In my experience, Toxikon is really easy to pass, almost too easy, whereas Nelson Laboratories is tougher, but very reliable. NAMSA is somewhere in between, and Apptec is probably on the tougher side. Again, this is just my experience. Your mileage may vary.
Generally, you should just figure out the minimum test protocol your Regulatory body will accept, and do that. As long as in reality your device is not harmful to humans, you can rely on the results.