I work for a commercial insurance company, and we base the taxonomy code for all claims processed for a provider based on what the credentialing paperwork was submitted with as the provider's primary taxonomy code. We don't go by what is on the claim, it all comes down to how the provider was credentialed with our organization. This applies to network and non-network providers because even with non-network providers we have to verify their licensure and eligibility to practice in our state and because we also have MA plans, we have to make sure they are not opt-out or excluded providers for Medicare.
I think you need to check with the payers you submit claims to most often, or at least the ones that the providers are contracted with and ask them if they go by the taxonomy code on the claim or is it based on how they were credentialed with the company. If they base it on how they are credentialed your organization will need to decide which taxonomy code should be primary for each provider who has more than one taxonomy code, you probably will want to select the one that they most frequently are practicing under. If you have a provider who has pulmonary disease and critical care, which services are they providing/billing more often, pulmonary or critical care?
Another thing to consider is that the primary taxonomy code on file with the payers may affect the patient's benefits. We see a lot of providers who have Internal Medicine and Cardiology taxonomy codes and if the Internal Medicine is listed first the patient will probably have a lower copay as they are usually considered PCPs, but it also can mean that the provider is on a fee schedule with lower allowances as a PCP. However, if they are really practicing in the area of cardiology, they may be losing out on the fee allowances because they are setup as a PCP, so while the patient might have a higher copay, if the provider is acting as a specialist they deserve to be reimbursed as such.