medical necessity and medical decision making (the third key component) are not the same thing. Convention has taught most coders to use mdm as the determining factor for an e&m code, however there is more to consider. If it's medically necessary to do a detailed hpi and detailed exam to arrive at a straightforward mdm, then the necessity would trump the mdm...and this is what cms is talking about. The majority of the time mdm will help determine necessity, but that is not necessarily correct 100% of the time.
An example might be with a patient coming in for a 6-month follow up of breast cancer. The provider details a history of the disease progress and current status. There will be at least 2 system reviews, and likely some medication history. The exam will be detailed....looking for symptoms of recurrance, late effects of treatement, etc. Labs may be ordered, but the patient is no longer undergoing treatment. With a single problem improved, and only one or two data points, you have straightforward mdm, a 99212 for an established patient. But is that appropriate for the amount of front end work that had to be done in order to arrive at that assessment? This is where medical necessity would come into play and the consideration of the nature of the presenting problem that required a detailed history and a detailed exam to arrive at the final diagnosis. In this case, splitting the difference and assigning a 99213 might be more appropriate.
Plugging the documentation into nice neat audit forms is not the only method for determining necessity...if only it were that easy! You have to determine the nature of the presenting problem, consider other related factors and evaluate the documentation to determine if it's relevant to the presenting problem. The audit tools that the contractors ask us to use are helpful and will allow you to come up with the best los most of the time, but looking at the encounter as a whole, rather than only measuring the bullet count is a crucial for determining medical necessity.