Early career physicians may feel overwhelmed by having to document the care they provide, but it's not as confusing as it used to be, according to Shreya Chandra, MD, CPC, FACP.
"Residency programs are getting a little more aware and trying to prep docs," said Dr. Chandra, who is physician director of coding and documentation at San Diego Kaiser Permanente. "Also, it's nice because coding has become more intuitive after the 2021 changes and the 2023 changes, so I think it is less daunting."
In 2021, Current Procedural Terminology (CPT) E/M codes for office or other outpatient care were updated to allow physicians to code a visit on the basis of medical decision making or total time spent, and in 2023, these changes were extended to inpatient, emergency, and observation care, among other settings. Visit codes can now be based on time spent (e.g., 20 to 29 minutes) or the level of medical decision making involved (defined by the number and complexity of problems addressed, the amount and complexity of data to be reviewed and analyzed, and the patient's risk of complications and/or morbidity and mortality).
"As an early career physician just starting out, you should develop a good basis of what E/M is and how you can accurately reflect your medical decision making," Dr. Chandra said.
She recently spoke with I.M. Matters about such basics of documenting a visit.
Q: How does documenting by medical decision making work?
A: In the past, E/M codes used to be based on bullet points, including the review of systems and subjective concerns. … Now, the history and exam portion includes what you as the physician feel is necessary for the care of the patient that day, but what actually determines your E/M calculation is usually going to be medical decision making or time.
For example, take the E/M codes for outpatient or office visits for established patients. Code 99213 indicates low risk, defined by the number of problems—two minor problems or one chronic problem—and one category for data amount and complexity, e.g., ordering tests or independently reviewing those tests. This would be constituted, timewise, as a 20- to 29-minute visit.
The code 99214 for an established patient indicates moderate medical decision making. Most of [our visits] seeing patients as internal medicine physicians will meet the criteria for this code, the reason being that most of our patients are moderate risk, meaning we are going to prescribe therapy for them, or we're going to assess social determinants of health. That would immediately get us to the moderate-risk category, which is defined as one new problem or two chronic problems. Most of us would say that describes all of our patients.
If you look at the time-based component for 99214, it's actually 30 to 39 minutes, so almost 40 minutes for a visit, and I would say most practicing internal medicine physicians don't typically have the time to have 40 minutes with every patient. As a result, we're meeting these criteria based on our medical decision making rather than by tracking time.
Finally, code 99215 now defines a high-risk visit, so anybody with one or more chronic diseases that constitute a severe risk to life and limb for whom you reviewed two out of three data categories. The classic example that I use is a patient who comes in for chest pain and ends up having EKG changes, and you emergently send them to the ED. That obviously would be a high-risk visit. The time component for this code, meanwhile, is 40 to 54 minutes. … Clinically, the patient should not be in your office for that amount of time.
Q: What's the advantage of this type of coding?
A: I feel like it's a little easier and more intuitive for doctors to understand, because it's really more about how many problems did they go over with a patient, and if you're above two chronic problems, you're probably a 99214, or 99204 if it's the first time you're meeting somebody. Most of the time, we can intuitively tell [which decision-making level is involved].
Q: When is it appropriate to use time-based coding?
A: That's most useful for those instances when you're seeing a new patient and you're going to be doing a lot of [previsit] chart work. That is the time you spent on the date of service. You might have a 30-minute visit with a new patient, but then you might actually do another 20 to 25 minutes after hours or before the patient comes in reviewing their outside records. As long as it's on the date of service, that counts, and you could potentially do time-based coding, especially if you spend extra time with the patient.
Another example of time-based coding, and maybe the easiest one, is if you're going to give any kind of IV treatment in the office. If you're going to give IV fluids, you're going to be doing time-based coding, because the patient is obviously going to be in your office for an hour or more, whereas if you did it by medical decision making, it would be considered a low-risk diagnosis to have a patient with dehydration, for example.
Q: What do physicians need to know about the new CMS code G2211?
A: CMS just introduced G2211 this year. The verbiage that CMS uses is that it's "meant to capture visit complexity inherent to the evaluation and management associated with medical care services that serve as a focal point of ongoing health services." Basically, it can be used by whichever clinician is providing that longitudinal care relationship for the patient, whether that be their primary care physician, their OB-GYN, or their HIV physician. It's an add-on code that you're going to use in addition to your regular E/M code to give you credit for what's involved in that longitudinal care. And it's very important to be using this code, because what we've seen is that when we don't tend to use [a code], it'll disappear, or the reimbursement for it will disappear. Even though it seems like the reimbursement for this code is not very high, when you start to get into per member per month, it can certainly help.
Q: What might seem most daunting about documentation for early career physicians?
A: Once you intuitively accept that you can trust your medical decision making and how to accurately reflect it, I think the only part that's a little bit daunting is making sure that you're documenting CPT codes and ICD-10 codes for patients' comorbidities. A lot of times, we don't understand why we're documenting comorbidities, but you have to code them because you actually are addressing them. You may not actively be making a change that day, but it certainly would go into your medical decision making.
For example, if a patient comes in for a joint injection and I look at their blood pressure and I know they're on antihypertensives, I need to tell the patient, "After this cortisone injection I'm giving you, your blood pressure might increase. Take an extra amlodipine for the next two days if you notice your blood pressure's running a little higher." Did I necessarily make a change in their chronic management? No, but the hypertension was included in my medical decision making, and I had a discussion with the patient. The goal is for internal medicine physicians to take credit for the work that we are doing.
Q: What are some common coding pitfalls to be aware of?
A: Modifier 25 is used to report an E/M service that took place on the same day that another service was provided to the patient by the same clinician. This becomes very confusing for physicians, as far as determining when they went above and beyond what a normal visit would be and when they can apply it.
It's pretty easy to remember to add modifier 25 when you do procedures. I think where people get confused and you have pitfalls is on physicals. What constitutes a physical is what would be normal age-appropriate screening for a patient: routine screening labs, routine vaccines, Pap smears, mammograms, colon cancer screening. What I tell physicians is if you're not really making any changes, if you're just performing a routine physical and reviewing chronic conditions, that may not constitute a modifier 25.
However, modifier 25 could apply if you see something on a physical that requires more extensive evaluation. The most common thing that we see now is atrial fibrillation. A patient comes in for a routine visit, you listen to their heart, and their heartbeat is a little irregular. You do an EKG and they have new-onset afib. Now you're going to institute more bloodwork, you're going to do an echocardiogram, you're going to do an evaluation, so that's when you would use the modifier 25. That's probably one of the more common ones that we miss.
Q: How should physicians handle their documentation being questioned by a coder?
A: Physicians need to remember that coders are there to try to help capture the complexity that's inherent to the visits. Physicians tend to overthink things. For example, in a patient with diabetes and CKD who is hypertensive, we as physicians spend a lot of time trying to figure out, "Was the CKD really caused by the diabetes or by the hypertension?" When you're a coder, you just say, "Pick one—patient has CKD, diabetes, or hypertension—and attach it to that code." That's what is going to give the clinician complexity, since they are obviously making medical decisions about all three conditions together. I tell clinicians that when they're being questioned by a coder, they have to remember that coders are not questioning your medical decisions. They're questioning how you reflect those medical decisions in your coding.
Don't overthink it. Remember that we're trying to capture the complexity that's inherent in our visits with our patients. Across the country, every physician group will tell you that our patient population is aging and our patients are more and more complex, and our less complex patients are being filtered out by phone visits, by email visits, any number of ways that they can achieve electronic care. We know that the patients who actually come into our office inherently have a higher level of complexity. That's what the coders are there to help you capture.