E/M Guidelines: Right Road, Wrong Direction Save

Last July, the Centers for Medicare & Medicaid Services (CMS) proposed new E/M guidelines for office visits with physicians. It issued the final guidelines on November 1, with a few tweaks. Implementation was delayed until 2021 (from 2019), possibly due to overwhelming opposition from physicians.

The good news: the agency acknowledges that fee-for-service is still an important feature of physician reimbursement. The bad news: the changes would make office visits even less attractive for clinicians, ostensibly pushing practices from disfavored "volume" to efficient "value," and ultimately to capitation.

The CMS Proposal

The new E/M guidelines would collapse payments for office visit codes 99212-99214 (established patients) to a single amount -- national average about $93 -- about halfway between the old 99213 ($74) and 99214 ($109) codes. The plan would be similar for new patients, 99202-99204. Physicians would receive a new flat rate, $135, between the old 99203 rate ($110) and the old 99204 rate ($167). Reimbursement for Level 5 visits would remain the same. Some subspecialists, such as cardiology endocrinology, rheumatology, and hematology/oncology, would receive a small add-on: $14. Certain as-yet-unspecified primary care services would permit a $5 add-on. (The agency is still working out another category of reimbursement, which is based on time.)

The Good News

It is good that CMS plans to eliminate the counting up of "elements" (symptoms, history, review of systems, physical exam) presently necessary to achieve a higher code. Documentation could be cut to that needed for a level 99212 visit: chief complaint, problem-focused history, problem-focused exam, and decision-making. CMS says this will reduce paperwork; the agency calculated that we would save 1.6 minutes per visit, or 51.2 hours per year for a doctor who sees eight patients a day. (Who does that?) Needless to say, such a "reduction" is trivial, especially given electronic health records with carry-forward or automatic populating.

Nevertheless, it is true that the history and physical template for ambulatory patient visits that we all learned in medical school is obsolete: it is based on visit for a single, acute problem, with symptoms and physical findings -- the medicine of the 1950s. Today, patients still have occasional acute problems. But as more patients survive multiple conditions that used to kill them quickly -- like HIV, heart disease, and cancer -- their diseases morph from acute to chronic.

When these patients come for a routine visit, they often appear stable, with no new symptoms or physical findings. However, there is still work to do: check the recent test results, which may require changes in care; read new findings of consultants, which may also affect management; look for behavioral health changes. Most important, what does the patient think: is he still good with current treatment or having problems? Does he have new concerns? Shared decision-making, a priority today, takes time. Little of this work is reflected in the old ambulatory template.

So much for the good news.

Unintended Consequences

The bad news is really bad. The only way a physician can make a living with a relatively low, fixed payment for most office visits is to see more patients. This focuses the clinician on patients with minor, self-limited problems or one or two straightforward chronic conditions like hypertension or hyperlipidemia. Why? Because to see more patients, visits must be shorter. Trying to rush visits for complicated, sick patients risks overlooking something important which might injure the patient or generate a malpractice claim.

So, ironically, this E/M change will likely lead to an increased volume of low-value visits -- exactly what CMS is trying to eliminate. This dismal prospect also will likely increase the flight of new physicians to specialties with procedures, which are more lucrative.

Most important, the new system will make it difficult to care for the complex patient with multiple problems -- congestive heart failure with renal insufficiency, cancer with heart disease, or any combination of diseases with cognitive impairment, anxiety, or major depression. These are the patients that require the 7 years or more of intensive medical training that only physicians have, as well as many visits, consultations, sometimes tests and imaging, and much time. Ninety-three dollars, even $107 per visit, simply will not cut it. (The restoration of a higher payment for 99215 in the final rule is an attempt to deal with this.)

Moreover, the key to successful treatment of serious problems today is a solid physician-patient relationship. Patients want a provider who cares about them and will not rush through a visit, limiting it to one problem where there are many. The best way to establish a good relationship is face-to-face visits, especially at the outset.

Finally, in today's fragmented system, continuity of care is also critical.

But it seems CMS is working from a different model. CMS appears to believe that the key to efficient care is a team, directed by a doctor -- or not -- performing the most valuable services as determined by executives with MBAs using artificial intelligence to develop algorithms for care. The management evaluates how the team is performing (including its "productivity"), and distributes bonuses and penalties accordingly. The organization ultimately delivering all the patient care would be paid by capitation.

IBM trotted out Watson (the computer that learned to beat Gary Kasparov at chess) to guide cancer treatment. The project did not go well: customers identified "multiple examples of unsafe and incorrect treatment recommendations" (subscription required).

The CMS model also does not appear very patient-centered, claims to the contrary notwithstanding.

A Better Direction

CMS is right to reconsider the physician visit, although the agency makes the wrong decision (i.e., to de-emphasize it). Instead, it should use the visit codes to focus physicians on the most complicated patients. CMS should pay a physician more to see a sicker patient with multiple chronic illnesses, and less to see a healthy one with a mild complaint. So, for example, Medicare could pay X for a patient with one straightforward routine complaint, then 2X for three chronic conditions, 5X for 10 chronic conditions, as well as add-ons for things like an acute complaint superimposed on the chronic ones, a significant deterioration, cognitive impairment, or frank dementia.

Obviously a real scale would be more complicated, but you get the idea. There could be any number of levels, probably more than five. Physicians are the only providers that are trained to handle multiple complex conditions in one patient, although midlevels like nurse practitioners or even medical assistants could perform routine tasks for such patients under physician supervision. Complex patients require more time and more work, and more coordination, so patient panels will need to be much smaller.

In fact, CMS makes such risk adjustments already to the capitation rates it pays to Medicare Advantage sponsors, in order to provide sponsors sufficient funds to care for each patient (efficiently). The CMS system is called Hierarchical Condition Categories (HCC). It should not be difficult to set up a similar system for E/M reimbursement.

A Note About Visit Notes

In medicine, visit notes, inpatient or outpatient, are traditionally used for billing: a physician documents the work for which he claims reimbursement. CMS has decided that the classic visit template is no longer useful for billing, so the agency shrinks it to a minimum. That's fine.

However, our medical system today is far more fragmented than it once was. Few people are followed by the same doctor for decades; instead, a patient passes through many offices, team members, and specialties over the course of his or her life. Moreover, we can no longer rely on patients for detailed histories. Poor communication, or missed communication, is now a major source of medical errors.

Communication should be the principal point of the progress note today. What would you want to know about a new patient before you took a significant action? That is what should be in the note -- the most important information for the next provider in the chain. Electronic records, designed for billing and workforce supervision, should be reinvented to promote communication, and be interoperable.

Such coding, reimbursement, and documentation reforms would take us in the right direction.