This article focuses on 2 ethical issues that the new CPT coding raises to higher stakes: the increased potential for conflict of interest in billing and the erosion of confidentiality.
The launch of the new Current Procedural Terminology (CPT) coding system in 2013 has thrown many psychiatrists into a quandary. It has challenged us to think in entirely new ways about describing what we do. Always available to us, the evaluation and management (E&M) codes were seldom used by psychiatrists, although they have been the essential codes used by other physicians since 1992. Now psychiatrists have no choice but to use them and the new add-on therapy codes. The challenges for psychiatrists include:
• Learning the complexities of the E&M Chinese-menu–style coding system
• Accepting the idea of an add-on “procedure code” for therapy
• Embracing the distinctions between “counseling” and “therapy” (defined differently in CPT )
• Disentangling “medical thinking” activity from “therapy” activity in order to code them separately
Besides the philosophical implications of this last point for further extending an already widening mind-body split, this new system sharply highlights ethical conundrums that have always lurked in fee-for-service medicine, though more subtly in the past. This article focuses on 2 ethical issues that the new CPT coding raises to higher stakes: the increased potential for conflict of interest in billing and the erosion of confidentiality.
Conflict of interest
Consider these 2 cases from my experience.
Case 1. In my fee-for-service practice, Joan started our last session with a question: “Last month’s bill had different CPT codes than the month before, and my insurance paid different amounts for the codes. We met for a half hour each time. Can’t you always use the same code, like last year, for our half-hour sessions? Can it be the one that gives me the higher reimbursement?”
During one appointment, I measured Linda’s blood pressure and weight, tested her balance when she told me about some dizziness, and discussed her stomach pain and the chronic pain in her neck. After reviewing the interim history (all the while performing a mental status examination), I “counseled” her about how her psychiatric medications might or might not be related to the physical symptoms, the optimal timing for taking them, and other choices of medication. We met for a total of 25 minutes. I coded 99214 (Table).
During another visit, after asking about a number of interim psychiatric symptoms, I did not check blood pressure or balance; I did order a lithium level test because I was concerned about a tremor. We then spent approximately 18 of 25 minutes going over cognitive techniques for managing her increasing anxiety at work. For that visit, I coded 90813 with an add-on psychotherapy code of 90833.
These code sets reimburse the patient quite differently. In the Baltimore, Maryland, suburbs where I practice, the average Medicare payment for 90814 averages $114 and 90813 + 90833 = $121 ($78 + $43). For patients with private insurance, the differences have ranged as high as $40.
Case 2. A colleague called me for an ethics consultation. She was treating patients with medication and therapy. In some cases, the E&M activity encompassed about 5 minutes, and the therapy about 45 minutes: the total time spent with the patient was 50 minutes. She took 10 minutes between cases. She used to bill 90807 before the code changes. “I so often feel there is more work to do in a session and have thought about taking more time. I’d like to spend 8 to 10 more minutes on the therapy, move right on to the next patient, and take a break every few hours to write notes. In the past, I’d get paid the same if I spent the extra 10 minutes. Now, if I spend those extra minutes on therapy, I can code the 90838 add-on rather than 90836 and get paid more by insurance. Now that there is an extra reward to me for doing so, is it ethical for me to make this change, which in the past would help patients more than me? Is that a conflict of interest?”
Previously, when I provided medication plus therapy in an outpatient setting (“medical psychotherapy” in the old CPT codes), I chose between 90805 for a half hour and 90807 for an hour-and I did anything that needed to be done in that time. There was no higher-rate code if I checked blood pressure or weight, or looked for a Romberg sign-procedures less clearly necessary at each visit with a psychiatrist than with an internist. Now, between the 9921x and 9083x add-ons, I have at least 15 choices of codes applicable to my usual half-hour or hour sessions. Multiple 9083x codes differ by merely 1 minute, such as 37 minutes of therapy for 90833 and 38 minutes for 90836. If I have set an hour up for a patient, I could easily stop or extend the work around that 38-minute mark. Also, just a few extra questions or assessments, such as measuring weight and blood pressure, can upcode the 9921x. So, insurance reimbursements now fluctuate far more from session to session.
What are our ethical challenges in this menagerie of new CPT codes? In the standard fee-for-a-session model, should we go the extra minute (or few) in outpatient treatment to get a higher payment rate for a patient or ourselves by upcoding the 9083x? Should we weigh and check blood pressure to be able to upcode the 9921x? That is useful information, but usefulness does not make it necessary. Is it even possible to accurately separate the alloy of psychology and medicine that constitutes psychiatry into component metals for the purpose of coding-distilling how much time we think like physicians (9921x) from how much time we think like other mental health professionals doing therapy (9083x). This was all amalgamated in the old “medical psychotherapy” codes. The new system asks us to make sometimes impossible distinctions. Other settings, such as inpatient hospitals or nursing homes, give additional opportunities to choose wider explorations of problems that would permit upcoding.
Charging for services is not an unethical practice, although the APA Principles of Ethics (ie, the American Psychiatric Association’s “The Principles of Medical Ethics With Annotations Especially Applicable to Psychiatry-Principle 2.6”) asks us to deploy “the utmost consideration for the patient and his or her circumstances” in our billing.1 Yet, there has always been an inherent conflict of interest as each physician determines the duration and frequency of treatment sessions and what is done in those sessions.
Our health care system has always tied financial rewards to these decisions. However, the core ethical principle of beneficence helps attenuate the conflict of interest-the patient’s medical needs outrank the economic needs of the physician. Beneficence has been a foundational ethical principle since the Hippocratic Oath: “What I do is for the good of the patient.” Beneficence is the bedrock of trust and allows protection of vulnerability in the doctor-patient relationship. That is why boundary violations are so particularly egregious: they involve the practitioner’s needs trumping those of the patient. Boundary violations are an exploitation of the patient that also breach another core principle-nonmaleficence (do no harm)-another Hippocratic idea.
However, the new CPT environment makes it harder than ever to honor beneficence. The increased granularity of coding choices is now more epistemically vague. Therefore, the opportunities to make decisions that might not be beneficent, which might be self-serving (deliberately or inadvertently), are far more numerous: 15 or more choices, not 2.
What’s a psychiatrist to do? Should we sit with a table of insurance reimbursement fees and try to upcode as much as reasonably possible? On a business level, clinicians benefit from maximum reimbursement; in fact, physicians from other disciplines often employ “coding technicians” to review documentation and maximize reimbursements, by giving careful and accurate attention to coding definitions to avoid undercoding. In a way, that approach-separating out billing consciousness from clinical consciousness by employing different personnel-makes the ethical “slippery slope” less steep. Or if done according to guidelines, it minimizes the conflict of interest. For practitioners who do not have the luxury of an in-house coder, that slope may be more precarious.
Our appointment times may need to become less rigid, more approximate, with an accordion quality (perhaps causing us to run perennially late, like many of our medical colleagues, when some patients require more time!). Perhaps that would allow us time to breathe more ethically in the new coding environment and, in the words of Maryland Secretary of Health and Mental Hygiene, Joshua Sharfstein, MD, we would be paid for “value not volume.”2 Flat salaries, such as those of the United Kingdom’s National Health Service, or capitations may obviate these ethical problems but come with ethical challenges of their own.
In his book The Humanities and the Profession of Medicine, Allen Dyer, MD, PhD, notes that the very definition of a professional is one who “professes a set of values.”3 That professing is not just public, in oaths and ethics codes, but is internal, in relationship to one’s self. In psychiatry in particular, where so much of our work is conducted in private, ethics involves an inner adherence to values such as beneficence, nonmaleficence, and respect for vulnerable persons. Indeed, one of the distinctions between ethics and law is internal commitment versus external policing.
So, to practice ethically in the new CPT coding environment, an explicit mindfulness is needed that small differences in questioning, examination, and time can have significant financial consequences. Inner interrogation about the motives for these choices must be a more scrupulous part of practice than ever before, constantly measuring those motives against the touchstones of clinical beneficence and nonmaleficence. Also, while patients may appreciate any financial benefits for themselves, this benefit is not generally included within the scope of “clinical benefits” understood to be the aim of beneficence.
A colleague in Maryland was recently called by the medical director of a patient’s managed care insurer for an audit. The chart was requested to help ascertain whether the CPT codes were correctly coded, based on the documentation, and whether the amount of care associated with each code was medically necessary.
Analogous to surgery, confidentiality is the “sterile field” of our operations, without which the procedures are at risk for poorer outcomes and greater complications. It may be more important and instrumental in psychiatry than in any other specialty because of the particularly vulnerable emotions and disclosures that arise in our work. Our scrupulous attention to confidentiality both encourages vulnerable people to show up for help in the domain of mental health-and fosters honest and complete disclosures. This is as true for medication therapy as it is for psychotherapy.
The new ramified CPT codes bring a renewed threat to the confidentiality of medical records and communications with third parties. With so many coding choices, there is more need for payers to audit records to determine whether the chosen code is justified. Although this has always been a potential issue, the less nuanced former codes did not give as much room for argument. Previously, coarse preauthorization review protocols for outpatient care were developed in many states, limiting payers’ access to charts and detailed information using simplified “treatment plan reviews.” New Jersey4 and Washington, DC,5 were some of the first jurisdictions to legislate this. Prior authorizations have typically specified numbers of inpatient days or outpatient visits. Now the question of what was done during each visit is more important than ever for authorization, since the system is no longer keyed to flat time-based fees, and there are so many more possible codes, each with different payments.
The Mental Health Parity provision of the Affordable Health Care Act could make authorization of numbers of visits obsolete, unless preauthorization of numbers of visits for other medical conditions becomes a new insurance practice. In the parlance of the insurance industry, “determination of medical necessity” might now be played out in a different way: greater attention to the codes that are being billed. This will require more detailed disclosure of clinical information to justify the myriad of coding possibilities, perhaps even session-by-session auditing: Were blood pressure and weight measurement and coordination of care required for 2 visits in a row? If therapy in the last session was 35 minutes (90833 add-on), why did this session need to be 40 minutes (90836 add-on)? Did the session include “counseling” or actual “therapy” (justifying a 9083x code)?
Electronic medical records will become a portal not only for potentially useful sharing of information between clinicians, but also for uploading clinical information to third parties for the purpose of billing audits. A myriad of recent news stories, such as the theft of electronic patient records at Howard University Hospital in Washington, DC, offers little reassurance regarding cybersecurity.6-8
If we remodel our work to the “accordion” approach for patient-encounter time as noted above, the confidentiality of stacked-up patients in the waiting room can become an increasingly sensitive issue; it is already a problem in large clinic waiting rooms.9 Indeed, many people in support groups I have addressed around the country tell me they try to avoid psychiatric clinic settings for just this reason.
The APA Principles of Medical Ethics devotes a good deal of print to the issue of confidentiality.1 It advises psychiatrists to protect psychiatric records with “extreme care.” We should “disclose only that information which is relevant to the . . . question at hand.” Many other forces are knocking at the door of confidentiality (eg, new gun-reporting laws, social networking). New CPT coding only adds to this threatening fray, and hence the need for the ethical psychiatrist to intensify vigilance, protection, and advocacy (both for the patient and on the policy and legislative levels through organizations such as the APA and the National Alliance on Mental Illness).
As we often teach our patients: change is hard, often needed, and sometimes inevitable. Although the new CPT coding changes might bring certain advantages, such as reflecting our role as physicians who use the same E&M codes intended for all physicians, they equally raise important ethical challenges that call for heightened attention. Indeed, the best safeguard to ethical practice is mindful attention to one’s own behaviors, measured against the venerable standards maintained by the carefully evolving ethos of our profession. When in doubt, consult a colleague, your local hospital ethics committee, or the APA ethics committee.
This article was published online ahead of print, on 3/11/2014.
Dr Komrad is Chair of Ethics for the Sheppard Pratt Health System in Maryland and a member of the APA Ethics Committee. He is also the author of You Need Help: A Step-by-Step Plan to Convince a Loved One to Get Counseling. He reports no conflicts of interest concerning the subject matter of this article.
1. American Psychiatric Association. The Principles of Medical Ethics With Annotations Especially Applicable to Psychiatry. Arlington, VA: American Psychiatric Press Inc; 2008.
2. Harder L. Can healthcare IT make Maryland healthier? Maryland Physician. June 2012.
3. Dyer AR. The Humanities and the Profession of Medicine. Research Park Triangle, NC: National Humanities Center; 1982.
4. New Jersey. Statute 45:14B-32.
5. Washington, DC. Official Code: § 7-1202.07.
6. Shultz D. Medical data breaches raising alarm. Washington Post. June 2, 2012.
7. Perrone M. GAO urges more Medicare plan oversight. Houston Chronicle. September 5, 2006.
8. Upton J. Michigan medical records accidentally posted on Web. Detroit Free Press. February 12, 1999.
9. Prowler ML, Neimark G. What’s lurking in your waiting room? Curr Psychiatry. 2008;7:63.