Psychosocial issues may exert substantial influence on glycemic control in diabetic patients.1-4 Psychological factors have been shown to increase the risks of poor glycemic control, “brittle diabetes,” and diabetic ketoacidosis. Depression has been identified as one negative influence of glycemic control among pediatric and adolescent patients.5 It has been cited as a possible contributing factor to sexual dysfunction in women with type 1 diabetes.6 And,depression has been observed to affect family members of patients with type 2 diabetes and may influence family dynamics toward the condition.7
This growing body of literature suggests that the inclusion of a clinical psychology component in the integrated diabetes care team is desirable. Psychological factors may significantly influence glycemic control.8,9 Indeed, the Education Recognition Program of the American Diabetes Association advocates this inclusion.10
However, the practical issues involved in such integration can be challenging. Problems in obtaining reimbursement for these services within a medical organization are one important set of obstacles. Psychological services are usually paid by insurers according to insurance codes and diagnostic groups (American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders codes) that differ from medical codes and diagnoses. These reimbursement patterns are often associated with distinct contractual arrangements from insurance contracts with medial care professionals. Therefore, psychological personnel and medical personnel are usually employed by different corporate entities. Thus, the actual degree to which clinical psychologists or other professional psychosocial professionals have been truly integrated into the multi-disciplinary diabetes care team is minimal.
Specialty programs of diabetes care may have a higher prevalence of patients who are referred for poor glycemic control and who have significant psychosocial problems.11 The need for integrated psychosocial support services in such programs is substantial. Our diabetes program reflects this need. It sees approximately 1,200 new diabetic patients per year, most of whom are referred for poor glycemic control. Therefore, we evaluated the mechanisms to integrate psychological counseling into our multidisciplinary care team in a fiscally responsible and feasible manner.
Reimbursement Issues of Integrating Psychosocial Services
Psychologists' services may be billed under medical codes by three possible methods. They may be billed incident to the care of the physician under evaluation and management (E&M) codes (99211-99215). They may also be billed under psychotherapy codes (90804-90809), if the patient has an established psychiatric diagnosis. Or, they may be billed under health and behavior assessment/intervention codes (96150-96154).
Payments under the E&M codes (99211-99215) are the standard payments for these codes. The same is true for mental health codes. Services rendered by a psychologist under these codes must be billed by the same criteria as billings for any other health professional. Billing under the health and behavior assessment codes are for evaluation of “the psychological,behavioral, emotional, cognitive, and social factors important to the prevention, treatment, or management of physical health problems.”According to Medicare,12 the specific use of each of the health and behavioral assessment codes depends on whether the code is for assessment of the patient or intervention and on whether the intervention is with the patient, the patient's family, or both(Table 1).
The psychotherapy codes (90804-90809) may only be used if the assessments and therapeutic interventions are carried out by a licensed psychiatrist or psychologist. Another requirement for the use of these codes, in contrast to the health and behavioral assessment codes, is that they be billed for a specific and established psychiatric diagnosis. The even-numbered codes in the sequence (90804, 90806, and 90808) are for outpatient psychotherapy services rendered without associated medical care services. Which of these codes is applied is determined by the length of the counseling session. Reimbursement for Medicare patients in Georgia ranges from $57.31 to $128.28 per session. The odd-numbered codes (90805, 90807, and 90809) are used when the office visit is associated with medical evaluation and management services.
The health and behavior assessment codes (96150-96154) apply to diabetic patients who are having difficulty adhering to their diabetes care program. To qualify for services under these codes, the patient must be capable of understanding the interventions and must be present for all interventions. Medicare lists a code (96155) for intervention with the family without the patient present; however, this is not a reimbursable code. Also, in some states, Medicaid may not recognize these codes. These codes cannot be applied for preventive counseling, such as smoking cessation. When patients with established psychiatric diagnoses are being counseled for their psychiatric condition, the psychotherapy codes should be used. The health and behavior assessment codes can also be applied to diabetic patients with known psychiatric diagnoses if the assessment or intervention at the visit being billed is primarily focused on the adherence of the patient to a medical treatment regimen, such as a diabetes care program.
The documentation of such services should clearly provide justification for the use of either the psychotherapy codes or the health and behavior codes. Documentation for the health and behavior assessment codes must establish the medical diagnosis for which the behavioral assessments and interventions were applied. The reason for the application of these assessments and interventions should be clear by descriptions of how the patient has failed to deal with the medical condition. The suspicions that problems in adherence are behavioral should be provided to justify initial assessments. Subsequent interventions should logically follow the results of assessments.
Application of Health and Behavior Codes in a Population of Diabetic Patients
We surveyed the use of these codes in insulin-requiring diabetic patients who failed to achieve hemoglobin A1c values ≤ 8.0% over a 1-year period. All of these patients were referred to our program by primary physicians because of poor glycemic control. All patients had completed a formal program of diabetes patient education and had seen either a diabetes nurse educator or both a diabetes nurse educator and a dietitian after they completed the formal program of diabetes education.
A synopsis of the psychosocial influences on poor adherence to the diabetes treatment program in 10 patients served under these codes is shown in Table 2. In every case, the patient was referred for psychosocial services after repeated episodes of severe hyperglycemia. In every case, the diabetes nurse educator attested that the patient had the knowledge and skills to adhere successfully to the diabetes treatment program. All patients but one were insulin-requiring diabetic patients.
The psychosocial causes of poor adherence shown in Table 2 were each deemed to be a primary or the primary influence on patient behavior toward the diabetes treatment regimen by the psychologist. Although psychiatric diagnoses had been established for three of these patients, the specific issues shown in Table 2 had never been documented in any of these cases.
Charges and Revenue From the Use of E&M and Mental Health Codes
We surveyed the use of these codes for 4 months, from June to September 2003 to determine the feasibility of engaging psychosocial services as a part of our multidisciplinary diabetes care team. During this period, the behavioral assessment and intervention codes were not yet being employed pending our study of their acceptability to our large regional insurers. Psychotherapy codes were billed in 9% of 141 contacts, and E&M codes were used for the remainder.
Table 3 shows that 5-9 months after services were billed, reimbursement was received for 72.3% of all charges. Approximately 9.6% of charges were deemed uncollectible, and 18.1%were still pending at the time this article was written. Since then, the majority of our charges have been submitted under the health and behavior assessment codes (96150-96154). While insufficient time has passed to determine reimbursement patterns under these codes versus the others,preliminary experience suggests that they are being well accepted by our area insurers.
Conclusions
Many providers of diabetes care recognize the potential benefits of including skilled psychosocial support services in the multidisciplinary program of care. Until now, the funding of these services has been problematic and unclear. We hope that the application of the three different sets of codes discussed above may provide integrated programs of care with some new options to include psychologists. While our experience is still preliminary, the relative reliability of these reimbursement procedures for our program seems promising. Health professionals interested in specialty diabetes care programs may wish to continue this dialog to expand our base of understanding of this exciting possibility.
Steven B. Leichter, MD, FACP, FACE, is co-director of the Columbus Research Foundation and President of Endocrine Consultants, PC, in Columbus,Ga., and is a professor of medicine at Mercer University School of Medicine in Macon, Ga. Elizabeth Dreelin, PhD, is the clinical psychologist at Endocrine Consultants, and Susan Moore is the head of billing at Endocrine Consultants in Columbus. Ga.