Nurse Case Management To Improve Glycemic Control in Diabetic Patients in a Health Maintenance Organization

The Diabetes Control and Complications Trial (DCCT) and the Kumamoto study [1, 2] showed that near-normal glycemic control reduces the development and progression of microvascular and neuropathic complications by approximately 50% in type 1 and type 2 diabetes mellitus. Additional analyses [3-5] indicate that therapy to achieve near normalization of blood glucose levels is cost-effective compared with other treatments. Thus, the American Diabetes Association has recommended that all persons with diabetes attempt to achieve near normalization of blood glucose levels [6]. This recommendation is not routinely followed in medical practice. In a 1989 national survey of physician practice behaviors in the United States, 64% of physicians agreed that achieving target HbA1c values is very important but only 18% reported that they ordered HbA1c tests every 2 to 3 months for patients with type 1 diabetes [7]. Although 98% agreed that patient education improves glucose control, only 55% reported that they routinely used a dietitian or a diabetes educator in patient care. Studies indicate that bringing clinical practice into line with scientific knowledge can be difficult. Methods used to achieve diabetes control in clinical trials are resource intensive. The American Diabetes Association currently recommends that patients with diabetes see their primary care physicians two to four times per year. Data from the National Health Interview Survey, a nationally representative survey [8], indicate that most patients with diabetes are seen by nonspecialists and that 69% of physician visits last less than 15 minutes. Algorithms for diabetes care exist but may be complex and difficult for physicians to follow, given patient load, diversity of patients seen, lack of information systems, and time constraints. Simple, low-cost methods of translating guidelines into clinical care are required. One solution may be to make greater use of personnel other than physicians. Nurse case management was an integral part of intensive therapy in the DCCT and has proven to be effective in reducing smoking and cholesterol levels after acute myocardial infarction [3, 9]. A nonrandomized study [10] of more than 700 patients with diabetes in a health maintenance organization suggests that nurse case management may be effective in improving metabolic control. Other studies [11, 12] show a strong association between algorithm-directed nurse interventions and improved glycemic control. To our knowledge, no randomized, controlled clinical trial of nurse case management in diabetes has yet been published. In a 12-month randomized, controlled trial, we compared a nurse case management model of diabetes care with usual diabetes management in a primary care setting. Methods Patients Our study was approved by the institutional review board of the Prudential Center for Health Care Research, and all patients gave written informed consent. Participants were recruited from two of the largest clinics within the Jacksonville Health Care Group, which is the largest provider of primary care services for the Prudential HealthCare HMO plan of Jacksonville, Florida. The Jacksonville Health Care Group is a group of 43 primary care physicians who provide care in eight clinics to more than 75 000 Prudential HealthCare plan members. Potential study participants were identified through a database used to support quality-improvement activities. Prudential HealthCare HMO members who had diabetes were included in the database if they had visited a physician for diabetes (International Classification of Diseases, 9th Revision, codes 250.0 to 250.9), had had a hospital claim processed for diabetes, had been seen by the utilization management nurse, or had been referred to an ophthalmologist for a diabetic retinal examination. This database is updated regularly. A list with each member's name, address, telephone number, medical record number, member identification number, age, sex, physician, and clinic was generated by merging the data from the database with enrollment information. In addition, a list of members who may have had diabetes was created by using pharmacy data. Adult members with diabetes who were potential study participants each received a recruitment call and were invited to schedule an appointment with a research assistant to discuss participation in the study. We made a total of 14 calls at different times and on different days before coding a member as unavailable. After consent was given and the eligibility assessment was completed, baseline information was obtained and an HbA1c test was ordered if the result of one given within the previous 60 days was not available. Patients were ineligible for the study if they had a recent HbA1c value less than 7.0%; had uncontrolled hypertension (blood pressure > 180/110 mm Hg); had unstable angina (class 4); had had a myocardial infarction in the past 3 months; had had two or more episodes of seizures; had alcoholism or drug abuse documented in the chart; had late-stage complications of diabetes or other chronic conditions, such as severe immunodeficiency or cirrhosis; were pregnant or were planning to become pregnant in the next 12 months; or were unable to perform self-management. Patients were randomly assigned in blocks to either the nurse case management (intervention) group or the usual care group. Randomization was based on a 1:1 allocation ratio and a block size of three. Each block contained six patients, three in each study group. This randomization scheme ensured that the desired allocation ratio-one intervention patient to one usual care patient-was maintained after sequential enrollment of every sixth patient. Outcome Measures Change in HbA1c value was the primary outcome measure. Decreased HbA1c values correlate directly with reduced risk for diabetes-related microvascular and neuropathic complications in type 1 and type 2 diabetes [1, 3]. We also assessed health-related quality of life by using four generic questions developed by the Centers for Disease Control and Prevention for the Behavioral Risk Factor Surveillance System (BRFSS) [13, 14]. These questions evaluate key conceptual domains of health-related quality of life: 1) patient-perceived general health status, 2) patient-perceived physical dysfunction during the previous 30 days, 3) patient-perceived mental dysfunction during the previous 30 days, and 4) patient-perceived functional incapacity during the previous 30 days for either mental or physical reasons. The BRFSS quality-of-life measures have been validated in a national sample of adults in the United States [15]. Patient-perceived health was found to be a good proxy indicator for chronic disease conditions. The other three domains further characterize general health functioning and quality of life [15]. In this analysis, we report findings related to the patient-perceived general health status domain. Intervention and Follow-up The nurse case manager was a registered nurse and a certified diabetes educator. She was trained to follow a set of detailed management algorithms under the direction of a board-certified family medicine physician and an endocrinologist who were responsible for all diabetes management decisions for patients in the intervention group but were not primary care providers for these patients. The algorithms were specific for type of diabetes and were developed by a multidisciplinary team on which endocrinology, family medicine, nursing, pharmacy, health services research, and epidemiology were represented. The algorithms progressively moved a patient toward improvement of glycemic control through adjustments in medication, meal planning, and reinforcement of exercise (Figure 1). Figure 1. Algorithm for management of type 2 diabetes mellitus. Patients assigned to receive nurse case management met with the nurse for an initial assessment, were instructed about a blood glucose monitoring schedule, and returned for a follow-up visit 2 weeks later. The initial visit with the nurse averaged 45 minutes. At the 2-week follow-up visit, the nurse reviewed the patient's blood glucose log; explained the algorithm step to which the patient had been assigned; and used this information as the baseline for subsequent medication adjustments, meal planning, and exercise reinforcement. Patients receiving nurse case management were also referred to a 5-week, 12-hour diabetes education program that included individual counseling by a dietitian, individual counseling by an exercise therapist, and group diabetes education classes. Subsequent in-person follow-up visits occurred quarterly. Patients in the nurse case management group who were taking insulin received weekly follow-up telephone calls. After the nurse reviewed the blood glucose log and discussed glucose values with the patient, medication regimens were adjusted as needed and meal planning and exercise were reinforced. Patients treated with oral agents or diet and exercise received follow-up telephone calls every 2 weeks. The nurse case manager met at least biweekly with the family medicine physician and the endocrinologist to review patient progress, medication adjustments, and other issues related to diabetes care. All medication adjustments or changes were communicated to the patients' regular primary care physicians. Patients assigned to receive usual care were given blood glucose meters and strips, were encouraged to discuss enrollment in the diabetes education class with their physicians if they had not done so in the past year, and continued to receive diabetes care and follow-up from their primary care physicians. The 5-week diabetes education program is a standard, free-of-charge benefit for all HMO members with diabetes. All Jacksonville Health Care Group primary care physicians participate in an annual diabetes care seminar and undergo regular peer review of their adherence to published diabetes care standards. Tests to meas