Imad M. El-Kebbi

Patient adherence improves glycemic control

Purpose The purpose of this study was to assess the influence of appointment keeping and medication adherence on HbA1c. Methods A retrospective evaluation was performed in 1560 patients with type 2 diabetes who presented for a new visit to the Grady Diabetes Clinic between 1991 and 2001 and returned for a follow-up visit and HbA1c after 1 year of care. Appointment keeping was assessed by the number of scheduled intervening visits that were kept, and medication adherence was assessed by the percentage of visits in which self-reported diabetes medication use was as recommended at the preceding visit. Results The patients had an average age of 55 years, body mass index (BMI) of 32 kg/m2, diabetes duration of 4.6 years, and baseline HbA1c of 9.1%. Ninety percent were African American, and 63% were female. Those who kept more intervening appointments had lower HbA1c levels after 12 months of care (7.6% with 6-7 intervening visits vs 9.7% with 0 intervening visits). Better medication adherence was also associated with lower HbA1c levels after 12 months of care (7.8% with 76%-100% adherence). After adjusting for age, gender, race, BMI, diabetes duration, and diabetes therapy in multivariate linear regression analysis, the benefits of appointment keeping and medication adherence remained significant and contributed independently; the HbA1c was 0.12% lower for every additional intervening appointment that was kept (P= .0001) and 0.34% lower for each quartile of better medication adherence (P= .0009). Conclusion Keeping more appointments and taking diabetes medications as directed were associated with substantial improvements in HbA1c. Efforts to enhance glycemic outcomes should include emphasis on these simple but critically important aspects of patient adherence.

Clinical Inertia Contributes to Poor Diabetes Control in a Primary Care Setting

Purpose The purpose of this study was to determine whether “clinical inertia”—inadequate intensification of therapy by the provider—could contribute to high A1C levels in patients with type 2 diabetes managed in a primary care site. Methods In a prospective observational study, management was compared in the Medical Clinic, a primary care site supervised by general internal medicine faculty, and the Diabetes Clinic, a specialty site supervised by endocrinologists. These municipal hospital clinics serve a common population that is largely African American, poor, and uninsured. Results Four hundred thirty-eight African American patients in the Medical Clinic and 2157 in the Diabetes Clinic were similar in average age, diabetes duration, body mass index, and gender, but A1C averaged 8.6% in the Medical Clinic versus 7.7% in the Diabetes Clinic (P < .0001). Use of pharmacotherapy was less intensive in the Medical Clinic (less use of insulin), and when patients had elevated glucose levels during clinic visits, therapy was less than half as likely to be advanced in the Medical Clinic compared to the Diabetes Clinic (P < .0001). Intensification rates were lower in the Medical Clinic regardless of type of therapy (P < .0001), and intensification of therapy was independently associated with improvement in A1C (P < .001). Conclusions Medical Clinic patients had worse glycemic control, were less likely to be treated with insulin, and were less likely to have their therapy intensified if glucose levels were elevated. To improve diabetes management and glycemic control nationwide, physicians in training and generalists must learn to overcome clinical inertia, to intensify therapy when appropriate, and to use insulin when clinically indicated.

Exercise preferences and barriers in urban African Americans with type 2 diabetes.

PURPOSE The purpose of this study was to determine physical activity preferences and barriers to exercise in an urban diabetes clinic population. METHODS A survey was conducted of all patients attending the clinic for the first time. Evaluation measures were type and frequency of favorite leisure-time physical activity, prevalence and types of reported barriers to exercise, and analysis of patient characteristics associated with reporting an obstacle to exercise. RESULTS For 605 patients (44% male, 89% African American, mean age = 50 years, mean duration of diabetes = 5.6 years), the average frequency of leisure activity was 3.5 days per week (mean time = 45 minutes per session). Walking outdoors was preferred, but 52% reported an exercise barrier (predominantly pain). Patients who cited an impediment to physical activity exercised fewer days per week and less time each session compared with persons without a barrier. Increasing age, body mass index, college education, and being a smoker increased the odds of reporting a barrier; being male decreased the chances. Men reported more leisure-time physical activity than women. Exercise preferences and types of barriers changed with age. CONCLUSIONS Recognition of patient exercise preferences and barriers should help in developing exercise strategies for improving glycemic control.

Diabetes in Urban African-Americans. IX. Provider Adherence to Management Protocols

OBJECTIVE Staged diabetes management should permit glycemic goals to be attained in a timely manner, but the success of such an approach requires conformity by health care providers. To test performance, we analyzed the adherence of practitioners to a protocol for staged management of NIDDM patients. RESEARCH DESIGN AND METHODS Records of patients treated at the Grady Memorial Hospital Diabetes Clinic were reviewed retrospectively over a 3-year period. For each patient, intensification of therapy was indicated if fasting plasma glucose was > 7.8 mmol/l and a prior HbA1c was > 7.0%. Protocols dictated a progression from dietary therapy alone to increasing dosages of sulfonylureas to increasing dosages of insulin. Patients were seen at bimonthly intervals. RESULTS During the 3-year period, 1,051 patient visits met protocol criteria for intensification. Adherence to the protocol improved significantly in the 3rd year compared with the first 2 years (30, 31, and 47% adherence in the 1st, 2nd, and 3rd years, respectively). Patients treated with diet alone were significantly less likely to have their therapy intensified than patients on sulfonylureas or insulin (intensification rates 25, 41, and 47%, respectively). In the management of patients treated with diet alone, practitioners were reluctant to intensify therapy at early visits, but were more likely to do so later, 19% of patients beyond goal range at the 2-month visit were started on pharmacological therapy vs. 28% at the 4-month visit, and 39% at the 6-month visit (P < 0.01). In contrast, there was no significant difference in the frequency of therapy intensification between early and late visits for patients on sulfonylureas or insulin. Practitioners appeared to base the decision to intensify on the fasting plasma glucose level more than on the most recent HbA1c. Age did not appear to be a significant factor in the decision to intensify. CONCLUSIONS Although staged management protocols constitute critical tools to achieve glycemic goals, the adherence of health care providers may be suboptimal. Special efforts may be needed to assure compliance.

Barriers to Diabetes Education in Urban Patients Perceptions, Patterns, and Associated Factors

Purpose This study explored patients’ perceptions of barriers to diabetes education among a mostly African American population of adults with diabetes. Methods A survey was conducted among 605 new patients attending an urban outpatient diabetes clinic. The questionnaire gathered information on issues patients believed would adversely affect their ability to learn about diabetes. The type and frequency of education barriers were evaluated, and variables associated with reporting an obstacle were analyzed. Results Average patient age was 50 years, diabetes duration was 5.6 years, body mass index was 32 kg/m2, and hemoglobin A1C was 9.1%. The majority (56%) were women, 89% were African American, and 95% had type 2 diabetes. Most respondents (96%) had received some prior instruction in diabetes care; however, 53% anticipated future difficulties learning about diabetes. The most commonly cited concerns were poor vision (74%) and reading problems (29%). Patients with a perceived barrier to diabetes education were older (P < .001) than were persons without a barrier, and they differed in both employment and educational status (both P < .001). In adjusted analyses, older age, male gender, being disabled, and having an elementary education or less were associated with a significantly increased likelihood of having a barrier to diabetes education, whereas having a college education decreased the odds. Higher hemoglobin A1C levels also tended to be associated with a greater chance of reporting an education barrier (P = .05). Conclusions A substantial number of persons anticipated a barrier to diabetes education. Interventions at multiple levels that address the demographic and socioeconomic obstacles to diabetes education are needed to ensure successful self-management training.

Little Time for Diabetes Management in the Primary Care Setting

PURPOSE This study was conducted to determine how time is allocated to diabetes care. METHODS Patients with type 2 diabetes who were receiving care from the internal medicine residents were shadowed by research nurses to observe the process of management. The amount of time spent with patients and the care provided were observed and documented. RESULTS The total time patients spent in the clinic averaged 2 hours and 26 minutes: 1 to 9 minutes waiting, 25 minutes with the resident, and 12 minutes with medical assistants and nurses. The residents spent an average of only 5 minutes on diabetes. Glucose monitoring was addressed in 70% of visits; a history of hypoglycemia was sought in only 30%. Blood pressure values were mentioned in 75% of visits; hemoglobin A1c (A1C) values were addressed in only 40%. The need for proper foot care was discussed in 55% of visits; feet were examined in only 40%. Although 65% of patients had capillary glucose levels greater than 150 mg/dL during the visit and their A1C averaged 8.9%, therapy was intensified for only 15% of patients. CONCLUSIONS During a routine office visit in a resident-staffed general medicine clinic, little time is devoted to diabetes management. Given the time pressures on the primary care practitioner and the need for better diabetes care, it is essential to teach an efficient but systematic approach to diabetes care.

The Improving Primary Care of African Americans with Diabetes (IPCAAD) project: rationale and design

African Americans have an increased burden of both diabetes and diabetes complications. Since many patients have high glucose levels novel interventions are needed, especially for urban patients with limited resources. In the Grady Diabetes Clinic in Atlanta, a stepped care strategy improves metabolic control. However, most diabetes patients do not receive specialized care. We will attempt to translate diabetes clinic approaches to the primary care setting by implementing a novel partnership between specialists and generalists. We hypothesize that endocrinologist-supported strategies aimed at providers will result in effective diabetes management in primary care sites, and the Improving Primary Care of African Americans with Diabetes project will test this hypothesis in a major randomized, controlled trial involving over 2000 patients. Physicians in Grady Medical Clinic units will receive (1) usual care, (2) computerized reminders that recommend individualized changes in therapy and/or (3) directed discussion by endocrinologists providing feedback on performance. We will measure outcomes related to both microvascular disease (HbA1c, which reflects average glucose levels over an approximately 2-month period) and macrovascular disease (blood pressure and lipids) and assess provider performance as well. We will compare two readily generalizable program interventions that should delineate approaches effective in a primary care setting as needed to improve care and prevent complications in urban African Americans with type 2 diabetes.

High Prevalence of Albuminuria Among African-Americans With Short Duration of Diabetes

These two conditions are considered variants of the same defect of the stimulatory guanine nucleotide-binding (Gs) protein of adenylate cyclase, which is necessary for parathyroid hormone and other hormones such as gonadotropin, beta-adrenergic agonist, and thyrotropin to use cAMP as an intracellular second messenger. We described two related women with apparent AHO and late-onset diabetes. Both patients had normal serum calcium levels, normal parathyroid hormone levels, and the characteristic somatic features of short stature, round face, obesity, and shortened fourth and fifth metacarpals and metatarsals, consistent with pseudo-PHP Both disorders, PHP and pseudo-PHP, can occur within the same family, and there is accumulating evidence that genomic imprinting is involved in the disease (1). Full phenotypic expression (AHO and parathyroid hormone resistance, as in PHP type la) occurs in maternally transmitted cases, whereas partial expression (AHO without parathyroid hormone resistance, as in pseudo-PHP) occurs when the gene is paternally transmitted. The pedigree of our patients showed genetic transmission from their father. Patients with type 2 diabetes have defects in insulin action, abnormal insulin secretion, and increased hepatic glucose production. Although precise pathways responsible for these defects have not been thoroughly identified, they are likely to be genetically heterogenous with mutations in several different genes that are able to cause hyperglycemia. Some reported genetic loci for type 2 diabetes have been mapped on chromosome 20q, chromosome 7p, chromosome 12q, chromosome 2, and so forth (2,3). In most cases of AHO, reduced levels of Gs protein a subunit (Gsa protein) have been found. A number of deactivating mutations in the gene for Gsa protein located on chromosome 20ql3 have been described for this disorder (1), but del(2)(q37) has also been described in some AHO patients (4) and thus explains the heterogeneity observed in this AHO disorder. PHP type la or pseudo-PHP is assumed to be a Gsa protein problem and this protein is encoded by chromosome 20ql3. 2-3. Occasionally, these disorders may be associated with resistance of diverse target tissues to hormones and neurotransmitters whose actions require stimulation of adenylate cyclase and thus open calcium channels. It should be considered whether this Gsa protein problem will lead to diabetes with insulin resistance. Certainly, either these pseudo-PHP women with type 2 diabetes have a mutation in the Gsa protein or nearby genome for its susceptibility to type 2 diabetes, or they represent just a phenomenon of coincidence. Further evaluation and collection of cases are necessary to define the possible role and interrelationship of pseudo-PHP and type 2 diabetes.

Use of a Glucose Algorithm to Direct Diabetes Therapy Improves A1C Outcomes and Defines an Approach to Assess Provider Behavior

Purpose The purpose of this study was to determine whether an algorithm that recommended individualized changes in therapy would help providers to change therapy appropriately and improve glycemic control in their patients. Methods The algorithm recommended specific doses of oral agents and insulin based on a patients medications and glucose or A1C levels at the time of the visit. The prospective observational study analyzed the effect of the algorithm on treatment decisions and A1C levels in patients with type 2 diabetes. Results The study included 1250 patients seen in pairs of initial and follow-up visits during a 7-month baseline and/or a subsequent 7-month algorithm period. The patients had a mean age of 62 years, body mass index of 33 kg/m2, duration of diabetes of 10 years, were 94% African American and 71% female, and had average initial A1C level of 7.7%. When the algorithm was available, providers were 45% more likely to intensify therapy when indicated (P = .005) and increased therapy by a 20% greater amount (P < .001). A1C level at follow-up was 90% more likely to be <7% in the algorithm group, even after adjusting for differences in age, sex, body mass index, race, duration of diabetes and therapy, glucose, and A1C level at the initial visit (P < .001). Conclusions Use of an algorithm that recommends patient-specific changes in diabetes medications improves both provider behavior and patient A1C levels and should allow quantitative evaluation of provider actions for that providers patients.