Objective: To compare angiotensin-converting enzyme (ACE) inhibitor use in patients with type 2 diabetes at 1 year and 3 years after guidelines were published.
Study Design: Retrospective database review.
Patients and Methods: The drug utilization review database of a state managed care plan was accessed to retrieve 2 random samples of 500 patients each. These patients had an International Classification of Diseases, Ninth Revision, Clinical Modification code for diabetes mellitus (250) and a National Drug Code for an oral hypoglycemic agent in both 1998 and 2000. Specific clinical modification codes, prescription claims, and diagnostic codes were obtained from patient profiles. Use of ACE inhibitors in 1998 and 2000 then was evaluated by using Pearsonâ€™s chi-square test.
Results: The proportion of patients with diabetes and hypertension who were taking an ACE inhibitor increased by 10 percentage points over the 2 years; however, ACE inhibitors were only used in 46% of those patients in 2000. A few of the patients receiving an ACE inhibitor had a contraindication to use of the agent. Microalbuminuria screening and glycosylated hemoglobin screening were found to have been conducted in only 4.6% and 54.6%, respectively, of the 496 patients in 2000.
Conclusions: The results of this study indicate that although ACE inhibitor use improved, fewer than 50% of patients received appropriate therapy. Awareness of and adherence to the recommendations in the guidelines need to be improved. Larger studies may be beneficial to determine more clearly the extent of this problem.
(Am J Manag Care. 2004;10(part 2):124-129)
Almost half (45%) of the patients newly diagnosed with end-stage renal disease and receiving hemodialysis had diabetes mellitus, according to The United States Renal Data System 2002 report.1 The incidence of end-stage renal disease secondary to diabetes has continuously increased over the past decade. Angiotensin-converting enzyme (ACE) inhibitors have been proven to slow the progression of renal dysfunction in diabetes patients with microalbuminuria.2-4 Furthermore, these agents have been found to decrease microalbuminuria in patients without hypertension.2,4-6 The American Diabetes Association (ADA) first recommended the use of ACE inhibitors for all type 2 diabetes patients with microalbuminuria in 1996.7 This also was a recommendation of the Joint National Committee on Detection, Evaluation, and Treatment of Hypertension in 1997 (JNC 6).8 Currently, ACE inhibitors are the preferred agents for diabetes patients with hypertension.2,9 However, even with several studies promoting the benefits of ACE inhibitors in diabetes patients and recommendations in national guidelines, use of these agents remains suboptimal.10-12
A state managed care plan database was accessed to evaluate any changes in the prescribing patterns for ACE inhibitors in the type 2 diabetic population at 1 year and 3 years after inclusion of recommendations for their use in the ADA and JNC 6 guidelines. Secondary outcomes measured included ACE inhibitor use in patients with contraindications, glycosylated hemoglobin and microalbuminuria screening, and correlations by age, sex, and eligibility to use ACE inhibitors.
International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM)
This study received approval from the managed care plan and a local institutional review board. The managed care plan is a statewide program that insures Medicaid-eligible patients, who account for approximately two thirds of the eligible patients, as well as uninsurable patients, who account for the remaining one third of the eligible patients. Medications are provided either at no cost to the patient or for $1 per prescription, depending on the patient's income. Eligible patients in the state managed care plan are equivalent to approximately one fourth of the population of Tennessee state (state total population 5.43 million in 1998 and 5.7 million in 2000). 13 The database houses an enrollment file as well as all inpatient, outpatient, home health, professional, dental, and pharmacy claims and encounters. Specific information (eg, laboratory results) is not included in the database. In 1998, approximately 1.46 million patients were enrolled in the state managed care plan. More than 45 000 of those patients had a diagnosis of diabetes, as indicated by code 250. In 2000, almost 1.54 million patients were enrolled in the managed care plan, and more than 56 000 of those patients had a diagnosis of diabetes.
Current Procedural Terminology (CPT)
Two random samples of patients were retrieved from the database, from which we gathered patient demographic information, codes, the Hierarchal Ingredient Code (HIC3) for prescriptions, and codes for medical services and procedures. The database had been cleaned to eliminate invalid data before evaluation. Patient identifiers, including patient name and social security number, were not gathered with the data to ensure compliance with the Health Insurance Portability and Accountability Act.
Patients included in the study had an code for diabetes mellitus (250) and a HIC3 code for an oral hypoglycemic agent (C4K, C4L, C4M, C4N, including biguanides, sulfonylureas, meglitinides, Ã¡-glucosidase inhibitors, and glitazones), and they were state residents. 14 An code for diabetes plus a HIC3 code for an oral hypoglycemic agent allowed only patients with type 2 diabetes to be included in the study; however, patients with type 2 diabetes who received only insulin therapy may have been excluded. Subjects consisted of 2 random samples of 500 patients each who met the inclusion criteria between January 1 and December 31, 1998 and between January 1, and December 31, 2000.
The sample size was limited to 500 patients for several reasons. First, this was a problem identification study for the managed care plan; its purpose was to evaluate compliance with evidence-based best practices. From this smaller sample, we could detect either discrepancies or agreement between clinical practice and guideline recommendations, and could determine whether there was a problem that required further evaluation by the managed care plan with a larger, more detailed study, or whether educational letters or inservice education for healthcare professionals might be in order. Use of ACE inhibitors in diabetes patients was chosen because reports in the literature indicated suboptimal use of ACE inhibitors in diabetes patients across the nation. 10-12 The sample is large enough to allow for an estimate of compliance with the guidelines. Second, the random sample allowed the 1998 and 2000 data to be compared and correlated. The data from 2000 would have been dependent on the 1998 data if all patients were included in the study. Third, the managed care plan granted permission to evaluate only a smaller sample of the population.
Patients under the age of 19 years were excluded from the study because ACE inhibitors have not been widely studied in children. The years 1998 and 2000 were chosen because 1998 was 1 year after publication of the ADA and JNC 6 guidelines recommending ACE inhibitor use in diabetes patients, and 2000 was the most current year with data in the managed care plan database. 7,8
Data including codes for hypertension (essential hypertension , hypertensive heart disease , hypertensive renal disease , hypertensive heart and renal disease ), end-stage renal failure (chronic renal failure ), renal artery stenosis (440.1), codes for pregnancy (cesarean delivery [59618-59622], vaginal delivery [59409-59410]), and the HIC3 code for ACE inhibitors (A4D) were collected and classified as dichotomous variables (0 = no code present; 1 = code present). 14 The following ACE inhibitors and combination ACE inhibitor products were included in the HIC3 code for ACE inhibitors: benazepril, benazepril with amlodipine, benazepril with hydrochlorothiazide, captopril, captopril with hydrochlorothiazide, enalapril, enalapril with diltiazem, enalapril with felodipine, enalapril with hydrochlorothiazide, fosinopril, fosinopril with hydrochlorothiazide, lisinopril, lisinopril with hydrochlorothiazide, moexipril, moexipril with hydrochlorothiazide, perindopril, quinapril, quinapril with hydrochlorothiazide, ramipril, trandolapril, and trandolapril with verapamil.
The HIC3 code did not distinguish among the specific agents, but identified any ACE inhibitor use. The presence of codes for hypertension and/or diabetes at any time during 1998 and 2000 was regarded as indicating that the patient had hypertension and/or diabetes during that time. The presence of 1 prescription claim for an ACE inhibitor during the year (1998 or 2000) was regarded as a positive indicator for ACE inhibitor use in the study. The above information was necessary to evaluate the primary end point of appropriate ACE inhibitor use.
End-stage renal failure, renal artery stenosis, and pregnancy, which are contraindications to ACE inhibitor use, were chosen to evaluate the appropriateness of this therapy. End-stage renal failure was chosen to rule out patients in whom an ACE inhibitor might have been avoided, as it was assumed that this popula- tion of type 2 diabetes patients may have decreased renal function. Angiotensin-converting enzyme inhibitors generally are not given to patients with a serum creatinine concentration greater than 3 mg/dL.2 They also should be avoided in patients with bilateral renal artery stenosis. Pregnancy is a contraindication to ACE inhibitor use as ACE inhibitors are classified in pregnancy category C. Other cautions and contraindications (hypersensitivity, hyperkalemia, cough, hypotension) were not included because ACE inhibitors could not be established as the cause through the database review. 2
Current Procedural Terminology
Additional data was gathered to evaluate secondary endpoints in the study. codes for glycosylated hemoglobin (83036) and microalbuminuria (82042-4) were collected to look at the appropriateness of nondrug therapy for diabetic patients, because all diabetic patients should have at least 1 of each of these tests yearly. Age, race, sex, continuous eligibility, dual eligibility, and county code were included to evaluate for correlations with ACE inhibitor prescribing.15 Continuous eligibility was defined by the managed care plan as no more than 1 break in coverage (lasting less than 45 days) in 1 year. Dual eligibility pertains to any patients who have another managed care plan in addition to the state managed care plan. Most patients with dual eligibility also were eligible for Medicare, which does not cover prescriptions. It was decided to include patients with dual eligibility so as not to exclude the patients with Medicare coverage. Urban and rural counties were defined by the state managed care plan.13 A 2-tailed Pearson's correlation was done to evaluate for correlations between ACE inhibitor use and the additional data that were gathered.
The patients were divided into 2 categories: (1) patients who had an code for both diabetes mellitus and hypertension and (2) patients who had an code for diabetes, but no code for hypertension. These divisions then were subdivided further by ACE inhibitor use as indicated by a National Drug Code. Contraindications to ACE inhibitor use (end-stage renal failure, renal artery stenosis, and pregnancy) were evaluated within the subdivisions. Angiotensin-converting enzyme inhibitor use then was compared between the years 1998 and 2000. The secondary data were evaluated in the population as a whole for each year by using SPSS (SPSS, Chicago, Ill) with chi-square and Pearson correlations.
Of the two 500-patient samples in 1998 and 2000, 3 patients in 1998 and 4 patients in 2000 were excluded because they were younger than age 19 years, leaving 497 and 496 patients, respectively. The 1998 and 2000 patient groups showed no statistically significant differences in demographics and were consistent with the demographics of the state's managed care plan population (Table). 13
Of the sample of patients with type 2 diabetes, 288 patients in 1998 and 317 patients in 2000 also had hypertension. In 1998, 239 patients received an ACE inhibitor. This number increased to 270 in 2000. There was a 10 percentage point increase in the proportion of patients with hypertension who were on an ACE inhibitor between 1998 and 2000 (36.2% in 1998, 46.2% in 2000). This was a statistically significant difference ( = .002) (Figure 1). The proportion of patients with a diagnosis of hypertension also increased approximately 5 percentage points between 1998 and 2000. The other groups (patients with hypertension who are not receiving an ACE inhibitor, patients without hypertension who are receiving an ACE inhibitor, and patients without hypertension who are not receiving an ACE inhibitor) showed no statistically significant differences in prescribing patterns between 1998 and 2000.
Patients receiving ACE inhibitors were evaluated for the presence of renal artery stenosis, end-stage renal failure, and pregnancy, which are contraindications to therapy. The numbers of patients with contraindications were too small to evaluate statistically. Seven patients in both 1998 and 2000 received an ACE inhibitor when they had a concurrent diagnosis of renal artery stenosis; however, unilateral versus bilateral renal artery stenosis was not differentiated by the diagnosis code. There was an increase in patients with endstage renal failure who received an ACE inhibitor: from 4 patients in 1998 to 9 in 2000. The number of pregnant patients receiving an ACE inhibitor decreased from 2 patients in 1998 to 0 in 2000. Data were gathered for the year as a whole; therefore, it is unknown whether the pregnant patients discontinued the ACE inhibitor after pregnancy was diagnosed.
Both microalbuminuria screening and glycosylated hemoglobin screening showed a statistically significant increase in 2000 compared with 1998 ( = .037 and = .002, respectively) (Figure 2). Even with this increase, however, the percentage of patients screened remained low. In 1998 and 2000, respectively, only 2.2% and 4.6% of the patients had a CPT code for microalbuminuria screening. Glycosylated hemoglobin screening increased from 44.9% in 1998 to 54.6% in 2000.
No correlations were found between ACE inhibitor prescribing and age (= 0.237 in 1998; = 0.117 in 2000), sex (= 0.034 in 1998; = -0.042 in 2000), dual eligibility (= 0.143 in 1998;= 0.000 in 2000), and continuous eligibility (= 0.077 in 1998; = -0.023 in 2000).
The percentage of patients on an ACE inhibitor in this study is similar to or lower than percentages reported in other quality-of-care studies evaluating ACE inhibitor use in diabetes patients.10-12 A study of 152 diabetes patients referred to a specific consultant nephrologist between 1991 and 1997 found that 44% of the patients had been on an ACE inhibitor before their referral. 10 Fortyeight percent of the diabetes patients with hypertension and no apparent contraindications were not prescribed an ACE inhibitor. This study evaluated both type 1 and type 2 diabetes patients. A second study evaluated a random sample of 5980 patients with type 1 or type 2 diabetes who were Medicare beneficiaries in 1993 and 1994 at ambulatory care sites in 3 states. 11 Thirty-six percent of the diabetes patients with hypertension in that study were on an ACE inhibitor. Scarsi et al reviewed the use of ACE inhibitors and renal outcomes in the Iowa Medicaid population with a random sample of 1000 patients with type 1 or type 2 diabetes and found that ACE inhibitor use increased from 38.1% of the patients in 1994 to 80.1% of the patients in 1998 ( < .001).12 Approximately 43% of the patients in the study who developed adverse renal outcomes had not been taking an ACE inhibitor. Scarsi et al conducted their study 3 years before and 1 year after the change in guidelines, using a database review of both type 1 and type 2 diabetes patients. The current study evaluated ACE inhibitor use 1 and 3 years after the change, using a database review of only type 2 diabetes patients. All the studies evaluated the quality of care diabetes patients received, and specifically looked at use of ACE inhibitors, which they found to be suboptimal in all cases.
As with any retrospective database reviews, this study has several limitations. The greatest limitation is that the database does not contain laboratory values; therefore, the patient may have a code for a glycosylated hemoglobin measurement or an code for hypertension, but the database does not contain the actual glycosylated hemoglobin or blood pressure values to assess disease control. This is also true for other factors that would have been beneficial to determine the appropriateness of ACE inhibitor use, including diagnostic and monitoring parameters such as microalbuminuria and serum creatinine concentration. The presence of a HIC3 code in the database does not indicate whether the patient has actually taken the medication and does not capture prescriptions written but not filled. A comorbidity index was not applied to assess differences in comorbidities according to demographic characteristics.
Because the data included in this evaluation are only from the years 1998 and 2000, it is unknown how long the patients may have been diagnosed with diabetes. Temporal relationships between the disease and ACE inhibitor prescribing, and between ACE inhibitor prescribing and contraindications, were not evaluated in detail because this was an initial look at the database to identify any problems. It is unknown whether a newer diagnosis might have been a confounding variable in the testing of glycosylated hemoglobin and microalbuminuria and in treatment with an ACE inhibitor. Further, it is unknown whether a patient had been prescribed an ACE inhibitor previously and failed to respond, or whether a patient actually took the prescribed ACE inhibitor, as only 1 prescription during the study year resulted in a positive indication for ACE inhibitor use. This study also does not look at other agents used in the treatment of patients with diabetes and hypertension, such as angiotensin receptor blockers or nondihydropyridine calcium channel blockers. Finally, as this was a pilot study, a small sample was collected.
Even with the limitations addressed above, use of ACE inhibitors in the study population was suboptimal, and glycosylated hemoglobin and microalbuminuria screening were done less frequently than recommended. Angiotensin-converting enzyme inhibitors have been proven to slow the progression of renal dysfunction in diabetes patients, and several guidelines recommend use of these agents in patients with both diabetes and hypertension.2-8 These agents also decrease longterm costs because of their positive renal effects.2-4
Clinical practice guidelines represent recommendations from expert panels on the current best practices in medicine that result in improved patient outcomes and decreased long-term costs. It might be assumed that publication of the guidelines results in healthcare practitioners following the new recommendations; however, the data presented in this study as well as in previous studies show that this may not be the case. This study illustrates an opportunity and a need for managed care plans to evaluate compliance of participating healthcare providers with best practices to promote improved prescribing in accord with evidence- based medicine and ultimately to decrease longterm costs.
This study showed low compliance with guidelines in a small sample. Future studies might include larger samples and evaluate the population in greater detail, including evaluation of temporal relationships between diabetes and hypertension diagnoses and the first ACE inhibitor prescription, prescription filling patterns, and other agents used in hypertensive diabetes patients when ACE inhibitors are not tolerated. Use of additional preventive measures such as yearly eye exams and nutrition counseling also might be evaluated. Furthermore, a study including examination of patient charts to retrieve laboratory values for renal function and glycosylated hemoglobin would be beneficial.
Although this study does identify increased use of ACE inhibitors in the hypertensive, diabetic population of the state managed care plan, the proportion of patients receiving an ACE inhibitor is still less than 50%. Microalbuminuria screening, recommended yearly for diabetes patients, and glycosylated hemoglobin screening, recommended approximately every 3 months for diabetes patients by the ADA, also are well below goal.2 Ideally, the patients with hypertension and no contraindications to therapy should be receiving a renal protective agent.2,10-12 The ADA included angiotensin receptor blocking agents as approved therapy in this population of patients in 2001, after this study was completed.
It is important to note that tight blood pressure and glucose control have been proven to be most effective in preventing complications, irrespective of the therapy used; therefore, although it is beneficial for diabetes patients to be on an ACE inhibitor, it also is important that blood pressure control is achieved.16
This study indicates that diabetes patients and healthcare practitioners need to improve compliance with the ADA and JNC 6 recommendations for appropriate prescribing of ACE inhibitors and screening for elevated glycosylated hemoglobin and microalbuminuria.
From the University of Tennessee Health Science Center, Memphis, Tenn. Address correspondence to: Erin M. Timpe, PharmD, University of Tennessee, Drug Information Center, 875 Monroe Ave, Suite 121, Memphis, TN 38163. E-mail: firstname.lastname@example.org.
USRDS 2002 Annual Data Report: Atlas of End-Stage Renal Disease in the United States.
1. United States Renal Data System (USRDS). Bethesda, Md: National Institutes of Health, National Institute of Diabetes & Digestive & Kidney Diseases, Division of Kidney, Urologic, & Hematologic Diseases; 2002. Available at: http://www.usrds.org/adr_2002.htm. Accessed August 27, 2003.
N Engl J Med.
2. Lewis EJ, Hunsicker LG, Bain RP, Rohde RD, for The Collaborative Study Group. The effect of angiotensin-converting enzyme inhibition on diabetic nephropathy. 1993; 329:1456-1462.
3. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. 2002;25(suppl):S33-49, S85-S89.
Arch Intern Med.
4. Ravid M, Lang R, Rachmani R, Lishner M. Long-term renoprotective effect of Angiotensin-converting enzyme inhibition in noninsulin- dependent diabetes mellitus. A 7-year follow-up study. 1996;156(3):239-240.
5. EUCLID Study Group. Randomised placebo-controlled trial of lisinopril in normotensive patients with insulin-dependent diabetes and normoalbuminuria or microalbuminuria. 1997;349: 1787-1792.
6. ATLANTIS Study Group. Low-dose ramipril reduces microalbuminuria in type 1 diabetic patients without hypertension. 2000;23:1823-1829.
7. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. 1996;19(suppl): S103-S106.
Arch Intern Med.
8. Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure. The sixth report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC VI). 1997;157: 2413-2448.
Am J Kidney Dis.
9. Bakris GL, Williams M, Dworkin L, et al. Preserving renal function in adults with hypertension and diabetes: a consensus approach. 2000;36:646-661.
Q J Med.
10. Dunn EJ, Burton CJ, Feest TG. The care of patients with diabetic nephropathy: audit, feedback, and improvement. 1999;92:443-449.
J Am Geriatr Soc.
11. Kell SH, Drass J, Bausell RB, Thomas KA, Osborn MA, Gohdes D. Measures of disease control in Medicare beneficiaries with diabetes mellitus. 1999;47(4):417-422.
12. Scarsi KK, Bjornson DC. The use of ACE inhibitors as renoprotective agents in Medicaid patients with diabetes. 2000;34:1002-1006.
13. HIT: Health Information Tennessee Web site. Available at: http://hitspot.utk.edu. Accessed August 27, 2003.
14. International Classification of Diseases, 9th Revision, Clinical Modification. 6th ed. Los Angeles, Calif: Practice Management Information Corporation; 2000:38-39, 181, 280, 293, 372-383.
Current Procedural Terminology CPT 2001.
15. American Medical Association. Standard edition. Chicago, Ill: AMA Press; 2000:236, 241-242.
16. UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes. UKPDS 38. 1998;317:703-713.