Can treatment with angiotensin-converting enzyme inhibitors in elderly patients with moderate to severe chronic heart failure be improved by a nurse-monitored structured care program? A randomized controlled trial☆☆☆★

Patients 

The inclusion criteria were: age 65 years or more; diagnosed CHF with Boston criteria score 8 or more13 at screening, followed by diagnostic work-up; and New York Heart Association (NYHA) classification III or IV at last admission. Exclusion criteria were: a large myocardial infarction (new Q wave or serum Creatine kinase-MB > 100 μkat.1−1) during the preceding 8 weeks; need of specific treatment at the department of cardiology; serum creatinine more than 300 umol/L; need of permanent nursing home care; serious or life-threatening other disease; and communication problems. After initial screening, patients with definite contraindications to ACE inhibitors were excluded (ie, significant aortic stenosis and renal artery stenosis).

The aim was to recruit 160 patients to the study. A study nurse regularly screened 5 medical wards for patients with CHF. This was done daily, except during weekends. Preliminarily eligible patients then were examined by 1 of the physicians participating in the study. The patients who met all criteria and gave informed consent then were recruited to the study.

Examinations 

The examinations that included all patients at baseline and after 6 months of follow-up included a questionnaire covering history of cardiovascular or other significant diseases, days in hospital, medication and other treatment, present symptoms according to the NYHA classification, a physical examination, and results from examinations, such as electrocardiography, chest radiography, echocardiography, radionuclide, or angiographic examinations. Serum electrolytes and creatinine were measured with the routine methods of the hospital. Creatinine clearence was calculated with a formula that was based on gender, age, body weight, and serum creatinine level.14 As with the CONSENSUS trial, examination of the ejection fraction was not mandatory.1 Ischemic heart disease as the cause of CHF was on the basis of previously documented myocardial infarction or hypokinetic wall motion at an echocardiographic examination.

Randomization and treatment groups 

A physician generated and a nurse executed the random assignment by opening a consecutively numbered, sealed envelope containing group assignments. The structured care treatment was based on a nurse-directed outpatient clinic under the responsibility of the study physicians, governed by a detailed program that was outlined in a manual.12 The program consisted of education about CHF and its treatment, advice on weight control, and efforts to increase concordance between caregivers (the nurses and the physicians) and the patients concerning prescribed treatment. A more specific aim was to institute treatment with enalapril in eligible patients with the aim of reaching a target dose of 10 mg twice daily. The starting dose was 2.5 mg once daily with the following uptitration steps: 2.5 mg twice daily (after 3 days), 5.0 mg twice daily (7 days later), 10 mg in the morning and 5 mg in the evening (2 weeks later), and 10 mg twice daily (2 weeks later). Follow-up visits with the study nurse were planned after each dose step and included measurements of blood pressure and serum concentrations of sodium, potassium, and creatinine. However, treatment could be individualized if deemed necessary by the responsible physician, who saw each patient at least every 3 months or more often if indicated. In the standard care group patients were managed in the primary care system in accordance with current clinical practice.

Follow-up 

A reexamination was scheduled for all patients after 6 months. This included the same examinations as at the baseline examination.

For patients who did not undergo the follow-up examination after 6 months (see subsequent), reasons for lack of reexamination and all other relevant information were obtained from the medical records or other sources. In patients who died during the follow-up period, information on drug treatment at least 1 week before death was recorded. Because the patients were treated with several types of ACE inhibitors, dose levels were recalculated as percentage of optimal daily doses, which were defined as follows: enalapril, 20 mg; captopril, 100 mg; lisinopril, 20 mg; ramipril, 10 mg; and cilazapril, 2.5 mg.

Statistical analysis 

The sample size calculation, which showed that 160 patients had to be randomized into the study, was based on the following assumptions: α = 0.05 and β = 0.20; ACE inhibitors would be used by 75% of the patients in the structured care group and by 25% less patients in the standard care group; and 30% of the patients would be lost to follow-up. The number of patients treated with ACE inhibitors who achieved doses in relationship to target doses at study closure and reasons for withdrawal and interrupted uptitrations were predefined endpoints. The analyses were performed according to the intention-to-treat principle.

Results are given as mean (± standard deviation [SD]) or number (percent). The 2 study groups were compared with Student t test (2-tailed) for continuous variables and with the χ2 test for discrete variables, and 95% CIs were calculated. Correlations were determined with Pearson correlation coefficient after log transformation of skewed variables. A multiple regression analysis was performed to examine the baseline characteristics, which were associated with the final ACE inhibitor dose. A P value of less than .05 (2-tailed) was regarded as indicating a statistically significant difference.

Treatment with ACE inhibitors 

As presented in Table II, about 40% of the patients were on treatment with ACE inhibitors at entry, with no difference between the groups.

Table II. Description of how ACE inhibitors were used in structured care and standard care groups, respectively

		Structured care (n = 70)	Standard care (n = 75)	Mean percentage difference (95% CI)	P value
	ACE inhibitor treatment at entry	28 (40%)	31 (41%)	−1 (−17 to 15)	.87
	Any treatment with ACE inhibitor	67 (96%)	63 (84%)	12 (2 to 21)	.021
	ACE inhibitor treatment at study closure	49 (70%)	47 (64%)	6 (−8 to 23)	.35
	Mean (SD) ACE inhibitor dose at study closure, expressed as percentage of target dose*	44 (43)	32 (35)	12 (−2 to 26)	.10
	Achieved target ACE inhibitor dose	18 (26%)	8 (11%)	10 (3 to 27)	.018
	*Target dose was defined as enalapril, 20 mg; captopril, 100 mg; lisinopril, 20 mg; ramipril, 10 mg; cilazapril, 2.5 mg.

Among the 130 patients who were given the opportunity to try an ACE inhibitor, this treatment had to be withdrawn in 34 patients (26%) because of causes presented in Table III. Of these 130 candidates for ACE inhibitor therapy, 70 patients (54%) were not uptitrated to the optimal dose, as specified in Table III. In 25 of 39 patients (64%) with suboptimal ACE inhibitor dose in the standard care group, no documented cause of interrupted uptitration was found compared with 3 of 31 patients (10%) in the structured care group (P < .001).

Table III. Causes of withdrawal or interrupted uptitration of ACE inhibitor treatment in all patients who tried such treatment (n = 130)

	Cause	Withdrawal of ACEI	Interrupted uptitration of ACEI
	Increased serum creatinine level*	9 (6.9%)	16 (12.3%)
	Hypotension*	7 (5.4%)	14 (10.8%)
	Cough*	5 (3.8%)	3 (2.3%)
	Patient decision*	3 (2.3%)	1 (0.8%)
	Other*	3 (2.3%)	7 (5.4%)
	Not stated*†	3 (2.3%)	28 (21.5%)†
	Confusion*	2 (1.5%)	0
	Dizziness*	1 (0.8%)	1 (0.8%)
	Hyperkalemia*	1 (0.8%)	0
	Sum of all*	34 (26.1%)	70 (53.8%)
	*Percentage of all patients who had tried ACE inhibitor treatment (n = 130). Each responsible physician decided what level for increased serum creatinine should be. †The cases with no given cause of interrupted uptitration of ACE inhibitor occurred more often in standard care group (53%) than in structured care group (5%), P < .01).

ACEI, ACE inhibitor.

The final ACE inhibitor dose correlated with the following baseline characteristics: calculated creatinine clearence (r = 0.34; P = .0026), log weight (r = 0.29; P = .0007), log age (r = −0.18; P = .045), and log blood hemoglobin (r = 0.18; P = .041) but not with systolic blood pressure or randomization to structured outpatient care (data not shown). Only calculated creatinine clearence contributed independently to the variation in final ACE inhibitor dose (β coefficient, 0.6; P = .019)