A retrospective, observational cohort analysis of a nationwide database to compare heart failure prescriptions and related health care utilization before and after publication of updated treatment guidelines in the United States.

BACKGROUND: The current American College of Cardiology/American Heart Association (ACC/AHA) clinical guidelines for heart failure (HF), published September 20, 2005, provide a summary of the best evidence for treatment, but these recommendations are not always reflected in clinical practice.

OBJECTIVES: The aims of this study were to compare 6-month prescribing habits in the United States before and after the publication of updated clinical guidelines for the evaluation and management of HF and the impact of these prescribing habits on health care resource use.

METHODS: This retrospective, observational cohort analysis used the Humana nationwide health insurance administrative claims database that includes -3.5 million covered members from all 50 states and Puerto Rico who are enrolled in a health maintenance organization, a preferred provider organization, or a Medicare plan. The data included demographics (age, sex, type of insurance, and geographic location), medical information with up to 9 diagnostic codes per encounter, codes for procedures and medical equipment, laboratory tests, and pharmacy-dispensed medications. HF medication prescriptions and health care utilization were evaluated for 2 cohorts: those identified from claims before guideline publication (January 1, 2005-June 30, 2005) and those identified from claims after publication (October 1, 2006-March 31, 2007). Patients were eligible if they were aged ≥45 years, had 12 months of continuous enrollment (6 months before and 6 months after the index date, defined as the date of diagnosis or hospitalization for HF), and had ≥1 claim for HF. The primary outcome was the proportion of patients who received prescriptions for HF medications individually or in combination. Secondary outcomes were adherence to medication, all-cause and HF-specific hospitalizations, and emergency department and outpatient physician visits.

RESULTS: The mean (SD) age in the before-publication cohort (n = 29,784) was 75 (11) years; in the after-publication cohort (n = 33,598), it was 74 (11) years (P < 0.001). Half of all patients in each cohort were female (50% [n = 14,796 and n = 16,803, respectively]); 9% (n = 2539) of the before-publication cohort and 7% (n = 2283) of the after-publication cohort were classified as having moderate to severe HF based on the baseline number of hospitalizations (P < 0.001). Fewer patients in the before-publication cohort received angiotensin-converting enzyme inhibitors (43% [12,811/29,784] vs 44% [14,776/33,598]; P = 0.01), β-blockers (37% [10,901/29,784] vs 41% [13,639/33,598]; P < 0.01), angiotensin receptor blockers (10% [3008/29,784] vs 13% [4378/33,598]; P < 0.01), or hydralazine (3% [865/29,784] vs 4% [1378/33,598]; P < 0.01). Among those with moderate to severe HF, there were no significant differences between groups in the use of combination therapy (ie, β-blockers with any combination of isosorbide dinitrate, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or hydralazine; all combinations, P = NS between cohorts), except for less use of a β-blocker with either an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker among the before-publication cohort (23% [595/2539] vs38% [875/2283]; P = 0.005). Therewere significantly fewer all-cause and HF-related outpatient visits, all-cause and HF-related hospitalizations, and all-cause and HF-related emergency department visits in the after-publication cohort (all, P < 0.001); however, the absolute differences in the proportions of patients in each cohort who required such services were relatively small.

CONCLUSIONS: Based on this analysis of real-world prescribing patterns, only small differences in prescribing practices were found before and after the 2005 publication of the ACC/AHA guidelines for HF treatment. Health care utilization was slightly, but significantly, reduced after publication of the guidelines.