Chronic heart failure (CHF) is a prevalent condition in the veteran population and is associated with a high mortality, a substantial expenditure of health care resources and a significant decrement in health-related quality of life (HRQoL). Compliance with medication prescribing guidelines for patients with CHF may be sub-optimal, and outcomes associated with use of different classes of medications may be influenced by patient characteristics.
The objectives are to: 1) determine medications used to treat CHF in veterans, including medication doses and patterns of multiple medication use in the same patient, and examine provider and patient factors associated with the use of such medications, especially as it relates to treatment guidelines; and 2) determine the relationship of medication patterns (including dosing and patient compliance) to HRQoL, mortality, and CHF-related hospitalizations.
The research design involves retrospective data analysis. The principal source of data is existing VA databases, and the principal type of analysis is outcomes assessment. Specific outcomes analyzed were hospitalizations, mortality and HRQoL. The principal population studied included veterans with a diagnosis (ICD-9-CM code) of CHF, outpatient care iat any six sites in VISN and decreased left ventricular (LV) function.
1) Patient distribution of LV function significantly differed by site, indicating that the profiling of process and outcome measures among sites should incorporate LV function for appropriate risk-adjustment.
2) Baseline (1999) use of angiotensin-converting enzyme inhibitors (ACEI) or alternatives was high (88.6%) and remained stable over several years. Use of beta-adrenergic blockers (BB) was 57.9% in 1999 and increased to 73.6% over the next several years, among survivors.
3) At baseline, use of ACEI (or alternative), target doses of ACEI, and BB were more likely in younger patients (<75 years old) and those with visits to a cardiologist (p<.05). The addition of BB over time was also associated with these characteristics (p<.05).
4) A risk-adjustment model for HRQoL, based on demographics and comorbidities, yielded R-squares of .22 and .21 for the physical and mental component summaries (PCS and MCS) of the Veterans Rand 36-Item Health Survey . Diagnostic categories with the most negative beta coefficients for PCS and MCS were consistent with the types of conditions generally associated with significant disability, lending support to the validity of this form of predictive model. An important finding, made possible through use of this model, was that the actual relationship between the degree of LV systolic dysfunction and HRQoL may be obscured by confounding effects of non-CHF patient characteristics.
5) Most medications were associated with a tendency towards lower PCS scores (indicating worse HRQoL). An exception was “statins”, in which such tendency was towards higher PCS. However, statistical significance was present for only several categories of medications and varied by the statistical model utilized. Use of digoxin, calcium channel blockers (other than amlodipine or felodipine),“target” doses of ACEI and high-dose loop diuretics were all associated with statistically lower PCS scores in at least one model. Because use of several of these drugs or dose intensities may be related to more severe existing symptoms, the observed relationship to lower PCS may be due, in part, to “confounding by indication”.
6) Consistent with prior randomized studies, use of ACEI and BB was associated with lower risk of both death and the combined death/hospitalization outcome at 24-month follow-up (p<.05). Statins were also noted to be associated with a lower risk of such adverse outcomes (p<.05), although a causal relationship cannot be inferred at this time.
Our study provides important insight into the phramacologic management of patients with CHF cared for in the Department of Veterans Affairs and how such management is asscoiated with clinical outcomes. In addition, the study has indentified issues that will be important to address in ongoing efforts to improve process of care and outcomes in patients with this condition.
External Links for this Project
- Rothendler J, Reisman J, Kazis L, Berlowitz D, Fincke G, Gaehde S, Glickman M. Prescribing of Beta-blockers in VA Patients with Chronic Heart Failure. Paper presented at: VA HSR&D National Meeting; 2004 Mar 17; Washington, DC. [view]