Medication Adherence and Decreased Kidney Function as Predictors of Hospitalization for Chronic Heart Failure in Elderly Patients

Abstract

Hospitalizations due to worsening of the clinical course of ambulatory care sensitive conditions (ACSC) are the significant and growing burden on health systems. Chronic heart failure (CHF) is one of the five leaders among chronic ACSC both in the frequency of hospitalizations and in their cost. Therefore, it is important to find predictors of the clinical course worsening which are easy to use in ambulatory settings.
 The aim of the research: to assess the effect of medication adherence and decreased renal function on the probability of hospitalization for CHF in the elderly during 6 months.
 Materials and methods. The prospective research included 111 ambulatory elderly patients aged from 60 to 74 years with CHF, stage II, arterial hypertension, stage II and chronic kidney disease, stage II-IIIA. During the first ambulatory visit a general clinical examination and ambulatory card analysis were performed. Creatinine level was checked with calculation of the glomerular filtration rate according to the formula of the Chronic Kidney Disease Epidemiology Collaboration (GFR EPI) to assess renal function. In addition, a 6-minute walk test (6MWT) was performed and a quality of life assessment was conducted according to Minnesota Living with Heart Failure Questionnaire. Adherence assessment on a Morisky Green Medication Adherence Scale as well as a collection of the information about hospitalization during the previous 6 months was done during the last visit. The variable "interaction" between decreased renal function and poor medication adherence was introduced separately into the logistic analysis. A logistic regression analysis was used to assess the influence of factors on the probability of hospitalization for CHF.
 Results. During the observation 21 patients were hospitalized for CHF (21/93, 23%). The analysis of the distribution of adherence level and GFR EPI ≤ 59,9 mL/min/1,73 m2 between groups of hospitalized and non-hospitalized patients revealed a predominance of non-adherent patients, as well as those with decreased renal function, in the group of hospitalized ones. Clinical characteristics and laboratory parameters, by which the differences were found between hospitalized and non-hospitalized patients, as well as categorical variables received by ROC-analysis - quality of life, GFR EPIcat, systolic blood pressure, 6MWTcat, and variable "interaction", were included in the univariate logistic analysis. According to univariate logistic analysis, the hospitalization probability increased with GFR EPI ≤ 59,9 mL/min/1,73 m2 (OR-6,73, CI 2,34-19,35) and decreased in the presence of medication adherence (OR-0,08, CI 0,02-0,37). In multivariate analysis (adjustment for lung crackles and functional class according to NYHA) or were 4,95 (CI 1,18-20,78) for GFR EPI ≤ 59,9 mL/min/1,73 m2 and 0,11 (CI 0,02-0,81) for medication adherence. During assessing the "interaction" variable it was found that in non-adherent patients with GFR EPI≤59,9 ml/min/1,73m2 the chance of inpatient treatment for CHF significantly increased by 6.82 times on average, with a maximum increase by 20 times.
 Conclusions. Medication adherence and decreased renal function in our research were shown to be significant predictors of hospitalization for CHF, which are simple to measure and assess and can be used in an outpatient setting. By increasing adherence level, especially in elderly patients with impaired renal function, it is possible to reduce the chance of inpatient treatment caused by CHF by 9.1 times.