Loop diuretic therapy with or without heart failure: impact on prognosis

Abstract

Background and aims: Many patients are prescribed loop diuretics without a diagnostic record of heart failure. Little is known about their characteristics and prognosis.

Methods: Glasgow regional health records (2009-16) were obtained for adults with cardiovascular disease or taking loop diuretics. Outcomes were investigated using Cox models with hazard ratios adjusted for age, sex, socioeconomic deprivation, and comorbid disease (adjHR).

Results: Of 198 898 patients (median age 65 years; 55% women), 161 935 (81%) neither took loop diuretics nor had a diagnostic record of heart failure (reference group), 23 963 (12%) were taking loop diuretics but had no heart failure recorded, 7844 (4%) had heart failure recorded and took loop diuretics, and 5156 (3%) had heart failure recorded but were not receiving loop diuretics. Compared to the reference group, five-year mortality was only slightly higher for heart failure in the absence of loop diuretics [22%; adjHR 1.2 (95% CI 1.1-1.3)], substantially higher for those taking loop diuretics with no record of heart failure [40%; adjHR 1.8 (95% CI 1.7-1.8)], and highest for heart failure treated with loop diuretics [52%; adjHR 2.2 (95% CI 2.0-2.2)].

Conclusions: For patients with cardiovascular disease, many are prescribed loop diuretics without a recorded diagnosis of heart failure. Mortality is more strongly associated with loop diuretic use than with a record of heart failure. The diagnosis of heart failure may be often missed, or loop diuretic use is associated with other conditions with a prognosis similar to heart failure, or inappropriate loop diuretic use increases mortality; all might be true.

Keywords: Diuretics; Ejection fraction; Epidemiology; Heart failure; Left atrium; Mortality.

Structured Graphical Abstract

Figure 1

Greater Glasgow & Clyde population classified by sex, age group, repeat prescription of loop diuretics, and a diagnosis of heart failure, based on the mid-year population estimate for 2012. GG&C, Greater Glasgow & Clyde; LD, loop diuretics

Figure 2

Transition diagrams show how many patients started in each of the four groups (left most boxes) and how many experienced subsequent events between 1 January 2012 and 31 December 2016. Percentages in the boxes are calculated with the baseline group size as the denominator, while transitions are calculated based on those eligible for each transition. Reasons for transitions include diagnosis of heart failure, initiation of loop diuretics, or death during follow-up

Figure 3

(A) Estimation of the cumulative initiation of loop diuretics for patients not already taking loop diuretics at baseline and the competing risk of all-cause mortality. (B) Estimation of the cumulative incidence of a diagnosis of heart failure for patients who did not have heart failure at baseline and the competing risk of all-cause mortality. LD, loop diuretics; HF, heart failure; yrs, years

Figure 4

Five-year event rates by patient-year at risk. (A) Hospital admission rate classified by the primary admission reason and baseline group determined by the presence or absence of a repeat prescription loop diuretics and a diagnosis of heart failure. Rates were adjusted by patient-year at risk for those eligible (i.e. not already in hospital or dead) to be admitted. The total number of admissions per group is reported above each column. Supplementary data online, Figures S12  and  S13, show similar data with the loop diuretic/HF group as a time-dependent covariate. (B) All-cause mortality classified by the underlying cause of death and the baseline group determined by the presence or absence of a repeat prescription of loop diuretic and a diagnosis of heart failure adjusted for patient-year at risk where the patient was under follow-up. Patients were censored at the last medical contact (blood test, prescription, etc.) date to account for patients who moved out of the region. The total number of deaths per group is reported above each column. Supplementary data online, Figure S14, shows similar data with loop diuretic/HF group as a time-dependent covariate. n, total number of deaths; N, total number of admissions

Figure 5

Five-year survival analysis from 1 January 2012 to end of follow-up classified by baseline group according to use of loop diuretics and diagnosis of heart failure. (A) Kaplan–Meier curves to compare survival patterns by baseline group. (B) Forest plot of hazard ratios with 95% confidence intervals for all-cause mortality by baseline group. The model was adjusted for age, sex, Scottish Index of Multiple Deprivation, a history of hypertension, coronary artery disease, peripheral arterial disease, diabetes mellitus, valve disease, atrial fibrillation or flutter, stroke, cancer, dementia, and the closest eGFR in the prior 2 years