Use of Valsalva's Maneuver to Detect Early Recurrence of Congestive Heart Failure in a Randomized Trial of Furosemide Withdrawal in Older Patients

Editors note: Estring was supplied by Pharmacia & Upjohn. Dr. Ouslander serves as a consultant to Pharmacia & Upjohn. To the Editor: Previous studies of diuretic withdrawal in older patients have cautioned against the risks of acute congestive heart failure (CHF).1–2 However, the clinical detection of CHF is difficult, particularly in older patients. In earlier studies, it was noted that measurement of the blood pressure (BP) response to a Valsalva's maneuver was useful for estimating left ventricular filling pressures and determining the presence of CHF.3–7 Pulmonary congestion causes a “square-wave” appearance of BP response, which is explained by an increased central blood volume acting as a reservoir that maintains left ventricular filling despite a decrease in venous return during strain (Figure 1).4 As part of a randomized, controlled, double-blind trial of furosemide withdrawal in older heart failure patients, we explored the use of repeated noninvasive measurements of BP response to Valsalva's maneuver in the early detection of recurrent CHF. . Finapres® measurement of arterial blood pressure response to Valsalva's maneuver in a heart failure patient before furosemide withdrawal (left panel), after developing recurrent congestive heart failure (middle panel), and after restart of furosemide (right panel). Valsalva blood pressure ratios are calculated by dividing the mean of the systolic blood pressure values, marked #, by the mean of the systolic blood pressure values, marked*. Thirty-two heart failure patients (mean age 75.1 ± 4.0 years) were randomly assigned in a 2:1 ratio to a withdrawal group and a continuation group. Six follow-up visits took place during 3 months after furosemide withdrawal. Heart failure symptom score8, body weight, and BP were determined on every occasion. Predefined criteria for unblinding and restart or dose augmentation of furosemide were determined by heart failure symptom scores with verified pulmonary congestion on chest X-ray, and/or repeated systolic or diastolic BP measurements >180 mm Hg or > 100 mm Hg. On every occasion, patients also performed a standardized Valsalva's maneuver. This was done in a semi-supine position (30° tilt), by blowing forcefully for 10 to 12 seconds into a blow-pipe connected to a pressure transducer. Patients were assisted in the maintenance of a straining pressure of 35 to 40 mm Hg by an on-line pressure monitor display. A tiny air leak in the mouthpiece ensured that the airway pressure was produced from the thoracic cavity. The Valsalva BP response was measured continuously using Finapres (FINger Arterial PRESsure, TNO-BMI, Academic Medical Centre, Amsterdam, The Netherlands). BP response was sampled at 100 Hz and stored digitally. Valsalva BP ratios were calculated by dividing the mean of the lowest three consecutive systolic BP values during strain by the mean of the first three values after the start of the Valsalva procedure (Figure 1).5,9 All Valsalva measurements were performed in triplicate separated by 2-minute resting periods. During the 3-month follow-up period, 16 patients were successfully withdrawn from furosemide. Three patients (2 from the withdrawal group, 1 from the continuation group) experienced an episode of recurrent CHF. Three patients from the withdrawal group restarted on diuretics for ankle edema (n = 2) and hypertension (n = 1). Table 1 compares the maximum increases in clinical parameters for patients successfully withdrawn (n = 16), patients continuing on furosemide (n = 10), and patients experiencing recurrent CHF (n = 3). For the latter three patients, the maximum increases in Valsalva BP ratio after withdrawal were significantly higher (0.20; 95% Cl, 0.13 to 0.27). Figure 2 shows the changes in Valsalva BP ratios for these three patients during follow-up compared with the upper limit of the 95% Cl for maximum increase in BP ratio for patients successfully withdrawn (dotted line). After treatment of CHF, Valsalva BP ratios in these three patients decreased, on average, by −0.17 (95% Cl, −0.07 to −0.27), and returned to baseline values (average change from baseline = 0.01). Eight patients demonstrated increased dyspnea and ankle edema. However, chest X-rays did not show pulmonary congestion, and Valsalva BP ratios did not change from baseline (maximum 0.01, 95% Cl, −0.02 to 0.05). Diuretics were not restarted, and these patients successfully completed the trial. Valsalva BP ratios did also not differ significantly from baseline (−0.01, 0.02, and 0.04, respectively) in the three patients restarted on diuretics for reasons other than CHF. . Changes in Valsalva blood pressure ratio from baseline following furosemide withdrawal in three patients experiencing recurrent CHF. Arrows indicate moment of recurrence of CHF and restart or augmentation of furosemide. The dotted line indicates the upper limit of the 95% Cl for maximum change in Valsalva blood pressure ratio for patients successfully withdrawn from furosemide. In the present study, Valsalva phase 2 systolic BP responses demonstrated obvious changes in patients experiencing recurrent CHF, whereas response remained unchanged in all other patients, including patients restarting diuretics without CHF and patients suspected of CHF on clinical grounds but not needing diuretics in the end. Finapres measurement of BP response to Valsalva's maneuver is an inexpensive, non-invasive, and safe bedside cardiac function test that is simple to perform, even for older patients, and can be repeated frequently. Our results suggest that assessment of Valsalva BP response is a useful additional clinical tool for the early detection of CHF in older patients following diuretic withdrawal, differentiating CHF from rebound ankle edema or physiological weight increase.