Remember me
ORIGINAL ARTICLE
Year : 2022 | Volume
: 8
| Issue : 3 | Page : 161-167
Prognostic value of resting heart rate and heart rate recovery in acute decompensated heart failure: A prospective cohort study
Mohammad Azharuddin Malik1, Raghav Bansal2, Sandeep Seth2, Neeraj Parakh2, Ambuj Roy2, Vinay Kumar Bahl2
1 Department of Cardiology, Aligarh Muslim University, Aligarh, Uttar Pradesh, India
2 Department of Cardiology, All India Institute of Medical Sciences, New Delhi, India
Correspondence Address:
Sandeep Seth
Department of Cardiology, All India Institute of Medical Sciences, New Delhi
India
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/jpcs.jpcs_68_22
Background: Acute decompensated heart failure (ADHF) remains a problem of public health concern. Better prognostication is needed to predict outcomes in these patients. Resting heart rate (HR) and HR recovery (HRR) have been proposed as prognostic markers for future events. Materials and Methods: Twenty-five patients admitted for ADHF were enrolled. Baseline demographic data and routine investigations were noted for all. After medical stabilization, before discharge, a 6-min walk test (6-MWT) was performed for all patients. Resting HR at the start, maximum HR achieved, chronotropic reserve, and hearing rate recovery was observed 1 min after completing the exercise. All patients were followed for the appearance of a primary composite outcome consisting of death, heart transplant, or hospitalization for HF. Results: Primary composite outcome occurred in 6 (24%) patients, including 3 (12%) deaths and 3 (12%) HF hospitalizations. The patients who had the primary outcome had a trend toward a shorter distance of walking of 6 min (154.33 ± 51.84 vs. 210.53 ± 78.62, P = 0.16). Resting HR was significantly higher among patients with primary outcome (113.00 ± 17.74 vs. 89.58 ± 16.25, P ≤ 0.01). HRR and chronotropic reserve were significantly lower in patients with primary outcome (7.33 ± 1.75 vs. 17.42 ± 9.47, P < 0.01 and 15.00 ± 11.93 vs. 34.94 ± 19.81, P = 0.02; respectively). A resting HR of 109/min, HRR of 9/min, and chronotropic reserve of 20/min had sensitivity/specificity for predicting the primary outcome of 83.33%/89.47%, 84.21%/83.33%, and 84.21% a/as 83.33%, respectively. Conclusions: Elevated resting HR, decreased HR reserve, and decreased chronotropic reserve are associated with hospital readmissions and decreased event-free survival in patients with ADHF. Resting HR is especially helpful in this regard given the simplicity and ease of its assessment.
Keywords: 6-min walk test, acute decompensated heart failure, chronotropic reserve, heart rate recovery, prognostic marker, resting heart rate
Heart failure (HF) can be defined as an abnormality in cardiac structure or function that causes the heart to fail to deliver oxygen at a rate that is proportional to the requirements of metabolizing tissues, despite normal filling pressures (or only at the expense of increased filling pressures).[1] HF is a major public health problem with an estimated lifetime prevalence of about 20%.[2] Hospitalization for HF is a powerful independent risk factor for death.[3] Mortality due to HF remains high, with hospital mortality in Europe ranging from 6% to 7% and in the United States from 3% to 4%.[3] Poor outcomes have also been shown after discharge, with 60–90-day mortality rates of 5% to 15% and readmission rates of 30%.[4] Contemporary Indian data among HF patients reported a 90-day mortality of 14.2% and hospital readmission rate of 8.4%.[5] Another registry data for acute decompensated HF (ADHF) from India showed hospital mortality to be 30.8%. With 6-month combined mortality/readmission rates of a staggering 39.5%.[6] Apart from being a prognostic marker, hospital readmissions also add to the overall cost of care of a heart failure patient.
For the prediction of in-hospital clinical events in patients admitted with HF, a risk prediction model based on the Acute Decompensated Heart Failure National registry data is available and widely used.[7] There is a growing realization of the role of heart rate (HR) in the prediction of clinical events in patients with HF.[8],[9],[10],[11],[12],[13] HR is also being studied to assess HF development in patients at risk of HF.[14],[15] Apart from resting HR, the chronotropic response (CR) to exercise and the recovery of HR after exercise are also helpful in the assessment of prognosis.[16],[17] Impaired CR to exercise is another prognostic marker useful in the prediction of clinical events.[18] These chronotropic parameters can easily be obtained at the bedside and could have additive prognostic value in addition to the already available parameters. No clinical risk score has utilized these variables to enhance the risk prediction value of the models.
The current pilot study is designed to evaluate the value of resting HR, HR reserve in predicting future clinical events in patients admitted with ADHF after discharge from a Tertiary Care Center in India.
Materials and MethodsThis was a prospective observational cohort study that recruited 25 consecutive patients admitted with ADHF at the time of discharge. All patients gave their informed consent in writing and the study was approved by the institutional ethics committee. Patients over 18 years of age with both HF with reduced ejection fraction and preserved ejection fraction were included in the study. Exclusion criteria included a history of acute coronary syndrome within the past 3 months, rheumatic heart disease, any other reversible etiologies of HF, significant liver or renal disease, patients dependent on the permanent pacemaker, and inability to perform a walk test were enrolled at discharge. Relevant history, physical examination, routine hematology, biochemistry, chest radiograph, echocardiography electrocardiography, and 6-min walk test (6-MWT) were performed in all patients.
Six-minute walk test
6-MWT was conducted according to the standard recommendation with an additional postexercise resting period till the return of HR to resting levels.[19] The following parameters were recorded: (i) Baseline HR/systolic blood pressure (SBP), postexercise HR and SBP; (ii) Peak HR; (iii) 6-min walk distance; (iv) HR recovery 1 (HRR1): Fall from the peak HR to HR 1 min after termination of exercise; (v) CR: HRmax – HRmin, % of HR reserve ((220-age)-baseline HR) achieved; and (vi) Pre- and post-exercise symptoms (dyspnea and fatigue) on Borg scale.
Postdischarge follow-up
After discharge, patients were followed for at least 1 month with scheduled outpatient visits and then, as and when required, up to 6 months, as dictated by the patient's symptoms and the overall course of the disease according to the discretion of the treating physician. The rest of the follow-up was telephonic or by outpatient visits. During follow-up visits, patients' symptoms Borg scale and New York Heart Association (NYHA) class were recorded, as well as data of and re-hospitalization recorded. In case of patient death during follow-up, an attempt was made to determine the cause of death by retrieving the records of the hospital where the patient died. Significant worsening of symptoms leading to rehospitalization or death was considered a clinical event and incorporated into a single composite primary outcome.
Statistical analysis
The present study was planned as a pilot study; therefore, a sample size of the convenience of 25 patients was taken. All data were compiled in IBM SPSS 26 software (IBM Corp, Armonk, NY), and it was subjected to statistical analysis with the help of the same. All categorical variables were compared using the Chi-square or Fisher's test, and all continuous variables were evaluated using the t-test or the Mann–Whitney U-test. The cutoffs for continuous variables were determined by the receiver operator characteristic (ROC) curve. An attempt was made to define the predictors of the primary outcome after discharge in patients with HF using Cox regression, and the event-free survival analysis was performed with Kaplan–Meier survival analysis. Statistical significance was accepted when the P < 0.05.
ResultsIn this study, 25 patients with ADHF were recruited after screening 63 consecutive patients. All patients completed 6 months of outpatient follow-up. Further telephonic follow-up was continued till the end of the study, leading to a mean follow-up of 246 ± 19 days. The primary outcome defined by the composite of death, cardiac transplant, or significant deterioration in symptoms leading to hospital readmission occurred in 6 (24%) patients. The primary outcome was due to death/heart transplant in 3 (12%) patients and readmission in the other 3 (12%). The mean days to readmission were 107.5 days with a range of 5–191 days. The consort diagram of the study is presented in [Figure 1].
[Table 1] shows the different categorical variables in patients with or without the primary outcome. There were no differences in age or sex distribution between the two groups. There was also no significant difference in the etiology of ADHF or risk factors such as smoking and diabetes. Serum sodium was slightly lower, and potassium was slightly higher in patients with the primary outcome, but none of these was statistically significant. Hemoglobin was significantly lower in patients with the primary outcome. There were also no significant differences in SBP at admission or NYHA class, presence of atrial fibrillation (AF), left bundle branch block, mitral regurgitation or tricuspid regurgitation (TR) on echo and mean left ventricular ejection fraction and TR gradient. Beta-blockers on discharge were prescribed in 15 of 19 patients (78.94%) who did not have a primary outcome compared to 1 of 6 patients (16.66%) who had the primary outcome. This difference was found to be statistically significant (P = 0.01). Overall beta-blockers were prescribed in 64% (16/25) of patients. There were no significant differences in the use of angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, diuretics, or aldosterone antagonists. Interestingly, hemoglobin levels were significantly different between the two groups. Those who suffered the primary outcome had a mean hemoglobin of 10.75 ± 0.67 g/dl, while those who did not have a higher hemoglobin of 12.33 ± 1.60 g/dl (P = 0.03).
Table 1: Distribution of clinical features in patients with or without primary outcomeThe 6-min walk distance was lower in patients with the primary outcome; however, this was not statistically significant, as shown in [Table 2]. Resting HR at discharge was significantly higher, with a mean difference of 23.42 b/min and a 95% confidence interval (CI) of 4.57–42.27 b/min. The HRR at 1 min was significantly lower with a mean difference of 10.08 b/min and 95% CI of 5.33–14.84. CR was also significantly lower in patients with the primary outcome, both as assessed by the difference between the maximum and minimum HR (CR).
Table 2: Study parameters in patients with or without the primary outcome*The ROC curve was plotted for HRR, resting HR, and CR to determine the area under the curves (AUCs) and the cutoffs. The cutoff values that maximized the Youden index (sensitivity + specificity - 1) were used as the cutoff values. The primary outcome was associated with HRR <9 beats/min, resting HR ≥109 b/min, CR <20b/min, CR% <25.5. Respective areas under the curve and P values are shown in [Table 3], with the curves shown in [Figure 2]. HRR had the best AUC of 0.877 (95% CI of 0.711–1.000).
Figure 2: ROC curve analysis of the studied parameters. (a) ROC curve of HRR for prediction of hospital readmission/death. (b) ROC curve of resting HR for prediction of hospital readmission/death. (c) ROC curve of CR for prediction of hospital readmission/death. ROC: Receiver operator characteristic, HRR: Heart rate recovery, HR: Heart rate, CR: Chronotropic reserve.
Table 3: Receiver operator characteristic table for various variables as a marker for prediction of primary outcomeEvent-free survival was evaluated with Kaplan–Meier survival analysis, and overall event-free survival and transplant-free survival were evaluated. The patients were then stratified with HRR, resting HR, and CR [Figure 3]. Log-rank test was used to assess the significance of the difference in survival by these variables. Overall event-free survival at 180 days was 78.8% with a 95% CI of 56.23%–90.68%; at 360 days, it was 73.95% with a 95% CI of 50.63%–87.47%. Event-free survival was significantly better for patients with HRR ≥9 b/min with P = 0.003. Similarly, event-free survival was also significantly better for patients with resting HR <109 b/min (P < 0.001) and CR <20 b/min (P = 0.003).
Figure 3: Kaplan Meier survival analysis. (a) Overall event free survival. (b) Difference in event free survival by heart rate recovery. (c) Difference in event free survival by resting heart rate. (d) Difference in event free survival by chronotropic reserve. Discussion Our study provides important information pertaining to the prognostication of patients admitted with ADHF. It shows the usefulness of HRR at 1 min postexercise as a useful parameter for the prediction of clinical events. Patients with HRR ≥9 beats/min had significantly lesser clinical events in terms of hospital readmissions, heart transplant, or death. Resting HR was also significantly associated with clinical events, with resting HR s <109 associated with better outcomes. CR, as assessed by maximal HR – Resting HR, was also found to be associated with clinical events with CR of ≥20b/min predictive of better event-free survival. The 6-min walk distance in our study was not found to be predictive of clinical events in patients admitted with ADHF at the time of discharge.
Patients with HF have significant sympathetic overactivity that is reflected in elevated levels of norepinephrine and epinephrine.[20] There is also a reduction in the parasympathetic outflow.[20] These are responsible for the higher resting HR and decreased HRR and chronotropic incompetence. In earlier studies as well, resting HR, HRR, and chronotropic reserve have been shown to correlate with clinical outcomes.[8],[9],[10],[11],[12],[13],[16],[17],[18] Benes et al. found that elevated resting HR correlates more with markers of increased myocardial stress like brain natriuretic peptide and markers of inflammation like C-reactive protein, while HRR correlates more with markers of neurohormonal response such as copeptin and norepinephrine.[21] This suggests different pathophysiological mechanisms for elevation in resting HRs which are correlated with increased myocardial stress in response to HF. Decreased HRR and impaired chronotropic reserve, on the other hand, correlate more with neurohormonal disturbances in patients with HF. The correlation between increased resting HR and worse clinical outcomes is explained by the increased myocardial stress and advanced state of HF in these patients. The correlation of decreased parasympathetic tone, which is reflected in impaired chronotropic reserve and worse clinical outcomes, is not well understood; however, it may be associated with neural regulation of pro-inflammatory cytokines. The vagal tone has an inhibitory control over the release of pro-inflammatory cytokines, which has been shown by attenuation of the level of Interleukin (IL)-6, tumor necrosis factor and IL-18 on vagal stimulation.[22] Therefore, decreased HRR and chronotropic reserve, which are markers of decreased parasympathetic tone, may result in worse clinical outcomes secondary to damage due to persistent pro-inflammatory state in patients with HF. However, neurohormonal regulation is very complex, and we are yet to translate the role of vagal tone in HF into clinically useful therapeutics.
In our study, we found resting HR to be a good marker for the prediction of readmission or mortality. It is a simple and easily measurable parameter. However, the reproducibility of the parameter may vary according to how it is recorded may be questionable. Nevertheless, a good correlation between the mean office resting HR and the average HR over a 24-h recording correlated well in the SHIFT Holter sub-study.[23] Vollmert et al. found that a resting HR at the discharge of more than 77 b/min predicted almost doubled mortality risk in a cohort of 78 patients with HF with reduced/mid-range ejection fraction.[11] Another contemporary study by Bahouth et al. suggested a cutoff 90 b/min for predicting 1-year mortality in HF patients with reduced ejection fraction. There was no statistically significant difference in outcomes between the groups having resting HR s of 70–90 b/min and <70 beats/min.[12] Our study suggested a higher cutoff of 109 b/min for predicting outcomes. The reason for these variables cutoff levels remains unexplained but may be related to the late presentation of patients in a sicker state when they receive medical care and hospitalization in our resource-limited setting. Further, we also included HF with preserved ejection fraction in our study population, making our study population different from the above-mentioned studies. Recently, extrapolatory analysis from the EMPEROR-Preserved trial has also suggested the role of resting HR as a prognostic marker in patients with HF with preserved ejection fraction.[13] Although HR has also been studied as a prognostic marker in patients with HF and AF,[24],[25] our study had included only three patients with AF and thus cannot draw any conclusions for this subset.
HRR 1 min after the 6-MWT had the best AUC of 0.87 on the ROC analysis to predict adverse outcomes among the parameters studied in our study. Cahalin et al. previously reported HRR after 6 MWT of >12 b/min to be a powerful predictor of outcomes in patients with HF.[16] The same study also reported that a walk distance of 6 min might not be useful for predicting HF outcomes. The results from our study are very similar and re-reiterate the strong prognostic role of HRR in HF patients. A recent small study demonstrated a better prediction with time to recovery to resting HR after 6 MWT than HRR.[17] However, this study included only NYHA Class II/III chronic HF patients and excluded acute HF patients.[17] Therefore, it should be noted that HRR may be a useful predictor of events in patients discharged after recovery from an acute HF episode but may not be useful in patients with chronic HF. Our study is novel in studying chronotropic reserve as a prognostic marker in patients recovering from acute HF, where it has not been studied before. An increase in HR during exercise parallels the sympathetic reserve at rest and thus may be a marker of autonomic influences over the heart. Chronotropic incompetence and reserve were found to predict coronary heart disease outcomes in healthy males from the Framingham heart study.[18] Our study predicted better event-free survival in patients with a chronotropic reserve of more than 20 b/min. The incremental benefit of this marker over HRR and resting HR is, however, questionable.
Among the other parameters studied in this study, the lack of a beta-blocker prescription at discharge and a lower hemoglobin level also predicted poor clinical outcomes at follow-up. Beta-blockers have been consistently shown to improve outcomes in chronic HF patients in several randomized controlled trials.[26],[27],[28] Although the continuation of beta-blockers at a steady dose as before during an acute HF episode is recommended, a new prescription during an episode should be considered only after stabilization.[29] Beta-blockers in naïve patients should be an objective at discharge after hospitalization for HF. In the present study, patients who did not have beta-blocker prescriptions at discharge fared poorly. The lack of beta-blocker prescription in these patients may be simply due to a more severe state of HF and the inability to tolerate the initiation of beta-blockers. Therefore, this may be an indirect marker of the prognosis rather than directly affecting the prognosis. Lower hemoglobin levels have been shown to be associated with poor HF outcomes, as it may be associated with lower oxygen-carrying capacity and thus increased demand for cardiac output. Our results are consistent with the available literature.[30],[31]
Our study has several limitations. First, the small sample size limited our overall events, and we did not have enough power, so we could not define any independent predictors of readmissions or death. We could not measure the levels of inflammatory markers or markers of increased myocardial stress to elaborate on the pathophysiological mechanisms behind the worse clinical outcomes associated with increased resting HR, decreased HRR, and decreased chronotropic reserve. Evaluation of predischarge resting HR, HRR, and chronotropic reserve together in the same population remains a major strength of the study.
ConclusionsOur study shows that elevated resting HR, decreased HR reserve, and decreased chronotropic reserve are associated with worse clinical outcomes. They are associated with higher hospital readmissions and lower event-free survival. Among the other studied parameters, the absence of beta-blocker use at discharge and the presence of anemia also predicted poor outcomes. All patients with HF should have an assessment of these parameters at the time of discharge to optimize their HF management and to stratify the risk group of patients with the highest risk of recurrent events for better resource utilization. Resting HR is especially helpful in this regard given the simplicity and ease of its assessment.
Ethics clearance
Appropriate ethical clearance was taken from the institutional ethics committee.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
Comments (0)