Table 2.
Studies examining the prognostic value of RDW in (a) the general population, (b) chronic, (c) acute and (d) advanced heart failure.
| Reference | Number of Subjects | Study Design | Outcome | Results | Conclusion |
|---|---|---|---|---|---|
| (a) Population-Based Cohort | |||||
| [25] | 17,533 | Retrospective (Mean follow up 11.2 years) |
Incident HF | Adj. HR 1.44, (95% CI 1.15–1.80) | RDW is associated with HF events in an apparently healthy middle-aged population. |
| [24] | 26, 784 | Retrospective (Mean follow up 15 years) |
Risk of hospitalization due to HF | Adj. HR 1.47, (95% CI 1.14–1.89) | Red cell distribution width was associated with long-term incidence of first hospitalization due to HF among middle-aged subjects. |
| (b) Chronic Heart Failure | |||||
| [38] | 6888 | Retrospective (Follow up 24 months) |
All-cause mortality and cardiovascular hospitalization | A lower Hb/RDW ratio was a predictor of mortality (Q1 vs. Q6: Adj HR 1.84 (1.63–2.08) | Hb/RDW ratio is a prognostic tool for predicting HF mortality and cardiovascular hospitalizations. |
| [32] | 169 HFpEF vs. 50 controls |
Prospective | Predictive value of deformation imaging combined with RDW | The associations of clinical and echocardiographic parameters with HFpEF were improved by adding RDW (p < 0.01) | RDW has an independent incremental predictive value for HFpEF. |
| [31] | 59 HFrEF vs. 40 controls |
Prospective | LV global longitudinal strain | RDW showed negative correlations with LV global longitudinal strain (r = −0.41, p = 0.001) | Elevated RDW is associated with poorer LV deformation assessed by speckle tracking echocardiography in HF patients with similar EF. |
| [30] | 1084 | Prospective | LVEDP, mortality | RDW was independently associated with high LVEDP (Adj. OR per unit change 1.14, 95% CI 1.0 to 1.29) and 5 year-mortality (Adj. HR 4.11, 95% CI 2.12 to 7.96) | RDW was independently associated with high LVEDP and with mortality. |
| [36] | 232 | Prospective (Follow up 12 months) |
Cardiovascular death and/or HF hospitalization | RDW > 14.45%, Adj. OR:3.894, (95%CI 1.042–14.55) | RDW is a better predictor of adverse outcome than several echocardiographic parameters. |
| [27] | 215 | Prospective (Mean follow up 24.2 months) |
All-cause mortality | Adj. OR: 2.963 (95% CI 1.066–6.809) | RDW may be an indicator in the risk stratification. |
| [35] | 350 | Retrospective (Follow up 12 months) |
All-cause mortality and HF hospitalization | Higher mortality and HF re-admission in patients with RDW > 14.5 (vs. RDW ≤ 14.5) (p < 0.001 and p = 0.004, respectively). Levels of RDW were associated with the presence of severe LV dysfunction (LVEF < 30%) * | Elevated RDW may be used as a prognostic tool among HF patients with the documented myocardial infarction. |
| [34] | 165 | Prospective (Follow up 24 months) |
All-cause mortality | Adj. HR 1.19 (95% CI 1.004–1.411) at 12 months | RDW is an independent predictor of mortality at 12 months, but it loses its significance during longer-term follow up. |
| [37] | 1021 (CHD vs. DCM vs. VHD) |
Retrospective (Mean follow up 21 months) |
All-cause mortality | The AUC of RDW for predicting mortality due to CHD and DCM was 0.704 (95% CI 0.609–0.799) and 0.753 (95% CI 0.647–0.860), respectively. The AUC of the RDW for predicting mortality from VHD was 0.593 | RDW is a prognostic indicator for patients with HF caused by CHD and DCM. |
| [29] | 85 HF vs. 107 controls |
Prospective | Peak VO2, VE/VCO2 slope |
RDW is an independent predictor for peak VO2 (β = −0.247, p = 0.035) and VE/VCO2 slope (β = 0.366, p = 0.004) | Higher RDW is independently related to peak VO2 and VE/VCO2 slope. |
| [28] | 118 | Prospective | Exercise capacity | Log[RDW] is associated with VO2peak (β = –0.277, p = 0.003) | Higher RDW is independently related to impaired exercise capacity. |
| [96] | 698 | Prospective (Median follow up 2.5 years) |
All-cause mortality HF hospitalization |
All-cause mortality HR (for RDW > 15.4%) 2.63, (95% CI 2.01–3.45) HF hospitalization HR (for RDW > 15.4%) 2.37, (95% CI, 1.80–3.13) |
RDW value is a risk marker for the occurrence of both death and hospitalization for HF in outpatients with chronic HF, independent of anemia. |
| [26] | 1087 | Retrospective (Median follow up 52 months) |
All-cause mortality | Adj. HR 1.12, (95% CI 1.05–1.16) | RDW has similar independent prognostic power to NT-proBNP. |
| [33] | 2679 | Retrospective (Median follow up 34 months) |
Morbidity and mortality | Adj. HR 1.17 per 1-SD increase, p < 0.001 | RDW is an independent predictor of morbidity and mortality. |
| (c) Acute Heart Failure | |||||
| [49] | 897 (≥65 years) |
Retrospective | All-cause mortality at 1 year | Adj. HR 1.41 (95% CI, 1.05–1.90) | A higher baseline RDW was associated with increased risk for 1-year all-cause mortality. |
| [42] | 2278 ED visits | Retrospective (Follow up 4 years) |
All-cause mortality at 30 days | AUC 0.723, (95% CI 0.693–0.763) | The prognostic assessment of acute HF patients in the ED can be improved by combining RDW with other laboratory tests. |
| [52] | 218 patients (71 diabetics) | Prospective (Follow up 1 year) |
All-cause mortality or rehospitalization for HF at 1 year | Diabetics: Adj HR: 1.349, (95% CI 1.120–1.624) Non-diabetics: Adj HR: 1.142, (95% CI 1.011–1.291 (βinteraction = −0.002; SE = 0.001; p = 0.042) between DM and RDW longitudinal changes |
RDW has similar prognostic significance (diabetic and non-diabetic) in HF patients. RDW longitudinal changes show significant difference in diabetic and non-diabetic patients. |
| [51] | 278 HFpEF patients | Retrospective (Follow up 3 years) |
Non cardiac mortality | Adj. HR 1.169, (95% CI 1.042–1.311) | RDW levels at admission independently predict non-cardiac mortality in acute HFpEF. |
| [50] | 402 | Prospective | All-cause mortality at 1 year | All-cause mortality of all patients increased with quartiles of rising RDW (χ2 18; p < 0.001). | High RDW predicts mortality in acute HF. |
| [47] | 128 | Prospective (Follow up 3 months) |
Cardiac death and/or readmission for HF | Adj. HR 4.610, (95% CI 1.935–10.981) | RDW and NT-proBNP are independent predictors of 90-day cardiovascular events in patients hospitalized with HF. RDW adds prognostic value to NT-proBNP. |
| [46] | 521 | Prospective (Median follow up 24 months) |
In-hospital mortality, All-cause mortality and HF readmission (long term) |
In-hospital mortality (for log RDW): coef. 5.21, p = 0.044, All-cause mortality and HF re-admission (long term): RDW (per SD increase, HR 2.19; 95% CI 1.92–2.50; p < 0.0001) |
Higher RDW values in acute HF at admission are associated with worse short- and long-term outcomes and RDW values are more prognostically relevant than hemoglobin levels. |
| [39] | 100 | Retrospective | Slow diuretic response | Adj. OR 1.47, (95 % CI 1.07–2.02) | High RDW at admission is a predictor of slower diuretic response. |
| [41] | 907 | Retrospective | All-cause mortality at 30 days |
Adj.HR 1.23, (95% CI 1.11–1.36) | RDW measured at ED is an independent predictor of early mortality. |
| [45] | 789 | Retrospective (Median follow up 573 days) |
All-cause mortality | Adj. HR 3.21, (95% CI 1.77–5.83) | Discharge RDW is an independent predictor of all-cause mortality in predominantly African American patients. |
| [48] | 205 | Retrospective (Follow-up 1 year) |
All-cause mortality | Adj. HR = 1.03 per 1% increase in RDW, (95% CI 1.02–1.07, p = 0.04) | RDW independently predicted 1-year mortality in acute HF. |
| [43] | 628 | Prospective (Median follow up 38.1 months) |
All-cause mortality | Adj. HR 1.072, (95% CI 1.023–1.124) | Higher RDW levels at discharge are associated with a worse long-term outcome, irrespective of hemoglobin levels. |
| [44] | 707 | Prospective (Median follow up 421 days) |
All-cause mortality | Adj. HR 1.06, (95% CI 1.01–1.11) | RDW provides incremental prognostic value to BNP in acute heart failure. The prognostic ability of RDW is independent of hemoglobin concentration. |
| [40] | 100 | Prospective | Relation between RDW and echocardiographic parameters | RDW was independently correlated with E/E (β-coefficient 0.431, p = 0.001) | RDW may be associated with elevated LV filling pressures in patients with acute HF. |
| (d) Advanced Heart Failure | |||||
| [59] | 409 patients with cf-LVADs |
Retrospective | All-cause mortality at 90 days | Adj. OR 1.16 for 1% increase, (95% CI: 1.04–1.31) | RDW is an independent predictor of 90-days mortality in cf-LVAD patients. |
| [53] | 367 | Retrospective (Mean follow up 4.4 years) |
All-cause mortality | Adj. HR 1.0492 (95 % CI 1.0247–1.0743) | RDW is an independent predictor of all-cause mortality in advanced HF patients with concomitant diabetes mellitus. |
| [62] | 173 | Retrospective (Mean follow up 45.5 months) |
All-cause mortality | Adj. HR 1.381 (95% CI 1.251–1.467) | RDW immediately before OHT is an independent predictor of all-cause mortality in heart transplant recipients. |
| [54] | 432 patients with ICDs | Retrospective (Follow up ≤ 5 years) |
First appropriate ICD therapy and death | Adj. HR 2.045 for RDW > 15.2 (95% CI 1.145–3.65) | RDW may be useful in risk stratification of patients selected for ICD implantation. |
| [58] | 188 cf-LVADs |
Retrospective (Follow-up ≥ 1 year) |
All-cause mortality | Adj. HR (for RDW > 18.1% vs. RDW < 15.7%) 4.61 (95% CI 1.74–12.21) |
Preimplant RDW is independently associated with postimplant mortality. |
| [61] | 44 | Prospective | Parameters associated with bone marrow dysfunction in patients with advanced chronic non-ischemic HF | Adj. HR 8.64 (95% CI 1.242–60.021) | RDW is an independent predictor of poor mobilization of CD34+ cells. |
| [60] | 37 patients with cf-LVADs |
Prospective (Median follow-up 136 days) |
Changes in laboratory parameters/biomarkers in patients who underwent LVAD implantation | median RDW (pre-implant) 16.7% vs. 16.5% (post-implant), p = 0.98 | RDW is elevated but does not change (pre- vs. post-LVAD implant). |
| [57] | 156 patients with CRTs | Retrospective (Median follow up 61 months) |
All-cause mortality | Adj. HR (baseline RDW) 1.33, (95%CI 1.16–1.53) HR (RDW 6 months after CRT implantation) 1.22, (95%CI 1.08–1.38) -HR (RDW 12 months after CRT implantation) 1.15, (95%CI 1.01–1.32) |
Baseline RDW levels, as well as RDW after CRT implantation, are independently associated with mortality in patients who undergo CRT implantation. |
| [55] | 233 patients with CRTs | Retrospective (Mean follow up 11.5 months) |
CRT response | Adj. OR 0.83, (95% CI 0.69–0.99) | Elevated RDW is associated with impaired reverse remodeling. |
| [56] | 66 patients with CRTs | Prospective (Follow up 6 months) |
CRT response | Adj. OR 1.435, (95 % CI, 1.059–1.945) | Elevated RDW is associated with poor CRT response. |
* In a multivariable logistic regression model, RDW was not found to be an independent predictor for re-hospitalization or mortality. Adj, adjusted; CRT, cardiac resynchronization therapy; cf-LVADs, continuous flow left ventricular assist devices; CHD, coronary heart disease; CI, confidence interval; DCM, dilated cardiomyopathy; ED, emergency department; HR, hazard ratio; HF, heart failure; HFrEF, heart failure with reduced ejection fraction; HFpEF, heart failure with preserved ejection fraction; ICD, implantable cardioverter defibrillator; LV, left ventricle; LVEDP, left ventricular end diastolic pressure; NT-proBNP, N-terminal pro b-type natriuretic peptide; OR, odds ratio; RDW, red blood cell distribution width; VHD, valvular heart disease.