Table 1 Summary of studies fulfilling the search criteria.

From: Bioimpedance-defined overhydration predicts survival in end stage kidney failure (ESKF): systematic review and subgroup meta-analysis

Author(s) Year Country RRT N Follow up Prim. Outcome Sec. Outcome N (Mort) QUIPS Criteria BIA Method Outcome of Study (with appropriate MVA outputs, if present)
End-Stage Kidney Failure Cohorts
Abad 2011 Spain HD+PD 164 6 yrs Mortality N/A 100 L M L L L M PA* PA < 8 (p < 0.01) and comorbidity (Charlson index) are independent predictors of mortality
Avram 2006 USA PD 177 15 yrs Mortality N/A 89 H H M M H H PA* PA (RR 0.54) and enrolment CRP were independent predictors of mortality.
Bebera-shvili 2014 Israel HD 91 3 yrs Mortality N/A 38 L L L L M L PA** Patients with greatest decline of PA had highest risk for mortality. 1 degree increase in PA (when treated as time varying variable) has a mortality HR of 0.61 (95% CI 0.53–0.71)
Bebera-shvili 2014 Israel HD 250 1.4 yrs Mortality N/A 64 L L L L M M PA*** As a continuous variable PA has a mortality HR of 0.72 (95% CI 0.54–0.96). Adjustment for MIS Score nullified of PA as mortality predictor (HR 0.75, 95% CI 0.54–1.03) but remained a predictor of hospitalisation risk (HR 0,76, 95% CI 0.63–0.92)
Caetano 2016 Portugal HD 697 1 yr Mortality N/A 66 M L L M L M OHI** OH/ECW >15% is an independent predictor of 1-year mortality during follow up (HR for mortality 2.22, 95% CI 1.29–3.79).
Chazot 2012 France HD 158 6.5 yrs Mortality Hypertension Unclear M L L L H M OHI** Giessen cohort patients with hyperdration (OH/ECW >15%) had worse cumulative survival than non-hyperhydrated patients (mortality HR 3.41, 95% CI 1.62–7.17).
Chen 2007 China PD 227 3 yrs Mortality N/A 58 L L M L L M ECWR** ECW/ICW is independent predicator of mortality in incident PD patients. Every 0.1 increase in ECW/ICW (time dependent) associated with mortality RR 1.37 (95% CI 1.10–1.70)
de Araujo 2013 Brazil HD+PD 145 1.3 yrs CV Events Mortality 13 M L M L L L PA PA is predictive of CV events in non-diabetics (HR 0.56, 95% CI 0.38–0.83) but not in diabetics (HR 1.01, 95% CI 0.60–1.70). Study has small number of endpoints.
Dekker 2017 Inter-national HD 8883 1 yr Mortality N/A Unclear L L M M M H OHI** Baseline pre-dialysis OHI/ECW >15% (overhydration 2.5–5L) is predictive of mortality (HR 2.62, 2.1–3.3), independent of comorbidity. Inflammation and FO in a dynamic cohort have additive effects on mortality in HD patients.
Demirci 2016 Turkey HD 493 2.3 yrs Mortality CV Mortality 93 L L L L M M BIVA** When adjusted for comorbidities, impedance ratio is independently predictive for all cause mortality (HR 1.13, 95% CI 1.04–1.23) and cardiovascular mortality (HR 1.15, 95% CI 1.03–1.27)
Di Iorio 2004 Italy HD 515 339 yrs^ Mortality N/A 75 L M M M L M PA* PA was an independent predictor of mortality in a HD population (RR 2.50).
Fan 2015 UK PD 183 1.7 yrs Mortality Technique Failure 37 L M M L M L ECWR** In PD patients, ECW is an independent predictor of mortality, including in cases adjusted for peritonitis episodes (HR 2.98, 95% CI 1.40–7.30). Log CRP also an independent predictor of mortality (HR 3.32, 95% CI 1.50–7.70).
Fein 2002 USA PD 45 0.6 yrs Mortality N/A 4 L M M M H H PA**** Univariate analysis revealed patients with PA < 6 had worse cumulative survival than those >6 (p < 0.01). No MVA present.
Fein 2008 USA PD 53 8 yrs Mortality N/A 21 H H L M M M ECWR* Enrolement BIA measures (Avram et al., ECM/BCM, RR 1.04/Fein et al., ECW/BSA, RR 1.50) were independent predictors of mortality. Avram et al. data included as same cohort*
Fiedler 2009 Germany HD 90 3 yrs Mortality Hospital admission events 36 M L M L M L PA** PA < 4 independently predicts mortality in HD patients (RR 2.34, 95% CI 1.06–5.14). Individual nutrition scores are superior to BIA in terms of prognostic utility.
Guo/Guo 2015 China PD 307 3.2 yrs Mortality CV Mortality 52 L L L L M L ECWR** In CAPD patients ECW/TBW >0.40 is an independent predictor of all cause mortality (HR 13.12, 95% CI 1.35–128.00) and PD technique failure (HR 10.34, 95% CI 1.88–57.02).
Hoppe 2015 Poland HD 241 2.5 yrs Mortality N/A 42 M H H M H H OHI Troponin and OH index predict mortality in 1 MVA, but when adjusted for other covariates, OH Index (continous variable) no longer an independent predictor of mortality (RR 1.12, 95% CI 0.92–1.37).
Huan-Sheng 2016 Taiwan HD 298 1 yr Hospital admission events CV Events 13 QUIPS not validated for assessing bias in RCT OHI No differences noted in all cause hospitalisation (HR 1.19, 95% CI 0.79–1.80), all cause mortality (HR 0.85, 95% CI 0.29–2.53) and fluid overload/cardiovascular event rate (HR 0,57, 95% CI 0.08–1.07) between the BIA and control groups.
Jotterand-Drepper 2016 Germany PD 54 6.5 yrs Mortality N/A 19 L L M L M H OHI** OHI/ECW >15% independently predictive of mortality (HR 7.82, 95% CI 1.10–29.07) in PD patients when adjusted for troponin values, CRP, the presence of heart failure and hypoalbuminaemia.
Kim 2015 South Korea HD 240 2 yrs Mortality Hospital admission events 50 M M L L M M OHI** When adjusted for comorbidities, OH/ECW >15% was an independent predictor of mortality (HR 2.58, 95% CI 1.16–5.75). Age was also an independent predictor.
Kim 2017 South Korea HD 77 5 yrs Mortality CV Events 24 L L H L M M ECWR** ECW/ICW ratio is an independent predictor of mortality (HR 1.12, 95% CI 1.01–1.25) and cardiovascular events (HR 1.09, 95% CI 1.01–1.18) when adjusted for multiple co-morbidities.
Koh 2011 Malaysia PD 128 2.2–2.3 yrs Mortality N/A 35 L M L L L M PA** PA is a independent predictor of mortality in HD patients (HR 0.39, 95% CI 0.27–0.57).
Maggiore 1996 Italy HD 131 2.2 yrs Mortality N/A 23 M L L L H M PA* When adjusting for age and other nutritional markers, PA is an independent predictor of mortality in HD patients (p < 0.01).
Mathew 2015 India HD+PD 99 2 yrs Mortality N/A 33 M L L L M L OHI** Absolute overhydration (>3.1L) is an independent predictor of mortality (adjusted OR 2.96, 95% CI 1.04–8.46).
O’Lone 2014 UK PD 529 4 yrs Mortality N/A 95 M M L L H L OHI+ECWR** Where OH/ECW and ECW/TBW values are in the top 30% for the cohort, both OH/ECW (HR 2.09, 95% CI 1.36–3.20) and ECW/TBW (HR 2.05, 95% CI 1.31–3.22) act as independent predictors of mortality.
Oei 2016 UK PD 336 2 yrs Mortality N/A 48 L L M L H M OHI**** Univariate analysis correlates overhydration with cardiac death (p < 0.05), but no further analysis noted.
Onofriescu 2014 Romania HD 131 3.5 yrs Mortality Adverse Events 9 QUIPS not validated for assessing bias in RCT OHI** RCT of BIA vs standard clinical care in determining ultrafiltration on HD. BIA group had survival advantage over standard clinical care group (HR 0.11, 95% CI 0.01–0.92). Study at risk of selection bias.
Onofriescu 2015 Romania HD 221 5.5 yrs Mortality CV Mortality 66 L L M L L L OHI** OH/ECW >17.4%, when adjusted for comorbidities, is independently predictive for mortality when LVEF (HR 2.29, 95% CI 1.08–4.89) and LVMI (HR 2.19, 95% CI 1.02–4.69) are adjusted for in the analysis.
Paniagua 2010 Mexico HD+PD 753 1.4 yrs Mortality CV Mortality 182 M L M L M L ECWR** ECW/TBW (OR 1171.33, 95% CI 3.35–409899.37) and NT-proBNP (OR 1.01, 95% CI 1.00–1.02) independently predictive of CV mortality but not all cause mortality (OR 84.64, 95% CI 0.52–13788.55) in dialysis patients.
Paudel 2015 UK PD 455 2 yrs Mortality N/A 72 L H M M L H OHI**** Univariate analysis revealed OH index predictive of mortality. Multivariable model used to assess SGA independent of OH.
Pillon/Chertow 2004 USA HD 3009 0–1.5 yrs Mortality N/A 361 M L L M M M BIVA** BIVA vector, per 100Ohm/m incremental increase, is independently predictive of mortality (RR 0.75, 95% 0.57–0.88).
Ponce 2014 Portugal HD 189 1 yr Mortality Adverse Events 20 QUIPS not validated for assessing bias in RCT OHI Univariate analysis revealed survival (p = 0.33) and event-free-survival (p = 0.17) equivalent between BIA and control groups. The study was terminated prematurely.
Pupim 2004 USA HD 194 3 yrs Mortality CV Mortality 50 M M L L M L PA* PA and Albumin independent predictors of cardiovascular mortality in MVA, although summary statistics from MVA not reported.
Rhee 2015 South Korea PD 129 2.1 yrs Residual RF Mortality 15 M H L M M H ECWR** In Korean PD patients with preserved RRF, ECW/TBW is predictive of mortality (HR 1.001, 95% CI 1.001–1.086) and, additionally, technique failure (HR 1.024, 95% CI 1.001–1.048).
Segall/Segall 2014 Romania HD 149 1.1 yrs Mortality N/A 43 L M M M L L PA** PA < 5.58 is independently predictor of mortality in HD patients (HR 2.15, 95% CI 1.16–3.99).
Shin 2017 South Korea HD 142 2.4 yrs Mortality CV Mortality 15 L M M M H M PA** PA is an independent predictor of all cause mortality (HR 0.56, 95% CI 0.33–0.97) and infection (HR 0.65, 95% CI 0.45–0.94) in HD patients, but not for cardiovascular mortality (HR 0.92, 95% CI 0.43–2.14).
Siriopol/Siriopol 2015 Romania HD 173 1.8 yrs Mortality N/A 31 L M M L H H OHI** OH/ECW >6.68% (HR2.93, 95% CI 1.30–6.58) and lung comet score (LCS>22; HR 2.72, 95% CI 1.19–6.16) independently predictive of mortality. Earlier study (2013) OH/ECW not predictive of mortality but was underpowered.
Siriopol 2017 Romania HD 285 3.4 yr Mortality N/A 89 L L M L M L OHI In combination overhydration (>6.9%) and high NT-proBNP levels independently predict mortality (HR 1.83, CI 1.02–3.54, whereas in patients with normal NT-proBNP levels overhydration is not a predictor (HR 1.34, 95% CI 0.67–2.68)
Tangvora-phonkchai 2016 UK HD 362 4.1 yr Mortality N/A 110 L L L M L L OHI** OH (%, as a continuous variable) is an independent predictor of mortality in MVSA (HR 1.15, 95% CI 1.03–1.28); co-morbidity, non-BIA nutritional indices, albumin and CRP also noted to be independent predictors of mortality.
Tian 2016 China PD 152 5 yrs Mortality N/A 44 L M L M M M ECWR When adjusted for inflammation (CRP), ECWR (a standard deviation away from the median) is not predictive of mortality in PD patients (HR 2.20, 95% CI 0.79–6.08)
Wizemann 2009 Poland HD 269 3.5 yrs Mortality N/A 86 L L L M L M OHI** OH/ECW>15% is an independent predictor of mortality in HD patients (HR 2.10, 95% CI 1.39–3.18).
Zoccali 2017 International HD 39566 1.4 yrs Mortality N/A 5866 L M L M L L OHI** Baseline OHI/ECW>15% (men)/13% (women) at baseline independent predictor of mortality (HR 1.26, 95% CI 1.19–1.33) and in patients with cumulative fluid overload over a 1  yr period irrespective of pre-dialysis systolic BP.
Heart Failure Cohorts
Alves 2016 Brazil N/A 71 2 yrs Mortality N/A 34 L M L L H M PA** PA < 4.8 independent predictor of mortality in following episodes of acute decompensated heart failure (HR 2.67, 95% CI 1.21–5.89). Ejection fraction also independent predictor of mortality (HR 0.94, 95% CI 0.89–1.00)
Castillo-Martinez 2007 Mexico N/A 242 N/A NYHA Class N/A Unclear L H M H H M PA + BIVA**** Univariate analysis demonstrated PA predicts severity of symptoms (indirect measure of risk of hospitalisation) in both HFSD + HFPSF.
Colin-Ramirez 2012 Mexico N/A 389 3 yrs Mortality NYHA Class 66 L L L L M M PA** Following adjustment for age, haemoglobin and diabetic status, a PA < 4.2 was independently predictive of all cause mortality (HR 3.08, 95% CI 1.06–8.99).
Doesch 2010 Germany N/A 41 5 yrs Cardiac MRI data CV Mortality 8 M M L M M L PA On univariate analysis, PA>5.5 correlated with CV survival, but not statistically significant (p = 0.13).
Sakaguchi 2015 Japan N/A 130 0.5 yrs Adverse Events CV Events 37 (2 deaths) L L L L M L ECWR** In acute decompensated heart failure, ECW ratio (measured/predicted) independent predictor of cardiac death/re-admission (HR 1.48, 95% CI 1.20–1.83).
Trejo-Velasco 2016 Spain N/A 105 0.9 yrs Mortality Readmission 19 M M L L H M BIVA** Hyperhydration (defined by BIVA readings>74.3%) was an independent predictor of adverse outcomes (HR 2.6, 95% CI 1.1–6.4),
  1. Individual patient cohorts listed according to author(s), year of publication (where multiple studies from the same cohort are identified, lead authors of each study and year of most recent study cited) and geographical location of cohort. For each cohort, the BI-OH markers, dialysis modalities, follow up period, number of patients within the cohort, primary/secondary outcomes, number of endpoints and summary of findings are provided. Summaries for each cohort are given, along with the appropriate BI-OH marker and its utility within the cohort to predict survival (denoted by the numbers of * by the BI-OH marker): * shows that the BI-OH marker is an independent predictor of the primary outcome (but does not report a hazard ratio/risk ratio/odds ratio and confidence interval), ** shows the BI-OH marker is an independent predictor of the primary outcome (and reports hazard ratio/risk ratio/odds ratio and confidence interval), *** shows the BI-OH is an independent predictor of secondary but not primary outcome and **** shows that BI-OH is a univariate predictor of primary outcome. QUIPS risk of Bias summaries are provided for each cohort, with QUIPS domains coded as “L” for low risk of bias, “M” as medium risk of bias and “H” as high risk of bias. QUIPS domains include SP = Study participation, SA = Study attrition, PFM = Prognostic factor measurement, OM = Outcome measure, SC = Study confounding and SAR = Statistical analysis reporting. Randomised controlled trials (RCT) were not quality appraised using QUIPS as this is not a valid method of appraising methodological quality in this study design. In one study (highlighted ^), the follow up period was reported ambiguously and may have reflected cumulative follow up.