Previous studies have suggested benefits of applying fractal analysis to intervals between R waves in electrocardiography as an additional prognostic marker. The aim of this study was to investigate whether fractal analysis can provide an independent predictor of cardiac mortality or all-cause mortality. Prognostic cohort studies reporting fractal heart rate variability results from 24-h Holter monitor recordings were selected for comparison. Populations were subdivided into four groups-post-myocardial infarction, left ventricular dysfunction, other cardiac, and non-cardiac patients-and analysed using ANOVA, Forest plots (using pooled mean difference), and Funnel plots. The most significant mean differences were recorded in short-term fractal self-similarity (α1) (-0.17, 95% CI [-0.21, -0.13], p < 0.00001) and the traditional measure called standard deviation of NN intervals (SDNN) (-13.31, 95% CI [-18.89, -7.73], p < 0.00001) between the deceased and survivor groups. Fractal measures of long-term fractal self-similarity (α2), 1/f scaling (β), and traditional heart rate variability measures of high frequency to low frequency ratio show promise. This review indicated that fractal measure α1 and traditional measure SDNN could be potential predictors of mortality, but require further assessment to determine appropriate thresholds for clinical significance and additional targeted prognostic studies to properly define their applicability as prognostic markers. Therefore, clinicians should interpret fractal and traditional measures with caution since such measures have yet to be fully described as biomarkers for clinical application.