. 2017 Feb 28;18:78. doi: 10.1186/s12882-017-0486-9

Table 2.

Grade of evidence for primary outcomes

Early RRT compared to Late RRT for aki
Outcomes	No of participants (studies)	Quality of evidence (GRADE)	Relative effect (95% CI)	Anticipated absolute effects
	Follow up			Risk with Late RRT	Risk difference with Early RRT (95% CI)
Overall mortality	1672 (10 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR 0.93 (0.75 to 1.15)	study population
Overall mortality	1672 (10 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR 0.93 (0.75 to 1.15)	414 per 1000	29 fewer per 1000 (from 104 fewer to 62 more)
Day 30 mortality	1301 (6 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR 0.85 (0.6 to 1.2)	study population
Day 30 mortality	1301 (6 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR 0.85 (0.6 to 1.2)	430 per 1000	65 fewer per 1000 (from 172 fewer to 86 more)
Day 60 mortality	1075 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.64 to 1.27)	study population
Day 60 mortality	1075 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.64 to 1.27)	473 per 1000	47 fewer per 1000 (from 170 fewer to 128 more)
Day 90 mortality	555 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.49 to 1.64)	study population
Day 90 mortality	555 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.49 to 1.64)	449 per 1000	45 fewer per 1000 (from 229 fewer to 287 more)
Overall ICU mortality	430 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.75 to 1.68)	study population
Overall ICU mortality	430 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.75 to 1.68)	350 per 1000	42 fewer per 1000 (from 87 fewer to 238 more)
Overall Hospital mortality	713 (6 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.7 to 1.68)	study population
Overall Hospital mortality	713 (6 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.7 to 1.68)	298 per 1000	24 fewer per 1000 (from 89 fewer to 203 more)
Dialysis dependence at day 90	539 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.51 to 2.22)	study population
Dialysis dependence at day 90	539 (3 studies)	⊕⊝ ⊕ ⊝ Low^a,b due to risk of bias, inconsistency	RR (0.51 to 2.22)	54 per 1000	3 more per 1000 (from 27 fewer to 66 more)

The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ration; GRADE Working Group of evidence

High quality: Further research is very unlikely to change our confidence in the estimate of effect

Moderate quality: Further research if likely to have an important impact on our confidence in the estimate of effect and may change the estimete

Low quality: Further research if very likely to have an important impact on our confidence in the estimate of effect and may change the estimete

Very low quality: We ae very uncertain about the estimate

^aHeterogeneity

^bThe risk was increased or decreased