Skip to main content

Table 2 The GRADE approach was used to assess the certainty of evidence

From: Clonidine for sedation in the critically ill: a systematic review and meta-analysis

Quality assessment

№ of patients

Effect

Quality

Importance

№ of studies

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations

Clonidine

Placebo

Relative (95% CI)

Absolute (95% CI)

Duration of mechanical ventilation (assessed with: days)

6

Randomized trials

Not seriousa

Seriousb

Not serious

Not seriousc

None

200

217

-

MD 0.05 days more (0.65 fewer to 0.75 more)

⨁⨁⨁ MODERATE

CRITICAL

ICU mortality

5

Randomized trials

Seriousd

Not serious

Not serious

Seriouse

None

14/164 (8.5%)

23/219 (10.5%)

RR 1.00 (0.50 to 2.02)

0 fewer per 1000(from 53 fewer to 107 more)

⨁⨁ LOW

CRITICAL

Dose of benzodiazepines

4

Randomized trials

Not seriousf

Seriousg

Not serious

Not seriousc

None

130

134

-

SMD 0.02 SD higher (0.34 lower to 0.39 higher)

⨁⨁⨁ MODERATE

CRITICAL

Dose of narcotics

4

Randomized trials

Not seriousf

Not serious

Not serious

Seriouse

None

130

134

-

SMD 0.26 SD lower (0.5 lower to 0.02 lower)

⨁⨁⨁ MODERATE

CRITICAL

Incidence of withdrawal symptoms from other sedatives

3

Randomized trials

Seriousi

Not serious

Not serious

Serioush

None

40/120 (33.3%)

49/124 (39.5%)

RR 0.91 (0.67 to 1.23)

36 fewer per 1000 (from 91 more to 130 fewer)

⨁⨁ LOW

IMPORTANT

ICU length of stay

6

Randomized trials

Seriousj

Not serious

Not serious

Not seriousc

None

233

240

-

MD 0.04 days more (0.46 fewer to 0.53 more)

⨁⨁⨁ MODERATE

CRITICAL

Incidence of clinically significant hypotension

4

Randomized trials

Not seriousk

Not serious

Not serious

Seriousl

None

31/178 (17.4%)

12/226 (5.3%)

RR 3.11 (1.64 to 5.87)

112 more per 1000 (from 34 more to 259 more)

⨁⨁⨁ MODERATE

CRITICAL

Incidence of clinically significant bradycardia

4

Randomized trials

Not seriousk

Not serious

Not serious

Very seriouse

None

14/178 (7.9%)

12/226 (5.3%)

RR 1.34 (0.45 to 3.98)

18 more per 1000 (from 29 fewer to 158 more)

⨁⨁ LOW

CRITICAL

  1. CI confidence interval, MD mean difference, RR risk ratio, SMD standardized mean difference
  2. aFour of six included studies had issues related to high risk of bias. However, excluding high risk of bias studies had no effect on overall pooled estimates
  3. bHigh I 2 value of 86%
  4. cPoint estimate shows no effect. Confidence intervals do not rule out slight harm or slight benefit, however, neither meet clinical significance threshold
  5. dAll included studies for this outcome were at high risk of bias
  6. eConfidence intervals do not exclude benefit or harm
  7. Confidence intervals do not exclude benefit or harm. Low number of patients and event numbers
  8. Two of three trials had issues related to high risk of bias. However, sensitivity analysis excluding high risk of bias trials did not alter results
  9. High I 2 value of 82%
  10. Confidence intervals do not exclude benefit or harm, with the benefit potentially meeting clinical threshold. Low number of patients
  11. fTwo of four trials had issues related to high risk of bias. However sensitivity analysis excluding high risk of bias trials did not alter results
  12. gHigh I 2 value of 55%
  13. hConfidence intervals do not exclude benefit or harm. Low number of patients
  14. iTwo of three trials had issues related to high risk of bias. However, sensitivity analysis excluding high risk of bias trials results in a potentially different outcome
  15. jFour of six trials had issues related to high risk of bias. However, sensitivity analysis excluding high risk of bias trials results in a potentially different outcome
  16. High I 2 value of 52%
  17. kThree of four trials had issues related to high risk of bias. However sensitivity analysis excluding the high risk of bias trials did not alter results
  18. lLow number of events
Back to article page

Critical Care

ISSN: 1364-8535