Skip to main content

Table 2 Direct, indirect, and network meta-analysis estimates of the effects of different preoxygenation methods

From: Preoxygenation before intubation in adult patients with acute hypoxemic respiratory failure: a network meta-analysis of randomized trials

Comparison

No. of trials

Direct estimate (95% CI)

Quality

Indirect estimate (95% CI)

Qualitye

NMA estimate (95% CI)

Quality

Lowest SpO2 during intubation (MD)

 HFNC vs. COT

3

− 1.64 (− 4.53, 1.25)

High

− 2.95 (− 8.23, 2.32)

Lowf,g

− 1.94 (− 4.48, 0.59)

High

 NIV vs. COT a

2

− 5.95 (− 9.38, − 2.53)

Moderateb

− 4.64 (− 9.58, 0.31)

Moderatef

− 5.53 (− 8.34, − 2.71)

Moderate

 HFNC vs. NIV

1

3.00 (− 1.01, 7.01)

Lowb,c

4.31 (− 0.17, 8.80)

Moderatef

3.58 (0.59, 6.57)

Moderate

 HFNC and NIV vs. NIV

1

− 3.10 (− 11.18, 4.98)

Moderateb

Not estimablek

–

− 3.10 (− 11.18, 4.98)

Moderate

SpO2 < 80% during intubation (OR)

 HFNC vs. COT

3

0.79 (0.32, 1.94)

Moderatec

0.44 (0.10, 1.95)

Very lowf,h

0.67 (0.31, 1.46)

Moderate

 NIV vs. COT

2

0.35 (0.13, 0.96)

Moderateb

0.63 (0.15, 2.60)

Moderatef

0.43 (0.19, 0.97)

Moderate

 HFNC vs. NIV

1

1.25 (0.42, 3.75)

Moderatec

2.23 (0.58, 8.60)

Lowf,g

1.58 (0.67, 3.69)

Moderate

 HFNC and NIV vs. NIV

1

0.16 (0.01, 1.80)

Moderatec

Not estimablek

 

0.16 (0.01, 1.80)

Moderate

Intubation-related complicationsi (OR)

 HFNC vs. COT

3

0.50 (0.27, 0.92)

High

0.44 (0.08, 2.53)

Lowj

0.49 (0.28, 0.88)

High

 NIV vs. COT

2

0.38 (0.07, 2.06)

Very lowb,d

0.44 (0.20, 0.96)

Moderateg

0.43 (0.21, 0.87)

Moderate

 HFNC vs. NIV

1

1.15 (0.70, 1.87)

Moderatec

1.30 (0.22, 7.77)

Very lowj,g

1.16 (0.72, 1.86)

Moderate

 HFNC and NIV vs. NIV

1

1.20 (0.31, 4.61)

Lowd

Not estimablek

 

1.20 (0.31, 4.61)

Low

Mortality (OR)

 HFNC vs. COT

2

0.90 (0.55, 1.46)

High

0.58 (0.29, 1.17)

Lowf,g

0.78 (0.52, 1.16)

High

 NIV vs. COT

2

0.68 (0.40, 1.14)

Moderateb

1.04 (0.53, 2.04)

Lowf,g

0.79 (0.53, 1.20)

Moderate

 HFNC vs. NIV

1

0.86 (0.54, 1.37)

Moderateb

1.32 (0.65, 2.70)

Lowf,g

0.98 (0.66, 1.45)

Moderate

 HFNC and NIV vs. NIV

1

0.78 (0.24, 2.55)

Lowd

Not estimablek

–

0.78 (0.24, 2.55)

Low

  1. COT conventional oxygen therapy (bag-valve mask or facial mask), HFNC high-flow nasal cannula, NIV noninvasive ventilation, MD mean difference, OR odds ratio, NMA network meta-analysis
  2. aThe median and interquartile range of lowest SpO2 extracted from Baillard et al. [5] were converted to mean and standard deviation using a published equation [8]
  3. bQuality of evidence for direct estimate rated down by one level for risk of bias
  4. cQuality of evidence for direct estimate rated down by one level for serious imprecision
  5. dQuality of evidence for direct estimate rated down by two levels for very serious imprecision
  6. eWe did not downgrade for intransitivity in indirect comparisons
  7. fQuality of evidence for indirect estimate rated down by one level for serious imprecision
  8. gQuality of evidence for indirect estimate rated down by one level for risk of bias
  9. hQuality of evidence for indirect estimate rated down by two levels for serious risk of bias
  10. iIntubation-related complications were defined as aspiration or new infiltrate on post-intubation chest radiograph, hemodynamic instability, and cardiac arrest
  11. jQuality of evidence for indirect estimate rated down by two levels for very serious imprecision
  12. kCannot be estimated because it was not connected in a loop in the evidence network
Back to article page

Critical Care

ISSN: 1364-8535