Skip to main content

Table 1 Four paradoxical observations among systematic review summaries of various VAP prevention interventions

From: Establishing the safety of selective digestive decontamination within the ICU population: a bridge too far?

Intervention

Ref

VAP incidence (per 1000 patients)

RR; 95% CI

n/N

Mortality incidence (per 1000 patients)

RR; 95% CI

n/N

Figure 1 symbol

  

Control

Intervention

  

Control

Intervention

   

Non-antimicrobial—RCCT

 Semi-recumbent a

[13]

316*

139§

0.44; 0.11–1.77

3/419

276*

240§

0.87; 0.59–1.27

2/307

 

 HME b

[12]

167*

155§

0.93; 0.73–1.19

13/2251

247*

257§

1.03; 0.89–1.2

12/1951

↓

 Probiotic c

[13]

309*

238§

0.7; 0.52–0.95

8/1018

214*

186§

0.84; 0.58–1.22

5/703

×€

 Protocolized MV wean d

[23]

NR

NR

NR

 

216*

166§

0.77; 0.39–1.5

2/513

 

Antimicrobial – CRT

 TAP ± PPAP

[56]

NR

NR

NR

 

See Fig. 1

1.00; 0.79–1.23‡

3/18335

 

Antiseptic – RCCT

 Chlorhexidine e

[16]

243*

180§

0.75; 0.62–0.91

18/2451

222*

242§†

1.09; 0.96–1.23

14/2014

↓

 Tooth brushing ± antiseptic f

[17]

259*

179§

0.61; 0.41–0.91

5/910

250*

210§

0.84; 0.67–1.05

5/910

×€

TAP ± PPAP – RCCT

 TAP ± PPAP

[56]

NR

NR

NR

 

NR

NR

0.85; 0.77–0.94‡

27/5699

 

 TAP + PPAP g,h,i

[6]

NR

NR

0.28; 0.2–0.38

16/3024

NR

NR

0.75; 0.65–0.87‡

17/4075

 

 TAP + PPAP g,i,j

[7]

417*

179§

0.43; 0.35–0.53

17/2951

303*

255§†

0.84; 0.73–0.96‡

18/5290

↓

 TAP (alone) h,i,k

[6]

NR

NR

0.34; 0.21–0.55

12/1735

NR

NR

0.97; 0.87–1.07

13/1783

 

 TAP (alone) i,j,k

[7]

324*

162§

0.50; 0.36–0.69

13/1848

305*

296§†

0.97; 0.87–1.07

15/3274

×€

  1. VAP Ventilator-associated pneumonia, MV Mechanical ventilation, RR Risk ratio, 95% CI 95% Confidence interval, n/N Number of participants/number of studies, NR Not reported, HME Heat and moisture exchanger, RCCT Randomized concurrent controlled trial, CRT Cluster randomized trial, MV Mechanical ventilation, TAP Topical antibiotic prophylaxis, PPAP Protocolized parenteral antibiotic prophylaxis
  2. *Paradox 1: median event rates for VAP and mortality among control groups of TAP (± PPAP)-RCCTs are generally high versus control groups of non-antimicrobial-RCCTs
  3. §Paradox 2: median event rates for VAP and mortality among intervention groups of TAP (± PPAP)-RCCTs are not unusually low versus either intervention groups of non-antimicrobial-RCCTs or versus intervention groups of anti-septic RCCTs
  4. †Paradox 3: median mortality event rates among chlorhexidine intervention groups are no higher than among TAP (± PPAP) intervention groups
  5. ‡Paradox 4: the mortality effect size for TAP (± PPAP)-RCCTs differs for studies with patients individually randomized versus cluster randomized, being equivalent to, respectively, a 15-percentage-point versus a zero-percentage-point difference between mortality in control and intervention groups
  6. aSemi-recumbent position (30° to 60° versus < 10°); Pneumonia is microbiologically confirmed VAP at > 48 h and mortality is ICU mortality at > 48 h
  7. bHME versus heated humidification: pneumonia measured at a median of 4 days (from Analysis 1.3 on page 65 of ref [12]) and mortality measured at a median of 8 days
  8. cProbiotic versus control: pneumonia is VAP measured at a median 37 days and mortality measured at a median of 35 days
  9. dProtocol-directed sedation management versus usual care for MV weaning
  10. eChlorhexidine (mouth rinse or gel) versus usual care: pneumonia is VAP measured at a median of 1 month and mortality measured at a median of 1 month
  11. fToothbrushing ± antiseptic: pneumonia is VAP measured at a mean of 1 month and mortality measured at a mean of 1 month
  12. gTAP + PPAP intervention is equivalent to SDD (selective digestive decontamination). Summary results for seven studies (1039 patients) with a duplex design (TAP + PPAP versus PPAP) were not shown
  13. hData is based on analysis of intention to treat data including outcomes for patients lost to follow-up through early mortality for eight studies
  14. iPneumonia is a respiratory tract infection and the mortality is at unspecified follow-up
  15. jData is based on the analysis of per-protocol data
  16. kTAP intervention is equivalent to SOD (selective oropharyngeal decontamination)