Next Article in Journal
Fear of COVID-19, Stress, Fear of Childbirth, Hardiness and Life Satisfaction in Polish Pregnant Women During the Pandemic
Previous Article in Journal
Measuring What Matters in Trial Operations: Development and Validation of the Clinical Trial Site Performance Measure
Previous Article in Special Issue
Non-Invasive Ventilation: When, Where, How to Start, and How to Stop
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
JCM
Volume 14
Issue 19
10.3390/jcm14196841
jcm-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Testing the Stable Unit Treatment Variance Assumption (SUTVA) Within Cochrane Reviews of Antimicrobial-Based Pneumonia Prevention Interventions Among Mechanically Ventilated Patients Using Caterpillar Plots

by
James C. Hurley
1,2,3
1
Melbourne Medical School, University of Melbourne, Ballarat, VIC 3350, Australia
2
Ballarat Health Services, Grampians Health, Ballarat, VIC 3350, Australia
3
Ballarat Clinical School, Deakin University, Burwood, VIC 3125, Australia
J. Clin. Med. 2025, 14(19), 6841; https://doi.org/10.3390/jcm14196841 (registering DOI)
Submission received: 12 August 2025 / Revised: 23 September 2025 / Accepted: 25 September 2025 / Published: 26 September 2025
(This article belongs to the Special Issue Weaning from Non-Invasive Mechanical Ventilation Approaches (NIV-NHF): Insights on How, When and Where)
Browse Figures
Review Reports Versions Notes

Abstract

Background/Objectives: Causal inference requires validating the stable unit treatment variance assumption (SUTVA). Whilst antimicrobial-based interventions, being topical chlorhexidine and topical antibiotics prophylaxis (TAP), appear effective in preventing ventilator-associated pneumonia (VAP) among ICU patients receiving mechanical ventilation (MV) within randomized concurrent controlled trials (RCCT), SUTVA has never been tested for this inference. Methods: Caterpillar plots of the VAP incidence proportions for control and intervention groups within RCCTs obtained from recent Cochrane reviews of antimicrobial-based VAP prevention interventions were derived using random effects methods to enable comparison versus the expert VAP incidence reference range (5 to 40%). Results: The summary VAP prevention effect size derived from three categories of 57 RCCTs of antimicrobial-based interventions was recapitulated. The VAP incidences of 24 control groups and 6 intervention group incidences were above, whereas only 1 and 6, respectively, were below the expert VAP incidence reference range (5 to 40%) (p < 0.001; chi-square = 17.42; df = 2). The results after excluding 18 low-quality studies were similar. Paradoxically, the 90% prediction limits in association with the summary control group incidences are each approximately 20 percentage points wider than for those associated with the intervention group summaries. Conclusions: Control group VAP incidences above and more dispersed versus the expert opinion VAP range are common within the Cochrane reviews of antimicrobial-based VAP prevention interventions. Recognition requires an arms-based analysis using caterpillar plots. The SUTVAs that underlie the inference of prevention from the effect size estimates are not valid.

1. Introduction

Interventions based on the use of topical antimicrobials, being either topical antibiotic prophylaxis (TAP) or topical chlorhexidine, toward preventing ventilator-associated pneumonia (VAP) have been widely studied within randomized concurrent controlled trials over the past 40 years [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62].
The most recent systematic reviews of the randomized concurrent controlled trials of these interventions used as prevention within populations of patients receiving prolonged mechanical ventilation suggest that they appear highly effective. Moreover, the study data on which this summary VAP prevention evidence is based are generally rated as either ‘high quality’ or ‘high-certainty evidence’ for both types of antimicrobial intervention [59,62].
The Stable Unit Treatment Variable Assumption (SUTVA) is required for valid inference from effect size estimates derived from randomized concurrent controlled trials and, by extension, from the summary effect size estimates within systematic reviews [63,64]. The SUTVA has two elements, consistency and spillover. However, testing these elements is not as simple as it requires reference to an external benchmark [65]. Neither element is considered within the scoring of study quality. In any case, the SUTVA elements can generally not be tested within any given randomized concurrent controlled trial or systematic review.
Several published potential range estimates are available against which to benchmark the incidence of VAP amongst the population of ICU patients receiving prolonged mechanical ventilation in the literature. For example, the most recently published systematic review states this incidence to be 5 to 40% of patients receiving prolonged mechanical ventilation [62].
The objective here is to test the SUTVA amongst the findings of Cochrane reviews of antimicrobial-based VAP prevention interventions [58,59,60,61,62]. This requires an arms-based analysis of the VAP incidence proportions within the control and intervention groups of these randomized concurrent controlled trials versus the most recently stated expert range.

2. Materials and Methods

2.1. Literature Search and Study Decant

This analysis is based on a previous search of the Cochrane database for systematic reviews of interventions studied with randomized concurrent controlled trials for the prevention of VAP among mechanically ventilated patients in the ICU [66,67]. From this search, five Cochrane reviews of randomized concurrent controlled trials of topical antimicrobial-based interventions were identified for which VAP counts were abstracted. The VAP counts in previous versions of each systematic review were compared with the most recent [58,59,60,61,62]. The categorization of randomized concurrent controlled trials according to intervention follows that in the most recent Cochrane reviews [59,62]. Note that a duplex randomized concurrent controlled trial is one in which the control group received protocolized parenteral antibiotic prophylaxis (PPAP).
The following randomized concurrent controlled trials were excluded: randomized concurrent controlled trials with fewer than 20 patients in total, and randomized concurrent controlled trials for which only per-protocol data were available. This being a review of the SUTVA within the systematic reviews, a search for randomized concurrent controlled trials in addition to those included in the systematic reviews was not undertaken. Also, given there is no possibility for spillover to occur within randomized concurrent controlled trials with a non-concurrent design, any studies with a non-concurrent design were excluded.
Ethics committee review of this study was not required as this is an analysis of published work, and each individual included study had been approved by an ethics committee.
The data was abstracted as intention-to-treat (ITT) data. Several discrepancies were identified in the VAP counts abstracted within the two most recent systematic reviews of TAP, which appear to relate to whether the data was abstracted from the study publications as per-protocol (PP) data, as in the most recent TAP ± PPAP systematic review [62], versus intention to treat (ITT), obtained by personal communication from the study authors, as in earlier reviews of the TAP ± PPAP randomized concurrent controlled trials [60,61]. There were trivial discrepancies between the two most recent Cochrane reviews of topical Chlorhexidine, which both provide ITT data for all except one randomized concurrent controlled trial [58,59].
The study quality scoring for the various domains, including those such as sequence generation and allocation concealment (selection bias), blinding of participants and providers (performance bias), blinding of outcome assessor (detection bias), incomplete outcome data (attrition bias), and selective outcome reporting (reporting bias) as scored in each Cochrane review, was extracted. As these were scored using a variable number of score elements in each systematic review, here, these are harmonized by conversion to a simple majority or minority score of the various quality elements, as previously harmonized [67].

2.2. Contrast-Based Analysis-Method

The study-specific and overall summary VAP prevention effect sizes and associated 95% confidence intervals were calculated for each category using random-effect methods of meta-analysis. The data was set for two group comparisons of binary outcomes using the ‘meta’ command in Stata 18 (Stata Corp., College Station, TX, USA) using the default restricted maximum likelihood [68].

2.3. Arms-Based Analysis and Caterpillar Plots

The data from the component control and intervention groups of randomized concurrent controlled trials were decanted from each randomized concurrent controlled trial with care to include the groups from the three-arm studies only once. The summary VAP incidence proportion data was derived using the ‘meta’ command in Stata 18 and the default restricted maximum likelihood. Caterpillar plots, being a forest plot with the group ordered by increasing the VAP incidence proportion, were derived [69]. The concept of a caterpillar plot and the use of benchmarks are illustrated in Figure 1. The associated 95% confidence interval, I2 heterogeneity estimates, and 90% prediction limits were calculated using random-effect methods of meta-analysis.

3. Results

3.1. Characteristics of the Randomized Concurrent Controlled Trials

There were 57 published randomized concurrent controlled trials identified by the search (Table 1, Table 2 and Table 3). One randomized concurrent controlled trial with fewer than 20 patients in total [56], one study with a non-concurrent design [57], and one randomized concurrent controlled trial with only per-protocol data available [17] were excluded.
The randomized concurrent controlled trials were streamed into three categories.
Three randomized concurrent controlled trials provide data for a second intervention arm, and data missing for the third arm from two randomized concurrent controlled trials was recovered from the original sources.
Most randomized concurrent controlled trials originated from ICUs in Northern Europe (n = 21), Southern Europe (n = 13), North America (n = 4), or South America (n = 4) (Figure 2). There were twelve randomized concurrent controlled trials from trauma ICUs, one from a cardio-thoracic ICU, and two randomized concurrent controlled trials from pediatric ICUs.
The randomized concurrent controlled trials were published between 1987 and 2019, with the interquartile range (IQR) being 1998–2010. The TAP and TAP + PPAP randomized concurrent controlled trials were mostly published in the late 1990s, and the topical chlorhexidine randomized concurrent controlled trials were mostly published two or three decades later. A majority quality score rating was achieved for 16 of 18 randomized concurrent controlled trials of TAP + PPAP versus 28 of 39 randomized concurrent controlled trials in the other 2 randomized concurrent controlled trial categories.
Among the TAP duplex and TAP + PPAP categories, there were at least 14 different TAP components within the study intervention regimens. There are also at least 6 different PPAP components within the study intervention regimens.

3.2. Contrast-Based Analysis

The summary effect sizes implied all three randomized concurrent controlled trial categories of intervention prevented pneumonia, although there was inequality in the summary effect sizes, with the TAP + PPAP category showing the strongest apparent prevention effect size (Table 4). The associated I2 measures of heterogeneity were all in the range of 60 to 70%.

3.3. Arms-Based Analysis

There were three indicators that there was more dispersion amongst the VAP incidences of the control versus intervention groups. This is visually apparent among the caterpillar plots (Figure 3). Second, the increased dispersion is also apparent in the 90% prediction limits, which were approximately 20 percentage points wider, in association with the control group versus intervention group summary incidences within each of the three categories of randomized concurrent controlled trial. Thirdly, the I2 was up to 5 percentage points higher in association with the control group summary VAP incidences than the intervention group summary VAP incidences. The results after excluding 18 low-quality randomized concurrent controlled trials were similar.
The VAP incidences of 24 control groups and 6 intervention groups were above the expert VAP incidence reference range (5 to 40%), whereas only 6 of 56 intervention groups across the 3 categories were below 5% (p < 0.001; chi-square = 17.42; df = 2) (Figure 3).

4. Discussion

The Stable Unit Treatment Variable Assumption (SUTVA) is required for valid inference from effect size estimates derived from randomized concurrent controlled trials. By extension, SUTVA is also required for valid inference from systematic review summary effect size estimates. The SUTVA has two elements, consistency and spillover.
In general, systematic reviews include high-quality randomized concurrent controlled trials with the aim that the summary effect size estimates derived are robust and unconfounded. Of note, neither element of SUTVA is considered within the scoring of study quality. Moreover, testing these elements requires an arms-based analysis of event rates together with reference to benchmarks external to the specific randomized concurrent controlled trial or systematic review in question. Benchmarking is rarely undertaken within a systematic review, even though a benchmark range together with a citation source might be mentioned in the systematic review introduction. Most typically, the analysis is limited to a contrast-based analysis by which the effect size estimates are derived without an arms-based analysis of event rates.
Spillover effects have never been considered within Cochrane reviews of topical antimicrobial interventions for preventing infections within ICU populations. Valid inference from randomized concurrent controlled trials presumes there is no spillover. Given, on the one hand, these interventions appear highly effective in preventing both VAP and mortality, whereas on the other hand, patients acquire infections while in the ICU from the ICU environment, the possibility that these interventions might have spillover effects warrant careful consideration, especially given their candidature as population-based interventions, such as in the ‘pneumonia zero” initiative [70].
The use of caterpillar plots is a novel approach to examining the incidence of the event rate of interest amongst the respective control and intervention arms of a randomized concurrent controlled trial within a systematic review [69]. The caterpillar plot is essentially a forest plot with the studies arrayed in order of increasing VAP incidence, together with external benchmarks. This allows a visual inspection of the pattern and a more robust means of identifying potential outlier studies that might disproportionately influence the summary result. Of note, a typical systematic review will have the studies arrayed by alphabetical order of the first author’s name. This traditional array is wasteful of study information as it fails to adequately characterize dispersion in the visual display.

4.1. Choice of Benchmark

The benchmark used here is the VAP incidence range as appearing in the most recently published systematic review [62]. There are several incidence range estimates that were available for selection (Table 5), and these have been derived from various original sources. The one chosen for use here is neither overly conservative nor liberal with respect to VAP incidence limits and can be considered a consensus benchmark amongst those that have been provided within the systematic reviews that were the source for the randomized concurrent controlled trials used in this analysis.
Other benchmarks for VAP incidence that could have been used were substantially more conservative (i.e., narrower). For example, a widely cited narrative review for VAP [80] with over 4300 citations in Google Scholar [as of August 2025] gives an incidence range of 8 to 28%.
Another potential benchmark is one derived from an international survey of 1873 patients receiving mechanical ventilation for > 48 h in eleven countries. In this survey, the VAP incidence was 15.6% globally. Across sites in the United States, Latin America, the Asia-Pacific, and Europe, the incidences were 13.5%, 13.8%, 16%, and 19%, respectively [86]. Another potential benchmark is one derived from 109 randomized concurrent controlled trials of non-antimicrobial methods of VAP prevention. From these 109 randomized concurrent controlled trials, the summary control group incidence was 23% (95% CI 20–26%; 95% prediction interval 5.5–52%; I2 86.5%) and the summary intervention group incidence was 19% (95% CI 17–21%; 95% prediction interval 5.1–52%; I2 82.9%) [87].
Regardless of which benchmark range might have been chosen, surprisingly, the summary 90% confidence or prediction limits derived here for each of the three categories of intervention groups are each more closely aligned with these expert range incidences than the 90% confidence or prediction limits derived from the control groups’ summary incidence.
The vagaries in defining and documenting the VAP endpoint have previously drawn comment. VAP is considered a somewhat subjective endpoint with the potential to allow the introduction of bias in low-quality randomized concurrent controlled trials without observer blinding. The basis for this suspicion arises from a previous analysis of 32 TAP ± PPAP randomized concurrent controlled trials, which found that as the certainty of study blinding and other study design quality parameters increased, the effect size for VAP prevention decreased, and the variance of the apparent TAP ± PPAP effect size increased [88]. This observation may also have resulted in the choice [62] of per-protocol data as appearing in the published version of the original randomized concurrent controlled trials versus the ITT data as had appeared within earlier versions of Cochrane reviews of the TAP randomized concurrent controlled trials [60,61].

4.2. SUTVA: Consistency Element

The consistency element of SUTVA is that the intervention is consistently received by the individuals assigned to receive it. For example, this element would be violated if some recipients within the intervention group of a randomized concurrent controlled trial received a dose of the intervention that was double that received by other individuals in the intervention group. This would also be violated if some randomized concurrent controlled trials within a systematic review investigated a double dose in comparison to other randomized concurrent controlled trials. This element would also be violated if there was uneven spillover or contagion of the infection of interest to the control group from other individuals within the study population, or an uneven spillover more apparent for groups of some randomized concurrent controlled trials in comparison to others within a systematic review.
This element has not been specifically considered here. However, there are at least a dozen different TAP regimens among the TAP randomized concurrent controlled trials alone. Moreover, for the consistency element, this would need to consider in addition factors such as dosage, route, and timing of administration of these interventions. Moreover, the timing of cessation of these interventions, which is often poorly documented within these randomized concurrent controlled trials, is also a major consideration for their use within ICU populations, given the potential for rebound colonization following the cessation of antimicrobial decontamination.
These factors would suggest that the consistency element of the SUTVA is invalid here. Any consideration of consistency for the TAP randomized concurrent controlled trials would require an appeal to a presumption that the TAP regimens achieve a state of ‘colonization resistance’ that is thought to be associated with achieving a state of ‘selective digestive decontamination’ as originally conceived. This has never been demonstrated.
The consistency element in relation to the use of topical chlorhexidine is somewhat less problematic, but even within the topical chlorhexidine randomized concurrent controlled trials, there are applications of topical chlorhexidine as gel and solution applied with or without the concomitant use of toothbrushing.

4.3. SUTVA: Spillover Element

The spillover element of SUTVA, that the treatment received by one individual should not affect the event of interest in other individuals, was anticipated in two postulates stated in the first study [89], being postulated as “…having heavily contaminated patients next to decontaminated patients might adversely affect the potentially beneficial results [postulate one]. Secondly, a reduction of the number of contagious patients by applying [selective digestive decontamination] SDD in half of them, might reduce the acquisition, colonisation and infection incidence in the not-SDD-treated control group [postulate two]” [89]. That antimicrobial interventions might mediate a population level of infection prevention within the ICU through a mechanism that alters the ICU microbiome to achieve colonization resistance [90] has never been demonstrated, although given the herd protection that occurs with some population-based vaccination interventions, this population effect of antibiotics used as prevention might appear plausible. However, there have been no attempts to test either postulate in the subsequent four decades.
Whilst the observed event rates on which the summary effect size estimates were derived are likely to be robust within high-quality studies, it remains possible that the event rate for the whole randomized concurrent controlled trial population (i.e., concurrent control and intervention groups) may be elevated by the antimicrobial study intervention as a result of rebound and spillover. Hence, there may be dual effects operating, an individual-level effect and a population effect.

4.4. Limitations

The main limitation to be considered is the question of randomized concurrent controlled trial inclusions. This being a review of the SUTVA within the systematic reviews of antimicrobial-based interventions, a search for randomized concurrent controlled trials not included in the systematic reviews identified by the search was not undertaken. Moreover, the randomized concurrent controlled trials were identified for inclusion from across more than one systematic review in the case of each of the three categories of intervention to avoid limiting the focus to a single systematic review.
Four large cluster randomized trials of each of TAP [57,91,92] and topical chlorhexidine [93] used for the prevention of acquired infection within this patient population were not included here. Only one of these large cluster randomized trials [57] was abstracted in one of the systematic reviews, and, in any case, none provided data for the VAP endpoint. These four cluster randomized trials are notable for two reasons. Firstly, they each included between 3000 and 9000 patients from at least 13 ICUs in each cluster randomized trial, a considerably larger number of patients than the total number included in all of the randomized concurrent controlled trials within the systematic reviews. Second, their results for the mortality endpoint differ strikingly from the summary estimate derived for the randomized concurrent controlled trials. For the randomized concurrent controlled trials, there is a significant 15-percentage point ICU mortality difference between control versus TAP intervention groups derived from 27 randomized concurrent controlled trials, versus a zero-percentage point ICU mortality difference between control versus TAP intervention groups derived from 3 large cluster randomized trials [94]. This difference between the results of cluster randomized trials versus randomized concurrent controlled trials is contrary to the direction of difference that would be expected if spillover from the TAP intervention had a favorable population effect in preventing ICU mortality among patients concurrent within the same ICU, as originally postulated [89].

4.5. Strengths

The caterpillar plot is a novel and simple method to enable the incorporation of external data, which, here, is the VAP benchmarks (Table 5), into systematic reviews. These plots enable a visual assessment of the SUTVAs that are crucial to inferences arising from any analysis.
Of note, the summary effect size estimates obtained here for the three broad categories of intervention are comparable, allowing for differences in the numbers of randomized concurrent controlled trials included, the specific methods of analysis (e.g., Mantel–Haenszel versus restricted maximum likelihood) and summary statistics used (e.g., risk ratios versus odds ratios), versus the five respective systematic reviews [58,59,60,61,62] from which the randomized concurrent controlled trials are drawn from.
Through including a broad selection of randomized concurrent controlled trials from within several systematic reviews, it is possible to state that the unusually high VAP incidence among the three categories of randomized concurrent controlled trials is not attributable simply to the per-protocol nature of the data, as listed in the most recent TAP systematic review [62].

4.6. Comparison of Findings

The effect size estimates here are comparable to those estimated in the source systematic reviews [58,59,60,61,62], even though the inferences here diverge from those in the sources. With respect to chlorhexidine mouthrinse or gel, the first systematic review (18 randomized concurrent controlled trials; 2451 participants) estimated that these reduced the risk of VAP compared to placebo or usual care from 24% to about 18% (risk ratio 0.75, 95% confidence intervals (CI) 0.62 to 0.91, I2 = 35%) [58]. The second systematic review estimated (with moderate-certainty evidence; 13 randomized concurrent controlled trials; 1206 participants) that these reduced the incidence of VAP compared to placebo or usual care from 26% to about 18% (risk ratio 0.67, 95% CI 0.47 to 0.97; I2 = 66%) [59].
With respect to TAP, the third systematic review estimated with TAP combined with PPAP, a strong significant reduction in VAP (odds ratio 0.35; 95% CI 0.29 to 0.41; n = 16) and TAP (including duplex randomized concurrent controlled trials) also estimated a strong significant reduction in VAP (odds ratio 0.56; 95% CI 0.46 to 0.68; n = 17) [60].
The fourth systematic review estimated a significant reduction with respect to TAP combined with PPAP (odds ratio 0.28, 95% CI 0.20 to 0.38; n = 16) and with respect to TAP (including duplex randomized concurrent controlled trials) (odds ratio 0.44, 95% CI 0.31 to 0.63; n = 17) [61].
Finally, the fifth systematic review estimated that, with respect to Topical plus systemic antibiotic prophylaxis, a significant reduction in VAP (risk ratio 0.43, 95% CI 0.35 to 0.53; 17 randomized concurrent controlled trials; 2951 participants; moderate-certainty evidence) and with respect to TAP (including duplex randomized concurrent controlled trials), which may reduce VAP (risk ratio 0.57, 95% CI 0.44 to 0.74; 19 randomized concurrent controlled trials, 2698 participants; low-certainty evidence) [62].
Of particular note in relation to the fifth systematic review is that this review provided illustrative population projections derived (by an unstated method) from the control and intervention groups of the included randomized concurrent controlled trials. For TAP (including duplex randomized concurrent controlled trials), this was 318 per 1000 (i.e., 31.8%) and 137 per 1000 (i.e., 13.7%), respectively. For TAP plus systemic antibiotic prophylaxis, this was 417 per 1000 (i.e., 41.7%) and 238 per 1000 (i.e., 23.8%), respectively. These population projections correspond to the control and intervention group summary VAP incidences derived here.
There are other paradoxical results in relation to both the prevention effect size and the incidences of VAP and other acquired infections such as blood stream infections versus the group mean length of stay among the randomized concurrent controlled trials included within the Cochrane reviews of antimicrobial interventions in this patient population, which implicate spillover contributing to concurrent patients within the ICU and spuriously inflating the apparent effect size of topical antimicrobial based interventions beyond simply being a direct effect limited to only the recipients within the ICU context [95,96].
If either the direct effect of chlorhexidine or the spillover (indirect) effect of TAP was thought to be harmful, it would be ethically difficult to further study them, except as de-adoption studies. Moreover, the logistics of estimating spillover effects are challenging, given that this requires a cluster randomized trial design with thousands of exposed patients in multiple clusters [97].
The SUTVA requires careful consideration in relation to drawing inferences from randomized concurrent controlled trials. Surprisingly, this may have been overlooked by editors of respiratory, sleep, and critical care journals within their recent guidelines on causal inference. These guidelines do not mention or consider SUTVA.

5. Conclusions

The possibility that harmful spillover occurred within randomized concurrent controlled trials of topical antimicrobial-based interventions negates the SUTVA and undermines the certainty rating of the VAP data observed in these randomized concurrent controlled trials. The techniques used in systematic reviews typically estimate the intervention effect as an individual-level effect. The use of the caterpillar plot here enables a test of the population effects, which might invalidate the SUTVA.
This spillover from the use of topical chlorhexidine or TAP likely increased the pneumonia incidence above the expected range for the control group incidences among patients receiving mechanical ventilation in the ICU. This spillover, as a population effect, would conflate the summary effect size estimate for these interventions as derived within the systematic reviews. The current guidelines for VAP prevention will likely need to be reviewed, considering this conflation, especially as they relate to the ‘pneumonia zero” initiative.

Funding

This research has been supported by the Australian Government Department of Health and Ageing through the Rural Clinical Training and Support (RCTS) program.

Institutional Review Board Statement

Ethics committee review of this study was not required, as this is an analysis of published work, and each individual included study had been approved by an ethics committee.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data analyzed during the current study are provided in Table 1.

Conflicts of Interest

The author declares no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
ARDSAcute respiratory distress syndrome
ICUIntensive care unit
IQRInter-quartile range
VAPVentilator-associated pneumonia
MVMechanical ventilation
PIPrediction interval
SUTVAStable unit treatment variance assumption
TAPTopical antibiotic prophylaxis
PPAPProtocolized parenteral antibiotic prophylaxis

References

  1. Bellissimo-Rodrigues, F.; Bellissimo-Rodrigues, W.T.; Viana, J.M.; Teixeira, G.C.; Nicolini, E.; Auxiliadora-Martins, M.; Passos, A.D.C.; Martinez, E.Z.; Basile-Filho, A.; Martinez, R. Effectiveness of oral rinse with chlorhexidine in preventing nosocomial respiratory tract infections among intensive care unit patients. Infect. Cont. Hosp. Epidemiol. 2009, 30, 952–958. [Google Scholar] [CrossRef]
  2. Berry, A.M.; Davidson, P.M.; Masters, J.; Rolls, K.; Ollerton, R. Effects of three approaches to standardized oral hygiene to reduce bacterial colonization and ventilator associated pneumonia in mechanically ventilated patients: A randomised control trial. Intern. J. Nurs. Stud. 2011, 48, 681–688. [Google Scholar] [CrossRef]
  3. Ćabov, T.; Macan, D.; Husedžinović, I.; Škrlin-Šubić, J.; Bošnjak, D.; Šestan-Crnek, S.; Perić, B.; Kovač, Z.; Golubović, V. The impact of oral health and 0.2% chlorhexidine oral gel on the prevalence of nosocomial infections in surgical intensive-care patients: A randomized placebo-controlled study. Wien. Klin. Wochenschr. 2010, 122, 397–404. [Google Scholar] [CrossRef] [PubMed]
  4. Chen, Q.L.; Ye, X.F.; Jiang, Y.Z.; Yan, M.Q. Application of new oral care method to orotracheal intubation. Fujian Med. J. 2008, 30, 155–157. [Google Scholar]
  5. DeRiso, A.J.; Ladowski, J.S.; Dillon, T.A.; Justice, J.W.; Peterson, A.C. Chlorhexidine gluconate 0.12% oral rinse reduces the incidence of total nosocomial respiratory infection and nonprophylactic systemic antibiotic use in patients undergoing heart surgery. Chest 1996, 109, 1556–1561. [Google Scholar] [CrossRef]
  6. Fourrier, F.E.; Cau-Pottier, H.; Boutigny, M.; Roussel-Delvallez, M.; Jourdain Chopin, C. Effects of dental plaque antiseptic decontamination on bacterial colonization and nosocomial infections in critically ill patients. Intensive Care Med. 2000, 26, 1239–1247. [Google Scholar] [CrossRef] [PubMed]
  7. Fourrier, F.; Dubois, D.; Pronnier, P.; Herbecq, P.; Leroy, O.; Desmettre, T.; Roussel-Delvallez, M. Effect of gingival and dental plaque antiseptic decontamination on nosocomial infections acquired in the intensive care unit a double-blind placebo-controlled multicenter study. Crit. Care Med. 2005, 33, 1728–1735. [Google Scholar] [CrossRef] [PubMed]
  8. Fu, T.; Zhong, Q.; Zheng, C. Bacteriostasis effect of oral administration of chlorhexidine on patients with mechanical ventilation and prevention and treatment of ventilator—Associated pneumonia. Chin. Nurs. Res. 2019, 33, 431–434. [Google Scholar]
  9. Grap, M.J.; Munro, C.L.; Hamilton, V.A.; Elswick, R.K.J.r.; Sessler, C.N.; Ward, K.R. Early, single chlorhexidine application reduces ventilator-associated pneumonia in trauma patients. Heart Lung 2011, 40, e115–e122. [Google Scholar] [CrossRef]
  10. Jacomo, A.D.; Carmona, F.; Matsuno, A.K.; Manso, P.H.; Carlotti, A.P. Effect of oral hygiene with 0.12% chlorhexidine gluconate on the incidence of nosocomial pneumonia in children undergoing cardiac surgery. Infect. Cont. Hosp. Epidemiol. 2011, 32, 591–596. [Google Scholar] [CrossRef]
  11. Koeman, M.; van der Ven, A.J.; Hak, E.; Joore, H.C.; Kaasjager, K.; de Smet, A.G.; Ramsay, G.; Dormans, T.P.; Aarts, L.P.; de Bel, E.E.; et al. Oral decontamination with chlorhexidine reduces the incidence of ventilator-associated pneumonia. Am. J. Respir. Crit. Care Med. 2006, 173, 1348–1355. [Google Scholar] [CrossRef]
  12. Kusahara, D.M.; Peterlini, M.A.; Pedreira, M.L. Oral care with 0.12% chlorhexidine for the prevention of ventilator-associated pneumonia in critically ill children: Randomised, controlled and double blind trial. Intern. J. Nurs. Stud. 2012, 49, 1354–1363. [Google Scholar] [CrossRef]
  13. Meidani, M.; Khorvash, F.; Abbasi, S.; Cheshmavar, M.; Tavakoli, H. Oropharyngeal irrigation to prevent ventilator-associated pneumonia: Comparing potassium permangenate with chlorhexidine. Intern. J. Prev. Med. 2018, 9, 93. [Google Scholar]
  14. Meinberg, M.C.; Cheade Mde, F.; Miranda, A.L.; Fachini, M.M.; Lobo, S.M. The use of 2% chlorhexidine gel and toothbrushing for oral hygiene of patients receiving mechanical ventilation: Effects on ventilator-associated pneumonia [Uso de clorexidina 2% gel e escovacao mecanica na higiene bucal de pacientes sob ventilacao mecanica: Efeitos na pneumonia associada a ventilador]. Rev. Bras. Ter. Intensiv. 2012, 24, 369–374. [Google Scholar]
  15. Munro, C.L.; Grap, M.J.; Jones, D.J.; McClish, D.K.; Sessler, C.N. Chlorhexidine, toothbrushing, and preventing ventilator-associated pneumonia in critically ill adults. Am. J. Crit. Care 2009, 18, 428–437. [Google Scholar] [CrossRef] [PubMed]
  16. Özçaka, O.; Basoglu, O.K.; Buduneli, N.; Tasbakan, M.S.; Bacakoglu, F.; Kinane, D.F. Chlorhexidine decreases the risk of ventilator-associated pneumonia in intensive care unit patients: A randomized clinical trial. J. Periodontal Res. 2012, 47, 584–592. [Google Scholar] [CrossRef]
  17. Panchabhai, T.S.; Dangayach, N.S.; Krishnan, A.; Kothari, V.M.; Karnad, D.R. Oropharyngeal cleansing with 0.2% chlorhexidine for prevention of nosocomial pneumonia in critically ill patients: An open-label randomized trial with 0.01% potassium permanganate as control. Chest 2009, 135, 1150–1156. [Google Scholar] [CrossRef] [PubMed]
  18. Scannapieco, F.A.; Yu, J.; Raghavendran, K.; Vacanti, A.; Owens, S.I.; Wood, K.; Mylotte, J.M. A randomized trial of chlorhexidine gluconate on oral bacterial pathogens in mechanically ventilated patients. Crit. Care 2009, 13, R117. [Google Scholar] [CrossRef]
  19. Sebastian, M.R.; Lodha, R.; Kapil, A.; Kabra, S.K. Oral mucosal decontamination with chlorhexidine for the prevention of ventilator-associated pneumonia in children—A randomized, controlled trial. Pediatr. Crit. Care Med. 2012, 13, e305–e310. [Google Scholar] [CrossRef]
  20. Tantipong, H.; Morkchareonpong, C.; Jaiyindee, S.; Thamlikitkul, V. Randomized controlled trial and meta-analysis of oral decontamination with 2% chlorhexidine solution for the prevention of ventilator-associated pneumonia. Infect. Cont. Hosp. Epidemiol. 2008, 29, 131–136. [Google Scholar] [CrossRef]
  21. Bergmans, D.C.; Bonten, M.J.; Gaillard, C.A.; Paling, J.C.; van der Geeest, S.I.; van Tiel, F.H.; Beysens, A.J.; de Leeuw, P.W.; Stobberingh, E.E. Prevention of ventilator-associated pneumonia by oral decontamination: A prospective, randomized, double-blind, placebo-controlled study. Am. J. Respir. Crit. Care Med. 2001, 164, 382–388. [Google Scholar] [CrossRef] [PubMed]
  22. Brun-Buisson, C.; Legrand, P.; Rauss, A.; Richard, C.; Montravers, F.; Besbes, M.; Meakins, J.L.; Soussy, C.J.; Lemaire, F. Intestinal decontamination for control of nosocomial multiresistant Gram-negative bacilli. Ann. Intern. Med. 1989, 110, 873–881. [Google Scholar] [CrossRef] [PubMed]
  23. Camus, C.; Bellissant, E.; Sebille, V.; Perrotin, D.; Garo, B.; Legras, A.; Renault, A.; Le Corre, P.; Donnio, P.Y.; Gacouin, A. Prevention of acquired infections in intubated patients with the combination of two decontamination regimens. Crit. Care Med. 2005, 33, 307–314. [Google Scholar] [CrossRef]
  24. Gastinne, H.M.; Wolff, M.; Delatour, F.; Faurisson, F.; Chevret, S. A controlled trial in intensive care units of selective decontamination of the digestive tract with nonabsorbable antibiotics. N. Engl. J. Med. 1992, 326, 594–599. [Google Scholar] [CrossRef]
  25. Georges, B.; Mazerolles, M.; Decun, J.-F.; Rouge, P.; Pomies, S.; Cougot, P.; Andrieu, P. Décontamination digestive sélective résultats d’une étude chez le polytraumatisé. Réanimation Soins Intensifs Médecin D’Urgence 1994, 3, 621–627. [Google Scholar] [CrossRef]
  26. Korinek, A.M.; Laisne, M.J.; Nicolas, M.H.; Raskine, L.; Deroin, V.; Sanson-lepors, M.J. Selective decontamination of the digestive tract in neurosurgical intensive care unit patients: A double-blind, randomized, placebo-controlled study. Crit. Care Med. 1993, 21, 1466–1473. [Google Scholar] [CrossRef]
  27. Pneumatikos, I.; Koulouras, V.; Nathanail, C.; Goe, D.; Nakos, G. Selective decontamination of subgloottic area in mechanically ventilated patients with multiple trauma. Intensive Care Med. 2002, 28, 432–437. [Google Scholar] [CrossRef]
  28. Pugin, J.; Auckenthaler, R.; Lew, D.P.; Suter, P.M. Oropharyngeal decontamination decreases incidence of ventilator-associated pneumonia. JAMA 1991, 265, 2704–2710. [Google Scholar] [CrossRef]
  29. Quinio, B.; Albanese, J.; Bues-Charbit, M.; Viviand, X.; Martin, C. Selective decontamination of the digestive tract in multiple trauma patients. A prospective double-blind, randomized, placebo-controlled study. Chest 1996, 109, 765–772. [Google Scholar] [CrossRef]
  30. Rodríguez-Roldán, J.M.; Altuna-Cuesta, A.; López, A.; Carrillo, A.; Garcia, J.; León, J.; Martínez-Pellús, A.J. Prevention of nosocomial lung infection in ventilated patients: Use of an antimicrobial pharyngeal nonabsorbable paste. Crit. Care Med. 1990, 18, 1239–1242. [Google Scholar] [CrossRef] [PubMed]
  31. Unertl, K.; Ruckdeschel, G.; Selbmann, H.K.; Jensen, U.; Forst, H.; Lenhart, F.P.; Peter, K. Prevention of colonization and respiratory infections in long-term ventilated patients by local antimicrobial prophylaxis. Intensive Care Med. 1987, 13, 106–113. [Google Scholar] [CrossRef] [PubMed]
  32. Wiener, J.; Itokazu, G.; Nathan, C.; Kabins, S.A.; Weinstein, R.A. A randomized, double-blind, placebo-controlled trial of selective digestive decontamination in a medical-surgical intensive care unit. Clin. Infect. Dis. 1995, 20, 861–867. [Google Scholar] [CrossRef] [PubMed]
  33. Chaari, A.; Zribi, E.; Dammak, H.; Ghadoun, H.; Chtara, K.; Sfar, S.; Bahloul, M.; Bouaziz, M. Does selective digestive decontamination prevent ventilatorassociated pneumonia in trauma patients? Am. J. Ther. 2014, 21, 470–476. [Google Scholar] [CrossRef]
  34. Ferrer, M.; Torres, A.; Gonzalez, J.; Puig de la Bellacasa, J.; el-Ebiary, M.; Roca, M.; Gatell, J.M.; Rodriguez-Roisin, R. Utility of selective digestive decontamination in mechanically ventilated patients. Ann. Intern. Med. 1994, 120, 389–395. [Google Scholar] [CrossRef]
  35. Hammond, J.M.J.; Potgieter, P.D.; Saunders, G.L.; Forder, A.A. Double blind study of selective decontamination of the digestive tract in intensive care. Lancet 1992, 340, 5–9. [Google Scholar] [CrossRef]
  36. Laggner, A.N.; Tryba, M.; Georgopoulos, A.; Lenz, K.; Grimm, G.; Graninger, W.; Schneeweiss, B.; Druml, W. Oropharyngeal decontamination with gentamicin for long-term ventilated patients on stress ulcer prophylaxis with sucralfate? Wien. Klin. Wochenschr. 1994, 106, 15–19. [Google Scholar] [PubMed]
  37. Lingnau, W.; Berger, J.; Javorsky, F.; Lejeune, P.; Mutz, N.; Benzer, H. Selective intestinal decontamination in multiple trauma patients: Prospective.; controlled trials. J. Trauma 1997, 42, 687–694. [Google Scholar] [CrossRef]
  38. Stoutenbeek, C.P.; Van Saene, H.K.F.; Zandstra, D.F. Prevention of multiple organ failure by selective decontamination of the digestive tract in multiple trauma patients. In Immune Consequences of Trauma, Shock and Sepsis. Mechanisms and Therapeutic Approaches; Faist, E., Baue, A.E., Schildberg, F.W., Eds.; Pabst Science Publishers: Berlin, Germany, 1996; Volume 2, pp. 1055–1066. [Google Scholar]
  39. Abele-Horn, M.; Dauber, A.; Bauernfeind, A.; Russwurm, W.; Seyfarth-Metzger, I.; Gleich, P.; Ruckdeschel, G. Decrease in nosocomial pneumonia in ventilated patients by selective oropharyngeal decontamination (SOD). Intensive Care Med. 1997, 23, 187–195. [Google Scholar] [CrossRef] [PubMed]
  40. Aerdts, S.J.; van Dalen, R.; Clasener, H.A.; Festen, J.; van Lier, H.J.; Vollaard, E.J. Antibiotic prophylaxis of respiratory tract infection in mechanically ventilated patients. A prospective, blinded, randomized trial of the effect of a novel regimen. Chest 1991, 100, 783–791. [Google Scholar] [CrossRef]
  41. Blair, P.; Rowlands, B.J.; Lowry, K.; Webb, H.; Armstrong, P.; Smilie, J. Selective decontamination of the digestive tract: A stratified, randomized, prospective study in a mixed intensive care unit. Surgery 1991, 110, 303–309. [Google Scholar]
  42. Boland, J.P.; Sadler, D.L.; Stewart, W.; Wood, D.J.; Zerick, W.; Snodgrass, K.R. Reduction of nosocomial respiratory tract infections in the multiple trauma patients requiring mechanical ventilation by selective parenteral and enteral antisepsis regimen (SPEAR) in the intensive care. In Proceedings of the XVII Congress of Chemotherapy, Berlin, Germany, 23–28 June 1991. [Google Scholar]
  43. Cockerill, F.R.; Muller, S.R.; Anhalt, J.P.; Marsh, H.M.; Farnell, M.B.; Mucha, P.; Gillespie, D.J.; Ilstrup, D.M.; Larson-Keller, J.J.; Thompson, R.L. Prevention of infection in critically ill patients by selective decontamination of the digestive tract. Ann. Intern. Med. 1992, 117, 545–553. [Google Scholar] [CrossRef] [PubMed]
  44. de La Cal, M.A.; Cerdá, E.; Garcıa-Hierro, P.; Van Saene HKGómez-Santos, D.; Negro, E.; Lorente, J.A. Survival benefit in critically ill burned patients receiving selective decontamination of the digestive tract: A randomized, placebo controlled, double-blind trial. Ann. Surg. 2005, 241, 424–430. [Google Scholar] [CrossRef] [PubMed]
  45. Finch, R.G.; Tomlinson, P.; Holliday, M.; Sole, K.; Stack, C.; Rocker, G. Selective decontamination of the digestive tract (SDD) in the prevention of secondary sepsis in a medical/surgical intensive care unit. In Proceedings of the XVII International Congress Chemotherapy, Berlin, Germany, 23–28 June 1991. [Google Scholar]
  46. Jacobs, S.; Foweraker, J.E.; Roberts, S.E. Effectiveness of selective decontamination of the digestive tract (SDD) in an ICU with a policy encouraging a low gastric pH. Clin. Intensive Med. 1992, 3, 52–58. [Google Scholar]
  47. Kerver, A.J.H.; Rommes, J.H.; Mevissen-Verhage, E.A.E.; Hulstaert, P.F.; Vos, A.; Verhoef, J.; Wittebol, P. Prevention of colonization and infection in critically ill patients: A prospective randomized study. Crit. Care Med. 1988, 16, 1087. [Google Scholar] [CrossRef]
  48. Krueger, W.A.; Lenhart, F.P.; Neeser, G.; Ruckdeschel, G.; Schreckhase, H.; Eissner, H.J.; Forst, H.; Eckart, J.; Peter, K.; Unertl, K.E. Influence of combined intravenous and topical antibiotic prophylaxis on the incidence of infections, organ dysfunctions and mortality in critically ill surgical patients. Am. J. Respir. Crit. Care Med. 2002, 166, 1029–1037. [Google Scholar] [CrossRef]
  49. Palomar, M.; Alvarez-Lerma, F.; Jorda, R.; Bermejo, B.; Catalan Study Group of Nosocomial Pneumonia Prevention. Prevention of nosocomial infection in mechanically ventilated patients: Selective digestive decontamination versus sucralfate. Clin. Intensive Care 1997, 8, 228–235. [Google Scholar] [CrossRef]
  50. Rocha, L.A.; Martin, M.J.; Pita, S.; Paz, J.; Seco, C.; Margusino, L.; Villanueva, R.; Duran, M.T. Prevention of nosocomial infection in critically ill patients by selective decontamination of the digestive tract. A randomized, double blind, placebo-controlled study. Intensive Care Med. 1992, 18, 398–404. [Google Scholar] [CrossRef]
  51. Sanchez-Garcia, M.; Cambronero, J.A.; Lopez-Diaz, J.; Gomez Aguinaga, M.A.; Onoro Canaveral, J.J.; Sacristan del Castillo, J. Effectiveness and cost of selective decontamination of the digestive tract in critically ill intubated patients. A randomized, double-blind, placebo-controlled multicenter trial. Am. Rev. Respir. Dis. 1998, 158, 908–916. [Google Scholar] [CrossRef]
  52. Stoutenbeek, C.P.; van Saene, H.K.F.; Little, R.A.; Whitehead, A. The effect of selective decontamination of the digestive tract on mortality in multiple trauma patients: A multicenter randomized controlled trial. Intensive Care Med. 2007, 33, 261–270. [Google Scholar] [CrossRef]
  53. Ulrich, C.; Harinck-deWeerd, J.E.; Bakker, N.C.; Jacz, K.; Doornbos, L.; De Ridder, V.A. Selective decontamination of the digestive tract with norfloxacin in the prevention of ICU-acquired infections: A prospective randomized study. Intensive Care Med. 1989, 15, 424–431. [Google Scholar] [CrossRef]
  54. Verwaest, C.; Verhaegen, J.; Ferdinande, P.; Schetz, M.; Van den Berghe, G.; Verbist, L.; Lauwers, P. Randomized, controlled trial of selective digestive decontamination in 600 mechanically ventilated patients in a multidisciplinary intensive care unit. Crit. Care Med. 1997, 25, 63–71. [Google Scholar] [CrossRef] [PubMed]
  55. Winter, R.; Humphreys, H.; Pick, A.; MacGowan, A.P.; Willatts, S.M.; Speller, D.C. A controlled trial of selective decontamination of the digestive tract in intensive care and its effect on nosocomial infection. J. Antimicrob. Chemother. 1992, 30, 73–87. [Google Scholar] [CrossRef]
  56. Tuon, F.F.; Gavrilko, O.; Almeida, S.; Sumi, E.R.; Alberto, T.; Rocha, J.L.; Rosa, E.A. Prospective, randomised, controlled study evaluating early modification of oral microbiota following admission to the intensive care unit and oral hygiene with chlorhexidine. J. Glob. Antimicrob. Resist. 2017, 8, 159–163. [Google Scholar] [CrossRef]
  57. De Smet, A.M.; Kluytmans, J.A.; Cooper, B.S.; Mascini, E.M.; Benus, R.F.; Van der Werf, T.S.; Van der Hoeven, J.G.; Pickkers, P.; Bogaers-Hofman, D.; Van Der Meer, N.J.; et al. Decontamination of the digestive tract and oropharynx in ICU patients. N. Engl. J. Med. 2009, 360, 20–31. [Google Scholar] [CrossRef] [PubMed]
  58. Hua, F.; Xie, H.; Worthington, H.V.; Furness, S.; Zhang, Q.; Li, C. Oral hygiene care for critically ill patients to prevent ventilator-associated pneumonia. Cochrane Database Syst. Rev. 2016, 10, CD008367. [Google Scholar] [CrossRef]
  59. Zhao, T.; Wu, X.; Zhang, Q.; Li, C.; Worthington, H.V.; Hua, F. Oral hygiene care for critically ill patients to prevent ventilator-associated pneumonia. Cochrane Database Syst. Rev. 2020, 12, CD008367. [Google Scholar] [CrossRef]
  60. D’Amico, R.; Pifferi, S.; Leonetti, C.; Torri, V.; Tinazzi, A.; Liberati, A. Effectiveness of antibiotic prophylaxis in critically ill adult patients: Systematic review of randomised controlled trials. BMJ 1998, 316, 1275–1285. [Google Scholar] [CrossRef]
  61. Liberati, A.; D’Amico, R.; Pifferi, S.; Torri, V.; Brazzi, L.; Parmelli, E. Antibiotic prophylaxis to reduce respiratory tract infections and mortality in adults receiving intensive care (Review). Cochrane Database Syst. Rev. 2009, 2009, CD000022. [Google Scholar]
  62. Minozzi, S.; Pifferi, S.; Brazzi, L.; Pecoraro, V.; Montrucchio, G.; D’Amico, R. Topical antibiotic prophylaxis to reduce respiratory tract infections and mortality in adults receiving mechanical ventilation. Cochrane Database Syst. Rev. 2021, 1, CD000022. [Google Scholar] [CrossRef] [PubMed]
  63. Rubin, D.B. Comment: Which ifs have causal answers. J. Am. Stat. Ass. 1986, 81, 961–962. [Google Scholar] [CrossRef]
  64. Holland, P.W. Statistics and causal inference. J. Am. Stat. Ass. 1986, 81, 945–970. [Google Scholar] [CrossRef]
  65. Stone, R. The assumptions on which causal inferences rest. J. R. Stat. Soc. Ser. B Stat. Methodol. 1993, 55, 455–466. [Google Scholar] [CrossRef]
  66. Hurley, J. Spillover effects on mortality within randomized concurrent controlled trials of antimicrobial-based infection prevention interventions among the mechanically ventilated patient population. A reappraisal of Cochrane review data. J. Hosp. Infect. 2025, 163, 1–9. [Google Scholar] [CrossRef] [PubMed]
  67. Hurley, J.C. Impact of selective digestive decontamination on respiratory tract Candida among patients with suspected ventilator-associated pneumonia. A meta-analysis. Eur. J. Clin. Microbiol. Infect. Dis. 2016, 35, 1121–1135. [Google Scholar] [CrossRef]
  68. StataCorp. Stata 18 Meta-Analysis Reference Manual; Stata Press: College Station, TX, USA, 2023. [Google Scholar]
  69. Hurley, J.C. Forrest plots or caterpillar plots? J. Clin. Epidemiol. 2020, 121, 109–110. [Google Scholar] [CrossRef]
  70. Álvarez-Lerma, F.; Palomar-Martínez, M.; Sánchez-García, M.; Martínez-Alonso, M.; Álvarez-Rodríguez, J.; Lorente, L.; Arias-Rivera, S.; García, R.; Gordo, F.; Añón, J.M.; et al. Prevention of ventilator-associated pneumonia: The multimodal approach of the Spanish ICU “Pneumonia Zero” Program. Crit. Care Med. 2018, 46, 181–188. [Google Scholar] [CrossRef]
  71. Apostolopoulou, E.; Bakakos, P.; Katostaras, T.; Gregorakos, L. Incidence and risk factors for ventilator-associated pneumonia in 4 multidisciplinary intensive care units in Athens. Respir. Care 2003, 48, 681–688. [Google Scholar]
  72. Edwards, J.R.; Peterson, K.D.; Mu, Y.; Banerjee, S.; Allen-Bridson, K.; Morrell, G.; Dudeck, M.A.; Pollock, D.A.; Horan, T.C. National Healthcare Safety Network (NHSN) report: Data summary for 2006 through 2008, issued December 2009. Am. J. Infect. Control 2009, 37, 783–805. [Google Scholar] [CrossRef]
  73. Ding, C.; Zhang, Y.; Yang, Z.; Wang, J.; Jin, A.; Wang, W.; Chen, R.; Zhan, S. Incidence, temporal trend and factors associated with ventilatorassociated pneumonia in mainland China: A systematic review and meta-analysis. BMC Infect. Dis. 2017, 17, 468. [Google Scholar] [CrossRef]
  74. Li, Y.; Liu, C.; Xiao, W.; Song, T.; Wang, S. Incidence, risk factors, and outcomes of ventilator-associated pneumonia in traumatic brain injury: A meta-analysis. Neurocrit. Care 2020, 32, 272–285. [Google Scholar] [CrossRef]
  75. Fagon, J.Y.; Chastre, J.; Vuagnat, A.; Trouillet, J.L.; Novara, A.; Gibert, C. Nosocomial pneumonia and mortality among patients in Intensive care unit. JAMA 1996, 275, 866–869. [Google Scholar] [CrossRef] [PubMed]
  76. Chevret, S.; Hemmer, M.; Carlet, J.; Langer, M. Incidence and risk factors of pneumonia acquired in intensive care units. Results from a multicenter prospective study on 996 patients. European Cooperative Group on Nosocomial Pneumonia. JAMA 1996, 275, 866–869. [Google Scholar]
  77. American Thoracic Society, Infectious Diseases Society of America. Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare associated pneumonia. Am. J. Respir. Crit. Care Med. 2005, 171, 388–416. [Google Scholar] [CrossRef] [PubMed]
  78. Reignier, J.; Mercier, E.; Le Gouge, A.; Boulain, T.; Desachy, A.; Bellec, F.; Clavel, M.; Frat, J.-P.; Plantefeve, G.; Quenot, J.-P.; et al. Effect of not monitoring residual gastric volume on risk of ventilatorassociated pneumonia in adults receiving mechanical ventilation and early enteral feeding: A randomized controlled trial. JAMA 2013, 309, 249–256. [Google Scholar] [CrossRef]
  79. Seguin, P.; Laviolle, B.; Dahyot-Fizelier, C.; Dumont, R.; Veber, B.; Gergaud, S.; Asehnoune, K.; Mimoz, O.; Donnio, P.-Y.; Bellissant, E.; et al. Effect of oropharyngeal povidoneiodine preventive oral care on ventilator-associated pneumonia in severely brain-injured or cerebral hemorrhage patients: A multicenter, randomized controlled trial. Crit. Care Med. 2014, 42, 1–8. [Google Scholar] [CrossRef]
  80. Chastre, J.; Fagon, J.Y. Ventilator-associated pneumonia. Am. J. Respir. Crit. Care Med. 2002, 165, 867–903. [Google Scholar] [CrossRef]
  81. Haley, R.W.; Hooton, T.M.; Culver, D.H.; Stanley, R.C.; Emori, T.G.; Hardison, C.D.; Quade, D.; Shachtman, R.H.; Schaberg, D.R.; Shah, B.V.; et al. Nosocomial infections in US hospitals, 1975–1976: Estimated frequency by selected characteristics of patients. Am. J. Med. 1981, 70, 947–959. [Google Scholar] [CrossRef]
  82. Pennington, J.E. Nosocomial respiratory infection. In Principles and Practice of Infectious Diseases; Mandell, G.L., Douglas, R.G., Jr., Bennet, J.E., Eds.; Churchill Livingstone: St. Louis, MO, USA, 1990; pp. 2199–2205. [Google Scholar]
  83. Chastre, J.; Fagon, J.Y. Pneumonia in the ventilator-dependent patient. In Principles and Practice of Mechanical Ventilation; Tobin, M.J., Ed.; McGraw-Hill: New York, NY, USA, 1994; pp. 857–890. [Google Scholar]
  84. Centers for Disease Control and Prevention. Monitoring hospital- acquired infections to promote patient safety: United States, 1990–1999. MMWR Morb. Mortal. Wkly. Rep. 2000, 49, 149–153. [Google Scholar]
  85. National Nosocomial Infections Surveillance (NNIS) System. National Nosocomial Infections Surveillance (NNIS) System report, data summary from January 1990–May 1999, issued June 1999. Am. J. Infect. Control 1999, 27, 520–532. [Google Scholar]
  86. Kollef, M.H.; Chastre, J.; Fagon, J.Y.; François, B.; Niederman, M.S.; Rello, J.; Torres, A.; Vincent, J.L.; Wunderink, R.G.; Go, K.W.; et al. Global prospective epidemiologic and surveillance study of ventilator-associated pneumonia due to Pseudomonas aeruginosa. Crit. Care Med. 2014, 42, 2178–2187. [Google Scholar] [CrossRef] [PubMed]
  87. Hurley, J.C. Visualizing and diagnosing spillover within randomized concurrent controlled trials through the application of diagnostic test assessment methods. BMC Med. Res. Methodol. 2024, 24, 182. [Google Scholar] [CrossRef]
  88. Van Nieuwenhoven, C.A.; Buskens, E.; Van Tiel, F.H.; Bonten, M.J. Relationship between methodological trial quality and the effects of selective digestive decontamination on pneumonia and mortality in critically ill patients. JAMA 2001, 286, 335–340. [Google Scholar] [CrossRef] [PubMed]
  89. Stoutenbeek, C.P.; Van Saene, H.K.; Miranda, D.R.; Zandstra, D.F. The effect of selective decontamination of the digestive tract on colonisation and infection rate in multiple trauma patients. Intensive Care Med. 1984, 10, 185–192. [Google Scholar] [CrossRef] [PubMed]
  90. Vollaard, E.J.; Clasener, H.A. Colonization resistance. Antimicrob. Agents Chemother. 1994, 38, 409–414. [Google Scholar] [CrossRef] [PubMed]
  91. The SuDDICU Investigators for the Australian and New Zealand Intensive Care Society Clinical Trials Group. Effect of selective decontamination of the digestive tract on hospital mortality in critically ill patients receiving mechanical ventilation: A randomized clinical trial. JAMA 2022, 328, 1911–1921. [Google Scholar] [CrossRef]
  92. Wittekamp, B.H.; Plantinga, N.L.; Cooper, B.S.; Lopez-Contreras, J.; Coll, P.; Mancebo, J.; Wise, M.P.; Morgan, M.P.; Depuydt, P.; Boelens, J.; et al. Decontamination strategies and bloodstream infections with antibiotic-resistant microorganisms in ventilated patients: A randomized clinical trial. JAMA 2018, 320, 2087–2098. [Google Scholar] [CrossRef]
  93. Dale, C.M.; Rose, L.; Carbone, S.; Pinto, R.; Smith, O.M.; Burry, L.; Fan, E.; Amaral, A.C.; McCredie, V.A.; Scales, D.C.; et al. Effect of oral chlorhexidine de-adoption and implementation of an oral care bundle on mortality for mechanically ventilated patients in the intensive care unit (CHORAL): A multi-center stepped wedge cluster-randomized controlled trial. Intensive Care Med. 2021, 47, 1295–1302. [Google Scholar] [CrossRef]
  94. Hammond, N.E.; Myburgh, J.; Seppelt, I.; Garside, T.; Vlok, R.; Mahendran, S.; Adigbli, D.; Finfer, S.; Gao, Y.; Goodman, F.; et al. Association between selective decontamination of the digestive tract and in-hospital mortality in intensive care unit patients receiving mechanical ventilation: A systematic review and meta-analysis. JAMA 2022, 328, 1922–1934. [Google Scholar] [CrossRef]
  95. Hurley, J. Estimating the herd effects of anti-microbial-based decontamination (ABD) interventions on intensive care unit (ICU) acquired bloodstream infections: A deductive meta-analysis. BMJ Open 2024, 14, e092030. [Google Scholar] [CrossRef]
  96. Hurley, J.C. Estimating the herd effects of antimicrobial prevention interventions on ventilator-associated pneumonia within ICU populations: A cluster randomized trial emulation using data from Cochrane reviews. J. Antimicrob. Chemother. 2025, 80, 1047–1058. [Google Scholar] [CrossRef]
  97. Hurley, J.C. Establishing the safety of Selective Digestive Decontamination within the ICU population. A bridge too far? Trials 2023, 24, 337. [Google Scholar] [CrossRef] [PubMed]
Jcm 14 06841 g001
Figure 1. Illustration of a forest plot (a) versus caterpillar plot (b) with study-specific effect sizes (blue symbols) with 95% confidence limits (blue lines) for individual studies and summary estimates with 95% confidence limits (green diamonds). In the forest plot, the vertical line indicates the summary effect size derived from the included studies, whereas in the caterpillar plot, the vertical lines indicate the benchmark range for the effect size as derived from external sources. In the caterpillar plot analyses here, the effect size is the VAP incidence proportion derived across all control and intervention groups, respectively.
Figure 1. Illustration of a forest plot (a) versus caterpillar plot (b) with study-specific effect sizes (blue symbols) with 95% confidence limits (blue lines) for individual studies and summary estimates with 95% confidence limits (green diamonds). In the forest plot, the vertical line indicates the summary effect size derived from the included studies, whereas in the caterpillar plot, the vertical lines indicate the benchmark range for the effect size as derived from external sources. In the caterpillar plot analyses here, the effect size is the VAP incidence proportion derived across all control and intervention groups, respectively.
Jcm 14 06841 g001
Jcm 14 06841 g002
Figure 2. Origins of randomized concurrent controlled trials of topical chlorhexidine or TAP included here.
Figure 2. Origins of randomized concurrent controlled trials of topical chlorhexidine or TAP included here.
Jcm 14 06841 g002
Jcm 14 06841 g003
Figure 3. Caterpillar plot with group specific incidence proportions (blue symbols) with 95% confidence limits (blue lines) for individual component groups and summary estimates with 95% confidence limits (red diamonds) of VAP incidence among control (a) and intervention (b) groups from antimicrobial-based pneumonia prevention randomized concurrent controlled trials drawn from five Cochrane reviews [58,59,60,61,62]. The TAP category includes the duplex randomized concurrent controlled trials. The expert range VAP incidence (5 to 40%) is indicated by the vertical red lines.
Figure 3. Caterpillar plot with group specific incidence proportions (blue symbols) with 95% confidence limits (blue lines) for individual component groups and summary estimates with 95% confidence limits (red diamonds) of VAP incidence among control (a) and intervention (b) groups from antimicrobial-based pneumonia prevention randomized concurrent controlled trials drawn from five Cochrane reviews [58,59,60,61,62]. The TAP category includes the duplex randomized concurrent controlled trials. The expert range VAP incidence (5 to 40%) is indicated by the vertical red lines.
Jcm 14 06841 g003
Table 1. Data extracted from randomized concurrent controlled trials of topical chlorhexidine.
Table 1. Data extracted from randomized concurrent controlled trials of topical chlorhexidine.
Patients (n) a
InterventionControl
AuthorYearRef.QS bStudy Intervention cnNnN
Chlorhexidine studies
Bellissimo-Rodrigues2009[1]2Chlx solution16641769
Berry2011[2]2Chlx solution + TB471178
Cabov2010[3]2Chlx gel117623
Chen2008[4]1Chlx solution16602860
De Riso d1996[5]2Chlx solution517317180
Fourrier2000[6]2Chlx gel5301428
Fourrier2005[7]2Chlx gel1311412114
Fu e2019[8]1Chlx solution7403740
Grap2011[9]1Chlx solution7211018
Jacomo f2011[10]2Chlx solution16871173
Koeman-Ch2006[11]2Chlx gel1312723130
Kusahara f2012[12]2Chlx gel + TB15461650
Meidani e2018[13]1Chlx solution6501550
Meinberg g2012[14]2Chlx gel + TB18281124
Munro2009[15]1Chlx solution ± TB389255100
Özçaka2012[16]2Chlx solution12292232
Panchabhai h2009[17]1Chlx solution (+PP)14881583
Scannapieco2009[18]2Chlx solution + TB14971249
Sebastian2012[19]2Chlx gel12411445
Tantipong2008[20]1Chlx solution + TB5581052
Abbreviations: QS = quality score; Chlx = chlorhexidine; TB = toothbrushing; PP = potassium permanganate. (a) All data are derived as intention-to-treat (ITT) except for Panchabhai [17], which is per-protocol data. (b) QS = quality score is derived as meeting the majority of quality score criteria in the original Cochrane review. (c) Study Intervention: Chlx = chlorhexidine; TB = toothbrushing; (+PP = potassium permanganate to control group). (d) De Rios [5] is a study from a cardiac ICU. (e) Data for the Fu [8], Meidiani [13] studies derived from Zhao 2020 [59]; all other data from Hua [58]. (f). Jacomo [10] & Kushara [12] are studies from pediatric ICUs. (g) Meinberg [14] is from a trauma ICU. (h) Panchabhai 2009 [17]; reclassified in Zhao’s 2020 systematic review as control group received oral care with potassium permanganate [59]. Only per-protocol data is available for those patients who received the complete protocol. This study is excluded from the current analysis.
Table 2. Data extracted from randomized concurrent controlled trials of TAP.
Table 2. Data extracted from randomized concurrent controlled trials of TAP.
Patients (n) a
InterventionControl
AuthorYearRef.QS bStudy Intervention cnNnN
TAP studies
Bergmans2001[21]2PGV98738139
Brun-Buisson1989[22]2PNeNa365668
Camus2005[23]1PT5313053126
Gastinne1992[24]1PTA2622033225
Georges d1994[25]2PNeA4311533
Koeman-TAP e2006[11]2PChlx1612823130
Korinek1993[26]2PTAV20963795
Pneumatikos d2002[27]1PTA5311630
Pugin1991[28]2PNeV4382441
Quinio d1995[29]2PGA19763873
Rodriguez-Roldan1990[30]1PTNeA1141117
Unertl1987[31]2PGA119920
Wiener1995[32]2PGNy830831
Duplex TAP studies
Chaari d,e2014[33]2 1031613
Ferrer1994[34]2PTA-Ctx + -Ctx7511150
Hammond1992[35]2PTA-Ctx + -Ctx2516230160
Laggner1994[36]1GA + Aug133434
Lingnau d,f1997[37]2PTA-Cip + -Cip389071177
Lingnau d,f1997[37]2PCipA + -Cip349071177
Stoutenbeek d1996[38]2PTA-C + -C249842
Abbreviations: QS = quality score; (a) All data are derived as intention-to-treat (ITT). (b) QS = quality score is derived as meeting the majority of quality score criteria in the original Cochrane review. (c) Study Intervention: P = polymyxin; T = tobramycin; A = amphotericin; Ne = neomycin, Na = nalidixic acid; Ny = nystatin; no = Norfloxacin; V = vancomycin; G = Gentamicin; Cip = ciprofloxacin; O = Ofloxacin; C = Cefotaxime; Ctx = Ceftriaxone; Ctz = Ceftazidime; Tr = Trimethoprim; Aug = aminopenicillin and clavulanic acid or ‘other appropriate regimens’. (d) Georges [25], Pneumatikos [27], Quinio [29], Chaari [33], Lingau [37], Stoutenbeek 1996 [38] are studies from trauma ICUs. (e) Data for the Koeman-TAP [11] and Chaari [33] studies derived from Minozzi [62]; all other data from Liberati [61]. (f) Lingau [37] provides a three-arm study with the data abstracted from the original source.
Table 3. Data extracted from randomized concurrent controlled trials of TAP + PPAP.
Table 3. Data extracted from randomized concurrent controlled trials of TAP + PPAP.
Patients (n) a
InterventionControl
AuthorYearRef.QS bStudy Intervention cnNnN
TAP + PPAP studies
Abele-Horn d1997[39]2PTA-C13582330
Aerdts1991[40]2PNoA-Ctx1282960
Blair1991[41]1PTA-Ctx1216138170
Boland d1991[42]2PTNy-Ctx14321732
Cockerill1992[43]2NyPG-Ctx4751275
de la Cal d,e2005[44]2PTA-C18532654
Finch1991[45]2PGA-C420724
Jacobs1992[46]2PTA-Ctx045446
Kerver1988[47]2PTA-Ctx5493147
Krueger2002[48]2PG-Cip91265149262
Palomar1997[49]2PTA-Ctx10502549
Rocha d1992[50]2PTA-Ctx7472554
Sanchez-Garcia1998[51]2PTA-Ctx3213160140
Stoutenbeek d2007[52]2PTA-Ctx62201100200
Ulrich1989[53]1PNoA-Tr7552657
Verwaest f1997[54]2PTA-C2219340185
Verwaest f1997[54]2OA-O3120040185
Winter1992[55]2PTA-Ctz3911792
Abbreviations: QS = quality score; (a) All data are derived as intention-to-treat (ITT). (b) QS = quality score is derived as meeting the majority of quality score criteria in the original Cochrane review. (c) Study Intervention: P = polymyxin; T = tobramycin; A = amphotericin; Ne = neomycin, Na = nalidixic acid; Ny = nystatin; no = Norfloxacin; V = vancomycin; G = Gentamicin; Cip = ciprofloxacin; O = Ofloxacin; C = Cefotaxime; Ctx = Ceftriaxone; Ctz = Ceftazidime; Tr = Trimethoprim; Aug = aminopenicillin and clavulanic acid or ‘other appropriate regimens’. (d) Abele-Horn [39], Boland [42], De la Cal [44], Rocha [50], and Stoutenbeek 1996 [52] are studies from trauma ICUs. (e) Data for the De la Cal [44] study derived from Minozzi [62]; all other data from Liberati [61]. (f) Verwaest [54] is a three-arm study with the data abstracted from the original source.
Table 4. Characteristics of studies.
Table 4. Characteristics of studies.
ChlorhexidineTAP/Duplex aTAP + PPAP
Number of studies b192018
Year (median)201019941992
     range (min–max)1996–20191987–20141988–2007
     IQR c2008–20121992–20011991–1998
Majority quality score d131516
Study results
     VAP prevention effect size0.530.460.30
     95% CI (n)0.35–0.80.33–0.640.22–0.42
     I270.766.765.2
Control group results
Group VAP—incidence
     • >40%7811
     • ≥5% to ≤40%11116
     • <5%100
     Total191917
Group size
     Median505057
     IQR32–7831–13047–140
VAP incidence
     Mean313240
     95% CI e21–44 (19)24–4030–50
     I292.788.792.4
     90% PI f6–7811–6413–74
Intervention group results
Group VAP—incidence
     • >40%321
     • ≥5% to ≤40%151614
     • <5%123
     Total192018
Group size
     Median585856
     IQR30–9231–9347–161
VAP incidence
     Mean191716
     95% CI e13–2712–2311–22
     I286.784.388.4
     90% PI f5–535–435–44
(a) A duplex study is defined as a randomized concurrent controlled trial where the control group patients received PPAP. (b) Number of eligible studies listed in the original Cochrane review. Note, several randomized concurrent controlled trials had several control or intervention groups. Hence, the number of groups does not equal the number of randomized concurrent controlled trials. Two effect sizes originate from two references each. (c) Data is median and inter-quartile range (IQR). (d) Majority quality score derived as meeting the majority of quality score criteria in the original Cochrane review. (e) 95% CI is 95% confidence interval. (f) 90% PI is 90% prediction interval.
Table 5. VAP benchmarks.
Table 5. VAP benchmarks.
Citation in SourceRef.Original Source
“…with a reported prevalence ranging between 6% and 52%…”[58], p. 6[71,72]
“…VAP is a relatively common nosocomial infection in critically ill patients, with pooled incidence from 23.8% to 36.0% in recent systematic reviews…”[59], p. 9[73,74]
“…incidence of pneumonia in such patients ranges between 7% and 40%…”[60], p. 1276[75]
“The incidence of pneumonia has been reported to vary from 7% to more than 40% in ICU patients”[61], p. 2[76,77,78]
“…ventilator-associated pneumonia (VAP) has been estimated to affect 5% to 40% of patients treated with mechanical ventilation for at least 48 h.”[62], p. 8[79,80,81]
“…ventilator-associated pneumonia (VAP) continues to complicate the course of 8 to 28%…”[80], p. 867 (and abstract)[81,82,83,84,85]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Hurley, J.C. Testing the Stable Unit Treatment Variance Assumption (SUTVA) Within Cochrane Reviews of Antimicrobial-Based Pneumonia Prevention Interventions Among Mechanically Ventilated Patients Using Caterpillar Plots. J. Clin. Med. 2025, 14, 6841. https://doi.org/10.3390/jcm14196841

AMA Style

Hurley JC. Testing the Stable Unit Treatment Variance Assumption (SUTVA) Within Cochrane Reviews of Antimicrobial-Based Pneumonia Prevention Interventions Among Mechanically Ventilated Patients Using Caterpillar Plots. Journal of Clinical Medicine. 2025; 14(19):6841. https://doi.org/10.3390/jcm14196841

Chicago/Turabian Style

Hurley, James C. 2025. "Testing the Stable Unit Treatment Variance Assumption (SUTVA) Within Cochrane Reviews of Antimicrobial-Based Pneumonia Prevention Interventions Among Mechanically Ventilated Patients Using Caterpillar Plots" Journal of Clinical Medicine 14, no. 19: 6841. https://doi.org/10.3390/jcm14196841

APA Style

Hurley, J. C. (2025). Testing the Stable Unit Treatment Variance Assumption (SUTVA) Within Cochrane Reviews of Antimicrobial-Based Pneumonia Prevention Interventions Among Mechanically Ventilated Patients Using Caterpillar Plots. Journal of Clinical Medicine, 14(19), 6841. https://doi.org/10.3390/jcm14196841

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop