Background
Health care associated infections among adults admitted to the intensive care, have been shown to increase length of stay, the cost of care, and in some cases an increased risk of hospital death [
1‐
4]. Daily bathing with chlorhexidine gluconate has been suggested as an effective intervention to reduce the risk of infection during an intensive care stay [
5‐
8]. However, the effectiveness of CHG bathing to reduce ICU infections has varied considerably among published trials, making the effectiveness of CHG bathing in ICU patients uncertain [
9], and possibly reliant on the underlying risk among the given ICU population [
9,
10]. Importantly, current opinion of the merit of CHG bathing to reduce hospital acquired infections, among adults admitted to intensive care, is divergent [
8,
9], suggesting a state of ‘clinical equipoise’ [
11].
A challenge when assessing the growing evidence of the effectiveness of an intervention, is that meta-analysis of accumulating data may obtain spurious statistical significance [
12‐
14]. This is thought to be due to an aggregate sample size from the accumulated published trials, lower than that expected to adequately assess effectiveness, may underestimate effect [
15]. In an attempt to overcome these potential pitfalls, a Trial Sequential Analysis (TSA) approach has been developed that attempts to address Type I and Type II error, with repeated significance testing of accumulating trial data [
16]. Therefore, this TSA meta-analysis was undertaken to summarize the current status of the evidence for the effectiveness of daily CHG bathing, among adult intensive care patients, to reduce various infections in the ICU; and, address the question - should we continue to attempt to assess effectiveness with further trials, or is current evidence adequate to recommend CHG bathing become common practice in the adult ICU?
Discussion
This trial sequential meta-analysis presents a summary of the current status of the estimated effectiveness of CHG-bathing to prevent infection among adults admitted to the intensive care. Routine bathing with CHG does not occur in the ICU setting, and TSA suggests that more trials are needed to address the current state of ‘clinical equipoise’. These future studies need to be conducted among a diverse group of ICU patients, including both surgical, medical and trauma patients if possible, to ensure generalisability of results to the majority of patients cared for in the ICU setting.
Previous reviews of daily CHG-bathing to reduce infections among adults admitted to the ICU have been undertaken [
7,
10,
25,
26], and suggest a benefit in CHG-bathing to reduce various hospital acquired infections. However, divergent results from two large, well planned and conducted, randomised cluster cross-over trials have not resulted in the widespread adoption of CHG bathing by nurses in the ICU. Specifically, the considerable variation in the baseline risk of infection in the ICU populations included in published trials, to date, has made generalisation difficult.
Our assessment of the current status of evidence of the effectiveness of CHG bathing to reduce the risk of infection in the intensive care suggests that more research in this area is needed. Specifically addressing two important short comings of current evidence, namely: (1) future trials need to include a diverse population of adults admitted to the ICU; and, (2) future trials need to be powered to add to the current cumulative data, to ensure clear statistical evidence of benefit, and avoid spurious results due to Type I error.
A potential strength of our meta-analysis is the use of TSA to avoid a spurious conclusion of the effectiveness of CHG bathing, and to address what needs to be done next due to the current ‘clinical equipoise’ of this intervention in the ICU setting. However, any systematic review and meta-analysis have a potential weakness of missing unpublished trials, and potential individual trial heterogeneity that is difficult to account for in analysis. We purposely omitted before–and-after trials that tend to overestimate effectiveness and contained significant within-trial-variation [
10], adding these trials would significantly increase the respective cumulative samples size, but would not add quality to any estimates of effect. A potential limitation of our meta-analysis is that mortality outcomes between CHG-bathing and comparison groups was only reported by two trials, and such an outcome has an important place in describing the ultimate burden of hospital acquired infection in the ICU setting. And, our meta-analysis lacks the data to answer the question – Do we need to bath all patients, or only those at the greatest risk, or already colonized? Further to this, the study by Boonyasrir et al [
27] didn’t use prepacked CHG impregnated clothes, instead CHG wash cloths were prepared by staff at the bedside, therefore the dose of CHG may have been unstandardized.
The lack of evidence for the benefit of daily CHG-bathing reported by Noto et al [
21] and positive results reported by Climo et al [
6], have been suggested to be due to a short ten-week intervention and control periods used in the study by Noto [
21], compared to the 6-months cross-over period used by Climo [
6]. The shorter ten-week cross-over period being considered to be insufficient to determine the true impact of CHG bathing on infection rates. Therefore, future trials should consider the optimal study design, and the optimal length of study intervention and control periods. The use of a randomised cluster cross-over design appears optimal and would be easily integrated into the current cumulative evidence. However, the contextual effect of the intervention period lowering the background rates of infection would need to be carefully considered, and for this reason a randomised stepped-wedge approach may also be an option [
28].
The results of our TSA meta-analysis have some important clinical implications for the wider ICU community of clinicians. Importantly, more research is need in this area, that specifically ensures trustworthy evidence of the effectiveness of daily CHG bathing to reduce infections among adults admitted to intensive care, and to ensure results are generalisable to a wider diverse population of ICU patients. The evidence needs to be of the highest quality, like any intervention, there are concerns regarding the safety of daily bathing of ICU patients with CHG, however our meta-analysis has not addressed this issue. Moreover, even though CHG bathing aims to reduce HAI, it has been suggested it may promote the emergence of chlorhexidine-resistance, and increase gram-negative organism bacteraemia [
9]. However, large studies have failed to support this hypothesis [
29]. Importantly, even a modest treatment effect should be considered in the context of the seriousness of some of these specific infections among ICU patients, that are costly and are potentially associated with increased patient morbidity and mortality.
Importantly, a previous met-analysis of CHG bathing by Afonso et al [
30] found that the effect of chlorhexidine gluconate-impregnated washcloth bathing may be unequal for Gram-positive BSIs versus Gram-negative BSIs and that this warrants further study.