Many studies have examined the importance of preventing the transmission of infectious diseases in the school environment, one such studied completed by Cramer et al
. 1999, indicated this item to be of great concern for the parent's of school-aged children [47
]. The most common infections transmitted in school environments are respiratory (influenza, pharyngitis etc) and diarrheal illnesses (i.e., Norwalk virus). Most of the infections occur at a constant low level but occasionally outbreaks do occur resulting in increased absenteeism and involvement of public health authorities. Since hands are the primary mechanism of transmission of these illnesses, proper hand hygiene techniques have been endorsed as the first defence at reducing the risk of transmission [1
]. In health care settings, the routine use of antimicrobial alcohol based hand gels has been endorsed as an alternative to handwashing when hands are not visibly soiled [48
]. Effectiveness in the hospital setting has not been easy to document given the relative low incidence of documented infections that can be specifically related to nosocomial transmission relative to the high number of handwashing opportunities in specific environments such as intensive care unit settings.
Can the evidence from these six trials reported here be used to promote this type program in elementary schools at the present time? This systematic review of antimicrobial rinse-free hand sanitizers for prevention of illness-related absenteeism in elementary school children is the first review, of the author's awareness, to assess this issue. Although randomized controlled trials are the study design least likely to provide biased estimates of effect, due to the nature of school-based interventions, the inclusion of both randomized and non-randomized cluster controlled trials was required [51
]. Of the six studies that met our inclusion criteria, three were non-randomized cluster controlled trials. Recent evidence indicates that non-randomized designs overestimate the effect of an intervention, thus the magnitude of the results should be interpreted with caution [52
Four of the six studies used an alcohol-based product, the other two using a benzalkonium chloride based disinfectant. The FDA in the United States has indicated that insufficient data exits to classify the latter compounds as safe and effective to use as antiseptic handwashes. They are also adversely affected in the presence of organic material such as food residues, which may be an issue in schools [53
]. Four studies were industry sponsored, and five were flawed due to the lack of sample size calculations. The five studies included were of low quality and methodologically weak. The only blinded randomized study using a placebo incorporated in this review was reported to be randomized and double-blinded, however, a description of the randomization technique was not discussed in the report and allocation concealment was unclear [41
]. Additionally, this study suffered from a large proportion of withdrawals and drop-outs, thus the results had to be cautiously interpreted. Current studies have indicated that poor quality studies are associated with exaggerated treatment effects [52
]. Although all studies reported statistically significant effects of the antimicrobial rinse-free hand gel in the experimental group, the aforementioned evidence suggests the reader should interpret these results cautiously. Thus, a clear delineation of the effectiveness of the intervention cannot be resolved from this review.
Several limitations were encountered when completing this review, the major one being the scarcity of high quality studies. Additionally, although content experts, primary authors and industrial companies were contacted, no grey literature was found. The possible existence of unpublished non-significant trials should not be discounted. The validity of performing a quantitative synthesis was considered, however based on a qualitative inspection of heterogeneity and estimates of intervention effectiveness this was not deemed appropriate. Sources of heterogeneity included study designs, population characteristics, intervention characteristics, case definition and primary outcome measure. Thus, sensitivity and subgroup analyses were not performed, and publication bias was not assessed quantitatively. Another limitation was the fact that one reviewer was used to do the broad screen of articles and review the two citations identified between September 2003 and the present time. This may have biased the results; however, it is believed that this reviewer would overestimate the citations to be included.