Transmission and control of infections in out-of-home child care

Research on effective strategies for reducing infections associated with out-of-home child care has lagged far behind the research on other aspects of these infections. This review highlights key points regarding the epidemiology and transmission of infections associated with out-of-home day care and summarizes and critiques published reports of interventional studies designed to reduce the frequency of common respiratory and gastrointestinal infections. EPIDEMIOLOGY AND TRANSMISSION OF INFECTION There are four general patterns of disease caused by infections associated with out-of-home child care, based on their frequency and degree of morbidity 1 : (1) children are infected at somewhat higher rates than staff and family contacts, but the associated morbidity is comparable (examples: viral upper respiratory and gastrointestinal infections); (2) children are infected at a higher rate than staff and family contacts, usually because of the absence of protective antibody, or they suffer more associated morbidity (examples: varicella, Escherichia coli O157:H7 infection with hemolytic uremic syndrome); (3) children, staff and family contacts are infected at similar rates, but children suffer less associated morbidity (example: hepatitis A virus infection); (4) children, staff and family contacts are infected at comparable rates but with less associated morbidity than the fetuses of pregnant adult or immunocompromised contacts (examples: cytomegalovirus and parvovirus infection). Many factors influence the risk of infection in child care settings. 2, 3 Infection rates are generally higher at large day-care centers than at small ones. Young children (i.e. <24 months) are at higher risk for infection in part because of their age and in part because they are likely to be cared for with other young children in diapers. The risk of infection is greatest among young children in their first 2 years of attendance and decreases with longer durations of attendance. Other epidemiologic parameters that influence risk include the prevalence of infection within the community; specific characteristics of the facility such as class size, staffing ratios and infection control procedures and practices; and the presence of chronic diseases among attendees. Respiratory viruses and gastrointestinal viruses and protozoa cause the majority of endemic infections associated with out-of-home child care and are transmitted by direct or indirect contact with respiratory secretions or stool (Table 1). Viruses such as cytomegalovirus and herpes simplex virus are also common and are likewise spread by direct or indirect contact with saliva (both), urine (cytomegalovirus) or skin or mucus membrane lesions (herpes simplex virus). Transmission of respiratory viruses, such as influenza virus, is common during seasonal outbreaks and occurs by droplet contact (i.e. by large respiratory droplets that travel ≤3 feet). Airborne infections are uncommon but can cause significant outbreaks affecting many or all children in an individual facility. Infections caused by foodborne pathogens are spread by common sources (i.e. contaminated food or water) and are very uncommon but also may result in outbreaks of disease in individual facilities. Table 1: Modes of transmission of microorganisms in out-of-home child careSTRATEGIES TO PREVENT AND CONTROL THE SPREAD OF INFECTION The American Academy of Pediatrics (AAP) and the American Public Health Association published a comprehensive set of guidelines for out-of-home child care in 1992, which included detailed recommendations for preventing and controlling the spread of infection. 4 The 1997 AAP Red Book 5 contains a revised, condensed version of these recommendations, including the following key points: (1) staff and children should practice handwashing and good personal hygiene; (2) proper facilities and equipment should be available, and procedures should be followed for diaper changing and toileting; (3) compliance with age-appropriate vaccination requirements must be ensured; (4) specific criteria should be used to exclude ill children with potentially communicable diseases; (5) contaminated surfaces and toys should be disinfected regularly and there should be appropriate environmental sanitation; (6) food should be prepared and stored properly; (7) pets should be housed and cared for appropriately; (8) good lines of communication should be maintained with parents, qualified health consultants and public health authorities. STUDIES EVALUATING INTERVENTIONS TO PREVENT AND CONTROL THE SPREAD OF INFECTION A number of interventional studies have evaluated strategies to reduce the incidence of respiratory and gastrointestinal infections in out-of-home child care. 6–17 In general these studies have used robust study designs and observed large numbers of children in diverse locations (Tables 2 and 3), enhancing the validity and generalizability of their results. Only one study has been conducted in day-care homes 15; the remaining studies were in day-care centers. 6–14, 16, 17 This body of literature can be divided into two groups. Table 2: Studies examining specific interventional to prevent and control the spread of infection in out-of-home child careTable 3: Studies examining interventional programs to prevent and control the spread of infection in out-of-home child careThe first group of studies examined the efficacy of specific interventions (Table 2). 6–9 Black et al. 6 examined the efficacy of handwashing using bar soap and paper towels on the incidence of diarrhea. Staff in two experimental centers were instructed to wash their hands before food preparation and after diapering and to assist the children with handwashing before eating and after using the toilet. Senior staff and the investigators monitored compliance “rigorously,” although there was no objective measure of compliance. Staff in two control centers received no instruction. The centers reportedly were similar in terms of their design and handwashing facilities. The investigators found a significant decrease in diarrheal episodes after implementation of the handwashing program in the two intervention centers compared with the two control centers. Although this study was conducted in the late 1970s, no subsequent study has examined the efficacy of handwashing as a single intervention. The effect of a waterless, alcohol-based hand rinse has also not been evaluated as a single intervention. Two studies in this first group used a cross-over design to examine the effect of paper vs. cloth diapers on contamination of environmental surfaces with fecal coliforms (Table 2). 7–9 Holaday et al. 8, 9 also cultured the hands of staff and children. The two studies reached differing conclusions on the efficacy of paper diapers: Van et al. 7 found that paper diapers decreased fecal coliform contamination, whereas Holaday et al. 8, 9 found there was no effect. The reasons for this discrepancy are unclear, although the latter study notes that they used cloth diapers that more closely resembled the design of paper diapers in that they were “modern, front-closure, all-in-one” diapers. Neither study examined the effect of diaper types on the incidence of diarrhea. The second group of studies examined the cumulative effect of interventional programs consisting of training courses, implementation of a variety of specific procedures and other miscellaneous interventions (Table 3). 10–17 The training courses covered topics such as transmission of infections, handwashing, diapering, food preparation and environmental and toy cleaning and disinfection. A detailed description or a reference for the training course was provided by only two studies. 11–13 Some studies used a single, one-time training course, 10, 11 whereas others provided an initial course with repeat courses or periodic updates and in-services. 12–17 In a few studies centers were provided with supplies for handwashing, including a waterless alcohol-based hand rinse and other supplies. 14–16 Before discussing the results of these studies, it is important to highlight a number of methodologic concerns. First, although there is a scientific or common sense rationale for many of the interventions used by the studies, data supporting the efficacy of the interventions in the day-care setting are limited to those discussed previously (Table 2). Second, the definitions of the primary outcomes, usually upper respiratory tract infection and/or diarrhea, are quite subjective. For instance stool consistency and frequency were often defined in comparison with the norm for the individual child. Third, the methods for assessing these outcomes were variable and included daily records maintained by day-care center staff, periodic telephone surveys of parents or written questionnaires mailed to parents (Table 3). Although individual staff members or parents may have applied the definitions accurately to children they knew well, it is unlikely that different staff or parents applied them similarly. The studies did not examine interrater reliability in the application of these definitions, nor was the frequency of infections correlated with other objective measures of disease (i.e. criterion validity), such as the frequency of visits to pediatric health care providers or use of over-the-counter medications. Fourth, although most studies included some form of compliance monitoring, only three reported the results of objective measurements of compliance. 12–14, 17 Fifth, only three studies were adjusted for potential confounding variables in their analyses. 11–14 Sixth, the measurement (i.e. Hawthorne) effect on outcomes in control centers is likely to be significant, as was documented in one study. 11 Finally only two studies used study design and analytic techniques to adjust for the effect of clustering (i.e. a child is likely to be at higher risk for developing diarrhea if another child in the center has diarrhea). 11–13 With these caveats what can be said about the results of these studies? Most studies reported reduced rates of infection associated with the implementation of their intervention programs (Table 3). However, the nature and magnitude of the effects of the interventions reported were inconsistent. The most methodologically rigorous studies reported lower rates of infection, but the effect was not present for both respiratory and gastrointestinal infections or for all groups of children. 11–14 Studies examining the incidence of both upper respiratory tract infections and diarrhea reported mixed results: in two studies the incidence of upper respiratory tract infection alone was reduced 11–16; in two the incidence of diarrhea alone was reduced 14,15; and in two the incidence of both infections were reduced. 12, 13, 17 Studies by Roberts et al. 12, 13 reported a decrease in upper respiratory tract infections only among children younger than 24 months of age and a decrease in diarrhea only among children older than 24 months of age. Most disconcerting is the finding of Uhari and Mottonen 17 that there was little improvement in the knowledge base of staff regarding the causality and spread of infections, despite the fact that the rates of upper respiratory and gastrointestinal infections were reduced. Logic would predict that more intensive programs would be associated with a larger magnitude of effect, but this is difficult to determine because there were no direct comparisons of different interventional programs. Studies in which the most intensive programs were used reported more impressive results (Table 3) 16,17; however, the studies were not as strong methodologically as others reporting less impressive results. Finally it is important that none of the studies documented its time and resource costs explicitly. Two studies reported that their interventional programs were associated with fewer physician visits, antibiotic courses and absent days, 16, 17 but the cost savings associated with these benefits was not quantified. SUMMARY The epidemiology of infections associated with out-of-home child care is well-known, but our understanding of cost-effective prevention strategies is limited. Existing studies indicate that multidimensional interventional programs can reduce infection rates, but these conclusions are limited by a variety of methodologic concerns. Important areas for future research are to determine the critical elements of effective intervention programs, whether improvements in infection rates can be sustained over long periods and the costs of implementing and sustaining these programs.