Infections acquired by patients in hospitals are a major cause of illness and death worldwide, and particularly so in hospitals with very limited resources. Simple hygiene measures, including improved hand hygiene (particularly with increased use of alcohol-based hand rub) by healthcare workers (HCWs) in physical contact with patients, are thought to be amongst the most effective ways of reducing such infections. However, in most hospitals hand hygiene is practiced poorly and improving hand hygiene behavior amongst hospital staff has been found to be difficult.

This study aims to evaluate the impact of a multimodal intervention to improve hand hygiene amongst health care workers in accordance with the WHO Guidelines on Hand Hygiene in Health Care. The research design is a prospective stepped wedged trial using all in-patient wards in a 1,000 bed hospital located in Northeast Thailand (Sappasitthiprasong hospital, Ubon Ratchatani). A stepped wedge design is a type of cluster randomized controlled trial which is appropriate when there are prior reasons to believe the intervention will be beneficial (as opposed to equipoise) and when it is impractical to deliver the intervention to all study units simultaneously. Both conditions hold here. The intervention will include educational visits to healthcare workers, improved reminders in the workplace, audit and feedback and social marketing with the aim of inducing behavioural and cultural change in relation to hand hygiene. The intervention will be delivered by the infection control team and the infection control ward nurses (ICWNs) who will receive additional training. The primary outcome will be directly observed hand hygiene compliance. Secondary outcomes will include incidence density of hospital-associated bloodstream and urinary tract infections with specified multiply antibiotic resistant bacteria, incidence density of hospital-associated bloodstream and urinary tract infections with non-multiply antibiotic resistant bacteria, total patient mortality rates, and ward-based consumption of alcohol hand rub and soap (a proxy for hand hygiene compliance).

Results from this study will be immediately generalisable to other resource-constrained hospitals in Thailand. The research will also have much wider international significance as there have been very few methodologically sound evaluations of the impact of hand hygiene interventions in hospitals. Moreover, the findings will provide essential information for subsequent work on economic evaluations of infection control interventions to determine under what circumstances such interventions are likely to be cost-effective.

Healthcare-associated infections(HCAIs) are a global problem and a major source of preventable morbidity and mortality worldwide, but particularly so in low and middle income countries. A recent review estimated the pooled prevalence of HCAI in developing countries (defined as Organisation Economic Co-operation Development (OECD) low or middle income countries) to be 15.5 per 100 patients, and highlighted the need to improve infection control practices. Rates of surgical site infection have been reported to range from 12 to 39% in developing countries compared to 2 to 5% in developed countries and rates of HCAI in neonates have been found to be 12 times higher in developing countries. Mortality due to such infections in developing countries is found to greatly exceed that in developed countries.

There is, however, evidence from quasi experimental research that concerted and sometimes low-cost interventions are able to greatly reduce the hospital transmission and prevalence of multi-drug resistant organisms (MDROs), resulting in improved patient outcomes. Improved hand hygiene(HH) is widely believed to be the single most effective intervention. However a 2010 Cochrane review has highlighted the lack of methodologically sound research evaluating interventions to improve HH: only four studies met the minimum inclusion criteria and only one such study came from a lower or middle income country (a randomised trial of an educational intervention for nurses in China). Moreover, a recent review has identified a need for stronger evidence linking improved HH compliance with reductions in HCAIs (currently such evidence is based only on observational studies).

While there are comprehensive hospital infection control guidelines at the proposed study site, Sappasithiprasong hospitals, a recent observational study our team carried out at this hospital has indicated poor HH compliance, below 5% for all health-care workers(HCWs) (manuscript in preparation). Previous data collected by our team has suggested hospital infection control guidelines may not have been implemented consistently due, in part, to resource constraints[5]. More recent qualitative research we have carried out at Sappasithiprasong hospital into the causes of poor hand hygiene compliance has identified several obstacles, and highlighted in particular a lack of knowledge about appropriate hand hygiene behaviour.

Available evidence suggests that hand hygiene promotion (HHP) with a multimodal intervention will be the most effective approach to improve HCW hand hygiene compliance[6]. While formal evaluation of such an intervention using a strong study design in lower and middle income countries is lacking, such interventions are relatively inexpensive and thought likely to be cost-effective (and potentially cost-saving as a result of reduced infections rates). Again, however, formal economic evaluations outside high income countries are lacking.

This approach promoted by the WHO guidelines identifies five moments for HH (before patient contact, before aseptic procedure, after body fluid exposure, after touching a patient, after touching a patient's surroundings). These guidelines aim to encourage good HH compliance in the real world and recommend a multimodal HHP strategy making use of up to five components: 1) system change (for example, changing systems to ensure that alcohol-based hand-rub is readily available wherever and whenever needed), 2) training and education, 3) observation and feedback, 4) reminders in the hospital, and 5) a hospital safety climate.

Adapting these guidelines to local conditions is important, and will make use of our initial qualitative research. This has identified important local barriers to achieving high levels of HH compliance.

Another major issue for improving HH behaviour in hospital HCWs is sustainability. While many studies have reported immediate improvements in HH following interventions, sustaining such improvements is much more challenging and another area where research has been very limited.

In conclusion, there is a need for methodologically rigorous research to evaluate the impact of a multimodal intervention based on the WHO recommendations both in the short and longer-term considering as outcomes both HH compliance, and preventable HCAIs. The need for such research is greatest in resource-constrained settings in lower and middle income countries where the burden of disease due to healthcare associated infections is the greatest. Findings from this research can also inform economic evaluations which are needed to determine the conditions under which such interventions are cost-effective.

Inclusion Criteria:

• All HCWs, having direct patient contact, in all in-patient wards including Intensive Care Units (ICU) and non-ICUs.

Exclusion Criteria:

• There are no exclusion criteria for the primary outcome.
• HCWs who refuse to sign the informed consent to participate self-administered questionnaire will be excluded from the secondary outcome assessing knowledge and beliefs of HCWs.