73133 GLOBAL SCALING UP HANDWASHING PROJECT: RESEARCH BRIEF Global Scaling Up Handwashing Project Handwashing Behavior Change at Scale: Evidence from a Randomized Evaluation in Vietnam September 2012 KEY IMPACT EVALUATION FINDINGS INTRODUCTION Preventable diseases resulting from poor hygiene behavior • Caregivers in treatment communities reported are responsible for a tremendous disease burden among higher exposure to handwashing messages, the world’s poor, especially infants and children under five. especially through the Vietnam Women’s Union Diarrhea alone accounts for 11% of child deaths worldwide— • The handwashing campaign led to modest more than HIV/AIDS and malaria combined. Washing hands improvements in handwashing knowledge, with soap has been shown to reduce diarrhea in young and may have impacted certain beliefs about children by as much as 48%, and is frequently referred to as handwashing among the most effective and inexpensive ways to avert child deaths. It has been called the “do-it-yourself� vaccine, yet • Caregivers reported washing their hands with despite its low cost and proven benefits, rates of handwashing soap more often, but observations in the home with soap are very low throughout the world. revealed low rates of handwashing at key times emphasized by the campaign Handwashing campaigns employing a range of methods are • Improvements in handwashing behavior reported common in developing countries, but little is known about the by caregivers do not appear sufficient to lead to effectiveness of these campaigns in getting people to wash impacts on child health or caregiver time savings their hands with soap. Few have been rigorously evaluated, and none on a large scale. Where evaluations have been done, they are often under trial conditions, with provision of soap and close follow-up of trial participants. The improvements in handwashing behavior reported by Beginning in 2009, the Water and Sanitation Program’s Global caregivers were not sufficient to result in impacts on child Scaling Up Handwashing project initiated a randomized health or reductions in time spent caring for sick children. controlled trial impact evaluation of a large-scale handwashing campaign in three provinces of rural Vietnam (Hung Yen, Television ads and face-to-face communication spread Thanh Hoa and Tien Giang). The results of the study show handwashing with soap campaign messages to mothers the campaign was successful in reaching its target audience and caregivers. The handwashing campaign messages and led to greater knowledge of good handwashing practice. were delivered through two main channels. These included Caregivers reported washing their hands with soap more often. television ads at the regional and national level that ran for a year However, when observed in the home rates of handwashing starting in January 2010, and interpersonal communication at the key times emphasized by the campaign were low. (IPC) activities that took place between January and 2 Handwashing Behavior Change at Scale: Evidence from a Randomized Evaluation in Vietnam Global Scaling Up Handwashing a combination of the TV ads and IPC activities (treatment) and another 70 were assigned to only the TV ads (control). Thus, the study evaluated the combined effect of the TV ads and IPC activities at the community level compared with the TV ads alone. A sample size totaling 3,150 households was visited in late 2009 before the intervention began, in July 2010 as a midpoint in the intervention, and in early 2011 after the intervention had ended. Over 94% of households participated in the three rounds of data collection. Caregivers in treatment communities reported higher exposure to handwashing messages, especially through Vietnam Women’s Union members teach women to wash hands the Vietnam Women’s Union. Caregivers in treatment with soap in the market. communities reported exposure to handwashing messages through three or more channels 56% of the time, whereas October 2010, carried out by specially trained members only 46% in the control communities reported the same. of the Vietnam Women’s Union, village health workers, Caregivers in the treatment communities were also nearly and teachers. Village activities were targeted at mothers, two-thirds more likely to have talked with someone from the grandparents, and women of childbearing age and Women’s Union about handwashing in the past month (56.4% included, among others, group meetings, household visits, in treatment vs. 34.7% in control; see Figure 1). Still, a large and distribution of promotional items. The communication proportion (45.9%) of the control group also reported exposure campaign was developed through formative research with to handwashing messages through three or more channels, the target audience. possibly the result of the increasing number of information sources on handwashing in Vietnam, such as marketing Messages aimed to dispel existing caregiver beliefs that campaigns of private soap companies or other public sector even clean-looking and clean-smelling hands can have hygiene messages related to Avian Influenza or H1N1 virus. germs, and to increase caregiver knowledge of the times for handwashing with soap that are critical for reducing pathogen The handwashing campaign led to improvements in transmission—after contact with feces and before touching handwashing knowledge, and may have impacted certain food. The campaign sought to promote handwashing as beliefs about handwashing. Knowledge about the best way something practiced by “good mothers� to ensure the welfare to wash hands increased substantially from a baseline of 79.4% of their children, and emphasized the need to make soap and to 97.7% at endline. However a similar increase was found in the water readily available for handwashing. Figure 1: Exposure to the Women’s Union and Other A randomized controlled trial impact evaluation established Campaign Channels the causal link between the handwashing campaign 20.9% 62.5% and behavioral and health outcomes. Rural Vietnamese communes1 within each intervention province were first paired 55.5% 56.4% 45.9% up by size and geographic location. Communes in each pair 34.7% were then randomly assigned to either a treatment or control group. A total of 140 communes were assigned to receive 1 A Vietnamese commune is an administrative sub-division of the district. Exposure to more Exposure to IPC The average population of the communes in this study is 7,577 people than 3 channels through Women’s Union (1,807 households) with a population range of 409 to 27,898 (172 to 5,531 households). Control Treatment www.wsp.org Global Scaling Up Handwashing Handwashing Behavior Change at Scale: Evidence from a Randomized Evaluation in Vietnam 3 control arm (97.3% at endline), so this is more likely the result Caregivers reported washing their hands with soap of a general trend towards greater knowledge. Caregivers in more often, but observations in the home revealed low the treatment group were more aware of the importance of rates of handwashing at the key times emphasized by handwashing with soap in diarrheal disease prevention (87.8% the campaign. When asked, nearly three-quarters (7.2% in treatment vs. 84.9% in control), and were almost 40% more more relative to control) of caregivers in the treatment arm likely to identify before food preparation as a critical time for reported washing their hands with soap after fecal contact, handwashing relative to the control group (40.5% in treatment and a little over 40% (14.4% relative to control) before feeding vs. 29.6% in control; see Figure 2). Additionally, some caregiver a child. However, when they were observed in the home beliefs about handwashing were impacted by the campaign, the actual rates of handwashing at these times were much such as the need to wash hands even though they have not lower and there were no statistically significant differences in touched unhygienic objects. observed handwashing rates between treatment and control groups (see Figure 3). Self-reported measures are generally Figure 2: Caregiver Knowledge of Critical Times for less reliable than observed measures, and as demonstrated Handwashing with Soap Emphasized by the Campaign here they are often subject to bias. In contrast to the reported rates, less than one-quarter of caregivers in treatment areas 77.4% 77.5% were observed to wash hands with soap after fecal contact 38.5% 39.5% and less than 10% before feeding a child. Improvements in handwashing behavior reported by After using After washing baby’s caregivers do not appear sufficient to lead to impacts on toilet bottom/changing diaper child health or caregiver time savings. Diarrhea prevalence 36.8% is 16.6% lower in the treatment group relative to the control (4.5% in treatment vs. 5.4% in control); however the difference 40.5% 39.8% 41.1% 29.6% is not statistically significant. Given the lack of observed impact on handwashing behavior we cannot attribute this difference to the handwashing campaign. Likewise, the evaluation found no reductions in time spent caring for sick Before preparing Before feeding/ food breast feeding baby children, which could in turn be used for money generating Control Treatment activities leading to increased household welfare. Figure 3: Self-reported vs. Observed Handwashing with Soap 7.2% 73% 14.4% 68.1% 41.4% 36.3% 31.1% 34.1% 24% 21% 7.1% 7.4% 4.6% 6.5% After fecal contact Before food preparation Before feeding a child Self-report Control Self-report Treatment Observe Control Observe Treatment www.wsp.org 4 Handwashing Behavior Change at Scale: Evidence from a Randomized Evaluation in Vietnam Global Scaling Up Handwashing Even under enabling conditions the intensity of these methods can be of high baseline knowledge and replicated at large scale. Related reading For the full report, please see: Claire access to soap and water, changing Chase and Quy-Toan Do, Handwashing handwashing behavior is difficult. The households in the communities Behavior Change at Scale: Evidence The handwashing campaign did not where this intervention was done were from a Randomized Evaluation provide soap and water to intervention found to be relatively advantaged when in Vietnam, World Bank Policy households and did not seek to only compared with other, more remote and Research Working Paper Series, 2012 improve knowledge of handwashing ethnically heterogeneous areas of rural (forthcoming). When published, the in the target audience. Formative Vietnam. Diarrhea prevalence is low in report will be available at www.wsp.org. research and baseline survey findings the households studied and child growth found these conditions were already in measures taken at baseline indicate Acknowledgments place. Despite this, handwashing with just a small proportion of children are The authors would like to thank the soap behavior in the target population clinically malnourished. project’s Global Task Team Leader, has not changed substantially as a Eduardo Perez; the project’s Global result of the intervention, which sought Furthermore, based on indicators of Impact Evaluation Team led by to influence the motivating factors and access to improved sanitation, safe Bertha Briceno; Almud Weitz, Regional barriers to handwashing such as beliefs water sources and safe drinking water Team Leader for WSP East Asia and treatment practices, the environmental Pacific region; and the country team in about the need for handwashing and fecal contamination that these children Vietnam, including Nga Kim Nguyen and placement of soap in an accessible are exposed to may be expected to Minh Thi Hien Nguyen, for their inputs. location. be small. Still, we cannot rule out the These results suggest that even under possibility that had the study been done About the project seemingly optimal conditions where on a more vulnerable population the Global Scaling Up Handwashing is a rates of handwashing observed would project by the Water and Sanitation knowledge and access to soap and have resulted in health impacts. Indeed, Program (WSP) focused on applying water for handwashing are not main innovative behavior change approaches constraints, behavior change campaigns handwashing with soap remains a to improve handwashing with soap that intend to reach a mass audience key preventive measure for areas with behavior among women of reproductive face tradeoffs in terms of intensity and a high burden of diarrheal disease age (ages 15–49) and primary effectiveness. In other words, large and malnutrition in children, especially school-age children (ages 5–9). It was declines in diarrhea found previously in where these other environmental health implemented by local and national highly-controlled pilot interventions may improvements are not in place. governments with technical support be difficult to achieve in practice unless from WSP in four countries: Peru, —Claire Chase and Quy-Toan Do Senegal, Tanzania, and Vietnam. For more information, please visit www.wsp.org/scalinguphandwashing. WSP is a multi-donor partnership created in 1978 and administered by the World Bank to support poor people in obtaining affordable, safe, and sustainable access to water and sanitation services. WSP’s donors include Australia, Austria, Canada, Denmark, Finland, France, the Bill & Melinda Gates Contact us Foundation, Ireland, Luxembourg, Netherlands, Norway, Sweden, Switzerland, United Kingdom, For more information, please visit United States, and the World Bank. www.wsp.org or email the authors at wsp@worldbank.org. The findings, interpretations, and conclusions expressed herein are entirely those of the author and should not be attributed to the World Bank or its affiliated organizations, or to members of the Board of Executive Directors of the World Bank or the governments they represent. © 2012 Water and Sanitation Program