Evaluating the effectiveness of a community-based hygiene promotion program in a rural Salvadoran setting

Abstract

There has been considerable progress in the reduction of diarrheal disease among children under five through health and nutrition interventions. However, diarrheal disease is still the second leading cause of child death worldwide. There is growing recognition that comprehensive hygiene behavior improvements should be integral to prevention efforts, but the effectiveness of different approaches for hygiene promotion is still being established. Hygiene risk practices vary across settings, suggesting that prevention strategies should be adapted to local contexts using community-based approaches. We planned, implemented, and evaluated a hygiene promotion intervention using the hygiene cluster framework. The two-year, multi-level intervention was implemented by local health promoters who were involved in identifying and addressing disease transmission risks at the household, school, and community levels. The intervention was evaluated using a quasi-experimental pretest-posttest design with repeated follow-up assessments to determine changes in hygiene knowledge and behavior. A household survey instrument was administered at three time points in the intervention (n = 480) and comparison (n = 271) communities to assess two hygiene knowledge and eleven hygiene behavior outcome variables. We used one-way analysis of variance with post hoc analysis using Tukey’s HSD for multiple comparisons to examine change and differences over time. We also fit a linear regression model to identify statistically significant differences. Study results demonstrated improvements in the areas of: knowledge of disease transmission and key times for handwashing, water container hygiene, sanitation practices, personal hygiene and food hygiene. The hygiene cluster framework is useful for hygiene promotion intervention planning and evaluation, and we recommended continued testing of this framework across contexts. We also recommend local community participatory approaches, as well as in-depth formative behavioral assessments by hygiene cluster that also consider environmental barriers to behavior change.

Keywords

behavior change communicable disease community-based research/participatory research evaluation health promotion prevention sanitation/hygiene Latin America

Introduction

Diarrheal disease contributes to nearly one in five child deaths (1,2), and is the leading cause of malnutrition in children under five years of age, resulting in failure to thrive (3). Considerable progress has been made with exposure prevention measures, including increasing access to potable water (4 –7), sanitation (8 –13), and promoting handwashing with soap (5,14 –19), as well as disease severity reduction measures, including improved nutrition, use of oral rehydration salts, breastfeeding, and micronutrient supplementation (3,20). Amidst this progress, there is growing recognition that comprehensive hygiene behavior changes should be an integral component of diarrheal disease prevention efforts (3,11,21 –24). Research has demonstrated that health promotion efforts can improve hygiene behaviors (25); however, the effectiveness of different approaches across settings is still being established (26).

Guidelines set forth by the USAID Environmental Health Project recommend implementing several diarrheal disease prevention components to yield the greatest impact, including the ‘hardware’ of improved water and sanitation and the ‘software’ of improved hygiene knowledge and practice (27). Yet the scientific literature is still mixed as to whether implementing several components yields the greatest effect. On one hand, some studies have shown that multiple interventions are more effective than a single intervention, such as handwashing or sanitation improvements alone (28,29). In others, multiple intervention effects do not seem to be additive (5,30). Further, as noted by Fewtrell et al. (5), prevention interventions attempting to target multiple disease transmission pathways risk being too scattered to be effective. This is complicated by intervention variability, making effectiveness comparisons difficult (5). Nevertheless, there is considerable consensus that certain hygiene-related behaviors across a set of domains, or clusters, should be targeted (31).

Community-based strategies engaging local residents tend to be more responsive to local needs, and offer local expertise for culturally- and context-appropriate risk reduction (33,34). As suggested by Figueroa and Kincaid (35), hygiene intervention models using community-based strategies are needed to achieve hygiene behavior improvements. For this study, we evaluated a community-participatory hygiene intervention guided by the hygiene cluster framework and implemented by local health promoters. The study goal was to determine the effectiveness of the two-year hygiene promotion intervention in improving hygiene knowledge and behavior across multiple hygiene domains.

Hygiene cluster framework

Since the mid-1990s, increased inter-disciplinary investigation has classified hygiene behaviors into domains, with implications for intervention planning and evaluation. The hygiene cluster approach is rooted in fecal-oral etiology, whereby microorganisms spread in contaminated food and water, or from person to person, as a result of poor hygiene or inadequate waste disposal (36). More than 80% of diarrhea cases worldwide result from fecal-oral contamination (17,33). Behavioral clusters involved in the disease transmission cycle span personal, domestic, food, sanitation, and water hygiene domains. Eisenberg et al. highlighted the need for interventions spanning multiple clusters to fully address complex disease transmission (34), and there is increasing evidence to support this notion (5). Responding to the call for further investigation (31,34), we evaluated the effectiveness of an intervention guided by this framework to determine achievable hygiene clusters behavior change, given the infrastructure- and resource-limited context, and whether implementing a participatory intervention with flexibility in design was feasible and could produce hygiene improvements.

Hygiene promotion intervention

The intervention was a two-year hygiene promotion effort in a rural Salvadoran community. Our George Washington University research team assembled a partnership between the local municipality and community development association (ADESCO), a Salvadoran rural water system association (ASSA), Peace Corps, the Salvadoran Ministry of Health (SIBASI), and Engineers Without Borders-DC Professionals chapter (EWB-DC). As a first step, we conducted participatory community assessments that identified structural and behavioral determinants of diarrheal disease: limited access to potable water and inadequate hygiene. Project phase one included the hygiene promotion intervention (described herein), followed by phase two water system construction. As in many Latin American countries, the predominant Salvadoran rural health services delivery strategy incorporates health promoters who address numerous areas, including ‘saneamiento ambiental’ (environmental hygiene) and infectious disease prevention (37). Building on this strategy, we recruited three local health promoters who received a five-day training in: diarrheal disease prevention; effective communication; leadership; time management; problem-solving; activity planning; human subjects protection; and health education techniques, such as group activities, games, role-playing, story-telling, and demonstrations. Intervention planners and trainers included one George Washington University research staff member, one master’s student, one undergraduate student, and a local Peace Corps volunteer. A SIBASI health promoter and physician supported training. The intervention was implemented at the household, school, and community levels to promote hygiene messages consistent with the hygiene cluster framework and informed by the work of Curtis and colleagues (33) (Table 1). Households were visited twice a month for 30–60 min to conduct hygiene promotion activities, and provide home modification assistance (creating latrine covers, modifying kitchen), skill-building demonstrations (handwashing, child care, hygienic cooking), and clinic referrals. Health promoters assessed each household’s hygiene risks and delivered tailored health promotion. Health promoters also implemented weekly 60-min school-based activities (grades 1–9) promoting recommended hygiene practices (role playing, games, contests, and storytelling), delivered parent presentations, and modified school hygiene infrastructure or protocol. At the community level, they organized hygiene campaigns and delivered messages at community gatherings and religious services.

Table 1.

Hygiene promotion message topics by domain.

Disease transmission	• Water-washed and water-borne illnesses • Diarrheal disease/parasitism transmission pathways • Diarrheal disease/parasitism prevention • Disease signs/symptoms • Home treatment and clinic referrals
Personal hygiene	• Hand-washing – key times, correct method • Wearing clean clothes, using shoes • Bathing frequency • Importance of using soap for handwashing, bathing • Oral Hygiene
Safe water management	• Water treatment – chlorination, UV radiation, boiling • Water storage – elevated storage, cleaning container, preferred narrow-neck container • Water handling – preventing re-contamination by pouring/using spout instead of dipping/scooping
Safe food management	• Kitchen hygiene – preventing cross-contamination when cooking, cleaning surfaces, using utensils and containers washed with soap, cleaning/covering dishes and utensils • Hygienic food handling, preparation and storage
Safe feces disposal	• Building a latrine – adequate depth, distance from home, covered • Maintaining a latrine – using cover, not depositing chemicals inside latrine • Latrine cleanliness – cleaning with soap, using cover, depositing toilet paper inside latrine • Child open defecation prevention
Trash disposal	• Types of trash: organic and inorganic • Adequate disposal of different types of trash • Maintaining cleanliness of riverbeds and streets

Methods

Study design and setting

We used a community-level, quasi-experimental pretest-posttest design with repeated follow up assessments to determine whether the intervention achieved improvements in hygiene knowledge and behavior. One community (pop. 3165) received the intervention while another served as a comparison group (pop. 3434). Both communities are rural, underdeveloped, and have agricultural-based economies, limited public services, and low educational attainment. The intervention community (IC) was selected due to high levels of diarrheal disease, and the comparison community (CC) was selected based on demographic, cultural, socioeconomic, and health care access similarities (Table 2).

Table 2.

IC and CC baseline demographics.

	IC(n = 1163) (%)	CC(n = 296) (%)
Gender
Male	46.3	49.6
Female	53.7	50.4
Age
0–5	8.5	7.7
6–12	17.7	15.8
13–18	15.2	11.2
19+	58.6	65.2
Occupation
Farmer	49.6	40.0
Vendor	4.9	1.4
Teacher	2.1	5.7
Unemployed	21.1	24.3
Day laborer	2.8	4.3
Homemaker	13.7	20.0
Other	5.8	4.1
Receive financial remittances
Yes	51.3	35.3
No	48.7	64.7
Education head household
No formal education	30.5	22.6
Primary school	49.2	56.5
Secondary school	6.7	6.4
College	5.3	9.7
Don’t know	9.3	4.8
Literacy head household
Can read only	8.2	1.5
Can write only	3.9	1.5
Can read and write	41.4	66.2
Cannot read or write	46.4	30.9
Housing	n = 389 homes	n = 71 homes
Electricity
Yes	84.1	95.7
No	15.9	4.3
Material of floors
Dirt	60.1	32.9
Tile	12.4	31.4
Concrete	17.3	32.9
Brick	8.5	2.9

Sampling and data collection

We administered baseline and two follow-up surveys with residents in the IC and CC, with households as the sampling unit. A minimum sample (n = 71) was determined based on effect sizes from similar studies and utilization of a table for two-tailed test on proportions with: power = 0.80, alpha = 0.05, P₁ = 0.10, and P₂ = 0.30. At baseline (T0) (April 2008), we collected the minimum sample in the CC (n = 71), but sampled every IC household (n = 289) to provide health promoters with a baseline hygiene assessment. For first and second follow-up waves (T1, T2) (April 2009 and 2010) we oversampled (n = 100 IC; n = 100 CC), with the exception of the IC at T2 (n = 91). We systematically sampled every third household along a randomly designated route, equally representing all regions. Surveys were orally administered by trained data collectors, and the survey response format included participant self-report and data collector observations. Instruments and protocol were approved by The George Washington University Institutional Review Board, and informed consent was obtained from study participants.

Measures

We evaluated the intervention using 13 dependent variables: two knowledge and 11 behavior variables across five hygiene clusters, including water [2], sanitation [4], domestic [1], personal [2], and food [2]. Table 3 provides detail on these measures.

Table 3.

Hygiene knowledge and practice outcome measures.

Hygiene domains	Measure	Cronbach’s alpha (# items)	Measure/item description
Hygiene knowledge	Disease transmission knowledge	0.72 (11 items)	Knowledge of diarrheal disease causes
Hygiene knowledge	Key handwashing times	0.75 (10 items)	Knowledge of key times for handwashing, such as before eating or feeding child and after using latrine or cleaning child’s bottom
Water hygiene practice	Adherence to hygienic water handling practices	0.55 (2 items)	Water container type used; how water is removed from container
Water hygiene practice	Adherence to hygienic water container practices	0.64 (2 items)	When and how water containers were last cleaned
Sanitation practice	Adherence to sanitary latrine cleaning practices	0.34 (6 items)	Cleaning of latrine seat, deposition of materials into latrine, cleaning of latrine floor
	Latrine cleaning frequency	1 item	Whether latrine was last cleaned today, yesterday, less than a week ago, more than a week ago or never
	Latrine covering adherence	1 item	Whether latrine had a cover
	Adherence to hygienic child defecation practices	1 item	Where small children defecate, including in the fields, patio, latrine or toilet
Domestic hygiene practice	Adherence to recommended trash disposal practices	1 composite item	Methods for trash disposal, including burning, burying, composting, recycling, trash can, throwing anywhere, putting in street or river bed
Personal hygiene practice	Adherence to recommended soap practices for personal hygiene	0.63 (7 items)	Soap usage at key times for handwashing, bathing and bathing children
Personal hygiene practice	Adherence to recommended soap practices for domestic hygiene	0.56 (2 items)	Soap usage for domestic purposes, including washing clothes and dishes
Food hygiene practice	Adherence to hygienic kitchen practices	0.73 (6 items)	Handwashing prior to food preparation and washing/storage of utensils and containers
Food hygiene practice	Adherence to hygienic food preparation practices	0.54 (3 items)	Proper washing/cooking of foods and proper food storage

Measures were developed de novo based on previous studies since we did not encounter measures appropriate for assessing this particular intervention tailored to the local context. We consulted studies by Almedom (38), Gorter et al. (31), and the report, Assessing Hygiene Improvement at the Household and Community Levels (27), to determine the overall measurement frame. Hygiene knowledge measures were modeled after Westaway and Viljoen’s (40) 40-item knowledge scale of diarrheal disease (2000) with four 10-item subscales: knowledge of diarrhea signs/symptoms, etiology, transmission and prevention, and an overall reliability of a = 0.73. The resultant knowledge variables were highly correlated, with a Pearson’s correlation of 0.818 (p < 0.000). Hygiene practice measures were developed based on Hygiene Improvement Framework indicators (32) and studies by Webb et al. (39), Pinfold and Horan (41), and Gorter et al. (31). For measures development, we used baseline survey data from the IC and CC (n = 360), and our approach involved factor analysis, and examination of inter-item correlation matrices and Cronbach’s alpha values. Higher measure scores indicate more correct hygiene knowledge or greater adherence to recommended hygiene practices. We evaluated whether the intervention resulted in improved hygiene knowledge and practice, evidenced by increased outcome measure means and statistically significantly different results when compared to the CC.

Data analysis

Baseline descriptive statistics were used to describe the sample, examine distribution of scores, and compare study communities using independent sample t-tests and Pearson’s chi-square analysis for continuous and categorical variables, respectively. For each outcome variable, in order to examine whether there were statistically significant differences between the six mean scores (three time points for two communities), we used one-way analysis of variance (ANOVA) with post hoc analysis using Tukey’s HSD for multiple comparisons. We standardized each dependent variable into standard deviation units. We examined the model F statistic for statistically significant differences among the six means, then identified differences through post-hoc comparisons. To determine whether the extent of change over time was statistically significantly different between communities, we fit a linear regression model using one dummy variable for community, two for time point and two community-by-time interaction terms. The criterion for statistically significant results was an alpha ≤ 0.05.

Results

Outcome variables

Results are presented below and are summarized in Tables 4 and 5.

Table 4.

Outcome variable mean scores for intervention and comparison communities.

	IC			CC
Variable (value range)	T0 (sd)	T1 (sd)	T2 (sd)	T0 (sd)	T1 (sd)	T2 (sd)
Knowledge disease trans. (0–5)	1.43 (1.03)	2.00 (1.36)	1.86 (0.66)	0.98 (0.63)	0.79 (0.48)	1.43 (1.00)
Knowledge handwashing (0–4.6)	1.43 (1.03)	1.82 (1.45)	1.68 (0.80)	1.09 (0.48)	0.68 (0.52)	1.15 (0.70)
Water handling (1.56–4.68)	3.11 (0.93)	3.33 (1.02)	3.19 (0.93)	3.31 (1.13)	3.22 (1.03)	3.60 (1.05)
Water container hyg (2.49–7.46)	7.12 (0.76)	7.09 (0.72)	7.27 (0.48)	6.96 (0.87)	5.85 (1.42)	6.52 (1.12)
Latrine cleaning (2.86–7.63)	4.80 (1.20)	5.16 (0.87)	5.08 (0.77)	4.82 (1.00)	5.35 (0.72)	5.21 (0.71)
When latrine last cleaned (1.03–5.13)	2.80 (0.98)	3.30 (0.79)	3.29 (0.77)	2.97 (1.03)	2.42 (0.99)	2.47 (1.08)
Latrine covered (2.05–4.10)	3.12 (1.03)	3.80 (0.74)	3.89 (0.62)	3.30 (1.00)	3.14 (1.03)	3.14 (1.03)
Child open defecation (4.05–8.10)	7.77 (1.12)	7.86 (0.97)	7.92 (0.83)	7.99 (0.67)	7.86 (0.97)	7.82 (1.03)
Trash disposal (1.18–4.71)	2.75 (1.07)	2.69 (0.90)	2.67 (0.91)	2.97 (0.99)	2.68 (0.93)	2.81 (1.03)
Soap use personal hyg (1.64–4.92)	3.24 (1.07)	3.30 (0.99)	3.00 (2.44)	2.60 (0.65)	2.22 (0.56)	2.44 (0.78)
Soap use domestic hyg (2.38–7.14)	6.71 (0.96)	7.05 (0.47)	6.75 (0.89)	6.67 (0.95)	6.51 (1.11)	6.64 (1.00)
Kitchen hyg (1.47–4.42)	2.64 (1.04)	2.84 (1.08)	3.21 (0.82)	2.45 (0.98)	2.31 (0.77)	2.75 (0.93)
Hygienic food prep. (1.46–4.37)	2.28 (0.89)	3.17 (0.98)	2.75 (1.04)	2.34 (0.80)	3.01 (1.05)	2.50 (1.00)

T0: Baseline.

T1: Follow-up 1.

T2: Follow-up 2.

CC: Comparison Community.

IC: Intervention Community.

sd: standard deviation.

Table 5.

Outcomes variable mean score comparisons.

Variable (value range)	IC T0-T1	IC T0-T2	IC v. CC T0	IC v. CC T1	IC v. CC T2	IC v. CC T0-T1	IC v. CC T0-T2
Knowledge disease trans.	0.56 (0.000)*	0.43 (0.002)*	0.45 (0.004)*	1.20 (0.000)*	0.43 (0.016)*	0.75 (0.000)*	−0.19 (0.917)
Knowledge handwashing	0.38 (0.006)*	0.24 (0.270)*	0.34 (0.070)	1.14 (0.000)*	0.52 (0.002)*	0.79 (0.000)*	0.18 (0.326)
Water handling	0.22 (0.415)	0.74 (0.990)	−0.20 (0.683)	0.11 (0.971)	−0.42 (0.054)	0.32 (0.117)	−0.21 (0.293)
Water container hyg	−0.03 (1.0)	0.15 (0.753)	0.16 (0.771)	1.24 (0.000)*	0.75 (0.000)*	1.07 (0.000)*	0.59 (0.001)*
Latrine cleaning	0.36 (0.115)	0.27 (0.188)	−0.20 (1.0)	−0.18 (0.80)	−0.13 (0.948)	−0.16 (0.406)	−0.11 (0.580)
Latrine last cleaned	0.50 (0.000)*	0.49 (0.000)*	−0.17 (0.785)	0.88 (0.000)*	0.81 (0.000)*	1.05 (0.000)*	0.98 (0.000)*
Latrine covered	0.68 (0.000)*	0.78 (0.000)*	−0.18 (0.693)	0.77 (0.000)*	0.78 (0.000)*	0.95 (0.000)*	0.94 (0.000)*
Child open defecation	0.09 (0.967)	0.16 (0.776)	−0.22 (0.547)	0.00 (1.0)	0.10 (0.983)	0.22 (0.252)	0.32 (0.100)
Trash disposal	−0.06 (0.99)	−0.8 (0.98)	−0.22 (0.561)	0.005 (1.0)	−0.14 (0.939)	0.22 (0.254)	0.08 (0.677)
Soap use personal hyg	0.06 (0.99)	−0.24 (0.28)	0.64 (0.000)*	1.08 (0.000)*	0.56 (0.001)*	0.44 (0.013)*	−0.08 (0.644)
Soap use domestic hyg	0.34 (0.034)*	0.04 (0.99)	0.04 (1.0)	0.99 (0.000)*	0.24 (0.528)	0.95 (0.000)*	0.20 (0.292)
Kitchen hyg	0.21 (0.446)	0.58 (0.000)*	0.19 (0.701)	0.53 (0.002)*	0.46 (0.013)*	0.35 (0.067)	0.28 (0.145)
Hygienic food prep	0.88 (0.000)*	0.46 (0.001)*	−0.04 (0.999)	1.3 (0.000)*	−0.26 (0.429)	1.35 (0.000)*	−0.21 (0.250)

Values in parentheses are p values.

Denotes statistically significant differences.

CC: Comparison community.

IC: Intervention community.

CC: Comparison community.

IC T0-T1, T0-T2: Comparison mean score by year, IC only.

IC v. CC T0, T1, T2: Comparison between community mean scores at time point.

IC v. CC T0-T1, T0-T2: Comparison between community mean scores over time period.

For both knowledge of disease transmission (F[5,746] = 24.02, p = 0.000) and key handwashing times (F[5,740] = 19.53, p = 0.000), we found strong evidence for differences between the six means. Post hoc analyses for disease transmission knowledge indicated that knowledge was higher at T1 and T2 in the IC, higher in the IC than the CC at T1 and T2, and increased more between T0 and T1 (but not T0 and T2). For knowledge of key handwashing times, post hoc analyses revealed that, in the IC, knowledge was higher at T1 than T0, but no differences were found between T2 and T0. Knowledge was higher in the IC than the CC between T0 and T1 but not between T0 and T2.

For adherence to recommended water handling (F[5,690] = 3.63, p = 0.003) and water container (F[5,702] = 34.02, p = 0.000) practices, we found strong evidence for differences between the six means. Post hoc analyses indicated that water handling adherence was higher at T1 and T2 in the IC, was higher in the IC at T1, and increased more from T0 to T1 compared to the CC. However, results were not statistically significant. For water container practices, adherence was higher at T2 (but not statistically significant), higher in the IC at T1 and T2, and increased more in the IC than the CC from T0 to T1 and T0 to T2.

For adherence to recommended latrine practices regarding cleaning method (F[5,723] = 6.51, p = 0.000), cleaning frequency (F[5,717] = 15.42, p = 0.000), and covering (F[5,729] = 16.70, p = 0.000), we found strong evidence for differences between the six means. Post hoc analyses indicated that in the IC, changes in cleaning methods adherence were not statistically significant at T1 or T2, were lower in the IC than the CC at T1 and T2, and did not increase more in the IC over time. For latrine cleaning frequency, adherence was higher at T1 and T2, and was higher and increased more in the IC than the CC from T0 to T1 and from T0 toT2. For latrine covering adherence, adherence was higher in the IC at T1 and T2, and increased more in the IC compared to the CC from T0 to T1 and T1 to T2.

For adherence to recommended child defecation (F[5,747] = 0.78, p = 0.57) and trash disposal (F[5,723] = 1.01, p = 0.410), analysis of variance did not provide evidence of differences among the six means and no post hoc analyses were conducted.

For adherence to recommended soap use for personal (F[5,732] = 29.21, p = 0.000) and domestic (F[5,744] = 11.60, p = 0.000) hygiene, we found strong evidence of differences among the six means). For personal hygiene soap use, post hoc analyses showed that in IC, there were no statistically significant differences in adherence to recommended practices at T1 and T2, and adherence was higher in the IC at T1 and T2, but adherence increased more in the IC between T0 and T1. For adherence to domestic hygiene soap use, post hoc analyses showed that in the IC, adherence was higher at T1 than T0, was higher in the IC at T1, and increased more in the IC from T0 to T1.

For adherence to recommended kitchen (F[5,734] = 9.92, p = 0.000) and food preparation (F[5,670] = 28.70, p = 0.000) practices, we found strong evidence for differences between the six means. For kitchen practices, adherence in the IC was higher at T1 and T2 (only statistically significant at T2) higher at T1 and T2 in the IC, and increased more in the IC from T0 to T1 and T1 to T2 (not statistically significant). For food preparation, adherence in the IC was higher at T1 and T2, higher in the IC at T1, and increased more in the IC between T0 and T1.

Discussion

Overall, study results are promising for future research using the hygiene cluster framework for intervention planning and evaluation. Contrary to assertions by Fewtrell et al. regarding the potential dilution of intervention effects when targeting multiple hygiene risk factors (5), this study demonstrated improvement in numerous hygiene domains.

Improvement in hygiene practices impinges on recognizing disease transmission pathways and responding with targeted prevention. The intervention achieved improvements in knowledge of disease transmission and key times for handwashing, with varying degrees of success for sustainability. We did not assess hygiene knowledge across all hygiene clusters, which could potentially shed more light on hygiene practice outcome results. Regarding hygiene practice outcomes, improvements were seen in water container hygiene, sanitation practices, personal hygiene and food hygiene outcomes.

We did not see the results we anticipated for some practice outcomes. For example, we included adequate trash disposal as a measure of domestic hygiene since we deemed it to be a considerable hygiene risk. Limited improvement was seen at the household level, possibly due to inadequate trash disposal infrastructure. However, results from a qualitative study (Andrade et al., unpublished), suggest that residents perceived considerable improvements in community cleanliness and trash disposal as a result of the program. Health promoters engaged residents in campaigns to improve community hygiene, including trash collection efforts. Potentially compelling results at the community level were not captured by household surveys.

Improving drinking water supply is also essential for reducing diarrheal disease. We did see improvement in water container hygiene at second follow-up, but not for water handling. When reviewing these outcomes, an important difference between the IC and CC should be noted. The majority of CC residents had access to piped water, while most IC residents used bucket-drawn wells or transported water from a neighbor’s well. The original study design planned to evaluate the impact of improved water and a hygiene intervention compared against improved water only. Water system construction was delayed, and this original study design was no longer feasible. Persisting differences is water source may explain limited improvement for some outcomes.

Despite limitations in infrastructure or resources, improvements in water container and food hygiene suggest that improved hygiene is still possible despite environmental constraints. Health promotion efforts focused on water storage and handling to minimize contamination until improved water infrastructure became available. Consequently, fewer households adopted narrow-neck water storage containers that are inconvenient for well water extraction, but did improve container hygiene. Food hygiene is also particularly important in the context of limited refrigeration access and in communities where dust and animal fecal matter are ubiquitous. Demonstrated improvements in food hygiene are promising, given resource challenges for such practices.

Consistent with Figueroa and Kincaid (35), and supported by Curtis et al. and Eisenberg et al. (33,34), engagement of community members in the planning, implementation and evaluation of the intervention was an asset. This participatory, community-based approach contributed to targeted hygiene messages, selection of health promotion strategies that were appropriate and tailored to individual households and schools, and adoption of new improved hygiene practices by participants.

This approach would be enhanced with the collection of detailed process data to facilitate conclusions about dosage and program outcomes. In this study, process data were collected by health promoters in paper format, including documentation of every household visit and school activity, and community event. However, limited resources for data monitoring resulted in record gaps that limited dosage estimation.

A participatory approach also enabled the research team to conduct a formative behavioral assessment to guide intervention planning. This assessment identified numerous priority hygiene risks; however, some important behavioral risks were not detected. Specifically, an important risk behavior overlooked during the formative assessment was deposition of used toilet paper outside of the latrine. According to findings from a separate qualitative study (Andrade et al., unpublished), significant improvements in toilet paper disposal are suspected. However, this practice was not captured by the survey.

Limitations

Some limitations should be noted when interpreting results. Randomizing participants to the intervention was not feasible, creating the possibility that differences between IC and CC could be attributed to pre-existing differences or outside factors. Some degree of inaccuracy in self-reported hygiene behaviors was possible due to recall or social desirability bias. Recall bias was minimized by limiting recall period, and socially desirable responses were minimized by administering survey items in unprompted question formats. Finally, we developed study measures de novo, resulting in a targeted program evaluation; however, there were insufficient resources to pilot test and validate measures, resulting in varying degrees of reliability.

Recommendations

Based on the findings on this study, we recommend that future research continue to apply the hygiene cluster framework for intervention planning and evaluation. Contextual factors and resources required for behavior change should be carefully considered during intervention formative research and planning, as well as potential barriers to behavior change. Using a community-based, participatory approach is a way to gain valuable community input during formative behavioral risk assessments, especially to identify private, unobservable behaviors such as latrine practices, or to adapt the intervention to fit local infrastructure and resource limitations. We recommend that future studies seek community input when conducting in-depth, formative assessments of hygiene clusters, especially for behaviors that are not readily observable.

In terms of measurement, we recommend future studies to assess knowledge in each cluster where hygiene practice will be targeted for change and that future studies also assess behavioral determinants for each targeted hygiene practice during formative research and program planning. Furthermore, we encourage future studies to collect data pertinent to every level where the intervention is planned to operate (individual, household, school, and community) in order to obtain a true estimation of intervention effects. In settings where resources for behavior change are not ideal, we recommend that hygiene interventions maintain flexibility in research design and in decisions regarding recommended practices. Specifically, in the case where wide-mouthed containers are preferred due to compatibility with water source and extraction methods (well water), covering such containers and hygienic water extraction should be the focus of prevention efforts rather than placing emphasis on the type of container used. In other words, instead of promoting a standardized recommendation that is backed by the scientific community (i.e. use of narrow-neck storage containers), practitioners should adapt the intervention to consider limitations in local resources and infrastructure, community preference, and compatibility with everyday lives of program participants.

Footnotes

Declaration of conflicting interests

None declared.

Funding

This research was funded by the American Public Health Association, Community-Based Field Research Grants for Behavioral Hygiene Promotion and Global Health, sponsored by Colgate-Palmolive and the George Washington University Milken Institute School of Public Health, Capital Connections Research Award.

References

UNICEF. Pneumonia and Diarrhea: Tackling the Deadliest Diseases for the World’s Poorest Children. UNICEF; 2012; Washington, DC.

Liu

Johnson

Cousens

Perin

Scott

Lawn

et al . Global, regional, and national causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. Lancet. 2012; 379: 2151–2161.

UNICEF/WHO. Diarrhoea: Why Children are Still Dying and What Can Be Done. UNICEF/WHO; 2009. Washington, DC.

Quick

Kimura

Thevos

Tembo

Shamputa

Hutwagner

et al . Diarrhea prevention through household-level water disinfection and safe storage in Zambia. Am J Trop Med Hyg. 2002; 66: 584–589.

Fewtrell

Kaufmann

Kay

Enanoria

Haller

Colford

Jr.

Water, sanitation, and hygiene interventions to reduce diarrhoea in less developed countries: a systematic review and meta-analysis. Lancet Infect Dis. 2005; 5: 42–52.

Arnold

Colford

Jr.

Treating water with chlorine at point-of-use to improve water quality and reduce child diarrhea in developing countries: a systematic review and meta-analysis. Am J Trop Med Hyg. 2007; 76: 354–364.

Clasen

Wolf-Peter

Rabie

Roberts

Cairncross

Interventions to improve water quality for preventing diarrhea: systematic review and meta-analysis. Br Med J. 2007; 34: 782.

Stanton

Clemens

Aziz

KMA

Rahman

Twenty-four-hour recall, knowledge-attitude-practice questionnaires, and direct observations of sanitary practices: a comparative study. Bull World Health Organ. 1987; 65: 217–222.

Han

Moe

Household faecal contamination and diarrhoea risk. J Trop Med Hyg. 1990; 93: 333–336.

10.

Baltazar

Solon

FS.

Disposal of faeces of children under two years old and diarrhoea incidence: a case-control study. Int J Epidemiol. 1989; 18(Suppl 2): S16–S19.

11.

Mertens

Jaffar

Fernando

Cousens

Feachem

Excreta disposal behaviour and latrine ownership in relation to the risk of childhood diarrhoea in Sri Lanka. Int J Epidemiol. 1992; 21: 1157–1164.

12.

Semba

Kraemer

Sun

De Pee

Akhter

Moench-Pfanner

et al . Relationship of the presence of a household improved latrine with diarrhea and under-five child mortality in Indonesia. Am J Trop Med Hyg. 2011; 84: 443–450.

13.

Jamison

Mosley

WH.

Disease control priorities in developing countries: health policy responses to epidemiological change. Am J Public Health. 2006; 81: 15–22.

14.

Luby

Agboatwalla

Feikin

Painter

Billhimer

Altaf

et al . Effect of handwashing on child health: a randomised controlled trial. Lancet. 2005; 366: 225–233.

15.

Luby

Agboatwalla

Painter

Altaf

Billhimer

Keswick

et al . Combining drinking water treatment and hand washing for diarrhoea prevention, a cluster randomised controlled trial. Trop Med Int Health. 2006; 11: 479–489.

16.

Curtis

Cairncross

Effect of washing hands with soap on diarrhoea risk in the community: a systematic review. Lancet Infect Dis. 2003; 3: 275–281.

17.

Huttly

SRA

Morris

Pisani

. Prevention of diarrhoea in young children in developing countries. Bull World Health Organ. 1997; 75: 163–174.

18.

Esrey

Feachem

Interventions for the Control of Diarrhoeal Diseases among Young Children: Promotion of Food Hygiene. Geneva, Switzerland: World Health Organization; 1989.

19.

Cairncross

Hunt

Boisson

Bostoen

Curtis

Fung

ICH

et al . Water, sanitation and hygiene for the prevention of diarrhea. Int J Epidemiol. 2010; 39: 193.

20.

Bhandari

Mazumder

Taneja

Dube

Agarwal

Mahalanabis

et al . Effectiveness of zinc supplementation plus oral rehydration salts compared with oral rehydration salts alone as a treatment for acute diarrhea in a primary care setting: a cluster randomized trial. Pediatrics. 2008; 121: e1279–e1285.

21.

Mathers

Stevens

Mascarenhas

Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risks. Geneva, Switzerland: World Health Organization; 2009.

22.

Prüss-Ustun

Kay

Fewtrell

Bartram

Unsafe water, sanitation and hygiene. In: Ezzati

Lopez

Rodgers

Murray

(eds). Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors. Geneva, Switzerland: World Health Organization; 2004: pp.1321–1352.

23.

Strina

Cairncross

Barreto

Larrea

Prado

MS.

Childhood diarrhea and observed hygiene behavior in Salvador, Brazil. Am J Epidemiol. 2003; 157: 1032–1038.

24.

Cairncross

AM.

Health impacts in developing countries: new evidence and new prospects. J Inst Water Environ Manage. 1990; 4: 571–577.

25.

Curtis

Kanki

Cousens

Diallo

Kpozehouen

Sangaré

et al . Evidence of behaviour change following a hygiene promotion programme in Burkina Faso. Bull World Health Organ. 2001; 79: 518–527.

26.

Cairncross

Shordt

Zacharia

Govindan

BK.

What causes sustainable changes in hygiene behaviour? A cross-sectional study from Kerala, India. Soc Sci Med. 2005; 61: 2212–2220.

27.

USAID. The Hygiene Improvement Framework: A Comprehensive Approach for Preventing Childhood Diarrhea. USAID; 2004. Washington, DC.

28.

Alam

Wojtyniak

Henry

Rahaman

MM.

Mothers’ personal and domestic hygiene and diarrhoea incidence in young children in rural Bangladesh. Int J Epidemiol. 1989; 18: 242–247.

29.

Haggerty

Muladi

Kirkwood

Ashworth

Manunebo

Community-based hygiene education to reduce diarrhoeal disease in rural Zaire: impact of the intervention on diarrhoeal morbidity. Int J Epidemiol. 1994; 23: 1050–1059.

30.

Esrey

Potash

Roberts

Shiff

Effects of improved water supply and sanitation on ascariasis, diarrhoea, dracunculiasis, hookworm infection, schistosomiasis, and trachoma. Bull World Health Organ. 1991; 69: 609–621.

31.

Gorter

Sandiford

Pauw

Morales

Perez

Alberts

Hygiene behavior in rural Nicaragua in relation to diarrhea. Int J Epidemiol. 1998; 27: 1090–1100.

32.

USAID. Strategic Report 8, Assessing Hygiene Improvement: Guidelines for Household and Community Levels. Environmental Health Project, Bureau for Global Health, Office of Health, Infectious Diseases and Nutrition. 2004; Washington, DC.

33.

Curtis

Cairncross

Yonli

Review: domestic hygiene and diarrhoea - pinpointing the problem. Trop Med Int Health. 2000; 5: 22–32.

34.

Eisenberg

JNS

Scott

Porco

. Integrating disease control strategies: balancing water sanitation and hygiene interventions to reduce diarrheal disease burden. Am J Public Health. 2007; 97: 846–852.

35.

Figueroa

Kincaid

Social, Cultural and Behavioral Correlates of Household Water Treatment and Storage. Center Publication HCI 2010-1: Health Communication Insights. Baltimore, MD: Johns Hopkins Bloomberg School of Public Health, Center for Communication Programs; 2010.

36.

Wagner

Lanoix

JN.

Excreta disposal for rural areas and small communities. Monogr Ser World Health Organ. 1958; 39: 1–182.

37.

Lewis

Eskeland

Traa-Valerezo

Challenging El Salvador’s Rural Healthcare Strategy, Volume 1. Washington, DC: The World Bank; 2000.

38.

Almedom

AM.

Recent developments in hygiene behavior research: an emphasis of methods and meaning. Trop Med Int Health. 1996; 1: 171–182.

39.

Webb

Stein

Ramakrishnan

Hertzberg

Urizar

Martorell

A simple index to measure hygiene behaviours. Int J Epidemiol. 2006; 35: 1469–1477.

40.

Westaway

Viljoen

Health and hygiene knowledge, attitudes and behaviour. Health and Place. 2000; 6: 25–32.

41.

Pinfold

Horan

NJ.

Measuring the effect of a hygiene behaviour intervention by indicators of behaviour and diarrhoeal disease. Trans R Soc Trop Med Hyg. 1996; 90: 366–371.