SEARCH

SEARCH BY CITATION

Keywords:

  • handwashing;
  • soaps;
  • behaviour;
  • social class;
  • Bangladesh
  • lavage des mains;
  • savons;
  • comportement;
  • classe sociale;
  • Bangladesh
  • Lavado de manos;
  • Jabones;
  • Comportamiento;
  • Clase social;
  • Bangladesh

Summary

  1. Top of page
  2. SummaryAssociations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du BangladeshAsociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Objectives  To explore the relationship of easy to collect handwashing indicators with socioeconomic status and reported respiratory disease among children <5 years of age.

Methods  We added several handwashing indicators to a population-based, cross-sectional study of respiratory illness in Dhaka, Bangladesh. We constructed a wealth index using 12 household characteristics analysed with principal component analysis to assess socioeconomic status.

Results  Of 6970 households, 92% had a bar of body soap, 41% had a place with water to wash hands inside the house, and 40% had soap present at the most convenient place to wash hands. Handwashing indicators were more common among households with higher socioeconomic status. Within each wealth quintile a place to wash hands within the household was strongly associated with the presence of soap at the handwashing location (odds ratios 13–70). In general estimated equation models that controlled for socioeconomic status, the presence of a place inside the house with water to wash hands was the only handwashing indicator significantly associated with a child in the household who reported cough or difficulty breathing in the preceding 7 days (adjusted odds ratio 0.95, 95% confidence interval 0.93–0.98, < 0.001).

Conclusion  Handwashing indicators were strongly influenced by socio-economic status and so would not be an independent measure of handwashing behaviour. Handwashing promotion efforts in urban Dhaka that include specific efforts to provide handwashing facilities inside the house are more likely to improve handwashing behaviour than interventions that ignore this component.

Associations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du Bangladesh

Objectifs  Explorer la relation entre la facilité de collecter des indicateurs de lavage des mains selon le statut socio-économique et les maladies respiratoires rapportées chez les enfants de moins de 5 ans.

Méthodes  Nous avons pris en compte plusieurs indicateurs de lavage des mains dans une étude transversale basée sur la population, pour les maladies respiratoires à Dhaka au Bangladesh. Nous avons conçu un indice de richesse des ménages basé sur 12 caractéristiques analysées avec l’analyse de composantes principales, afin d’évaluer le statut socio-économique.

Résultats  Sur 6970 ménages, 92% possédaient un pain de savon corporel, 41% disposaient d’un endroit avec de l’eau pour le lavage des mains à l’intérieur de la maison, et 40% avaient du savon présent à l’endroit le plus approprié pour se laver les mains. Les indicateurs de lavage des mains étaient plus fréquents dans les ménages avec le statut socio-économique le plus élevé. Au sein de chaque quintile de richesse, un endroit pour se laver les mains à l’intérieur de la maison était fortement associéà la présence de savon à l’emplacement pour le lavage des mains (rapport de cotes «odds ratio»: 13 - 70). En général, pour les modèles d’équation d’estimation qui tiennent compte du statut socio-économique, la présence d’un emplacement à l’intérieur de la maison avec de l’eau pour se laver les mains était le seul indicateur de lavage des mains significativement associés avec un enfant dans le ménage ayant rapporté une toux ou des difficultés respiratoires au cours des 7 précédents jours (Odds ratio ajustés: 0,95; intervalle de confiance 95%: 0,93-0,98; p < 0,001).

Conclusion  Les indicateurs de lavage des mains étaient fortement influencés par le statut socio-économique et donc ne seraient pas une mesure indépendante du comportement de lavage des mains. Les efforts de promotion du lavage des mains dans les zones urbaines de Dhaka qui prennent en compte des efforts particuliers en vue de fournir des installations de lavage des mains à l’intérieur de la maison, sont plus susceptibles d’améliorer le comportement de lavage des mains que les interventions qui ne tiennent pas compte de cette composante.

Asociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 

Objetivos  Explorar la relación entre indicadores para el lavado de manos, fáciles de recolectar, con el estatus socioeconómico y la enfermedad respiratoria reportada en niños menores de cinco años.

Métodos  Hemos añadido varios indicadores de lavado de manos a un estudio croseccional, basado en la comunidad, sobre la enfermedad respiratoria en Dhaka, Bangladesh. Con el fin de evaluar el estatus socioeconómico, hemos construido un índice de riqueza utilizando 12 características de los hogares analizados mediante análisis de componente s principales.

Resultados  De 6970 hogares, 92% tenían una pastilla de jabón de manos, 41% tenían dentro de la casa un lugar con agua en donde lavarse las manos, y 40% tenían jabón en el lugar donde se lavaban las manos. Los indicadores de lavado de manos eran más comunes en hogares con un mayor estatus socioeconómico. Dentro de cada quintil de riqueza el tener un lugar donde lavarse las manos estaba fuertemente asociado con la presencia de jabón en el lugar de lavado de manos (OR 13-70). En modelos generales con ecuaciones estimadas controlando para el estatus socioeconómico, la presencia de un lugar dentro de la casa con agua para lavarse las manos era el único indicador de lavado de manos asociado significativamente con la presencia de un niño en la casa que reportase tos o dificultades respiratorias en los 7 días precedentes (OR ajustado 0.95, intervalo de confianza del 95% 0.93, 0.98, p < 0.001).

Conclusión  Los indicadores de lavado de manos estaban fuertemente influenciados por el estatus socio-económico y por lo tanto no serían una medida independiente del comportamiento de lavado de manos. Los esfuerzos realizados en Dhaka urbana para promover el lavado de manos que incluyen esfuerzos específicos para proveer facilidades para el lavado de manos dentro de la casa, tienen más posibilidad de mejorar el comportamiento de lavado de manos que aquellas intervenciones que ignoran este componente.


Introduction

  1. Top of page
  2. SummaryAssociations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du BangladeshAsociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Diarrhoea and respiratory disease are leading global causes of death in early childhood (Bryce et al. 2005). Focused intervention studies consistently demonstrate that handwashing with soap reduces diarrhoea (Curtis & Cairncross 2003). There is also evidence that handwashing with soap reduces respiratory disease (Luby et al. 2005; Rabie & Curtis 2006).

Methods to promote handwashing that target millions of people living in communities with high levels of child mortality are quite different from the intensive methods usually involving repeated household visits that work quite well in efficacy trials of few thousand people. A critical barrier to scaling up handwashing promotion is limited evidence that methods which are able to be delivered at large scale actually change handwashing behaviour. An important reason for the limited evidence is the absence of practical low-cost methods that validly measure handwashing practices. Reported handwashing behaviour consistently overestimates observed behaviour (Stanton et al. 1987; Cousens et al. 1996; Manun’ebo et al. 1997). Structured observation of handwashing is expensive and the presence of an observer risks modifying handwashing behaviour. Microbiological methods are expensive, and probably because of rapid recontamination of hands in highly contaminated environments (Sobel et al. 1998) have not efficiently identified persons who regularly wash their hands with soap (Luby et al. 2007). Spot checks for the presence of soap and water are inexpensive and easy to collect, but their association with genuine handwashing behaviour is incompletely defined (Ruel & Arimond 2002). Without valid, affordable measures of handwashing it is difficult to evaluate and optimize handwashing promotion interventions.

Child health including childhood respiratory infections are strongly associated with socioeconomic status (Hussain et al. 1999; Rahman & Shahidullah 2001; Hawker et al. 2003; Schell et al. 2007). A number of factors that affect childhood respiratory disease including nutrition, immunization and indoor air pollution are also strongly associated with socioeconomic status (Perry et al. 1998; Niport 2005; Dasgupta et al. 2006). To evaluate programs targeted at changing specific behaviours, it is important to separate changed behaviour that results from the intervention program from changed behaviour that resulted from a change in socioeconomic status. Although some measures of hygiene in low income countries have been reported to be negatively associated with socioeconomic status (Curtis et al. 1995; Armar-Klemesu et al. 2000), we were able to locate only two published reports that explored the relationship between indicators of handwashing and socioeconomic status. Gorter et al. (1998) observed 50 separate indicators of hygiene practices in structured observation among families living in a rural community in Nicaragua and reported that families that owned a radio and whose mother had >3 years of education were significantly more likely to wash hands before preparing a baby’s bottle than households without a radio or with less maternal education. Among 966 households in seven villages in rural Bangladesh, Bhuiya and colleagues reported that mothers living in households that owned more consumer goods were more likely to report handwashing after defecation than mothers living in households that owned fewer items (Bhuiya et al. 1990).

We added handwashing indicators to a population based cross sectional study of respiratory illness in Dhaka, Bangladesh. The primary findings from the larger study will be published later when all of the components have been analysed. The objective of this analysis was to explore the relationship between handwashing indicators that are easy to collect to socioeconomic status and to reported respiratory disease among children under 5 years of age.

Methods

  1. Top of page
  2. SummaryAssociations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du BangladeshAsociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study population

The larger study was designed to enrol children who lived in the nearby urban catchment area of two hospitals in greater Dhaka, Dhaka Shishu Hospital and Shishu Sahayesta Foundation Hospital. Study workers identified neighbourhoods by listing patients under the age of 5 years, who were admitted for treatment of pneumonia to the study hospital. Study workers selected names at random from the admissions list, and if the house was located within 60 min travel time to the hospital field researchers travelled to the patient’s household. The field team noted the patient’s house, identified the closest house, and the next closest house to this house, until five houses were skipped. At the sixth house, if a child under the age of 5 years was living in the house or if a child born alive, and less than 5 years had died within the last 12 months, the field workers sought consent for enrolment in the study. After collecting data, field workers went to the next closest household without backtracking to identify additional participants. Between May and August 2006 study workers identified 70 patient households for community starting points and sought 100 households from each of these starting points. Because one of the components of the study was a child mortality assessment (data not presented here), there was an aggressive effort to enrol all eligible households. Field workers visited neighbourhoods repeatedly including during the evening and on weekends and made appointments, if necessary, to return at a time that was convenient to study subjects to maximize participation.

Data collection

Trained field workers administered a questionnaire that included questions on household assets, respiratory symptoms among children living at home, and soap purchasing and handwashing behaviour. The workers also conducted brief direct observation on the presence of soap, sanitary facilities, and a place to wash hands.

Data management

Questionnaires were reviewed by supervisors. Field researchers returned to households to collect missing information. Supervisors revisited 5% of households to double check information collected by field workers. Information from the written questionnaires was entered into an electronic database.

Data analysis

Because of the broader objectives of the parent study that supported this analysis, the field team collected data on respiratory symptoms including repeated coughing, difficulty breathing, nasal congestion and runny nose in the 7 days preceding the interview as well as visits to a health care provider in the preceding 2 months or hospitalization for respiratory symptoms in the preceding year. On exploratory analysis the visits to health care providers and hospitalization were strongly associated with measures of wealth, and so did not appear to be reliable marker for severe respiratory disease (data not shown). Thus, we focused on respiratory symptoms in the 7 days preceding the interview. Different combinations of respiratory symptoms yielded similar results, so we selected either cough or difficulty breathing, the two symptoms we had data on that are included in the WHO clinical case definition for pneumonia (Gove 1997) as the respiratory outcome variable.

We calculated the prevalence of household characteristics based on all of the enrolled households and respiratory disease morbidity upon all of the children under the age of 5 years living in these households. We constructed a wealth index using principal components analysis (Vyas & Kumaranayake 2006). Contributions to the wealth index included number of household rooms, elements of household construction, ownership of specific durable goods, maternal education and cooking fuel. We excluded water connection and sanitary facilities as variables for constructing the wealth index because some of the handwashing assessment indicators were directly dependent on having running water in the home and we were interested in the indirect effect of wealth on indicators of handwashing (Houweling et al. 2003). Specifically, we were evaluating the relationship between handwashing indicators and respiratory disease, and were using the wealth index to control confounding due to socioeconomic status. If the wealth index included elements of the handwashing indicator, it could overestimate the amount of confounding in the relationship between the handwashing indicator and respiratory disease. We used the first factor from the principal components analysis, as this is believed to best capture economic status (Houweling et al. 2003).

We divided households into quintiles based on their wealth score. Because many households had identical wealth index scores, the number of households in each quintile is not identical. We assessed trends in household characteristics across wealth quintiles using chi square for linear trend for categorical variables and simple linear regression for continuous variables.

The analysis was exploratory. We evaluated the relationship between indirect measures of handwashing and respiratory disease morbidity first by calculating crude odds ratios. We tested for effect modification by evaluating the homogeneity of odds ratios across wealth quintiles using the Woolf test for homogeneity. We calculated adjusted odds ratios and 95% confidence limits using general estimated equations (Hanley et al. 2003). The general estimated equation models accounted for multiple observations within each neighbourhood cluster using an exchange correlation matrix and adjusted for socioeconomic status using the wealth index calculated by principal component analysis for the household.

Ethics

Adult respondents in participating households provided informed consent. The study protocol was reviewed and approved by the Ethics Review Committee of the International Centre for Diarrheal Disease Research, Bangladesh.

Results

  1. Top of page
  2. SummaryAssociations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du BangladeshAsociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

From the 7000 identified households that met the enrolment criteria, field workers completed interviews in 6971 (99.6%). Households had a mean of five members. 26% of fathers and 34% of mothers had not completed primary education. 55% had lived at their present location for <3 years and 29% reported a monthly household income < US$ 70.5 (Table 1).

Table 1.   Participating household characteristics, Dhaka, Bangladesh 2006
CharacteristicnPercent*/ mean
  1. *Some categories do not sum to 100% because of rounding.

Mean number household residents 4.96
Mean number of sleeping rooms 1.8
Education of father of the youngest child (%)6971 
 No schooling96414
 Some schooling, but did not finish secondary269539
 Finished secondary (SSC)321446
 Unknown981
Education of mother of the youngest child (%)6971 
 No schooling114817
 Some schooling, but did not finish secondary120450
 Finished secondary (SSC)235234
 Unknown120
Occupation of father of the youngest child (%)6971 
 Salaried employee317446
 Shopkeeper/merchant182326
 Employed on daily wages143821
 Other3876
 Unemployed1492
Proportion of mothers earning income127018
Monthly household income (%)
 ≤70.5 US$204429
 >70.5–≤ 141 US$234834
 > 141 US$247436
 Refused/don’t know1051.5
Own
 Radio138020
 Refrigerator249836
 Mobile phone405558
 Television512474
 Electric fan674797
House construction
 Cement floor648993
 Brick walls593685
 Tin roof411159
 Natural gas connection624490
 Piped drinking water inside house296543
 Hygienic toilet inside house304944

Twelve indicators were included in the principal component analysis of socioeconomic status (Table 2). Whether or not the house had a cement floor and whether or not the household used natural gas for cooking fuel were the household characteristics that explained the most variance among households. The first component captured 34% of the total data variability. The wealth index divided households fairly evenly across a continuum with some clumping of values at the highest end (Figure 1).

Table 2.   Indicators used for creating a wealth score, and how they were coded
IndicatorCodingLoadingSDCoefficient*
  1. *Coefficient = Loading/SD × 100.

 1. Floor type1 = Mud/wood/bamboo/brick and soil, 2 = cement0.525.254207
 2. Wall type1 = Bamboo/straw/wood/mud, 2 = tin, 3 = concrete0.601.528114
 3. Roof type1 = Polythene/bamboo/straw/wood, 2 = tin, 3 = concrete0.707.496143
 4. No. of living rooms1 = 1, 2 = 2, 3 = 3, 4 = 4+0.7141.03469
 5. Own fan0 = No, 1 = Yes 0.271.176154
 6. Own radio0 = No, 1 = Yes 0.222.39856
 7. Own TV0 = No, 1 = Yes 0.576.441131
 8. Own cycle0 = No, 1 = Yes 0.291.29798
 9. Own refrigerator0 = No, 1 = Yes 0.759.480158
10. Own mobile phone0 = No, 1 = Yes 0.743.493151
11. Cooking fuel category1 = Wood/coal/kerosene/electric heater, 2 = natural gas0.501.306164
12. Mother’s education0 =  No schooling0.7411.50149
1 =  Some primary
2 =  Primary
3 =  Some secondary
4 =  Secondary
image

Figure 1.  Distribution of household wealth index scores (n = 6970).

Download figure to PowerPoint

About 91% of household respondents (n = 6361) reported usually washing their hands only with water, while 61% of respondents (n = 4232) reported using soap to wash hands for special purposes including after defecation and before eating food (Table 3). 95% of households (n = 6601) reported purchasing body soap within the last month. Field workers observed that 92% of households (n = 6401) had body soap in their households, 41% (n = 2285) had a place with water available to wash hands inside their house and 40% (n = 2770) had soap present at the place to wash hands. Both reported and observed indicators of handwashing with soap were consistently more common among households in wealthier quintiles (Table 3).

Table 3.   Household soap use indicators by wealth index, Dhaka, Bangladesh 2006, (= 6971)
Soap use indicatorHouseholds by wealth index quintile
Overall (n = 6971) %First (Poorest) (n = 1436) % Second (n = 1364) % Third (n = 1384) % Fourth (n = 1455) %Fifth (Richest) (n = 1332) %P-value for trend‡
  1. *Special purposes such as cleaning hands after defecation, after washing bottom of the child, before and after eating, and before feeding a child.

  2. †Calculated based on the number of bars of hand soap of different sizes purchased in the preceding 30 days, multiplied by the weight of the bars, and divided by the number of persons in the household.

  3. ‡Chi-square for trend for categorical variables, simple linear regression for continuous variables.

Reported behaviour
 Usually washes hands only with water for general purposes (%)919898958779<0.001
 Usually washes with body soap for special purposes* (%)613346587791<0.001
Most days body cleaning materials are
 Water (%)607569595638<0.001
 Body soap (%)402530414462<0.001
 Laundry soap (%)0.40.80.70.30.10.20.27
 Purchased body soap in last 1 month (%)9594969694930.03
 Average amount spent (per capita, among those 95% of households that purchased soap in last 1 month) in last 15 days for body soap US$0.0520.0410.0490.0520.0590.063<0.001
 Per capita soap consumption (g/day)† (among those households (95%) that purchased soap in last 1 month) 1.891.531.751.931.992.31<0.001
Observation
 Any soap/detergent (%)9895989999.699.8<0.001
 Body soap (%)927889959899<0.001
 Body soap that appeared to have been recently used (%)917789959799<0.001
Most convenient place with water to wash hands for the household
 Inside house (%)4148287296<0.001
 Outside house < three-steps (%)76101082<0.001
 Outside house >three and <10 steps (%)19233227101<0.001
 Outside house and > 10 steps (%)26424332101<0.001
 No place at all (%)726730.40<0.001
 Soap present at most convenient hand washing location (%)4048256994<0.001

Households with a place for residents to wash hands inside their house were much more likely to have soap present at the place of handwashing than households without a place to wash hands inside (Table 4). The farther away the place of handwashing was from the house, the less likely soap was available (chi square for trend P < 0.001). Even within wealth quintiles there was a consistent relationship between the proximity of the most convenient place to wash hands and the presence of soap (Table 4). The strength of association between the location of the most convenient place to wash hands and the presence of soap varied by wealth quintile (Table 5) (Woolf test for homogeneity P < 0.001). The strongest association (odds ratio = 70) was in the poorest quintile, but even in wealthier quintiles the location of the most convenient place to wash hands was strongly associated with presence of soap.

Table 4.   Presence of soap at the most convenient place to wash hands by location of this handwashing station and wealth quintile (n = 6971)
Place of handwashing (n)Proportion of households with soap present at handwashing station Households by wealth quintile
Overall (n = 6971) %First (Poorest) %Second %Third %Fourth %Fifth (Richest) %Chi-square for trend P-value by wealth
Inside house (2885)856248688595<0.001
Outside house & within three steps (503)301113295687<0.001
Outside house & between three and 10 steps (1303)82582735<0.001
Outside house and beyond 10 steps (1777)32331150<0.001
No place available (503)00
Chi-square for trend across soap locations<0.001<0.001<0.001<0.001<0.001<0.001 
Table 5.   Relationship between presence of soap at most convenient place to wash hands and place of hand washing stratified by wealth index (n = 6971)
Wealth Index Quintile (n)Location of the most convenient place to wash handsOdds ratio (95% confidence interval) inside vs. outside house
Inside houseOutside houseUnavailable
Q1 (1436)62% (36/58)2% (23/1009)0% (0/369)70 (35–137), < 0.001
Q2 (1364)48% (53/110)5% (54/1166)0% (0/88)19 (12–30), < 0.001
Q3 (1384)68% (263/389)9% (82/954)0% (0/41)22 (16–30), < 0.001
Q4 (1455)85% (890/1050)30% (119/399)0% (0/6)13 (10–17), < 0.001
Q5 (1332)95% (1217/1278)61% (33/54)13 (7–23), < 0.001
All (6971)85% (2459/2885)9% (311/3582)(0/504)Crude Odds Ratio 61 (52–71), < 0.001 Woolf test statistic, P < 0.001

In univariate analysis children living in households that reported usually washing their hands with soap, that were able to show a bar of body soap to the field investigator, that had a place with water to wash hands inside their house, or that had soap present at the most convenient place to wash hands were significantly less likely to have cough or difficulty breathing the preceding 7 days than children living in households without these characteristics (Table 6). In univariate analysis wealthier households were less likely to report a child in the household with cough or difficulty breathing in the preceding 7 days (Table 7).

Table 6.   Relationship between hand wash indicators and cough or difficulty breathing among children under the age of 5 years in the preceding week (n = 7992)
Soap use indicator Proportion (%) of children with this indicator with cough or difficulty breathing in the last 7 daysProportion of children with this indicator without cough or difficulty breathing in the last 7 daysCrude odds ratio Adjusted* odds ratio (95% CI)
  1. *Adjusted for wealth using PCA of asset score and for neighbourhood clustering using general estimated equations.

  2. †General purposes such as cleaning house, preparation of foods, serving foods.

  3. ‡Special purposes such as cleaning hand after defecation, cleaning hand after washing bottom of the child, washing hand before having food, washing hand after having food, feeding for child, washing hand after cleaning dirt.

  4. §Calculated based on the number of bars of hand soap of different sizes purchased in the preceding 30 days, multiplied by the weight of the bars, and divided by the number of persons in the household.

  5. ¶Above versus below median.

Reported behaviour
 Usually washes hands with soap for general purposes†6 (127)9 (551)0.680.996 (0.961–1.033), P = 0.84
 Usually washes with body soap for special purposes‡57 (1142)62 (3694)0.841.02 (0.996–1.04), = 0.11
 Usually uses soap for bathing/body cleaning 35 (688)41 (2441)0.770.99 (0.97–1.01), = 0.15
 Purchased body soap in last 1 month (%)95 (1893)95 (5674)1.11.02 (0.98–1.06), = 0.42
 Average amount spent per capita (all households included) in last 15 days for body soap (US$)0.0500.0481.06¶1.02 (0.997–1.035), P = 0.10
 Per capita soap consumption (g/day)§1.831.860.95¶1.01 (0.99–1.03), P = 0.46
Observation
 Had any soap/detergent98 (1955)98 (5897)1.031.05 (0.97–1.1), = 0.21
 Had body soap91 (1803)92 (5524)0.841.02 (0.98–1.05), P = 0.38
 Had body soap that looked recently used90 (1793)91 (5478)0.881.02 (0.99–1.06), = 0.19
 Most convenient place is inside house with water to wash hands for the household 32 (631)45 (2710)0.570.95 (0.93–0.98), P < 0.001
 Most convenient place is outside house with water to wash hands for the household 58 (1149)48 (2899)1.51.02 (0.99–1.04), P = 0.19
 No place with water to wash hands for the household 11 (209)7 (394)1.671.06 (1.02–1.10), P = 0.007
 Soap present at the most convenient place with water to wash hands32 (643)43 (2568)0.641.0 (0.97–1.02), P = 0.70
Table 7.   Risk of child in the household with cough or difficulty breathing in the last 7 days by wealth index (n = 7992)
Child’s Household Wealth index Quintile (n)Children with cough or difficulty breathingOdds Ratio (95% Confidence Interval), handwashing location inside versus outside house
All % (n)Handwashing location inside house % (n)Handwashing location outside or no convenient location % (n)
Q1 (1681)33 (557)31 (22)33 (535)0.90 (0.54–1.51), = 0.694
Q2 (1532)29 (437)24 (30)29 (407)0.78 (0.51–1.20), = 0.266
Q3 (1532)24 (360)21 (91)25 (269)0.81 (0.62–1.07), P = 0.134
Q4 (1661)21 (342)18 (215)27 (127)0.59 (0.46–0.76), P = 0.000
Q5 (1586)19 (293)18 (273)30 (20)0.50 (0.29–0.87), = 0.013
All (7992)25 (1989)19 (631)29 (1358)0.56 (0.51–0.63), P = 0.000
Chi square for trend< 0.001< 0.001< 0.001 

In the general estimated equation models that controlled for the confounding of the wealth index, the only handwashing indicator significantly associated with a child in the household who reported cough or difficulty breathing in the preceding 7 days was the presence of a place with water to wash hands inside the house (adjusted odds ratio 0.95, 95% confidence interval 0.93–0.98, P < 0.001) (Table 6). The protective effect of a place to wash hands inside the home was seen across all wealth quintiles (Woolf test for homogeneity P-value > 0.20).

Discussion

  1. Top of page
  2. SummaryAssociations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du BangladeshAsociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Several measures of handwashing including reported handwashing behaviour, presence of soap in the home, and reported soap consumption were closely associated with household wealth. These findings are consistent with work in other settings that found indicators of hygiene were less common among poorer households (Bhuiya et al. 1990; Curtis et al. 1995; Gorter et al. 1998; Armar-Klemesu et al. 2000). Handwashing promotion programs are often evaluated by comparing baseline and post-intervention assessments and assuming that any change in handwashing is due to the intervention (UNICEF 1999; Saade et al. 2001). Recent efforts to promote handwashing through mass media (Curtis et al. 2007) prevent enrolling local groups unexposed to the intervention to control for secular trend and confounding. The present study findings suggest that changes in handwashing behaviour risk being falsely attributed to handwashing promotion when they actually resulted from economic development of a community. This suggests that these handwashing indicators should not be used to assess the effectiveness of a handwashing promotion intervention unless wealth is carefully controlled for.

A place to wash hands within the household was strongly associated with the presence of soap at the handwashing location. This effect was particularly strong in the poorest households (OR 70), but was present in all wealth groups. A handwashing station outside of the household often means a location that is shared by more than one family. If the handwashing station is shared, then a family that invests in soap risks losing some or all of that investment to their neighbour. The hospital literature on handwashing promotion suggests that an environment that facilitates hand hygiene with a minimum of inconvenience increases hand hygiene practices (Kaplan & Mcguckin 1986; Graham 1990; Bischoff et al. 2000). This is consistent with behaviour change theory and interventions to promote healthier behaviours for a wide range of health issues that stress the importance of a physical environment that is facilitative for a desired behaviour (Story et al. 2002; Gielen & Sleet 2003; Saelens et al. 2003). An observational study in Kyrgystan found an increased rate of handwashing after latrine use among households that owned a washstand (Biran et al. 2005). Taken together this evidence suggests that if soap is not immediately available at the handwashing station, then handwashing with soap is less likely to occur. Thus handwashing promotion efforts that include specific efforts to provide handwashing facilities inside the house are more likely to improve handwashing rates than interventions that ignore this component (Cairncross & Cliff 1987). The ‘tippy tap’ has been proposed as a low cost approach for constructing a handwashing station when running water is not available (Watt 1988).

An important barrier to progress in improving the effectiveness of handwashing promotion is difficulties in measuring handwashing behaviour. In the present analysis the location of the most convenient place to wash hands was the one indicator of handwashing associated with a health outcome independent of socio economic status. This indicator is easy to collect through simple observation and so is not prone to courtesy bias in reporting. Future research should evaluate this indicator in other settings.

These data provide some additional evidence of the potential contribution of handwashing in preventing respiratory disease in low income countries. A meta-analysis summarizing studies conducted in developed countries on the impact of handwashing promotion programs concluded that various reported interventions on average reduced respiratory illness by 16% (Rabie & Curtis 2006). A study in Karachi, Pakistan noted a 50% reduction in pneumonia among children under the age of 5 years who lived in neighbourhoods that received intensive handwashing promotion (Luby et al. 2005). The present study found that after controlling for socioeconomic status, children under age of 5 years who lived in households that had a place inside their house to wash hands had a 5% reduced risk of respiratory symptoms in the preceding 7 days. This is a lower magnitude of reduction than noted in the other studies, but in contrast to the other literature it is in the absence of a handwashing promotion program.

There are important limitations to the conclusions that can be drawn from this study. First, the strong relationships between socioeconomic status and handwashing indicators identified in these urban Bangladeshi communities may be different in other settings, with different income level, income distribution, cultural attitudes toward hand hygiene and the availability and type of household water supply. However, optimum handwashing practice requires resources and so we would expect handwashing behaviour to vary depending on wealth especially in low income settings.

The second limitation was that this study was not specifically designed to test a priori hypotheses on the relationship of handwashing indicators to socioeconomic status and respiratory symptoms. This means that several elements that would be of interest to explore including evaluating the relationship of socio-economic status to additional handwashing indicators especially structured observation and microbiological hand contamination, and the relationship of handwashing indicators to reduced gastrointestinal disease were not included. Indeed, the study was exploratory and the study conclusions should be viewed as preliminary.

The third limitation is that the association between respiratory illness and a place to wash hands in the home may be due to residual confounding, i.e. that a place to wash hands in the home may be an additional marker for higher socio-economic status, and so its association with respiratory disease may not be independent of wealth. However, principal component analysis of household characteristics is a robust technique for measuring relative levels of socioeconomic status. Moreover, when the analysis was repeated using two additional measures in the wealth index, the location of the principal source of drinking water, and the type of latrine, the results of the analysis were largely unchanged; specifically, having a place inside the house with water to wash hands was the only handwashing indicator associated with respiratory symptoms independent of socioeconomic status (data not shown). Thus, the available evidence suggests that their was a modest effect of a place to wash hands in the home that reduced the prevalence of respiratory illness independent of socio-economic status.

Measuring handwashing is difficult. Researchers have raised many concerns with both the validity and practicality of available handwashing indicators. The additional concern raised by this analysis is that several handwashing indicators are compromised by strong confounding by socioeconomic status. New techniques to measure handwashing behaviour that address some of the many shortcomings of the current measures would substantially improve the ability to assess and optimize handwashing promotion efforts.

Acknowledgements

  1. Top of page
  2. SummaryAssociations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du BangladeshAsociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was funded by the Department of Health and Human Services National Vaccine Program Office (NVPO) through the United States Agency for International Development Global Bureau’s Global Research Activity Cooperative Agreement with Johns Hopkins University Bloomberg School of Public Health, and the Government of Bangladesh through IHP-HNPRP. ICDDR, B acknowledges with gratitude the commitment of NVPO and IHP-HNPRP to the Centre’s research efforts. The authors thank Mamun-Ar-Rashid, Iffat Sharmin and Kazi Faisal Alam who supervised the field team in data collection and Pavani Ram and Lisa Danquah who provided helpful comments on an earlier draft of the manuscript.

References

  1. Top of page
  2. SummaryAssociations entre indicateurs de lavage des mains, de la richesse, et les symptômes de maladies respiratoires infantiles dans les zones urbaines du BangladeshAsociaciones entre indicadores de lavado de manos, riqueza y síntomas de enfermedad respiratoria en niños en Bangladesh urbana. 
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Armar-Klemesu M, Ruel MT, Maxwell DG, Levin CE & Morris SS (2000) Poor maternal schooling is the main constraint to good child care practices in Accra. Journal of Nutrition 130, 15971607.
  • Bhuiya A, Streatfield PK & Meyer P (1990) Mother’s hygienic awareness, behaviour and knowledge of major childhood diseases in Matlab, Bangladesh. In: What We Know About Health Transition: The Cultural, Social and Behavioural Determinants of Health, Vol. 1 (ed. JCaldwell) Health Transition Centre, Canberra, pp. 462477.
  • Biran A, Tabyshalieva A & Salmorbekova Z (2005) Formative research for hygiene promotion in Kyrgyzstan. Health Policy Plan 20, 213221.
  • Bischoff WE, Reynolds TM, Sessler CN, Edmond MB & Wenzel RP (2000) Handwashing compliance by health care workers: the impact of introducing an accessible, alcohol-based hand antiseptic. Archives of Internal Medicine 160, 10171021.
  • Bryce J, Boschi-Pinto C, Shibuya K & Black RE (2005) WHO estimates of the causes of death in children. Lancet 365, 11471152.
  • Cairncross S & Cliff JL (1987) Water use and health in Mueda, Mozambique. Transactions of the Royal Society of Tropical Medicine and Hygiene 81, 5154.
  • Cousens S, Kanki B, Toure S, Diallo I & Curtis V (1996) Reactivity and repeatability of hygiene behaviour: structured observations from Burkina Faso. Social Science and Medicine 43, 12991308.
  • Curtis V & Cairncross S (2003) Effect of washing hands with soap on diarrhoea risk in the community: a systematic review. The Lancet Infectious Diseases 3, 275281.
  • Curtis V, Kanki B, Mertens T et al. (1995) Potties, pits and pipes: explaining hygiene behaviour in Burkina Faso. Social Science and Medicine 41, 383393.
  • Curtis VA, Garbrah-Aidoo N & Scott B (2007) Ethics in public health research: masters of marketing: bringing private sector skills to public health partnerships. American Journal of Public Health 97, 634641.
  • Dasgupta S, Huq M, Khaliquzzaman M, Pandey K & Wheeler D (2006) Who suffers from indoor air pollution? Evidence from Bangladesh Health Policy Plan 21, 444458.
  • Gielen AC & Sleet D (2003) Application of behaviour-change theories and methods to injury prevention. Epidemiologic Reviews 25, 6576.
  • Gorter AC, Sandiford P, Pauw J, Morales P, Perez RM & Alberts H (1998) Hygiene behaviour in rural Nicaragua in relation to diarrhoea. International Journal of Epidemiology 27, 10901100.
  • Gove S (1997) Integrated management of childhood illness by outpatient health workers: technical basis and overview. The WHO Working Group on Guidelines for Integrated Management of the Sick Child. Bulletin of World Health Organisation 75(Suppl. 1), 724.
  • Graham M (1990) Frequency and duration of handwashing in an intensive care unit. American Journal of Infection Control 18, 7781.
  • Hanley JA, Negassa A, Edwardes MD & Forrester JE (2003) Statistical analysis of correlated data using generalized estimating equations: an orientation. American Journal of Epidemiology 157, 364375.
  • Hawker JI, Olowokure B, Sufi F, Weinberg J, Gill N & Wilson RC (2003) Social deprivation and hospital admission for respiratory infection: an ecological study. Respiratory Medicine 97, 12191224.
  • Houweling TA, Kunst AE. & Mackenbach JP (2003) Measuring health inequality among children in developing countries: does the choice of the indicator of economic status matter? International Journal of Equity Health 2, 8.
  • Hussain A, Ali SM & Kvale G (1999) Determinants of mortality among children in the urban slums of Dhaka city, Bangladesh. Tropical Medicine and Internal Health 4, 75864.
  • Kaplan LM & Mcguckin M (1986) Increasing handwashing compliance with more accessible sinks. Infection Control 7, 408410.
  • Luby SP, Agboatwalla M, Feikin DR et al. (2005) Effect of handwashing on child health: a randomised controlled trial. Lancet 366, 225233.
  • Luby SP, Agboatwalla M, Billhimer W & Hoekstra RM (2007) Field trial of a low cost method to evaluate hand cleanliness. Tropical Medicine and Internal Health 12, 765771.
  • Manun’ebo M, Cousens S, Haggerty P et al. (1997) Measuring hygiene practices: a comparison of questionnaires with direct observations in rural Zaire. Tropical Medicine and Internal Health 2, 10151021.
  • Niport (2005) Bangladesh Demographic and Health Survey 2004. National Institute of Population Research and Training, Mitra and Associates, Dhaka, Bangladesh and Calverton, MD, USA.
  • Perry H, Weierbach R, Hossain I & Islam R (1998) Childhood immunization coverage in zone 3 of Dhaka City: the challenge of reaching impoverished households in urban Bangladesh. Bull World Health Organisation 76, 565573.
  • Rabie T & Curtis V (2006) Handwashing and risk of respiratory infections: a quantitative systematic review. Tropical Medicine and Internal Health 11, 258267.
  • Rahman MM & Shahidullah M (2001) Risk factors for acute respiratory infections among the slum infants of Dhaka city. Bangladesh Medical Research Council Bulletin 27, 5562.
  • Ruel MT & Arimond M (2002) Spot-check observational method for assessing hygiene practices: review of experience and implications for programmes. Journal of Health, Population, and Nutrition 20, 6576.
  • Saade C, Bateman M & Bendahmane DB (2001) The Story of a Successful Public–Private Partnership in Central America: Handwashing for Diarrheal Disease Prevention. The World Bank, Washington.
  • Saelens BE, Sallis JF, Black JB & Chen D (2003) Neighbourhood-based differences in physical activity: an environment scale evaluation. American Journal of Public Health 93, 15521558.
  • Schell CO, Reilly M, Rosling H, Peterson S & Ekstrom AM (2007) Socioeconomic determinants of infant mortality: a worldwide study of 152 low-, middle-, and high-income countries. Scandinavian Journal of Public Health 35, 288297.
  • Sobel J, Mahon B, Mendoza CE et al. (1998) Reduction of fecal contamination of street-vended beverages in Guatemala by a simple system for water purification and storage, handwashing, and beverage storage. The American Journal of Tropical Medicine and Hygiene 59, 380387.
  • Stanton BF, Clemens JD, Aziz KM & Rahman M (1987) Twenty-four-hour recall, knowledge-attitude-practice questionnaires, and direct observations of sanitary practices: a comparative study. Bull World Health Organisation 65, 217222.
  • Story M, Neumark-Sztainer D & French S (2002) Individual and environmental influences on adolescent eating behaviors. Journal of American Diet Association 102, S40S51.
  • UNICEF (1999) Towards Better Programming: A Manual on Hygiene Promotion. Water, Environment and Sanitation Technical Guidelines Series. UNICEF, New York.
  • Vyas S & Kumaranayake L (2006) Constructing socio-economic status indices: how to use principal components analysis. Health Policy Plan 21, 459468.
  • Watt J (1988) The Tippy Tap: a simple handwashing device for rural areas. Journal of Tropical Pediatrics 34, 9192.