Accounting for water quality in monitoring access to safe drinking-water as part of the Millennium Development Goals: lessons from five countries
Rob ES Bain a, Stephen W Gundry a, Jim A Wright b, Hong Yang b, Steve Pedley c & Jamie K Bartram d
a. Water and Health Research Centre, Merchant Venturers Building, University of Bristol, Bristol, BS8 1UB, England.
b. Geography and Environment, University of Southampton, Southampton, England.
c. Robens Centre for Public and Environmental Health, University of Surrey, Guildford, England.
d. The Water Institute, University of North Carolina, Chapel Hill, United States of America.
Correspondence to Stephen W Gundry (e-mail: email@example.com).
(Submitted: 12 August 2011 – Revised version received: 28 November 2011 – Accepted: 29 November 2011 – Published online: 20 January 2012.)
Bulletin of the World Health Organization 2012;90:228-235A. doi: 10.2471/BLT.11.094284
The United Nations Millennium Development Goals (MDGs) encapsulated a global agreement to tackle the pervasive health, social and economic effects of poverty. Their influence on national policy and development practice has been profound. From their first formulation, the MDGs included a target for access to safe drinking-water. After several revisions, this target, designated Target 7c, is now to reduce by half, between 1990 and 2015, “the proportion of the population without sustainable access to safe drinking-water and basic sanitation”.1
The World Health Organization (WHO) and the United Nations Children’s Fund (UNICEF) report progress towards this target through their Joint Monitoring Programme for Water Supply and Sanitation.2 However, the functioning of the Joint Monitoring Programme and the appropriateness of some of the indicators currently used to monitor access to safe drinking-water and the pace of improvement have been questioned.3,4 A key concern has been the inclusion of the word safe in the target and whether or not the data on water quality available are suitable for monitoring access to safe drinking-water up to 2015, as well as for providing a retrospective estimate of access at baseline in 1990.
In the 1990s, WHO and UNICEF categorized households as obtaining drinking-water from either an improved or an unimproved source, with only water from an improved source being regarded as safe. This approach is still used today by the Joint Monitoring Programme. The decision on whether a source is classified as improved or unimproved is based on expert judgement of the likelihood that a particular type of source provides safe drinking-water (Table 1). Consequently, this approach assesses access to specific types of water sources but not the quality of the water sources.
Table 1. Classification of drinking-water source types as improveda or unimproved, Joint Monitoring Programme for Water Supply and Sanitation, 201121
Further, the indicator used for monitoring progress towards the achievement of MDG Target 7c is the “proportion of households using water from an improved source”.1 This indicator conflates the requirement for access with that for safety. Although this measure has been criticized, its use was perhaps inevitable given the need for a single percentage figure that could be used as a target and given the limited availability of data on water quality that could be backdated to 1990, the baseline year for Target 7c. In 1990, there was no international system for collating data on water quality. However, the information available today suggests that there were large gaps in data for peri-urban areas in low-income countries and for small, often rural, supply systems in low-, middle- and high-income countries. Moreover, data comparability between countries was poor since different parameters were assessed using a range of sampling and reporting methods.5
Currently, progress on the drinking-water component of MDG Target 7c is judged as being “on track”. The most recent projections suggest that only 9% of the world’s population will be using drinking-water from an unimproved source in 2015 — slightly better than the target requirement of 12%.6 However, assessing the safety of drinking-water according to whether or not it comes from an improved source is likely to overestimate both the proportion of the population with access to safe drinking-water at baseline and progress towards Target 7c because many improved sources will not provide safe water, particularly in developing countries. Conversely, the proportion of unimproved sources that actually do provide safe drinking-water is likely to be small.7,8
In 2010, WHO and UNICEF released data on the quality of water sources in five countries that had participated in the Rapid Assessment of Drinking-Water Quality (RADWQ) project.9–13 The availability of these data offered the opportunity to determine how information on water quality affects Joint Monitoring Programme estimates of the proportion of the population without access to safe drinking-water.
Taking water quality into account was expected to have two main consequences. First, since some of the water sources categorized as improved at baseline in 1990 were actually unsafe, the proportion of the population without access to safe drinking-water at that time will have been greater than estimated. Second, since some of the new water sources installed since 1990 and categorized as improved were actually unsafe, the subsequent reduction in the proportion of the population without access to safe drinking-water will have been smaller than estimated.
The combined effect of these two factors on estimates of progress towards achieving MDG Target 7c will vary between countries according to the mix of improved source types in use and the actual safety of each source type. Thus, for countries with predominantly safe, well maintained, piped systems, taking water quality into account probably leads to only small adjustments in the estimates. In contrast, the adjustments are likely to be substantial for countries where a large proportion of improved water sources are poorly maintained.
The aim of this study was to determine how accounting for RADWQ data on water source quality affects Joint Monitoring Programme assessments of progress towards achieving MDG Target 7c in five countries: Ethiopia, Jordan, Nicaragua, Nigeria and Tajikistan.
Data on drinking-water safety
Between October 2004 and April 2005, the RADWQ project conducted pilot studies in eight countries to assess drinking-water safety and to determine how the Joint Monitoring Programme could be modified to take drinking-water safety into account. Each survey considered a nationally representative sample of approximately 1500 water sources, with an emphasis on the types of improved water source in widespread use.14 Each water source was assessed using a checklist of potential risk factors and tested for a restricted set of water quality parameters. Summary reports of the pilot studies carried out in Ethiopia, Jordan, Nicaragua, Nigeria and Tajikistan are available.2
Water sources were tested for the presence of thermotolerant coliform bacteria, fluoride, arsenic and nitrate compounds. Coliform bacteria serve as an indicator of the possible presence of waterborne pathogens, which cause more disease than any other drinking-water contaminant. Among chemical contaminants, fluoride and arsenic are believed to have the greatest effect on public health globally.15 They both occur naturally in groundwater in certain geological settings. Although the presence of nitrates is thought to pose a smaller threat to public health globally, it is still a concern. Moreover, as nitrates are primarily anthropogenic in origin, these compounds provide another perspective on drinking-water safety.
For each of the five countries, we used RADWQ data on drinking-water safety to determine the percentage of each type of water source that complied with WHO guidelines on thermotolerant coliforms (i.e. percentage microbial compliance) and the percentage that complied with guidelines on thermotolerant coliforms, arsenic, fluoride and nitrates (i.e. percentage overall compliance).16
Adjustment for compliance
The Joint Monitoring Programme bases its estimates of access to improved water sources on data from selected national censuses and nationally representative household surveys. The proportion of the population with access to an improved water source is derived by categorizing the water source types reported in national data as either improved or unimproved using the classification shown in Table 1. Then, for each country, the figures for access derived from individual surveys and censuses are plotted over time, separately for rural and urban populations, and least squares linear regression is used to fit temporal trend lines. When required, the trend line is extrapolated for up to 2 years beyond the available data points, after which it is assumed that the figure does not change for another 4 years. The Joint Monitoring Programme uses these trend lines to estimate the proportion of the population with access to an improved water source in any given year.
To ensure that we were applying the methodology correctly, we replicated the procedures used for the five study countries for the data contained in the Joint Monitoring Programme country reports. The figures we obtained for Ethiopia, Jordan, Nicaragua and Nigeria were checked against the published figures;17 the figures for Tajikistan were checked against revised figures provided by the Joint Monitoring Programme.
Our first step in investigating the effect of RADWQ drinking-water safety data on Joint Monitoring Programme estimates of access to safe drinking-water was to match the source types reported in the RADWQ project with those reported in the household surveys and censuses used by the Joint Monitoring Programme. In some instances, we encountered difficulties in matching the two sets of source types (details of these difficulties and how they were resolved are available from the corresponding author on request). To estimate the proportion of the population with access to safe drinking-water, we adjusted the proportion with access to water from each source type using estimates of the percentage microbial compliance and percentage overall compliance for each source type reported by the RADWQ project. Although Joint Monitoring Programme estimates of the proportion using specific source types were available for rural and urban areas separately, RADWQ compliance figures were for rural and urban areas combined. Moreover, improved sources used by only a small proportion of the population were not assessed in the RADWQ project (Table 2). In the absence of RADWQ data, we followed the approach used by the Joint Monitoring Programme and assumed that such sources were 100% compliant.
Table 2. Compliance of drinking-water sources with WHO guidelines on contamination in five countries, Rapid Assessment of Drinking-Water Quality project, 2004–2005
The revised figures for access to safe drinking-water for different source types were summed to give separate estimates for the proportion of the urban and rural population with access to safe drinking-water in each country. Trend lines were fitted to each separately, in accordance with Joint Monitoring Programme methods. Each trend line was then used to estimate the proportion with access to an improved water source that met our microbial and overall compliance criteria in 1990, 1995, 2000, 2005 and 2008. In addition, for these years, the number of individuals with access to safe drinking-water in each country was calculated using United Nations population data obtained from the Joint Monitoring Programme web site.
The MDG targets for each country except Tajikistan were recalculated using revised estimates of the proportion of the population with access to safe drinking-water at baseline in 1990. For Tajikistan, the 1995 figure was the earliest available. This process produced three sets of figures for each country for both the MDG target and for the proportion of the population with access to safe drinking-water in any year: (i) the original Joint Monitoring Programme figures; (ii) the Joint Monitoring Programme figures adjusted for percentage microbial compliance; and (iii) the Joint Monitoring Programme figures adjusted for percentage overall compliance.
Revised 2008 safe drinking-water estimates
Taking RADWQ data on water source quality (Table 2) into account substantially reduced the estimated proportion of the population with access to safe drinking-water in 2008 in four of the five study countries. After adjusting for percentage overall compliance, the estimated proportion of the Ethiopian population with access to safe drinking-water in 2008 fell by 11% (Fig. 1). The reductions in three other countries were similar: 16% in Nicaragua, 15% in Nigeria and 7% in Tajikistan. In addition, Fig. 1 also shows that this adjustment had the greatest effect in Nicaragua and Nigeria, mainly because the microbial compliance of improved water sources was low. In Ethiopia, microbial contamination of springs and protected dug wells was the main reason for the reduction; in Tajikistan, it was microbial contamination of piped supplies. In Jordan, the main source of water is a piped supply provided by public utilities, which was almost entirely free of microbial and chemical contamination (Table 2). Consequently, adjustment for percentage overall compliance resulted in only a slight reduction in the estimated proportion of the population with access to safe drinking-water. Overall, microbial contamination was the principal reason that improved water sources in all five countries were judged unsafe by the RADWQ project. Chemical contamination had only a limited effect.
Fig. 1. Estimated percentage of the population with access to safe drinking-water in 1990a and 2008 in five countries and the Millennium Development Goal target for 2015, by assessment method
JMP, Joint Monitoring Programme for Water Supply and Sanitation; RADWQ, Rapid Assessment of Drinking-Water Quality; WHO, World Health Organization.
a For Tajikistan, the baseline year was 1995.
b JMP: the percentage with access to safe drinking-water was estimated using the original JMP figures.
c JMP-MC: the percentage with access to safe drinking-water was estimated using the original JMP figures adjusted for the percentage microbial compliance (MC) with WHO guidelines on drinking-water contamination with thermotolerant coliform bacteria derived during the RADWQ project.
d JMP-OC: the percentage with access to safe drinking-water was estimated using the original JMP figures adjusted for the percentage overall compliance (OC) with WHO guidelines on drinking-water contamination with thermotolerant coliforms, arsenic, fluoride and nitrates derived during the RADWQ project.
Revised Millennium Development Goal targets
Fig. 1 shows the revised figures for the proportion of the population with access to safe drinking-water at baseline in 1990 in the five study countries and the revised MDG targets for 2015 based on these figures. Since adjustment for microbial and chemical contamination resulted in lower baseline values, the target increase in the proportion of the population that should have access to safe drinking-water by 2015 was raised. Consequently, less progress was made between baseline and 2008 than previously estimated. In addition, progress was reduced further because some improved water sources installed after baseline did not comply with guidelines on contamination and were reclassified as unsafe.
Adjustment for microbial and chemical contamination meant that in all countries except Jordan the estimated proportion of the population needing to gain access to safe drinking-water between 2008 and 2015 to reach the 2015 target increased: the absolute increase was 8% in Ethiopia, 7% in Nicaragua, 8% in Nigeria and 4% in Tajikistan. Table 3 lists the size of the population without access to safe drinking-water in the five study countries between 1990 and 2008, as estimated using the original Joint Monitoring Programme figures and using these figures adjusted for percentage microbial compliance and percentage overall compliance, respectively. Adjustment for percentage overall compliance increased the estimated number of people without access to safe drinking-water in 2008 in Ethiopia and Nigeria, the two most populous countries, by 8.9 and 22 million, respectively.
Table 3. Estimated population without access to safe drinking-water in five countries, by assessment method, 1990–2008
Our revision of the estimated proportion of the population with access to safe drinking-water involved several assumptions that could have led to systematic underestimation or overestimation. These assumptions and their likely effects are detailed in Table 4 (available at: http://www.who.int/bulletin/volumes/90/3/11-094284).
Table 4. Effect of study assumptions on the estimated percentage of the population with access to safe drinking-water
Notably, we assumed that the best estimate of the percentage of each type of water source that complied with WHO guidelines in 1990 was the same as that observed in the RADWQ project in 2004 and 2005 because equivalent data on water source quality were not available for 1990. However, our analysis did take account of the mix of improved water source types in use at baseline in each of the five countries, as recorded in Joint Monitoring Programme inventories. This ensured that we correctly accounted for a major influence on estimated overall access to safe drinking-water. In our view, the variation over time in the quality of water from each source type is likely to have a smaller effect on estimated access to safe drinking-water than the variation in the mix of improved water source types in use. Since comprehensive data on water quality were not available for different source types in each reporting period, we believe that the best estimates were obtained by extrapolating RADWQ data on compliance rates to both earlier and later periods.
In addition to assuming that the percentage of each type of water source that complies with WHO safety guidelines remains constant over time, we also assumed that the percentage is the same for urban and rural areas. The effect of these two assumptions on our analysis and on progress towards achieving MDG Target 7c will vary between countries.
One consequence of the lack of data on changes in water source quality over time is that the target of increasing the “proportion of households using water from an improved source” in MDG Target 7c encourages the installation of new improved water sources but does not provide an incentive for maintaining the quality of existing sources. Modifying the target to include both water source quality and the type of source could lead to improvements in existing sources as well as to the installation of new sources.
For water sources other than reliable piped water into the home, contamination between the source and the point of use is known to be significant and has led to increasing interest in household water treatment and safe storage.18 We did not account for this type of contamination or for the effect of home water treatment in our study because of the conceptual and methodological difficulties in doing so.19
The RADWQ project assessed one unimproved water source: vehicle tankers in Nigeria. These were used by 2.5% of the population in 2008 and their percentage overall compliance was 62%. However, we followed the standard Joint Monitoring Programme methodology and treated these tankers as an unimproved water source in our analysis since the sustainability of access was questionable.
During our study, we found it difficult to match the water source types reported in the various surveys from which the Joint Monitoring Programme obtained its data with those reported by the RADWQ project. We resolved these difficulties by adopting the same approach as the Joint Monitoring Programme, where possible, or by disaggregating water source types into broad groups using data from a more recent survey that provided a sufficiently detailed classification. In addition, we observed discrepancies between the data on access to water sources used by the Joint Monitoring Programme and those reported by the RADWQ project, particularly for the estimated coverage of boreholes in Ethiopia and Nigeria.
We noted substantial differences between countries in the level of compliance of certain types of improved water sources with WHO water safety guidelines. For example, in Ethiopia the percentage overall compliance reported in the RADWQ project for water from piped supplies from public utilities was higher than that from boreholes (80.4% versus 65.6%, respectively; Table 2), whereas in Nigeria compliance was lower for water from piped supplies than from boreholes or tube wells (77.0% versus 86.0%, respectively). These data highlight the potential for substantial water quality improvements.
In conclusion, we found that taking water quality into account substantially reduced estimates of the proportion of the population with access to safe drinking-water at baseline in four of the five study countries. Although this resulted in lower revised values for MDG Target 7c in 2015, the difference between the revised baseline and target values increased. Progress towards the target was further impeded because some improved water sources installed after the baseline year were likely to be unsafe. Countries in which a high proportion of improved water sources is poorly maintained would be most affected.
Our analysis suggests that the criterion used by the Joint Monitoring Programme to monitor progress towards MDG Target 7c, namely, whether drinking-water comes from an improved or unimproved source, is likely to lead to substantial overestimation of the number of safe sources compared with criteria that include water quality. An earlier study in Madhya Pradesh, India, led to a similar conclusion.20
A way of monitoring water supplies that records details of both access to water sources and their safety, rather than the composite parameter used at present, would be a substantial improvement. However, water quality assessment would place greater demands on national and international monitoring systems and on the data collection methods on which they rely.
For the future, we recommend, first, that the usefulness of data on water quality currently available from national monitoring programmes be assessed by comparing these data with data from the RADWQ project and, second, that the feasibility and cost of including selected water quality parameters in household surveys be evaluated.
We thank Rolf Luyendijk at UNICEF and Federico Properzi at WHO for their comments on the analysis and manuscript.
This research was funded by the Bill & Melinda Gates Foundation under grant reference 48599 to the University of Bristol.
Stephen W Gundry is named as the sole inventor on European patent number 1960104 entitled: Apparatus for determining the presence of a contaminant in a sample of water or other fluid. Patents for this apparatus have been granted or are pending in a number of other countries. Stephen W Gundry and Rob Bain are named as coinventors on International Patent Application number PCT/GB2010/050728 entitled: Apparatus for testing the quality of a fluid sample.
- Goal 7: ensure environmental sustainability [Internet]. New York: United Nations; 2011. Available from: http://www.un.org/millenniumgoals/environ.shtml [accessed 30 November 2010].
- WHO/UNICEF Joint Monitoring Programme for Water Supply and Sanitation [Internet]. Geneva: World Health Organization & United Nations Children’s Fund; 2011. Available from: http://www.wssinfo.org/ [accessed 30 November 2010].
- Easterly W. How the Millennium Development Goals are unfair to Africa. World Dev 2009; 37: 26-35 doi: 10.1016/j.worlddev.2008.02.009.
- Satterthwaite D. The Millennium Development Goals and urban poverty reduction: great expectations and nonsense statistics. Environ Urban 2003; 15: 179-90 doi: 10.1177/095624780301500208.
- Ince ME, Smith MD. Using rapid assessment techniques in development of WSPs. Presented at the Water Safety Plans: Global Experiences and Future Trends Conference; 2008 May 12–14; Lisbon, Portugal.
- The Millennium Development Goals Report 2010. New York: United Nations; 2010.
- Kravitz JD, Nyaphisi M, Mandel R, Petersen E. Quantitative bacterial examination of domestic water supplies in the Lesotho Highlands: water quality, sanitation, and village health. Bull World Health Organ 1999; 77: 829-36 pmid: 10593031.
- Guillemin F, Henry P, Uwechue N, Monjour L. Faecal contamination of rural water supply in the Sahelian area. Water Res 1991; 25: 923-7 doi: 10.1016/0043-1354(91)90139-H.
- Tadesse D, Desta A, Geyid A, Girma W, Fisseha S, Schmoll O. Rapid assessment of drinking-water quality in the Federal Democratic Republic of Ethiopia: country report of the pilot project implementation in 2004-2005. Geneva: World Health Organization & United Nations Children’s Fund; 2010. Available from: http://www.wssinfo.org/fileadmin/user_upload/resources/RADWQ_Ethiopia [accessed 30 November 2010].
- Ince M, Bashire D, Oni OOO, Awe EO, Ogbechie V, Korve K, et al. Rapid assessment of drinking-water quality in the Federal Republic of Nigeria: country report of the pilot project implementation in 2004-2005. Geneva: World Health Organization & United Nations Children’s Fund; 2010. Available from: http://www.wssinfo.org/fileadmin/user_upload/resources/RADWQ_Nigeria [accessed 11 November 2010].
- Properzi F. Rapid assessment of drinking-water quality in the Hashemite Kingdom of Jordan: country report of the pilot project implementation in 2004-2005. Geneva: World Health Organization & United Nations Children’s Fund; 2010. Available from: http://www.wssinfo.org/fileadmin/user_upload/resources/RADWQ_Jordan [accessed 30 November 2011].
- Aldana JM. Rapid assessment of drinking-water quality in the Republic of Nicaragua: country report of the pilot project implementation in 2004-2005. Geneva: World Health Organization & United Nations Children’s Fund; 2010. Available from: http://www.wssinfo.org/fileadmin/user_upload/resources/RADWQ_Nicaragua [accessed 30 November 2011].
- Aliev S, Shodmonov P, Babakhanova N, Schmoll O. Rapid assessment of drinking-water quality in the Republic of Tajikistan: country report of the pilot project implementation in 2004–2005. Geneva: World Health Organization & United Nations Children’s Fund; 2010. Available from: http://www.wssinfo.org/fileadmin/user_upload/resources/RADWQ_Tajikistan [accessed 30 November 2011].
- Rapid assessment of drinking water quality: a handbook for implementation. Geneva: World Health Organization & United Nations Children’s Fund. Available from: http://www.bvsde.paho.org/CD-GDWQ/Biblioteca/Manuales_Guias_LibrosDW/RADWQ/RADWQ%20handbook.pdf [accessed 30 November 2011].
- Hunter PR, MacDonald AM, Carter RC. Water supply and health. PLoS Med 2010; 7: e1000361- doi: 10.1371/journal.pmed.1000361 pmid: 21085692.
- Guidelines for drinking-water quality, fourth edition. Geneva: World Health Organization; 2011.
- Progress on sanitation and drinking-water: 2010 update. Geneva: World Health Organization & United Nations Children’s Fund; 2010.
- Wright J, Gundry SW, Conroy R. Household drinking water in developing countries: a systematic review of microbiological contamination between source and point-of-use. Trop Med Int Health 2004; 9: 106-17 doi: 10.1046/j.1365-3156.2003.01160.x pmid: 14728614.
- Clasen TF. Household water treatment and the Millennium Development Goals: keeping the focus on health. Environ Sci Technol 2010; 44: 7357-60 doi: 10.1021/es1018674 pmid: 20804134.
- Godfrey S, Labhasetwar P, Wate S, Pimpalkar S. How safe are the global water coverage figures? Case study from Madhya Pradesh, India. Environ Monit Assess 2011; 176: 561-74 doi: 10.1007/s10661-010-1604-3 pmid: 20635198.
- WHO/UNICEF Joint Monitoring Programme for Water Supply and Sanitation [Internet]. The drinking-water and sanitation ladders. Geneva: WHO/UNICEF JMPWSS; 2011. Available from: http://www.wssinfo.org/definitions-methods/watsan-ladder [accessed 30 November 2010].