May 2014 Update of the Sawyer Filter Bibliography

March 18, 2014 · 2 comments

Below are links to third-party publications (grey literature and peer-reviewed) on the Sawyer Filter that were shared by members of the International Network on Household Water Treatment and Safe Storage following a request from Ryan Rowe, Daniele Lantagne and Carolyn Meub.

There is also a response from Sawyer Products about comments made in Derek Baker’s review of the Sawyer Filter.

May 2014 Update

A Cluster Randomized Controlled Trial to Reduce Childhood Diarrhea Using Hollow Fiber Water Filter and/or Hygiene-Sanitation Educational Interventions. Am J Trop Med Hyg. 2014 May 27.

Lindquist ED et al.

Safe domestic potable water supplies are urgently needed to reduce childhood diarrheal disease. In periurban neighborhoods in Cochabamba, Bolivia, we conducted a cluster randomized controlled trial to evaluate the efficacy of a household-level hollow fiber filter and/or behavior change communication (BCC) on water, sanitation, and hygiene (WASH) to reduce the diarrheal disease in children less than 5 years of age. In total, 952 households were followed for a period of 12 weeks post-distribution of the study interventions. Households using Sawyer PointONE filters had significantly less diarrheal disease compared with the control arm during the intervention period, which was shown by diarrheal prevalence ratios of 0.21 (95% confidence interval [95% CI] = 0.15-0.30) for the filter arm and 0.27 (95% CI = 0.22-0.34) for the filter and WASH BCC arm. A non-significant reduction in diarrhea prevalence was reported in the WASH BCC study arm households (0.71, 95% CI = 0.59-0.86).

Annotated Bibliography on the Sawyer Filter
Baker, Derek. (2013) Merits and Limitations of the Sawyer Filter. (PDF, 132KB)

“The Sawyer Squeeze Filter for outdoor travelis convenient for water treatment. I think it would be great as a light‐weight,occasional‐use device. For disaster relief it would also make sense to distribute Sawyer filters to make drinking water safer until permanent infrastructure can be reestablished.However as a long‐term water treatment device for remote rural areas where there is no supply chain and no one to pay for the replacement filter when the donated one clogs or breaks, I think it would be unsuitable.”

Brune, Lia, et al. (2013) Monitoring and Evaluation of a Point-­‐of-­‐Use Water Treatment Pilot Project in the Peruvian Amazon. (PDF, 1.9MB)
CONAPAC is currently piloting a point-­‐of-­use (POU) water treatment system that incorporates a Sawyer membrane filter in three communities located on the Amazon and Napo river, The small population and distance between the homes of these communities has made the installation of water treatment plants an inefficient solution. As a result, these communities were selected for a pilot project to test the effectiveness of POU treatment systems in the area.

CAWST. (n.d.) Household Water Treatment and Safe Storage Product Sheet: Sawyer Point One (TM) Filter. (PDF, 152KB)
The Sawyer Point One® filter is a gravity membrane filtration technology that uses hollow fibre membranes to remove pathogens. It has a pore size of 0.1 microns, making it effective for removing bacteria, protozoa and helminths. The Point One® filter does not remove viruses (see Sawyer Point Zero Two Product Sheet for virus removal).

CAWST. (n.d.) Household Water Treatment and Safe Storage Product Sheet: Sawyer Point Zero Two(TM) Purifier. (PDF, 142KB)
The Sawyer Point Zero Two® filter is a gravity membrane filtration technology that uses hollow fibre membranes to remove pathogens from water. It has a pore size of 0.02 microns, making it effective for removing viruses, bacteria, protozoa and helminths.

Kohlitz, Jeremy, et al. (2013) Assessing reported use and microbiological performance of a point-of-use household water filter in rural Fiji. Journal of Water, Sanitation and Hygiene for Development, 3(2), pp 207–215. (Abstract/order info)
A non-governmental organisation has distributed point-of-use water filtering units in the Western Division of Fiji. We sought to understand filter utilisation and water quality: both water flowing directly out of filters and stored water. We surveyed 270 households and 6 schools on filter use and performed hydrogen sulphide bacterial indicator testing on 24 water samples directly from filters and 37 stored water samples. Our response rate was 95%. Of these, only half (52%) reported consistently filtering their drinking water. Very few (8%) reported consistent use when preparing kava, a traditional drink. Factors associated with limited filter use included lost or broken filter parts (22%) (p < 0.05) and perception of source water quality as 44% of respondents who believed their source water was safe to drink reported consistent filter use compared to 68% of respondents who did not (p < 0.01). Bacterial indicator testing using hydrogen sulphide paper-strips showed that most water samples directly from the filter (71%) and from storage vessels (76%) were contaminated. Limited levels of use and high levels of contamination in both water directly from the filter and stored water raise serious questions as to the benefit of the filter even as an interim water quality solution in this setting.

MAP International. (2012) Project 350 Plan: Part 1 of 4 – Safe Water for Rural Communities of Ecuador. (PDF, 5.2MB)
MAP International launched a Pure Water Pilot Project in 2010 to test the effects of filters on the health of rural communities. Communities from each of the three regions of Ecuador, the Coast, the Andes, and the Amazon,were included in this Pilot Project. In each community, ten families were carefully selected to use the filters, and to be community leaders. Each family received one filter, reaching a total of 50 families.

Various. (n.d.) In Their Own Words: Comments from Around the World on the Use of Sawyer Filters. (PDF, 43KB)
An excerpt:  “School attendance was about 72%.  After we installed Sawyer filters into the school, the attendance increased to 90%” – Dr. Feroz Ismail, HASWA. HASWA has been providing clean water to families in Pakistan since the flooding occurred. The sharing model we use provides clean water for the children during the school day and at the end of the day each child brings a two liter bottle of clean water home to share with the family.

Bookmark and Share

Leave a Comment

{ 2 comments… read them below or add one }

Derek Baker April 1, 2014 at 1:13 pm

Response to Sawyer Products by Derek Baker April, 2014
Response to: Letter from Sawyer Products Inc. by John T. Smith available at:
Letter from Sawyer Products Inc. comment on the: Merits and Limitations of the Sawyer filter by Derek Baker 2013, available at:
Biosand Filter Effectiveness and Maintenance
Sawyer Products’ remark: “Mr. Baker neglects to point out that sand filters take time to form bio layers, are slow to filter water once the bio layer is formed, and have questionable periods as to when the bio layers are working. Sand filters have to have the bio layers cleaned and re-grown periodically – during which time the water is not safe to drink, and needs to be maintained by a trained person.”
The biosand filter (BSF) is the common name used for a sand filter designed for household use. The biosand filter involves four mechanisms to remove microbial contaminants from water; mechanical trapping, adsorption, natural die-off, and predation. The predation mechanism results from biological processes that naturally develop in the filtration sand in the first few weeks of use. The other mechanisms begin working immediately – suspended solids are trapped on top of the sand, and microbes become attached to the sand grains where they naturally die-off. Within the first 30 days of operation (of a 10+ year life-span), the user is advised not to drink the water from the filter without first disinfecting it.
As with the Sawyer filter, the biosand filter (BSF) needs to be cleaned of the material it has removed from the water. The BSF is small enough to ‘clean in place’, using a technique that restores the flow rate without the need to remove the sand and without any special devices. When the flow rate becomes too slow, the user is advised to ‘clean’. the top 1 – 2 cm of sand manually by filling the reservoir with about 4 L of water, removing the diffuser, swirling the top of the sand with the palm of the hand to stir up the dirt, then scooping out the dirty water with a cup and dumping it into a soak pit (or the bushes). The steps in this ‘swirl-and-dump’ technique are repeated as many times as necessary to restore the flow rate. Normally it requires 15-20 minutes for cleaning the BSF in this manner.
Sawyer Products Inc. concerns about the removal effectiveness during biolayer formation and cleaning of the biosand filter appear to be overstated. Research by Jenkins et al. (2011) 1 found that disturbance of the biolayer by cleaning had no measurable effect on virus removal and only a modest reductive effect on bacterial and turbidity removal, based on measurements taken seven days or more after the disturbance. In another study (Napotnik & Jellison, 2011) 2, the biosand filters were cleaned a total of 72 times – in 22 cases (31%) the bacteria removal effectiveness (E. Coli) was lower in the next sample tested after the cleaning; in the other 50 cases (69%) of the cleanings the next water test showed as good or better bacteria removal.
The remark that “Sand filters… needs to be maintained by a trained person.” is a false statement. The biosand filter does not need to be maintained by a trained person. The cleaning process mentioned above is the only maintenance required, and it is performed entirely by the user.
1 Jenkins, M. W., Tiwari, S. K. and Darby, J. (2011) Bacterial, viral and turbidity removal by intermittent slow sand filtration for household use in developing countries: Experimental investigation and modeling. Water Research 45(18) 6227-6239.
2 Napotnik, J., and Jellison, K., Optimizing the Biosand Filter Lehigh University, presented at Disinfection 2011 conference.
Sawyer Filter Effectiveness and Maintenance
Sawyer Products’ state: “…the pore size through the [Sawyer filter] fiber walls that are 0.1 micron absolute. This makes the fibers extremely easy to clean with as much as over 99% regeneration of original filtering capabilities after backwashing. He [Mr. Baker] wrongly assumes, based on no knowledge of the filter’s fibers that the filters will need replacing. The Sawyer fibers are robust and in this application they can last decades. Therefore no “replacement supply chain” is needed.”
 The independent study of the Sawyer Filter in Fiji (Kohlitz et al. 2013)3 surveyed 270 households and 6 schools on filter use with a 95% response rate. This paper states:
“only half (52%) reported consistently filtering their drinking water…Factors associated with limited filter use included lost or broken filter parts (22%) (p 100 CFU/100ml, classified as dangerous by the WHO). (Average turbidity removal rate: 53%, Average E. coli removal rate: 50%).
“The filters which produced water of poor quality looked as though they were fully functional during field inspection, and they showed no signs of breakage or malfunctioning” (PWW, 2013)5.
 In the Kohlitz et al. 20133 study bacterial indicator water testing was carried out on 24 water samples directly from filters and 37 samples from stored water:
”Bacterial indicator testing using hydrogen sulphide paper-strips showed that most water samples directly from the filter (71%) and from storage vessels (76%) were contaminated.”
Dr. Colleen Chan of the British Columbia Institute of Technology6 has suggested that this water quality issue may result from two concurrent problems in the filters:
1. Biological clogging of the membrane which cannot be removed by simple backwashing. The filter membrane can be clogged not only physically by dirt and other sediments, but also biologically. A biofilm can form and grow on the membrane. When a biological layer forms on the surface of the filter membrane, it is not easily removed by backwashing with water. The biolayer tends to re-establish quickly.
2. Membranes are often fragile and prone to breakage. As the membranes become clogged biologically, users need to use more force to backwash the filters, resulting in breakage. Cracks in the filter membrane enable water to pass through the filter at a normal flow rate, without being filtered by the small pores. There is no visual indication when cracks form.
Scanning electron microscope (SEM) images and pictures comparing a new Sawyer Filter and a poorly functioning filter from a community in Honduras by (Hopper, C. et al, 2013)7 showed membranes of a used Sawyer filter have been clogged with bacteria, and membranes have ruptured. Sawyer Products state that “These contrasts to a HFM filter where if water is flowing, that water is safe. “ This is wrong. Water will flow in a Sawyer filter with ruptured membrane. The problem is that users can’t tell if the membrane is ruptured. The flow rate of a ruptured Sawyer can be similar to a functioning Sawyer, and users will continue to think that the filter is working, but in fact, the water is contaminated.
Further field research is required on Sawyer filters in order to determine whether the growth of a biological layer is causing clogging of the filter membrane, followed by breaking of the filter tubes. This is an issue which potentially compromises the effectiveness of the filters. Until more research is completed, it is recommended that water quality testing of all Sawyer filters be undertaken regularly to check for membrane functionality, and that chlorine be used when backwashing the filter to remove any biological material from the filter.
3 Kohlitz, Jeremy, et al. (2013) Assessing reported use and microbiological performance of a point-of-use household water filter in rural Fiji. Journal of Water, Sanitation and Hygiene for Development, 3(2), pp 207–215. Abstract provided in WASHplus Household Drinking Water Quality Update.
4 MAP International. (2012) Project 350: Part 3 of 4 – Safe Water for Rural Communities of Ecuador.
5 A series of pilot studies by Pure Water for the World (PWW), Honduras, on implementation of the Sawyer Point One® filter. The studies involved water quality testing, inspections and user interviews for filters installed in the communities of El Naranjal (8 filters), San Ramon (32 filters), San Francisco de las Quebradas (38 filters) and Santa Rosa (22 filters) in Trojes, Honduras (PWW, 2013).
6 Personal communications with Dr. Colleen Chan, Feb 2014
7 Hopper, C., Stewart, B., & Robbins, E. (2013) University of Maine and Bangor High School’s STEM Academy, Investigation of Six Sawyer Filters Removed from San Francisco, Honduras.

Virus Removal
Sawyer Products’ remark: “…while we do not make a claim of viral removal the very same science that allows the sand filters to remove viruses also allows the Sawyer PointOne filter to remove viruses as well. Water borne viruses clump in particulate to keep from being exposed to UV light. This natural occurring process makes it easy for the sand filter, as well as the PointOne filter, to remove them. But the big difference is the Sawyer filter is removing them on day one, not 200 days later.”
Response: Unfortunately virus removal is not simply removing particulates with viruses attached. Viruses also ‘swim alone’. Viruses are negatively-charged and very tiny. For example, rotaviruses are found worldwide, causing major diarrhea-associated hospitalization and 600,000-850,000 deaths per year8. The size of rotaviruses is 0.06 – 0.08 microns, substantially less than “the pore size through the fiber walls that are 0.1 micron absolute“.
In the biosand filter, research (Elliot et al. 2011)9 found that the most probable mechanism of virus removal is predation: “production of microbial exoproducts such as proteolytic enzymes or grazing of bacteria and higher microorganisms on virus particles”. Since this mechanism involves biological processes, these require time to develop. Virus removal in the biosand filter generally does not begin before approximately 2 weeks of use and then increases more or less exponentially reaching 99% after 150 days and over 99.99% after about 200 days according to one study (Bradley et al. 2011)10.
The Sawyer filter and the biosand filter differ in an important area; the Sawyer filter is a membrane filter while a biosand filter is a granular depth filter. The Sawyer filter depends on a single pass through a thin membrane to filter pathogens from water. In the biosand filter, the water filtration takes place as the water moves through a sand bed over 20 inches deep. With a single membrane filter, rupturing of the membrane (i.e. the walls of the hollow fiber) would mean that some amount of contaminated water would bypass the membrane and flow out of the filter as ‘filtered’ water. If the hollow fiber membrane should become ruptured, bacteria and other pathogens would be detected in the ‘filtered’ water. In the biosand filter, the granular packing of the filter media prevents this type of short circuiting from occurring.
8 Microbiology and Immunology On-line, University of South Carolina School of Medicine
9 Elliott, M.A., et al., Virus attenuation by microbial mechanisms during the idle time of a household slow sand filter, Water Research (2010), doi:10.1016/j.watres.2011.05.008
10 Bradley, I., Straub, A., Maraccini, P., Markazi, S. and Nguyen, T. H. (2011) Iron oxide amended biosand filters for virus removal, Water Research, 45(15), 4501 – 4510.
Please contact Derek Baker at if you have and questions or comments.
*** End of Document ***


Darrel Larson April 10, 2014 at 12:20 am

My organization, Give Clean Water, is the one referred to in this report. The report was a directive from the Fiji Ministry of Health and done through a Doctoral student with the intent of assessing the benefits of the Sawyer filter. When we first started working in Fiji, we had no official partnership with the Fiji Ministry of Health. We worked with locals in Fiji to install the very first Sawyer filters in several villages. We learned so much from the report listed above. The report made us pay attention to the need for a more formal partnership with the Ministry of Health to ensure continuing education and follow ups were done in each of the villages. The report also addressed the need to establish spare parts solutions in order for the filters to have sustainable use.

As a result of the findings of this report, the Fiji Ministry of Health assessed that they should make an official partnership with Give Clean Water to ensure continuing education and follow up would happen in the villages that received Sawyer filters. On September 18th, of 2013, Give Clean Water signed a Memorandum of Understanding (MOU) with the Fiji Ministry of Health. This contract gives the full backing of the Ministry of Health to accompany Give Clean Water on data collection, installation and follow up trips to the villages to inspect filters, and give continuing education on the use of the filters. Local Health workers who live in the villages now conduct regular follow up visits to each family as a part of their Health Worker responsibilities. The villages in which this process has been established have experienced significant successes in filter use and maintenance. Give Clean Water has also worked to equip the health workers with spare parts, such as replacement syringes for back flushing, to ensure filters can be maintained properly.

Much of what the above report documented were issues mostly focused on end user education, community partnerships, and spare parts…not the effectiveness of the Sawyer filters. Bacteria counts observed were primarily determined to have been “user error” in nature.

As an organization, we are committed to continuous improvement in our follow up process, our communication with the Fiji Ministry of Health, and the sustainability of our project in Fiji and around the world. We will always be learners. Sustainability is not an easy task. It requires lots of work, improvements where necessary, and a commitment to collecting data that documents the health changes the Sawyer filters are bringing.

My personal experience is that the Sawyer Filters are improving the health of the people who live in the rural villages in Fiji. Our follow up data as part of our MOU support this as well. As with any filter device, the end user plays a big role in its success and sustainability. It doesn’t matter if it’s a bio sand filter, a ceramic filter, or a membrane filter, all of which are good, the effectiveness of each will largely depend on working WITH people locally, continuously improving processes, and providing real data on successes and best practices.


Previous post:

Next post: