When you design new interventions, it’s vital to get out into the field to pilot and prototype them on a small scale first. Doing so can throw up all sorts of issues that may be easy to overlook, but if left unaddressed could undermine the impact of your program.
We recently took to the field in Bangladesh where we are running a project with international development organisations BRAC and BIGD to reduce coronavirus transmission. Here, we’re installing 1,000 public handwashing stations at busy locations such as outside mosques, markets, schools and bus stands, to improve access to hygiene facilities in a context where many households do not have soap and running water at home.
Fig 1. An image of one of BRAC’s public handwashing stations.
To better understand whether our handwashing stations were ‘fit for purpose’, we collected and analysed data from multiple sources on a subset of stations including user feedback, qualitative interviews, direct observation by enumerators and technical prototypes. From this we gleaned 4 key insights into how people behaved at stations which allowed us to make crucial iterations to the station design. These adaptations will improve our chances of reducing coronavirus transmission when the full set of stations is rolled out over the next few months.
1. Make stations easy to use for everyone
Enumerators from our partner D2 observed how people used a set of 14 pilot stations across a 10-day period. User feedback revealed that the foot pedals were often too firm and too high, making them difficult to operate, particularly for children and the elderly.
In response to this feedback, BRAC has altered the construction of the foot pedals in two ways. First, the pedal height has been reduced from 6-10 inches above the floor down to 4 inches. Second, the left and right pedals (for soap and water respectively) have been moved closer to one another in the centre of the sinks so that people can easily use their strongest foot to operate both pedals rather than switching legs.
Fig 2. Foot pedals on the pilot stations (left) were too high off the ground and spread too far apart. BRAC adjusted the design ready for the full-scale roll out (right).
2. Beware of unhelpful social norms
The enumerators also recorded the number of users at each of the sinks, as well as detailed information on handwashing practises for each user (e.g., whether they used soap and the duration of handwashing).
Overall, the quality of handwashing practises was high. Almost all users of the stations washed their hands with soap, and 75% washed their hands for at least 11 seconds.
However, despite being positioned at busy locations, each station was only used about 10 times per hour on average and queues were small or non-existent. The lack of queues may create an unhelpful social norm, giving the impression that people don’t use the stations, which may in turn deter use.
We’re designing a package of interventions to address this norm. For example, we’re going to explore whether it helps to record the cumulative number of handwashes at the station on a large board (much like how charities use “fundraising thermometers” to record cumulative donations). This board would show potential users that other people use the station even though queues are not visible. We will test the interventions by conducting a large-scale randomised controlled trial (RCT) in the field over the next few months.
3. Make station monitoring cheap & scalable
It’s important to monitor station use. But having an observer sit by all 1,000 stations all hours of the day is neither cheap nor scalable.
We therefore prototyped a new data collection method at the pilot stations. We attached clickers to the foot pedals, which count the number of times each pedal is pressed.
Video 1. The original clicker mechanism: the foot pedal constantly presses on the clicker when not in use.
In most cases the clicker counts were highly correlated with the number of uses recorded by the enumerators. However, in a few instances the clickers broke. In response, BRAC’s team of engineers went back to the workshop to design a new mechanism (see below), which minimises contact between the pedal and the clicker to improve durability.
Video 2. The new clicker mechanism minimises contact between the pedals and the clicker, to improve durability.
4. Cultural norms impact people’s behaviour in unanticipated ways
BIGD conducted a series of interviews with people living in the areas surrounding the pilot stations. This included over 200 short phone surveys, as well as 25 in-depth interviews and 8 focus group discussions.
These interviews revealed that some station locations, such as close to markets and mosques, are male-dominated areas that women are much less likely to visit. Indeed, this placement may inadvertently communicate that the stations are intended for men. In addition, some women reported feeling shy or uncertain about washing their hands in public. This was reflected in the data too: women comprised only 13% of total uses of the stations.
The gender gap in usage is hard to address because women spend considerably less time outside their homes than men, particularly in rural areas. However, we plan to modify signs which point to the stations to make it clear that they are intended for women as well as men. In addition, BRAC will promote female participation through in-person demonstrations in the communities surrounding the stations.
What next for the handwashing stations?
By piloting and prototyping an initial set of 14 stations, we have been able to pre-empt functional issues, iterate on the design, and identify barriers to usage which we can address through additional interventions in the full-scale roll out. This builds on work we previously wrote about, where we used an online RCT to test variants of the instructional handwashing posters placed behind the sinks on the stations (see also our case study on the Covid-19 Hygiene Hub).
We are now ready to roll out the full set of 1,000 stations, and test what works to increase usage with a field RCT. By increasing the uptake of public handwashing stations we hope to slow the spread of coronavirus and improve public health in a setting where access to hygiene facilities is limited.