A do-it-yourself abacus to diagnose infant pneumonia, a leading killer of children? Sound too good to be true? Must have just been invented on a cell phone, right?

No, the breath abacus is decades old and was invented in India.

Most experts in maternal and newborn help know about Abhay and Rani Bang and their Society for Education, Action and Research in Community Health (SEARCH) program in Gadchioroli, India. Numerous articles quote the stunning success of the Bangs' work in training relatively low-skilled health providers to diagnose and treat common killers of infants. A 2011 profile in The Lancet notes that the trained workers have a case fatality is equal to or lower than in small and medium sized healthcare facilities. (The Bangs have been profiled in TIME) [http://www.time.com/time/magazine/article/0,9171,1124299,00.html] as well.

When the Bangs started their work, they soon found that pneumonia was responsible for some 40 percent of infant deaths. Yet, the traditional birth attendants could not execute some critical tasks like breath counting because of their limited literacy and numeracy skills. Yet the health workers knew how to count to twelve.

The Bangs invented a simple low-cost breath abacus (pictured here) to train the health workers to count breaths. The top row of beads was for infants older than two months, the bottom row for infants for infants younger than two months. An hourglass timer told the health workers when a minute was up. Quite simply, the health worker starts the hourglass timer and then counts breaths, pushing the bead over after every 12 breaths. If they push the RED bead over before the minute is up, they diagnose pneumonia. The Bangs themselves stated that they were literally able to turn low-literate health workers into 'doctors' capable of diagnosing pneumonia with extreme precision.

This is a stunningly simple technology about which we have many questions. Where has the breath abacus spread? How can it be replicated in-country? Would funds be best spent inventing a new microfluidic device to detect ARI or better spent rapidly deploying the abacus which has now been around for decades? Does our obsession with new gadgets and Android devices cause us to overlook extremely effective solutions that already exist? What do you think?