Sweat trickles down the back of Aaron Heerboth’s gray T-shirt as he looks at the endless line of patients before him. A “sadhu,” or holy man, with decades of dreadlocks swirled on his head holds his swollen jaw. Three barefoot ladies in faded flowered saris grimace as they cough. Dozens of pilgrims with various ailments queue behind them. At the front of the line, an elderly man with broken plastic sandals carries his limp granddaughter in his arms.
Heerboth watches as an Indian doctor shines a light into the unconscious girl’s dark brown eyes. Her grandfather says she has a fever after bathing in the Ganges when temperatures were near freezing. The doctor quickly determines she’s got flu-like symptoms, is dehydrated and possibly contagious. While he notes her ailments in a logbook, a nurse leads the grandfather to lay the girl down on a nearby hospital bed, and then gives her an injection to help with the dehydration.
Hours later, the girl’s diagnosis is visible in an app that can be seen by Heerboth’s colleagues in New York, Boston, Delhi and Mumbai. She is one data point among thousands. Researchers can see her age, chief complaint, diagnosis and prescribed medications on line graphs and pie charts. On their screens, they are watching the pulse of the Maha Kumbh Mela, the world’s largest religious gathering.
Such mass gatherings can be hotbeds of disease transmission. From cholera to bird flu, researchers are studying how diseases spread at such events, in the hopes of preventing a future pandemic.
But in this case, Heerboth and his team have a broader goal. Every 12 years — over just a few weeks — the Kumbh’s network of hospitals, roads, homes and businesses is built anew. For the team, the Kumbh is a megacity under a microscope, one that can be monitored in a top-down way that would be impossible for an existing city.
What’s more, the Kumbh is a model of what’s to come. The population of urban areas around the world is climbing, and most of that growth is taking place in Asia and Africa. By 2050, the United Nations expects two-thirds of the world’s inhabitants to be city dwellers. The scientists studying the Kumbh are using it as a test bed for the future. How will Earth’s swelling cities provide health care to the millions of new residents?
Medicine and Metrics
Even before sunrise, the beating of drums and methodic chants begin. Pilgrims carry cooking pots, bags of rice and bedrolls on their heads as they push their way to the river for a holy dip. The leader of one group wears fluorescent orange pajamas and holds a trident in the air — a signpost for those traveling with him. Ladies clench each other’s saris to stay together, and small children ride on their parents’ shoulders. One teenage boy who has become separated from his family frantically screams out their names, his cries drowned out by loudspeaker announcements from a nearby telephone pole.
In size and duration, the Kumbh dwarfs all other mass gatherings. The grounds where it is held, at the convergence of the Ganges and Yamuna rivers, transform for three months every 12 years into a teeming megacity. With its roughly 70 million pilgrims, the Kumbh is larger than any single city on Earth. From empty floodplains sprouts a metropolis with a population bigger than Tokyo, Jakarta and Delhi combined.
Pilgrims come from every corner of India and the globe to bathe in the rivers, where Hindu mythology says an ancient scuffle between gods and demons led to the spilling of amrit, the nectar of immortality. Hindus believe that bathing in these waters washes away their sins and brings them a step closer to immortality.
And while they’re there, hundreds of thousands of them will need medical attention. That’s where Satchit Balsari spotted a need.
Balsari, a Mumbai native, is an emergency room physician at Weill Cornell Medical College in New York. Tall and articulate, Balsari’s big brown eyes light up when he talks about the future of health care in India.
In 2008, while he was a medical resident at New York-Presbyterian Hospital, Balsari and two colleagues were struck by how little data existed about emergency rooms in non-Western countries. Records of the number of patients or the patients’ ailments were poorly maintained, and almost always on paper charts. It was next to impossible to match supply with demand to improve care or anticipate staffing needs.
To remedy that problem, the young doctors teamed up with programmers to create a software tool. The software, called EMcounter, served as an easy form to collect “who, what and when” data for every patient visiting an emergency room. The team tried out the software at a hospital in Chennai, India, where it succeeded in making the ER run more efficiently and improved the doctors’ training.
A few years later, when Balsari heard that Harvard’s School of Public Health would be coming to the Kumbh, he wondered if a digital surveillance counter like EMcounter could be useful. “When I thought about this large mass gathering and the challenges faced by it, I realized that there was not a way to track diseases in real time,” says Balsari.
Balsari selected researchers from Harvard, Cornell, Yale and the University of California, San Francisco, as well as medical institutions across India to create a surveillance system of health care at the Kumbh.
The data would help researchers better understand disease distribution over time and in a large population, as well as how the Kumbh hospitals use their resources. What medicines were used most? How could doctors best serve the patient population? What were the patients’ most common ailments? When did patient numbers peak, and when did they subside?
Balsari knew that his desktop software wouldn’t work for this situation. Doctors at the Kumbh would face power outages, an unreliable Internet connection and an overwhelming number of patients. So he and his team set about re-creating EMcounter as an iPad app that would store data locally and upload it to a remote server when an Internet connection was available.
In the fall of 2012, Balsari approached his student Aaron Heerboth to be the project’s field supervisor. Although Heerboth and Balsari previously worked together on global health projects in India, this would be far more challenging than any they had experienced.
On the Ground
Heerboth arrived at the Kumbh in late January 2013 as it neared its peak. The festival had been up and running for a few weeks, but the city looked like it had always been there. Rows of tents formed a multicolored checkerboard pattern, the dirt lanes in between lined with vendors. Some sold bottles of holy water from the Ganges, while others haggled with potential buyers over beaded bracelets, candles and flowers for offerings. A parade float carrying sadhus barreled recklessly through the crowded, potholed dirt road. On the other side of the street, people gathered around a brightly painted elephant that pulled coins from gawking onlookers’ hands.
But inside one of the nearby hospital tents, tempers were wearing thin. People ran around harried. A line of patients stretched out the door. Doctors handed hastily written notes to medical volunteers, many of them sporting dark circles under their eyes from late nights transcribing those notes. Heerboth tried to help, standing nearby to assist the doctors, but the line of patients grew longer and the doctors’ visits shorter, often ending with a haphazardly written prescription and a packet of pills shoved at a patient.
As the crowds swelled to nearly 30 million people on the largest bathing day, two weeks after Heerboth arrived, the morale of the doctors and medical students hit rock bottom. It was taking hours for them to log the patient information into the iPads, and the rate of user error was high. A group of volunteers threatened to quit. The entire project, years in the making, was on the line with just two weeks left of the Kumbh.
On short order, Rishi Madhok, the software’s lead programmer, flew to Allahabad. After touring the hospitals, Madhok and Heerboth sat down to talk programming at a bar outside the strictly alcohol-free Kumbh. They quickly realized the answer: Their survey tool could replace the hospitals’ paper record-keeping entirely. Instead of having volunteers laboriously transcribe data, it could be digital from the outset.
Within days, Madhok revised the software to allow doctors to input patient data digitally. Upon diagnosis, the screen would prompt them with a multiple-choice selection of treatment options. Prescriptions entered in the software were instantly shared with the pharmacist a few meters away, theoretically reducing prescription errors. And the data finally began to flow.
Care By the Numbers
In all, over 280,000 people received medical treatment at the Kumbh — twice as many as some of the busiest hospitals in America would see in a year. Heerboth and his team used their digital monitoring system to record data on almost 50,000 of these patients, making theirs the largest-ever digital medical dataset from a mass gathering.
The most common patient complaints were sore muscles or joints, fever and cough. Although waterborne outbreaks can be a risk when people are densely clustered, only a small number of people reported having diarrhea. And fortunately, there were no major outbreaks of infectious diseases, such as cholera.
The only outbreak detected was a spike in patients complaining of cough and respiratory symptoms around the busiest bathing day. This likely indicated a minor viral outbreak spreading through the crowd.
Deeper in the data, however, were some intriguing findings about the hospitals themselves. One of the most alarming findings had to do with medications prescribed. Ninety percent of patients who visited the smaller satellite hospitals received medication. Some drugs were relatively harmless, such as acetaminophen or multivitamins, but many other patients were given short courses of antibiotics without a proper diagnosis or complete instructions to the patient.
“The overprescribing of antibiotics that we’re seeing at the Kumbh is endemic to clinics across India,” says Balsari. Those practices have already made India one of the global hotspots for drug-resistant “superbugs.” Balsari believes clinical guidelines are needed to better guide doctors’ prescribing behaviors at future Kumbhs.
The data also pointed to ways that hospitals could run more efficiently. The busiest satellite hospital saw more than six times as many patients as the least busy hospital, even though they were identically staffed. “A sector hospital could have 18 free beds, be fully stocked with medicine and have zero patients,” Heerboth says. “Another more highly trafficked hospital may be packed with patients and be out of medicine and supplies.”
Based on this data, Balsari worked with the Indian government to improve the allocation of resources at the smaller Nashik Kumbh, held over the summer of 2015.
Heerboth spent that summer training health care providers for the Nashik Kumbh. Instead of tediously transcribing paper records, doctors in 37 of the 55 hospitals at the Kumbh used a mobile tablet-based disease monitoring system. These systems provide real-time resource utilization data to medical staff at the event, administrators in Mumbai and researchers back at Harvard, who can then help Indian officials match supply and demand of personnel, pharmaceuticals and equipment, while also monitoring for epidemics.
And the researchers took their project one step further, screening a subset of patients for diabetes, hypertension and oral cancers. Since many pilgrims come from rural areas that may not even have a doctor, the Kumbh represents a huge opportunity not just for medical treatment, but for preventive care. The people who are screened will have an option to receive automated text messages over a period of months, reminding them to follow up with doctors or to take medications as prescribed. In time, the researchers hope, the frequent religious gatherings in India could take on a secondary role as massive public health interventions.
Cities of the Future
As people continue to move to cities, the Kumbh provides a useful testing ground for the provision of medical care to the world’s millions of new urban residents. In India alone, the urban population will likely double in the next 30 years, to 700 million. City planners would be wise to heed lessons about where to place hospitals, how to staff them and what kinds of ailments to anticipate.
Even better, the researchers say, would be if urban hospitals conducted their own data-driven planning. Simple digital medical records and well-designed apps could allow medical systems to quickly respond to the changing populations of their cities.
And just as cell phones came to many parts of India before landlines ever did, medical technology could also “leapfrog” over the era of paper record-keeping straight to digital. Perhaps, Balsari says, in the absence of a national system of medical records, Indians’ medical records could be individually stored in their phones.
The solutions, in other words, might look nothing like what’s worked in other places. And that’s the idea.
“There are enormous bureaucratic hurdles that have to be overcome to implement projects like this on the ground and change the face of health care in India and other countries in South Asia,” Balsari says. “What we are doing here is a first step in how to meet the needs of the future.