In February, EPM editor Logan Plaster traveled to Allahabad, India with a global health team from Harvard to study healthcare delivery at the Kumbh Mela – the largest human gathering in history.
Last month marked the end of India’s Kumbh Mela, a Hindu festival billed as the world’s largest human gathering. Over the course of the 55-day festival, as many as 100 million ascetics and pilgrims traveled by train, car and foot to perform a bathing ritual in the Ganges river in the city of Allahabad. Some came for a single dip while others settled for weeks, inhabiting a temporary tent camp that is arguably the largest pop-up mega city ever erected.
Just how big is the Kumbh? The number of people present on the busiest bathing days – about 30 million – is roughly the population of Shanghai and New York City combined. But instead of living in dense high rises, the nomadic pilgrims of the Kumbh reside in tents on a fair ground that is 7.5 square miles – an area only slightly larger than the footprint of the Atlanta Airport and roughly a quarter the size of Manhattan. Making matters even more challenging is the unique fact that the Kumbh Mela is completely temporary. In a dry river bed that is submerged for part of the year, officials line out wide avenues, pontoon bridges and rows upon rows of street lights. By the end of March, the entire city will have been dismantled. By the time the monsoons arrive, almost the entire area of the Kumbh will be reclaimed by the rising rivers.
In February I had the privilege of traveling to the Kumbh with a team from Harvard’s FXB Center for Health and Human Rights to study how healthcare is delivered in this unique and challenging environment. Specifically, given the prevalence of many communicable diseases in rural India, how could local authorities monitor illnesses in a way that allowed early warning of disease outbreak? What we found was both impressive and delicate – an orderly, seemingly well stocked system always just one disaster away from being massively overwhelmed.
One of the goals of our trip was to perform an experiment. Would it be possible – and helpful – to deploy an electronic record system to help health clinics record and collate complaints so that they can be tracked over time? Would this be a sustainable way to track important changes in disease presentation (like diarrhea, to pick an ever-present threat) and create an early warning system for outbreaks? As it stood at the Kumbh—and in much of India—such a warning sign would come anecdotally, and only after hundreds had fallen ill.
The first step was to gain a comprehensive understanding of the healthcare system at the Kumbh. The health facilities at the festival are impressive by local standards, but overextended and underutilized by any American perspective. The grounds are divided into 10 sectors with one health clinic per sector.
They’re clean, well stocked and staffed 24/7 by rotating physicians. According to Dhruv Kazi, a cardiologist and healthcare economist on our team who was born in Bombay, this is good representation of what the Indian government can accomplish when it so desires. Each day between 500 and 800 patients arrive and are seen – briefly – by one of the physicians on duty. While that number might seem high compared with the daily census at any American emergency department, it is only the slimmest fraction of the Kumbh’s population. Why these health clinics have such low utilization per capita remains unclear.
Clinic doctors come from government clinics from around the state and are assigned to the Mela for two months apiece. The doctors work in 8-hour shifts, have no official days off, and sleep in tents that are pitched adjacent to the clinic. Each hospital has a pharmacy with over 90 drugs that are provided free of charge.
At the center of this pop-up health system is a central hospital, where patients can be seen by a range of specialists, including orthopedics, surgery, and obstetrics. There is a 100-bed inpatient unit and a 2-bed ICU. Diagnostic tools such as X-ray, ultrasound and electrocardiograms are available. Amazingly, during our visit to this hospital, it was anything but overwhelmed. There were many empty beds and there were virtually no queues. In light of the millions upon millions of pilgrims camping nearby, one could only conclude that the relative tranquility of the hospital had much more to do with a lack of utilization than an inherent efficiency or readiness. How would trauma be handled on a larger scale? How was the sector hospital prepared to surge in size in the case of a real emergency?
Connecting these hospitals was a fleet of more than 100 ambulances which were responsible for transferring patients from the sector hospitals to the central hospital. The ambulances, like the doctors who staff the hospitals, were drafted from community health centers across the state. Each ambulance arrives with its own driver, who is then provided with accommodation at the Mela.
“The ambulances themselves appear to be new and well maintained, with clean stretchers to transport patients and a hand-held radio device for communicating between ambulances and with central dispatch,” said Kazi. “Each ambulance carries an oxygen tank, a host of emergency medications, and four disaster kits: for drowning, burns, bomb blasts and stampedes. It is evident that a reasonable amount of thought has gone into designing each of the kits, but there are no paramedics (which is typical in India) and a physician must accompany seriously-ill patients. It appears that an ambulance makes 5-6 trips a day.”
Yet, while the facilities at the sector hospitals may have been well stocked, health records were nearly non-existent. As our team observed, after a one-glance patient encounter, the doctor quickly scrawled down age, sex and a chief complaint. These notes were mostly illegible, largely incomplete and essentially useless. It’s understandable given the strain on each doctor, but it made syndromic surveillance all but impossible.
To address this issue, Harvard’s team created a simple iPad-based electronic medical record that tracks chief complaints and prescriptions and then deployed an enthusiastic team of Indian medical students to gather the data from four clinics each day. The iPads were linked to a web-based portal that synced and collated the data, ran simple analytics, and provided real-time results.The building blocks—a few iPads, a cloud-based application and a dozen student volunteers—are elegantly simple and manageable. But thanks to the proliferation of internet connectivity across India, these tools could allow rural clinics to “leapfrog” from handwritten charts to a portable, web-based system accessible on any mobile device. This would give previously unconnected clinics the benefits of real-time syndromic surveillance without the burden of a resource-intensive electronic health record system, something American physicians have struggled under for years.
“First, we want to show that it is feasible to use low-cost technologies to gather quality data in a resource-scarce setting,” says Kazi. “The fieldwork for the project is being done with a small team of passionate (and remarkable) students wielding a handful of iPads. If we can do it, the government certainly can too.”
As of this printing, the Harvard team has gathered more than 30,000 patient records, an impressive number by any research standards, and arguably the largest public health dataset ever gathered on a transient population. Their findings have been stable and predictable; most complaints are of cough and cold, and most prescriptions are for anti-inflammatory drugs, like ibuprofen.
Prior experience might suggest that generating quality data in resource-scarce settings is prohibitively expensive and that ad hoc planning is therefore unavoidable. By collating and analyzing data from over 30,000 patients, the Harvard team turned that assumption on its head. With current smartphone and tablet technology and cell phone coverage, even the poorest, most remote medical systems can employ a cloud-based electronic medical records that spot outbreaks before they happen and save thousands of lives.
Dawn at the Kumbh
Dawn is perhaps the best time to witness the ancient ritual of bathing where the Ganga, or the Ganges river, meets the Yamuna. Together with a colleague, I descended from the hill where our tent was perched. He was a cardiologist from San Francisco who completed our team of four from Harvard’s FXB Center for Health and Human Rights. Harvard researchers have been here for weeks to understand the logistics, economy, and population control of one of the largest gatherings of humans ever.
We heard the Kumbh long before we entered its hazy, golden streets. In fact, if you close your eyes anywhere in the river valley where this pop-up mega city has been erected, you can hear the constant, occasionally thunderous hum—car horns, public announcements and sacred song punctuated by the occasional blast of fireworks. But don’t close your eyes for too long. Cars and motorbikes speed down muddy makeshift roads made of endless connections of steel plates. One must keep their wits about them to walk safely on the Mela’s bustling avenues.
The crowds are thick but subdued near the water, some anticipating and others savoring the memory of the morning’s sacred dip. The morning sun is full and low on the horizon, shrouded in a haze of smog. A family gathers at the water’s edge to light a paper diya, a handmade paper boat bearing a small, lit candle. Their prayers complete, they launch the offering into the Sangam, the confluence of the holy rivers. A long-haired Sadhu, or religious ascetic, plunges fastidiously into the shallows again and again, drawing the attention of a gaggle of foreign photographers. A woman squats shivering on the bank and tries to cover her cold, wet shoulders with a dry sari.
The crowds are quiet, attentive to the task at hand. I, too, keep silent, feeling more than ever that I am in another’s world. I put my camera away and give what I hope is a friendly nod to a boy selling diyas made of large leaves. He knows I am a stranger, but his smile bridges the gap and welcomes me all the same.