(Summer 2014) Sara Dolcetti: Medical Diagnostic for Childhood Pneumonia in India

Sara Dolcetti (G, Master of Science &  Master of Business Administration)

Sara spent the summer in India, accelerating the development of a low-cost mobile phone-based thermal imaging device to screen for pneumonia, the leading cause of death in children less than 5 years of age.  To this end, she was awarded a Beck-Vest fellowship in 2014. Sara’s work involved determining the clinical and operational environment for children presenting with symptoms of pneumonia. Her goal is to help build a simple tool for healthcare workers in the developing world to use and evaluate its effectiveness. She partnered with physicians at the Lata Medical Research Foundation in India for their clinical and operational expertise in local public and private healthcare facilities.


Sara Dolcetti


Project Context: 

Thermal Imaging Diagnostic for Childhood Pneumonia in Resource-Poor Settings

The problem: Pneumonia (PNA) is the primary cause of death in children less than 5 years of age and accounts for 18% of total deaths worldwide in this age group.  Approximately 25% of these deaths occur in India, where many regions suffer from a lack of healthcare workers and cost-prohibitive diagnostic equipment.  Because early stage pneumonia is difficult to diagnose, due to a variety of pathogenic causes, it is often treated too late or as if it were Malaria. To postpone treatment and to use malaria medication likely exacerbates the condition, increasing the risk of death and the resistance to medication. Our hypothesis: Diagnosing pneumonia at an earlier stage could help reduce false positives, false negatives, mortality rates and system costs. The device: Massachusetts Institute of Technology (MIT) and The Consortium for Affordable Medical Technologies (CAMTech), headquartered at MGH, is developing a thermal imaging device capable of being fixed to the cell phone of a healthcare provider.   Thermal imaging offers a novel, cost-effective approach to PNA screening.  Using CAMTech’s cell phone ‘app,’ providers can quickly generate a thermal image which, if asymmetrical, can indicate the presence of pneumonia (Figure 1).

Sara Dolcetti

Figure 1: Thermal imaging as a diagnostic device (Source: CAMTech Proposal)

Our team: Our team is comprised of myself (a joint Master of Science and Master of Business Student at MIT), MIT and Harvard professors, an MIT Technology and Policy student, a Harvard Medical student and an MIT Engineering Systems Division Masters student. Our affiliates and partners are: Physicians from Massachusetts General Hospital (MGH), The Consortium for Affordable Medical Technologies (CAMTech) and the Lata Medical Research Foundation in Nagpur, India




From Delhi to Nagpur and monsoons in-between

Something feels amiss. As we make our way through security checkpoints, I look around and immediately see what it is. At noon on this Tuesday in July, the Delhi airport is unlike the Delhi I’ve experienced thus far: it’s almost barren! Stepping onto the plane, I find it difficult to bid ‘Alavidā’ to New Delhi – with its kind people, delicious food and fascinating history. I look forward, however, to saying ‘Namastē’ to Nagpur. There, my colleague and I will meet our sponsor and her team at the Lata Medical Research Foundation and begin our field work. I have learned a great deal about the Indian healthcare system and cultural norms from them thus far and am keen to imbibe the environment once there. Tomorrow, we meet with the Lata team in India to discuss our research plan in person and to gather any final input and advice. Our first priority is to ensure that we understand the system which cares for children presenting with symptoms of pneumonia and the children (and their families) themselves. If we don’t do this, we risk applying a ‘solution’ to an insignificant (or even worse, to the wrong) problem. The challenge will be to do this within the less-than three week timeframe, with translators, in monsoon season and still get the data we need in order to continue our research when we return home. Additionally, while we have extensively researched prior to arriving, I am mindful that we’ve likely formed a series of assumptions through that research which we will need to identify and then challenge. Working closely and aiming for quick iterations with the expert medical team here in India will be critical to getting to the key issues efficiently and effectively. The monsoon rains embrace our plane as we arrive in Nagpur. We’re ready to begin!



Entrance to a rural Primary Health Center

  Hub and Spoke

And begin we did! The LATA team immediately immersed us in the healthcare system and local Nagpurian culture. In the span of a few hours we got an in-depth review of the healthcare system, vetted our plan by the team and Dr. Patel and started to work through their databases. The Indian Healthcare System follows a hub and spoke model: ASHA workers –> Sub Centers –> Primary Health Centers –> Rural / Sub-district Hospitals –> General Hospitals –> Tertiary Medical Centers (e.g. Medical Colleges). Our plan is to sample from each center and level of care in both the private and public system. After a full day of prepping for our field analyses, Sanjeev and Ambar took us out for dinner in a traditional restaurant – typical of the Maharashtra state. Delicious? Yes. Spicy? Yes. Milk consumed to douse the spiciness? Priceless.

                PHC Outside Sara Dolcetti PHC Outside #2 Sara Dolcetti

Primary Health Center (PHC) and Residences of medical team (adjacent to the PHC)

  Sara Dolcetti

Corinne and I with Ranaya, Medical Officer at a Primary Health Center in the Nagpur Cluster

At the heart of it all: Starting with the patient

One of our objectives is to follow a patient through the healthcare process. We aim to understand/identify:

  1. What decisions are made when and by who
  2. Barriers to accessing care
  3. The prevalence of children who exhibit symptoms of / treated for pneumonia at different points in the process

To this end we went to the village of Hatodi to meet the local auxiliary nurse midwife  and anganwadi as they led a mother-and-child vaccination clinic. (Auxiliary nurse midwives (ANM) primarily deliver babies and care for children. Anganwadi’s are government-trained employees who live in and serve one village. They provide child and antenatal care as well as pre-school education.)

We noted many things during our time there. One of particular interest concerned barriers to accessing care. Through our conversations that day and since then, we found that perhaps the largest barrier to treatment existed at this level. Within the home environment, parents would benefit from being able to identify that their child has a serious medical condition and requires treatment.  Particularly in rural settings – where a parent would need to leave their family/work to bring a child to a healthcare center  – parents would wait for their child to ‘get better’ before going to a clinic to receive care. Once at a clinic they would have access to antibiotic treatments and, depending on the facility, diagnostics/treatment equipment (e.g. oxygen tanks).


Nurse with smart phone Sara Dolcetti

Auxiliary Nurse Midwife using her smartphone during a vaccination session in the village Hatodi

Sara Dolcetti

Mother and child vaccination session in the village Hatodi

DSC09035Mother and child handbook IMG_1835

In front of the vaccination clinic in Hatodi

Moving up the chain

From there we went made our way up the chain: sub- centers, primary health centers, general / rural hospitals, sub district hospitals and tertiary hospitals. Relevant to our analysis, we found that patients could enter, exit, transfer to/from any point in the chain – which changed the way we thought of the care pathway. At the lower levels of the chain (i.e. sub- centers, primary health centers), pneumonia is primarily diagnosed by its symptoms. Doctors follow WHO guidelines to interpret symptoms and, from those we shadowed, use limited tools (i.e. primarily use stethoscopes and thermometers) to do so.


 “In children under five years of age, who have cough and/or difficult breathing, with or without fever, pneumonia is diagnosed by the presence of either fast breathing or lower chest wall in-drawing where their chest moves in or retracts during inhalation“

– (WHO Pneumonia Fact Sheet 2013)

Treatment was consistent across the PHCs we visited. The doctors generally prescribe first generation antibiotics and monitor the child over its course to see if the child recovers. If the child does not recover, the doctor refers the child to a hospital.  It was difficult for our team to assess during our time there whether or not patients followed through on the referral process and if not, if they then chose to visit a private health center (due to a widespread belief that private centers provided ‘better’ care – an opinion we heard often during our visits).

Sara Dolcetti

 Comparing notes with a Medical Officer in Kondhali


Corrine and I with the pediatric nursing team


X-ray machine and images  at the General Hospital in Bhandara


Near the entrance of a general hospital

Private vs Public

Given that the majority of health-care spending in India is within the private system, we prioritized visiting private healthcare centers and pharmacies. At the private centers, we noted a difference in care, diagnostic procedures and prescription behavior. Doctors spent more time with each patient (~8 min versus ~1.5 min), had a larger breadth of tools at their disposal (access to computers, x-ray images etc.), asked more questions to the children and their parents, and primarily prescribed second/third generation antibiotics.

Many of the citizens we spoke with believe that the private system provides better care than the public system. Many women – irrespective of income level – said that they would go to a private center if they knew their child was very sick.

graph india

Figure 2: Public and private healthcare spending as a % of GDP (Source: The Economist)

Sara Dolcetti

Private dispensary  (pharmacy) in Kondhali village


 Houses within the Kondhali village


Given that pneumonia is the primary cause of death in children less than 5 years of age in India, proactively addressing this disease can help to save the lives of children. Arguably the most effective form of ‘addressing’ this disease is to increase preventative measures. Parallel to these efforts are to implement solutions that promote earlier and more accurate diagnostic capabilities.  The work our team has done in this area is preliminary. However, from our limited time on the ground – through observations, research and discussions – we see the following as the primary ways to reduce the number of childhood deaths due to pneumonia:

  1. Improve sanitation and promote basic hygiene practices to limit the spread of the virus
  2. Provide adequate nutrition to strengthen immunity within vulnerable populations
  3. Enable early and accurate diagnosis capabilities by: increasing community health workers’ ability to identify PNA;  providing diagnostic devices to community care workers to enable earlier detection; equipping sub centers and/or primary health centers with diagnostic devices (as they currently do not have diagnostic equipment).

This last point speaks to the benefit of introducing a diagnostic device – like a thermal imaging device- to enable widespread detection.

By introducing such a device our goal is to: encourage earlier detection (and thereby treatment);  reduce the number of false positives and therefore reduce the over prescription of antibiotics; and ultimately reduce both mortality and instances of drug  resistance.

Another potential use of the tool would be to replace x-ray machines at the hospital level. If the medical environment allows for this type of transformation, this would lead arguably lead to greater net savings than simply introducing the tool at the clinic level.

Our next steps are to: adjust the model; incorporate the data we collected (i.e. prevalence rates, cost information, throughput probabilities etc.); and run the model under several scenarios (i.e. placing the device at different points in the chain, assuming different demand levels and referral rates in the private/public system).  To do this, we will continue to engage with members of the LATA foundation. Our work has just begun.

On a personal note, I was inspired by the members of the LATA team. With limited resources they created an extensive and rigorous research program with data collection initiatives that will help to understand and improve the region’s current healthcare status. Moreover, they have created an organization where members (including numerous healthcare and community workers ) are united under, and energetically work to realize, a common mission.


A light-hearted moment with the nursing team at the General Hospital in Bhandara

Thank you 

As with any complex project, this work owes much to many others.

First, to the members of the LATA foundation for generously giving of their time and expertise to explain/show the intricacies of the healthcare system around Nagpur. I would be remiss if I did not also thank them for graciously taking care of us while we were in India.

To the sponsors of the grant The Becky Vest Fellowship– thank you. This funding made this project possible.

To Dr. Stan Finkelstein M.D., Dr. Patricia Hibbard M.D. and Dr. Jarrod Goentzel for your expertise and guidance in shaping this project.

To Dr. Alison Hynd and members of the MIT Public Service Fellowship office – thank you for your support and coordination to ensure my time in India was a safe and seamless one.

Lastly, and most importantly, to all the citizens and healthcare workers we spoke to during our field visits. Thank you for your enthusiasm in helping us to understand the healthcare system. We greatly appreciate and value your perspectives.


Corrine and I with the LATA Medical Foundation Team

« All Posts