by Ruchit Nagar
To begin, I just want to point out that it’s well past midnight, and we have to go to the field tomorrow morning to test our baseline survey instrument. More importantly, I like to sleep in, so it’s going to be rough tomorrow morning.
So they say, whoever “they” are, that time flies when you are having fun. I’m not sure where these past two, no, three months in Udaipur have gone, let alone the past 18 months since Khushi Baby was born. From marker board to prototype to partnership to funding to implementation and now to evaluation, KB has had its triumphs and faced its hiccups.Naturally, the process has been highly engaging, but regrettably it hasn’t left much time for me to put my reflections down on paper.
Thus far we have trained a team of 7 surveyors and 6 General Nurse Midwives, fine-tuned and operationalized our software, engaged with multiple IRBs (in the US and India), solidified the stake of our partner NGO and implementer, Seva Mandir, and collected baseline social network data from nine villages in the Badgaon block of Rajasthan. We’ll be starting our 9 month RCT on August 3rd with a baseline study of immunization rates, come hell or high water (which becomes less of a figure of speech in the monsoon season), and we look forward to closely supervising the process to ensure data is being collected according to guidelines in our protocol.
But my intention here is not to give a laundry list of updates. The point of this reflection is to share what I have learned and my vision for how KB will continue to grow during and beyond the RCT about to get underway. Here’s a serious picture of me thinking to set the mood:
- The right team matters; could not be more fortunate to have Shahnawaz, Sanjana, Logan, Preethi, Alam, and with me on the ground. Praneeth has been holding down our technological fort from back home, while Kantilal, the Badgaon GNM, has been riding to the last mile with a vaccine box strapped to his bike and a sense of duty strapped to his chest. We’ve also been very lucky to have outstanding advisors: Professors Bo Hopkins and Joe Zinter (who have fathered Khushi Baby), Martin Klein, Nicholas Christakis, and Allen Wilcox (CEO of VillageReach). Seeing the team bring different backgrounds, skillsets, and new ideas has energized me to engage more deeply with the project. At the same time, I have come to see the team as one that is constantly morphing, yet a closely knit and unified family. The team around me works around the clock, on weekends, and wakes up with a drive to get hands on with the project. A flat structure within our group allows for everyone’s voice to be heard and for our discussions to go further. For us, an individual’s value add to the team is defined by the derivative of their contribution curve, not primarily by the area under it. My co-founder role or 18 months with the project is not of consequence. We hope this framework will allow us to push pass founder’s paralysis, but regrettably that means that we have to bid farewell to members who in the past helped provide the integral building blocks for where we are today but no longer have the capacity to do so. While both the financial sustainability and long-term commitment of a volunteer team present a challenge, I am encouraged by how far we have been able to come and how much we have been willing to sacrifice in spite of it all. Our momentum at KB is driven by the team’s desire each morning to tackle something new and creative from the process of design to evaluation.
- Things move slowly in India; or we at KB are moving too quickly – relativity certainly is not confined to physics. The sense of urgency and commitment to deadlines is continual struggle with partner institutions. After meeting with Ajay Mehta, Chairman of the Board of Seva Mandir, my perspective has seen a paradigm shift. Ajay openly acknowledged the weaknesses of Seva Mandir and urged us to think not simply from a scientific or technocratic lens with our intervention/innovation – although I would argue we are also incorporating an anthropological one. His point was to highlight that an intervention is only as strong as the system that can implement it. From that lens, it’s important to recognize what factors are within the implementing body’s control, and which factors must be accounted for in the design itself. In doing so, we must not only design an app and necklace that is robust enough to be useful without regular syncing or wifi, but also we must design a process that can be implemented by an organization that sees high turnover of the implementing and managing staff and which currently has limited data management resources.
- Regulatory boards present a veritable bureaucratic challenge with east/west philosophical differences. The danger in discounting an anthropological lens in safety and ethics determinations is compromising the local adoption of the intervention, which may confound or mitigate its ultimate measured effect. We’ve thought deeply about the advantages of the necklace as a form factor, asked hundreds of mothers, and come to the opinion that medical history worn by the patient may serve as an important visual catalyst. With children customarily wearing these necklaces, the risk involved in introducing a new necklace should be weighed in balance with the existing cultural context. While we have faced delays in the approval process, it’s reassuring to have official approval from the local IRB and the implementing NGO. Moreover, the delays have forced us to probe deeper into our study design and methodology and to both justify and think of new alternatives to our existing form factor. On the whole, the experience has been necessary but exhausting. There must be some unspoken saying among those engaged in research that goes along the lines of, to put it very nicely, “the IRB always gets in the way”. Just as the guild of magicians chooses not to reveal their secrets, I too should not reveal this shared research sentiment, but I do wonder at times about the opportunity cost of valuable investigation due to over-caution, for which bright lines vary by culture and context.
- It’s important to constantly think about the essentials when designing a system. It may seem obvious, but large scale schemes often leave out “innovations” not because they are undiscovered, but rather because they are non-essential or not yet shown to be cost effective. When meeting with BMGF and in a conversation with the Dean of IIHMR Jaipur, we were pushed to isolate each essential component of our system. Is there a need for the patient to even hold a digital copy if data syncing is a requirement in any mHealth system? How much do we actually help an identification process by scanning when it can occur through fingerprinting or lookup of a few ID fields? What happens when twins arrive at a camp and our app doesn’t support biometric identification? A simpler, non-NFC phone that doesn’t rely on a patient copy of data would also save costs. I would argue at its core, a patient copy of data adds a needed level of redundancy, an ability for field verification that a patient interacted with provider, a source of clinical information in the common case of lack of regular connectivity to the central server, and that’s yet to even get into the socio-technical nature of the visual signal of the pendant or community engagement through our budding voice reminder system (which I believe will ultimately replace unclear SMS schemes for illiterate populations). But I do believe that, more so than the “X factor of the necklace”, the crux of this project is about patient ownership of their data and creating a better touchpoint between health care provider and patient.
- Designing sound study methodology in the face of bureaucratic and on-the-ground limitations is a cakewalk, and by that I mean, it is an uncomfortable process in which one must tread carefully to minimize the messiness of the trail created. It has been exciting to learn about the essentials of impact evaluation: cluster randomization, 30 cluster sampling for baseline, multiple regression for survival analysis, and theory of change. We have wrestled with multiple strategies for determining something as simple as the accuracy comparison of the KB app vs. the existing logbook. What is the real gold standard? In this case we are using syringes and empty vaccine vials. But then one must ask how must the correlation be made from the logbook/KB app tally vs. the respective counts? Certainly we must consider both under and overreporting instances as well as the specific composition of the vaccines reported compared to those which were used. Overall the metrics can get as muddled as the data sources themselves. With our limited sample we still are pushing to ask as many questions as possible, and that in a rigorous manner: accuracy, efficiency, cost, and level of program benefit from novel features introduced by our system. In doing so we can’t be rushed to start, despite our eagerness, and we must engage in regular course-correction to our design based on local feedback.
- It’s important to have fun. Life is an adventure. I’ve been lucky enough to ride on the roof of a Jeep in the middle of rural Rajasthan or on motorcycle for 160 km to the village Undithal, all the while passing through the green shoulders of the Aravali Mountains. It’s scenic, but it also starts to hurt after a while, especially on the unpaved roads. Alam and I have had fun with the kids in these villages, teaching them how to wash their hands, then literally giving them baths at the handpump, and finally seeing them climb up wavering tree branches to retrieve clusters of jamun for us to eat. Speaking of food, Sanjana and I had the privilege of trying goat’s milk chai in Jhadol. Not only am I soaking in the cuisine, but I am also picking up the language. My Hindi has improved by training our surveyors and watching the occasional Bollywood movie. My ability to hike has improved after visiting villages like Bedakudi, where we trekked for 4 hours to meet 3 mothers, 2 of whom had the KB necklace. It really is the coolest feeling in the world to see a mother bring a child to a follow-up immunization camp with a KB necklace, to see her smile as she ties the string around the child’s neck, and to have her ask me to take a photo of her with her child.
- After much convincing, I’ve gotten out of the cave and visited Ahmedabad, Jaipur, Agra, and Delhi. Maybe I haven’t sought out leaving because Udaipur has become a second home. Vineeta and family at Balaji’s have welcomed us as family by providing warm, home-cooked food every day. Sanu and Sanju have taken us in and around Udaipur, giving their local lens to the landscape. As a result, if Drake were to ask me “when’s the last time you did something for the first time”, I would have no difficulty producing a list, and I must give due credit to the team – the family – with me.
I’ll save the cliché sentiments and the already wordy post with a brief glance into the future:
- Biometry (an ongoing challenge for infants’ fingerprints) and encryption
- True name-based, personalized voice call reminders coupled with educational information
- KB for antenatal care, emergency medicine, chronic disease management – full and comprehensive electronic records for unsynced, off-grid frontline health workers
- Integrating network targeting approaches to identify referral champions, who can supercharge awareness dissemination about camps using the necklace as a social hook
- Scale up (the foolish dream of all pilots) and integration with a government EHR system that currently handles hundreds of thousands to millions of mothers and children each year (for example, PCTS – check out their use of Comic Sans on their website – or a more mature MCTS)
- A wearable health product line: bracelets, necklaces, etc; each culturally tailored for its local population
- A global health wearable campaign to end the immunization gap
Sure there are lingering concerns:
Is this just another mHealth app that will suffer from pilotisis? Does the project have sex appeal but no substance? Will our study have impact at least if not eventually published in a high impact journal? Can this scale? Was Esha Chhabra of Forbes right in her blanket criticisms of our about-to-be-tested system.
Yes, I too want answers, but I can settle for lessons learned. Let’s see what the RCT shows. Like all things in life, longitudinal studies, take time. For example, Banerjee and folks published their findings on lentils and immunizations 6 years after starting their baseline. We are up for a similar, lengthy journey, and in that time, this Khushi Baby still has plenty of room to learn and grow. I do have an answer to one question though, and it’s a question I like to ask to others, so I’ll end with this: what drives you to get out of bed in the morning?