Learning from Comet: rural clinics and home-care aren’t ready for phototherapy
Last year we field tested Comet—a compact and radically low-cost phototherapy device designed for rural clinics and micro-hospitals—in three countries, India, Kenya, and Nepal. With Comet, D-Rev was seeking to catch severely jaundiced babies living far away from urban referral centers that treat severe jaundice, or hyperbilirubinema. Once a baby with severe jaundice is identified, she needs to be treated as soon as possible to prevent brain damage or death. Our thinking then, based on fieldwork, was that if we could develop a device suitable for rural and clinic settings, we could prevent negative health outcomes and relieve the burden on already crowded urban hospitals by reducing patient referrals.
It’s now a year later and it is important to discuss the results of our field trials. Despite enthusiastic response from users, D-Rev will not pursue its original target market and product specifications of Comet. Based on our findings, supported by USAID’s Saving Lives at Birth seed grant, we do not believe that a device that targets rural clinics will succeed in the near future.
True to D-Rev’s model (see our design process diagram below), need identification and iterative design are critical first steps. They are not, however, a means to an end, and are as critically important as every other step in the design process. Findings from these periods can be formative during all other stages of the process and are revisited often.
A simplified version of the iterative design process used for Comet. User and product requirements were developed, then tested using the field trials as a foundation for user feedback and business development prototyping.
We installed 13 prototype Comet devices at clinics in three countries, and collected data over three to six months. Seven units reported back usage data; the other six prototypes were not used. As NPR’s Marketplace would say, let’s do the numbers:
- Countries where Comet was field tested – 3
- Units Installed – 13
- Units reporting back usage data – 7
- Total collective amount of time units were available for use – 812 days
- Total collective amount of time units were used – 126 days
- Percent of time the units were receiving power – 40%
From our field testing and qualitative user feedback we learned of and observed several roadblocks that determined the adoption and effective use of phototherapy in rural and small clinic settings.
Inpatient care is required for phototherapy. Few rural clinics keep mothers and newborns overnight, and this is very difficult to change. Jaundice peaks 2-3 days after birth, whereas most mothers are released after delivery within 24 hours. Infants that are severely jaundiced need typically 2-3 days of mostly uninterrupted treatment.
If the baby has severe jaundice, she probably also has other health problems that need treatment—and rarely can clinics treat those too. Comorbidities, the presence of more than one condition, are common. Introducing overhead or underside phototherapy without also providing a compatible warming device in rural areas (where passive warming like kangaroo mother care (KMC) are common) increases the likelihood of hypothermia. Surface contact devices that could work with blankets or KMC increase the likelihood of infection and cross-contamination.
Clinics that can’t screen or diagnose for jaundice won’t treat the baby. Newborns, particularly those with darker pigmentation, can be suffering from dangerously high level of bilirubin and show minimal outward sign. Screening and diagnostics are prerequisite to treatment. The gold standard diagnosis for severe jaundice is a blood test, which is rarely possible in a rural setting. There are other screening methods, but we saw no evidence of their use—primarily because of their high cost. Even with phototherapy, but without screening, medical staff prefers to refer infants to hospitals—even if those hospitals are hours away.
Power is unreliable. Power is required for phototherapy, warming, lighting, and other instruments. Device-specific power solutions, like a battery pack for phototherapy, are short-sighted and incomplete because lack of electricity indicates bigger infrastructure problems than the need for phototherapy. For instance, we’re asking clinicians to switch from KMC to active warming when phototherapy is in use. If the power goes out and phototherapy can continue to be run on battery power, how will the baby stay warm? How will the clinic’s lights stay on?
Staff training is needed. Medical staff that has not been managing infants with jaundice require training to develop screening and diagnostics methods, treatment protocols, and comorbidity management. Community workers who do outreach also need to be trained to monitor babies and educate parents. These activities are high-touch and require major upfront investment.
Here is an example of a hard reality that we encountered. One clinic in Nepal who received a Comet prototype had no reliable power source with which to operate Comet. And, in order to use Comet the clinic had to purchase an incubator to warm the baby simultaneously since their previous form of warming was Kangaroo Mother Care, which is incompatible with most forms of phototherapy. Therefore, the use of Comet in this setting required the clinic to purchase solar panels for its roof to power both Comet and the brand new incubator in the case of a power outage, which often lasted 12-16 hours per day. The cost for these peripherals was more than tenfold the cost of the Comet unit. So, phototherapy delivery was not the problem. The ecosystem in which it needed to operate wasn’t yet robust enough to support its treatment.
Comet treated 137 newborns from October 2013 to March 2014. Of those infants, 117 would not have received effective care without Comet present. At least three of the 117 newborns would have died or been left with permanent brain damage as a result of their ineffective care. To be clear, these results are from Comet’s use in larger clinics and hospitals—not our target areas, which had to delay installation per the challenges discussed above.
After we reviewed the field data and discussed our observations we concluded that introducing phototherapy to rural clinics and hospitals will not have measurable impact and could, in fact, have negative consequences. Our hypothesis going into trials was that rural clinics would reduce the incidence of kernicterus by moving care closer to rural families and home births. From a product development standpoint, we believed that rural clinics had drastically different needs than those of larger hospitals. The needs of clinical settings where jaundice is not currently managed cannot be addressed with just a phototherapy device. In order to enable effective jaundice management in clinics, it requires a comprehensive training program, including diagnostics protocols and setup, an affordable, reliable power source for the entire clinic, and a context-specific warming solution in addition to Comet, among other peripheral devices.
So, where do we go from here? And what did we learn? One thing to be clear about is that Comet was not a failure. In design, everything is information for the next iteration. User-testing for Comet focused our attention on areas to improve upon in its distribution and supply chain for Brilliance. We are using information from these field tests to inform future work, particularly better serving the patient populations with the greatest need. So stay tuned.
For more information about Comet, check out this blog post: User-obsessed in rural Nepal by Michael Hong, October 2013