Personalized medicine, technology optional

So a friend of mine just had a baby in London.  His first comment:  “People who complain about NHS have no idea how health care is for the rest of the world.”

One of the things that impressed him the most was the proactiveness of the health care system.  Within the first three weeks after the baby came home, a midwife came to visit five times.  Each time, she did basic check-ups on the baby (weight, etc.), and then asked the mom about any issues with sleeping, breastfeeding, and how she was doing.  An assistant from the baby’s general practitioner also called to talk about scheduling vaccinations.

I’ve never had a baby, so I don’t have any personal experience as a point of comparison.  But compared with my experiences with the U.S. health care system, his narrative includes an incredible amount of initiative on the part of the provider.  It also demonstrates a practice of public health principles.  Breastfeeding and vaccinations are two pillars of UNICEF’s “GOBI” mantra during Jim Grant’s leadership in the 1980s.  GOBI stood for:  Growth monitoring, Oral rehydration therapy (as treatment for diarrhea), Breastfeeding, and Immunizations (aside: great free book available online about James Grant that should be a “must-read” for all aspiring public health leaders).

 By focusing all its efforts on just promoting these four proven practices, UNICEF strove to drastically increase child survival in developing countries.  The emphasis on public health in wealthier countries was tempered by the availability of high-end hospitals and doctors.  But nonetheless, especially as healthcare costs increase and money is tight, public health is an easy place to look to increase the return on dollars spent.

But what struck me about the story was the use of data.  Obviously the hospital where my friend’s baby was born has an information management system with intelligence built in.  It signals the midwives when a home visit is required. Maybe it even tracks the home address and assigns the case to someone already headed to that neighborhood.  There’s a trigger for scheduling immunizations.  Technology has enabled busy clinics to personalize care in extremely effective ways.

The only other group of people I know who receive such personalized care are BRAC’s clients.  Within 48 hours of delivery, our community health workers will make a home visit to assess the baby’s health, talk about breastfeeding, and provide general support as needed.  She’ll come back again for another 2-3 check-ups over the next six weeks.  When the government’s annual immunization campaign takes place, a BRAC health worker will work with the families to ensure that all infants are covered.  It works pretty well—our maternal, neonatal and child health program covers 25 million people, and we’ve helped Bangladesh reach a measles immunization rate of 85% (which is the indicator used at the Millennium Development Goal target for reducing child mortality).  It works: maternal and infant mortality have dropped significantly.

Of course, in BRAC’s case, there is no beautiful database with bells, whistles and notifications.  The system is built on people and relationships.  One community health worker looks after one village.  At any given time, there are maybe 5-10 pregnancies.  She can remember that.  When there’s a delivery, she calls her supervisor, a community health worker with specialized training on maternal and child health, and asks her to conduct a home visit.  Two weeks later she reminds her to come again.  The supervisor relies intensely on the memories of the village-level community health workers, who serve as an informal information management system at the grassroots level.  Many would say that this can’t work at scale, but I would argue that there’s ample evidence that it can, give proper motivation, training, and support.  The beauty of it is its simplicity—it works at any scale and can be implemented almost anywhere.  You could imagine a school, a church, or an employer, creating support systems where coverage is divided to a unit that one person can easily manage by memory. 

The biggest problem: unlike the NHS system, the memory system has information that’s distributed across all members.  No one can see the big picture.  There’s great dependence on the frontline staff.  If they don’t share their information, nothing happens.

Reading an article in the Guardian earlier this week, I saw another example of this type of large-scale, people-based information system working efficiently at a significant scale: the incredible lunch-box delivery system in Mumbai that delivers over 200,000 hot lunches a day. Mr.  Arvind Talekar, spokesperson for the Nutan Mumbai Tiffin Box Suppliers, said:

"We keep it simple. I have seen that many times technology doesn't work. So, it is best not to use it because if it fails, then we can't deliver the lunch boxes on time. We are illiterates and we prefer to keep a mental note. We don't depend on gadgets."

There are many good reasons to move towards an electronic and centralized management information system. And I am personally excited to see development organizations invest in better information systems.  But as the BRAC example shows, you don’t necessarily need a fancy system to provide personalized care, and building a fancy system may not change things at the frontlines (if everything was already working smoothly).  It’s easy to assume personalized and informed service delivery requires sophisticated inputs, but really, the system is just mimicking the human processes, at a much larger scale.

What other assumptions are we making about what’s “impossible” without “real” information management systems?