A Doctor on Every Block
Last year, the New York Times asked some kids — If Joe Biden called you to ask what the government should spend money on, what would you tell him? This one by Prudence Lipkin caught my eye because it is rather unusual for an eight-year-old to bring up suggestions that are so pragmatic. Perhaps it was the long drive to the hospital when she was ill, or the long wait at the ER or watching her parents fret over the bill. Obviously, something prompted her to say this and no matter her reasons, she articulated a need that she thinks is important enough for the President of our country to address.
In 2015, we were launching Google health search in India and I was in Delhi for the event. I had read an article on the Mohalla (neighborhood in Hindi) Clinics that were being built by the Delhi government and managed to get a bureaucrat friend to set up a visit for me. The clinic was in Peeragarhi, a refugee colony on the outskirts of the city. A small modular shed on a concrete platform, it housed a visiting room with an examination table, a pharmacy, a small lab, and a receptionist to keep order. The waiting area was outside.
The clinic, staffed by a doctor, a pharmacist, a lab technician, and a trained midwife, served a population of about 10,000 people. They saw a couple of hundred patients every day — pregnant women, children, and people of all ages. The doctor was busy but you could see that she was deeply connected to the community, serving as a counselor as well as a conduit to government-provided services.
The doctor we spoke to said they were able to take care of more than 90% of the cases with a stock of about 100 medications and about 50 lab tests. They also triaged patients whom they were unable to help and referred them to a specialty hospital.
Being able to address the needs of most people and saving them the time and trouble of traveling long distances for consults, medications, and tests with such limited infrastructure and services is a remarkable statistic. Anyone who has traveled in India will tell you that going by public transport is difficult when you are well, but when you are sick it is just not a good idea.
As a networking technologist, it struck me that this little modular clinic was similar to a content cache on the web. Let me explain.
Many of us are binge-watching shows these days and a lot of data is being sent across the internet to our devices. Let’s say the latest series of a popular show is released and lots of people worldwide want to watch it right away. If this data were all to be served from servers in a single location, it would need massive capacity and would consume very high network bandwidth. It would also be slow for the user and expensive for a service provider.
To solve this problem, Internet Service Providers use a content distribution network (CDN). A CDN consists of a set of servers that store the original content and caches that are distributed geographically. Very simply, the way it works is if the movie you want to watch is in the nearest server, it is delivered from there, if not, the request is directed to the central server. If the content requested can be served by the cache it is called a hit and if it is not, a miss. The main parameters that are optimized in a CDN are the quality of service provided to the user, server/network capacity usage, and overall cost.
The little Mohalla Clinic is much like a healthcare service cache with an excellent hit ratio of more than 90%. Not only does it provide convenient, essential care to most of the people that come there, but it also triages people who need more complex services and sends them to the right specialists at other facilities. A set of Mohalla Clinics, supported by a network of secondary and tertiary clinics and hospitals would comprise a Healthcare Distribution Network. And similar to a Content Distribution Network, the overall parameters that would be optimized are quality of care for patients, capacity utilization for secondary and tertiary facilities, and the overall cost of providing care.
With convenient access to quality care, people wouldn’t have to take time off to travel long distances when they are ill, elderly people with chronic conditions could be monitored regularly, and maternal and child care provided proactively ensuring the health and well-being of the entire population.
This seems like a good thing for everyone — doctors, patients, and hospitals and the service metrics support this strategy. So why don’t we see more of this model of care? Perhaps the parameters that are being optimized in our healthcare systems are not the right ones.
Today there are very few independent primary care or family practices in most cities in the US, although rural areas still have some remaining. Care is provided in large clinics or hospital settings by doctors who are typically employees. The relationship between the doctor and patient is in most cases transactional.
Hospital chains continue to increase their footprint by consolidating independent practices and this trend is expected to accelerate due to the pandemic-related financial hardships. Although large institutions have increased their focus on patient convenience and care delivery, they are still primarily controlled by insurance-driven constraints and financial performance expectations. Healthcare costs on aggregate continue to go up and care access and quality have not increased significantly.
It was not always like this. In the early years in the US, when the population was largely in small towns, there were very few physicians. They either traveled around the country, worked in cities among the poor or middle classes, or attended to rich families. They went to visit patients in their homes and took care of the entire family, from delivering babies to fixing broken bones, treating all manner of illnesses, and finally helping at the deathbed. They had a relationship with the family, and were committed to serving their communities.
However many of these general physicians did not have formal medical training and they were not held to any standards of care. It was only during the early part of the 20th century, that medical education was standardized and all physicians were required to have a basic degree.
Later in the 1930s, there was an emphasis on extended graduate education and specialization. As a result, many doctors chose to specialize and consequently, the number of general physicians declined over the following decades.
The primary care specialties — pediatrics, internal medicine, and family medicine, however, grew and flourished through the 70s, 80s, and mid-90s. Right up to 2000, the bulk of healthcare delivery occurred in these independent practices and the statistics were similar to those of the Mohalla Clinic: Of the people who consulted a primary care physician, less than 5% had to be referred to a specialist or admitted to a hospital.
The use of electronic health records (EHRs) became mandatory for certain healthcare providers in the United States as a result of the 2009 HITECH Act. The implementation of the EHR proved to be too burdensome for small doctors’ offices. The billing burden and all the rules and regulations of the insurance companies made things even more complex. Many physicians did not want to deal with it and becoming part of the hospital infrastructure enabled them to get rid of this administrative burden. Real estate in some locations became too costly for small doctor offices and they preferred to have an office in a hospital.
Studies of several hospital systems revealed that the primary motivation for hospital acquisition of physician practices was financial and competitive. Limiting the referral to their own systems was lucrative for the big hospital chains. While the costs of integrated practices were supposed to be lower, the increased leverage and clout enabled them to get more money from insurance companies.
This had several consequences. It was inconvenient for patients who had to drive longer distances to the hospital since their local clinic was closed. The doctors had to limit the time they spent with patients because they were expected to meet capacity goals and their schedules were strictly controlled by their employers. There were many doctors who used to provide sliding scale or pro bono services to their indigent population who were no longer able to do so.
All in all, it seems like the parameters that are being optimized now have strayed a long way from the most important ones which are:
- Best and most convenient care for patients
- Best utilization of doctors’ time and hospital resources
- Lowered costs on the entire system
Going back to my content caching example, it was a no-brainer to implement a content distribution network because the objective function was clear. It is not that the objective function for the healthcare system is not clear: many doctors believe this is the right approach to medicine. However, hospital systems are not maximizing these objectives. Each one is focused on its own profitability and doing a local optimization rather than a global one.
I have left out a lot of details and nuances of how we got here, but today we have a very complicated, expensive and inefficient healthcare system with suboptimal care and outcomes. In the catacomb of complexity the two main protagonists in the system, the patient and the physician, have been relegated to the sidelines and a whole host of other entities — insurance companies, hospital administrators, pharmacy benefit managers, private equity companies have taken over and they are making decisions that support their own financial goals. We need to bring the focus back on the physician and the patient and optimize the real objectives: the overall well-being of the patient, the best use of physician time and hospital resources and the total cost of care.
