Rural healthcare in India is fragmented and tedious, given various infrastructure challenges and a shortage of qualified doctors. These problems are compounded for the poor, who make less and have a very high opportunity cost for travel to seek good care. In remote areas, unqualified providers fill the gaps. This research aims to determine the value of the informal social network among these providers. In particular, how and why providers connect on medical information or business, and determine what about it can be utilized in public health efforts, especially around four disease areas: tuberculosis, pneumonia, kala azar (visceral lieshmaniasis) and childhood diarrhea.
This research is proposed to help WHP, a global health non-profit that uses telemedicine and a social franchise model to bring care to the poor. WHP is currently working on establishing 400+ franchisees in India’s poorest state, Bihar. A social network analysis would allow WHP to better understand the existing informal networks among providers.
Social Network Analysis would help inform:
• Public health efforts focused on the four disease areas – what links already exist and how can they be exploited?
o How often do rural providers communicate with one another on specific diseases and/or treatments?
• How structural holes can be strengthened or bottlenecks removed so that there are no areas “off the grid” vis-à-vis updated medical information.
• Which providers (Eigenvector) can be most useful in initiating a public health effort that involves all providers in the area?
o Can those with formal training assist or teach those who have no formal training?
• What providers are completely isolated from the rest of the providers in the area (cliques and factions)? Why?
• Which providers would be the best candidates as franchisees?
• Do providers desire more communication with others? How does this vary by level?
Network to be surveyed:
All providers in [TBD] block of the state of Bihar, within 10 km of the central market village. Levels of providers are defined as:
• Rural healthcare provider (trained or untrained, uncertified, working out of own home)
• Rural clinic-based healthcare provider (trained or untrained, uncertified, practice attached to home)
• Board Certified Doctor (MBBS or higher)
• Chemist shop owner (Board certified)
• Pharmaceutical distributor (> Rs10,000 per day / US$250 per day)
• Pharmaceutical stockist (
• Pharmaceutical company representative (Field staff)
A list of all providers in the area will be provided by advance field staff for the research team.
Attributes of network nodes:
• Name
• Age
• GPS location of practice
• Marital status
• Type and level of medical qualification (see above list of levels)
• Size of practice (patients per week)
• Size of practice (income per week, in Rupees)
• WHICH of the following four diseases the has treated in the past month (may select any number from 0-4) and how many instances:
o TB
o Pneumonia
o Childhood diarrhea
o Kala Azar (visceral leishmaniasis)
Network information (will check boxes based on a list of the area’s providers):
• If the provider has referred a patient to the other provider listed
• 1-5 times in the past month
• 5+ times in the past month
• If the provider received a referral information from the provider listed
• If the provider delivered or provided medical information regarding one of the four diseases to the provider listed
• If the provider received medical information regarding one of the four diseases
• If the provider delivered or provided business information to a provider
• If the provider received business or other professional information from another provider
Important information to analyze:
• Density: How interconnected are the providers at each level? Often, if a patient presents symptoms that don’t go away, they will be referred to providers at a higher level. This indicates that providers consider their competitors to be at their own level of training, but this does not preclude communication on the same level. Understanding the network density will help to separate the patient flow (largely one direction, from bottom to top) from communication flow (which may be more complex).
• Eigenvector: Which providers communicate with multiple other providers.
• Bridges and bottlenecks: If providers are bridges between geographically distant providers. Likewise, if providers are bottlenecks, or essentially serve as monopolies of information. I expect that there is no interaction between private and public doctors, but if one or two are interacting, this would be the channel through with public health information and education could be more effectively disseminated.
o In particular, the “weak ties” that may exist between providers as a result of competition could be exploited with new information and trainings.
I look forward to further thoughts on this proposed research.
No comments:
Post a Comment