Major Constraints in regard to Frontline Health Systems in LMICs

Over the past decade, discussion of integrated care has become more widespread and prominent in both high- and low-income health care systems (LMICs). The trend reflects the mismatch between an increasing burden of chronic disease and local health care systems.

Healthcare challenges in LMICs have been the focus of many digital initiatives that have aimed to improve both access to healthcare and the quality of healthcare delivery.

  • What are the biggest constraints when considering frontline health systems in low- and middle-income countries?
  • Which aspects of these systems require the most work?
  • How can they benefit from advances in A.I. technology?

Hi @a1m2r3h4, @RahulJindal, @LeeStein, @stepet and @creativiti - Please share your thoughts on the biggest constraints in regard to Frontline Health System in LMICs. Thanks.

Unmet (and unlimited) demand for the scarcest healthcare resource, clinical expertise, results in delayed and inconsistent care that contributes to 20% greater odds of death. While any licensed clinician can theoretically prescribe any of the tens of thousands of medications and diagnostic tests a patient may need, we are well past the point where the escalating “complexity of modern medicine exceeds the capacity of the unaided expert mind.” The result is sub-specialization fragmentation with limited care coordination. Over 25 million in the US alone have deficient access to specialty care, with a projected shortage of over 100,000 physicians by 2030. (Let alone the masses in low and middle income countries who cannot even access basic medical care.)

Advances in clinical practice that gain efficiencies are among the few ways to simultaneously improve all ends in the “iron triangle” of healthcare quality, cost, and access. Most advances will otherwise cause trade offs or increase disparities that impact underserved populations, for there is no quality without access.

Electronic consultation systems allow clinicians to request specialty advice through messages, teleconferencing, or online crowdsourcing. These illustrate the potential to treat patients with up to 71% fewer in-person specialist visits. This eliminates patient time and travel to initiate a treatment plan (days vs. months), which could be lost completely when up to 40% of specialty referral requests are not completed. A recent literature review of such systems found improvements in process measures such as timely access to care across a range of studies, but also noted the need for further research on how best to design such systems to ensure improvements in clinical outcomes. Despite benefiting patients and providers, existing (electronic) consult systems remain fundamentally constrained by the availability of human experts.

In the face of ever escalating complexity in medicine, integrating informatics solutions is the only credible approach to systematically address challenges in healthcare. Tapping into real-world clinical data streams like electronic medical records with machine learning and data analytics will reveal the community’s latent knowledge in a reproducible form. Delivering this back to clinicians, patients, and healthcare systems as clinical decision support will uniquely close the loop on a continuously learning health system. Our research group seeks to empower individuals with the collective experience of the many, combining human and artificial intelligence approaches to medicine that will deliver better care than what either can do alone.

@jonc101 - Thanks Jonathan for your inputs on the major constraints faced by Frontline Health Systems. I just wanted to understand little more on how we can equip frontline health workers, who are working in remote places so that they can take informed decision and be able to provide timely primary care to vulnerable population?

E-Consults and systems like Project Echo are one example to use telecommunications to disseminate expertise, but still require a human expert to answer on the other end. We’re trying to develop next-generation tools with full automation and digital delivery for fully scalable distribution to frontline health workers and patients so they can invoke support systems on-demand.

@jonc101 - Thanks Jonathan for your inputs. This is really helpful.

I spent 3 days with the Project ECHO (https://echo.unm.edu/about-echo/ourstory) in Albuquerque, New Mexico. I think it has outlived its novelty and utility. It brings specialists to local communities and perpetuates the “specialist” model creating dependence on them. What is required is “task-shifting” in LMIC, in which motivated local people with less qualifications will do the work of physician providers. In our model, we have shown that general surgeons can safely perform kidney transplants and Ophthalmologists can perform corneal transplants (https://link.springer.com/article/10.1007%2Fs00268-019-05093-w; Microeconomic Benefit of Corneal Transplantation in a Developing Country via Public-Private Partnership Model - PubMed).

Therefore, I believe that the biggest constraints are dependence on medical providers in LMIC and lack of task-shifting. Several models of task-shifting have been developed and need further validation. One example is the SEVAK model in Gujarat, India (www.sevakproject.org) which utilizes village level high school students to train them in measuring and following hypertension and diabetes, by making life-style changes rather than reliance on medications. This model was also tested in Guyana, South America, with initial success (https://academic.oup.com/milmed/article/180/12/1205/4160666).

@RahulJindal - Thanks for sharing your perspective and the resources on the major constraints to FLHS.

Hi @Lizzy_2020, @PHall, @acavaco and @marschenrj - You may have thoughts on the major constraints to Frontline Health Systems. Join the discussion to share your thoughts. Thanks.

In rural Wyoming, constraints along with financial are having to travel hours to see a provider. There is a general lack of awareness of preventative and alternative treatments. With limited access to specialists and primary care providers along with the lack of insurance, folks may only see a provider when it is an emergency. Educating the consumer about digital medicine and AI could yield great benefits. Using mobile clinics equipped with AI to reach the most remote areas would be the most beneficial.

@marschenrj - Thanks Janet for sharing your experience and views on major constraints to FLH Systems.

Hello @gajewski, @angelfoster, @Budoff, @timothymusila and @andwhite - It would be great to hear your views on the major constraints to FLH systems. Please join the discussion.

Hello @SArora, @anitasmoore, @Fatima and @aroamer - Join the discussion to share your thoughts major constraints to Frontline Health systems.
Hi @Shabbir - We would love to hear your thoughts on how exponential technology can help resolve all/few of the constraints listed above.

@jonc101 Thank you so much for your insights on the limits of clinical expertise. Are there any particular research studies you recommend we look into? Specifically regarding the adverse effects of delayed and inconsistent care? We’re very interested in learning more about this, so thank you in advance for your guidance!

@RahulJindal Thank you for your feedback, we truly appreciate it. I’m curious to learn more about the “task-shifting” model you mentioned. Does the training typically require human experts to be involved, or are digital tools being used to accomplish these goals? If not, what kinds of opportunities do you think there are for using digital tools (especially artificial intelligence) for implementing task-shifting to frontline health care systems in LMICs? Also, what do you think are the limits to task-shifting?

@marschenrj Thank you for your feedback. Are there any A.I.-enabled mobile clinics that you know of – especially ones located in low- and middle-income countries? We would love to learn more about their operations and how effective they’ve been in administering health care to remote locations. Thank you!

@RahulJindal thank you for your thoughts about Project ECHO. Task-shifting is core to the ECHO Model, and exactly what we do to build capacity amongst frontline health workers in low-and-middle-income countries. By democratizing knowledge - bringing the right knowledge to the right place, at the right time, we build the capacity amongst healthcare workers, shifting what has traditionally been done by doctors to nurses, nurse practitioners, Community Health Workers, etc. Learning effective triaging, deciding who needs to see a physician and who could be well-served by a Community Health Worker, is part of this knowledge. By partnering with The WHO, the CDC, Ministries of Health, and other on-the-ground experts, we are able to build local capacity.

Unlike other healthcare workforce “pipeline” training models where knowledge only flows one way, the ECHO model is a platform that allows all user-participants to co-create knowledge through group discussion, peer interaction, and engagement of local expertise so that all teach and all learn, and our collective understanding of how to disseminate and implement best practices across diverse economic and cultural contexts grows.

Our pilot project in rural China has identified specific pain point that can benefit from AI enabling village doctors to do more primary care things. It seems similar to “task-shifting”. But, in order to achieve scale, we have designed an AI-powered Internet-based primary care team model. Feasibility study was published on Lancet conference last year.

Thank you, Sanjeev, @SArora for this excellent summary of the philosophy of the ECHO project, which is funded by many Federal agencies. The point I am making, which you also made, “peer interaction and engagement of local expertise” is well taken. However, our model (www.sevakproject.org) differs significantly in that peer interaction should be without the intermediary of a physician. True task-shifting is when the physician removes him/herself from “directing” or “hand-holding” the community worker and let them make decisions relying on their own knowledge or by interacting with their peers (community workers). In other words, the effect of “safety net” makes the community worker feel dependent on the physician and this would prevent them from being truly independent.

I also wonder who decides “bringing the right knowledge to the right place, at the right time” for a community worker in a remote village in India, Africa or South America? Who makes a decision on what is correct methodology or treatment pathway? Should we in the USA make such decisions?

Artificial intelligence could potentially factor in by allowing the community worker ask questions and interact with ie. Amazon’s Alexa or google. I envisage an ‘open source’ fund of knowledge which is not under the domain of a university, Federal agency or a government.

@jonc101 yeh, automation is key for large scale impact!

We can use covid-19 pandemic to study the major constraints in LMICs. There is an urgent need to come up with technology solutions to ease these constraints. For example, doctors may not have the latest knowledge of care and treatment for covid-19 since there is no cure yet and latest clinical research moves very fast in advanced countries (e.g. NIH sponsored data-driven studies).
So, two problems we can look into:

  1. Rapid dissemination of latest guideline and best practices does not happen normally. How ca AI help remove this bottleneck?
  2. Real-time clinical collaboration across hospitals and countries does not happen normally because too much resources require and international interoperability of clinical data is difficult. Is there new AI solution for this?

XPrize has formed a covid pandemic alliance. The alliance is probably also looking at the major constraints on covid patient care in LMICs, such as the lack of rapid dissemination and real-time clinical collaborations in current global clinical infrastructure. If we can have some cross-talk, that may help identify the constraints amplified in covid pandemic.

Thank you @SArora, @RahulJindal and @ajchenx for sharing amazing insights into the major constraints of FLH Systems.
@ajchenx - We would like to read more about the AI-powered Internet-based primary care team model you’ll have designed. Is it possible for you to share link to this resource.