Pneumonia Diagnostics

Today I learned that it requires many cups of coffee for even the most brilliant minds to formulate global health policy solutions. In the last couple of days I was able to attend a technical consultation for pneumonia diagnostics in which a group of experts discussed the issue of a gold standard for pneumonia diagnostics (involving pulse oximeters and respiratory counters) and evaluation of new devices. The ultimate goal is to have a device that will differentiate between viral and bacterial pneumonia. Although I was not originally invited to the meeting, a spot fortuitously opened up, and so there I was—the random undergrad in an inter-agency meeting full of experts in their respective fields! It is a strange situation to be gaining more experience and knowledge and yet to feel less experienced and knowledgeable than ever before.

Here are some key points from the meeting that I thought were interesting:

Problems to consider:

• The working environment of pulse oximeters: Many health facilities face the issue of “queue clearing”. They are way understaffed for the volume of patients and are forced to take what are considered safe shortcuts to get through all of the people who need to be seen. These shortcuts often include asking fewer questions while prescribing more broader spectrum treatment that may not be optimal. This means that whatever diagnostic that is recommended needs to be suited to this type of environment—in other words it needs to be faster and simpler than what is already done, or it may not be a feasible solution in practice. There is also a wide range of temperatures and humidity that need to be taken into account from a technical perspective.
• The issue of getting oxygen to those patients who are diagnosed: Unfortunately, even patients who are diagnosed with pneumonia may not be able to receive oxygen therapy depending on the resources of the health facility.
• Reusable probes for pulse oximeters for neonates are hard to come by: Manufacturers of course have an incentive for making disposable probes as opposed to reusable ones, and this causes a financial issue for re-stocking those probes.
• Gold standard device for pulse oximetry is not an option: Since each device can vary by plus or minus 2%, there is no way to have a perfect device as a reference. In addition, the technical specifications of the devices do not always comply with what is given by the manufacturers.
• Procurement: Even though the overall pulse oximeter may last quite a while, the sensors do not have a long life and need to be replaced frequently. This industry makes more profit off of the sensors than the whole device.

Looking at it now, that is a whole lot of negative points, but it was a lot easier to take note of these issues than to summarize the solutions and future steps being carved out by these experts in order to establish a suitable reference point for devices and then test new devices based on this standard.

One interesting topic that came up was the new trend of crowdsourcing data; in other words, using a lot of data from a lot of different places to potentially create devices that integrate different parameters and information to help with decision points in the clinical setting. Examples of data mining can be seen in various fields, especially in research, where large data sets can vastly improve analyses. Current issues with this mostly involve getting people to share data. Attendees of the meeting pointed out that unfortunately, all developers are in competition with not another—they are not collaborating. By nature, they constantly have to compete for grants, publications, patents, etc., so there is no real incentive for sharing data that costs money and other resources to obtain. Nevertheless, there is a huge potential for this method in the future of diagnostics, so a diagnosis does not depend on a single reading or expert, but rather is based on probabilistic analysis for decision making done by a device.

Local Production and Technology Transfer (LPTT)

It is now week 4 of my internship, and although I have had very little success with my French skills, I have made some decent progress on my work on the Local Production and Technology Transfer project.

The objective of this project that I am involved in is to evaluate the capacity of countries for the local production and manufacturing of crucial medical devices. One of the first steps was to disseminate a survey to the countries to determine barriers to access to medical devices that are specific to individual countries. It includes everything from development to commercialization, and this information is important for the later stages of the project which will involve in-country workshops later this summer in Tanzania, Ethiopia, South Africa, and Nigeria. As other members of my unit are rushing to work out the logistics for these capacity building workshops, I am mostly working on the Phase II report, which previous interns from Rice have also contributed to in the last couple of years. This entails coordinating the contributions from our consultants for the publication as well as analysing the results of the survey and adding my own contributions to the report (to be reviewed by the real experts of course).

I’ve found that the results of the survey often reflect my own experience working on scaling-up access to the bCPAP (e.g. challenges of data collection, human resources, etc.), but I am also learning that the issue of increasing access to medical devices is more complicated than I can even really process at this point in my experience. There are so many different stakeholders with various expertise that need to be involved, from the development of the products, to regulatory approval, to quality manufacturing, to procurement, to usage, that it seems a small miracle to me that we have had success in getting these devices to the countries that need them the most. And there is so much more work ahead of us, but the great thing about WHO is being able to bring all these different stakeholders together to work towards the same ultimate result.