top of page

Leadership

Issue/Problem

​

Over the summer, I shadowed a Neurologist and noticed a weak point in the treatment strategy used for patients that decreases cost-effectiveness of the treatment plan not only for neurology, but for the entire healthcare field. The neurologist I shadowed specialized in pediatric epileptic disorders. We would go through several patients a day and for each patient, the neurologist would take into account the type of epileptic disorder and prescribe the most generic form of medication to the patient. The patient would then start on their “trial-and-error” period, meaning they would try out the medication on a trial basis for a few months and then report back to the neurologist. Depending on the side effects and/or efficacy of the medication, the doctor would then decide whether to continue prescribing that same medication or whether they should switch to a different drug to see if it was a better match. This would then continue until a good match was found. Therefore, finding the right medication to try to ease the epileptic symptoms could take several years using this “trial-and-error” approach. During this time period, these kids would have to potentially suffer from side effects or spend months trying out a drug that may not even be effective.

​

This trial-and-error approach to treatments is a common approach in many healthcare fields, not just in epileptic disorders. This is a problem because every person is different. Each person has a unique genetic background, gender, family history, lifestyle, and many more factors that make it almost impossible to predict how a patient might respond to a drug. 

​

One of the key factors that drives this ineffective approach is the idea that a patient’s treatment plan is decided from just one perspective. Although a doctor will be most knowledgeable and is very capable of making treatment decisions for their patient, perhaps there is another way to approach a treatment plan. I believe a goal can be reached quicker and more efficiently when diverse talents and minds work towards it. Therefore, the idea of “personalized medicine”, a concept I learned during my research at UT Southwestern, may in fact be capable of being more than just an abstract futuristic idea.

Solution/Recommendation

​

Personalized medicine can be achieved through the efforts of various minds working towards it. As I have experienced through my summer research experiences in Baltimore and Dallas, mentioned in Key Insight 1, there are hundreds of approaches and disciplines working together across the nation trying to cure cancer. The same is true of other diseases, so why not utilize various expertise areas to try to create a better approach to developing a treatment plan?

​

Although there is rationale behind the current trial-and-error treatment method, I believe this method does not consider enough factors. For example, computers are capable of sorting through vast amounts of data and therefore would be a smart way to help the medical field make logical decisions. Computers and algorithms can also help sort through previous mistakes made by doctors and ensure that the same mistakes are not made anymore.

​

Currently, one of the biggest steps towards personalized medicine being taken is using stem cells derived from a patient’s blood to test the drugs before giving the drugs to the patient (as learned in BMEN 389 as well as at a seminar given during my internship at UT Southwestern). However, this approach is several decades away from being a feasible plan. Using patient-derived stem cells will require many years of clinical trials and will require many ethical regulations to be implemented into the FDA’s drug approval process. Therefore, I believe using a computer algorithm to predict a patient’s outcome is a more reasonable plan that can be implemented in the meantime while the stem cell approach is being worked on.

Detailed Plan

​

A large database capable of storing vast amounts of information will have to be created. The NIH tends to fund long-term medical projects such as this. Therefore, assembling a team to collaboratively work together would be a realistic goal. Computer scientists will work together to create an algorithm that can sort through data and that can store data differentiated through different categories such as family history, side effects, age, gender, etc. This computer program should be able to recognize any patterns in data and determine if there are any correlations in factors. This program should also have a library of medications available and be able to create a list of the factors that best respond to each medication and which factors respond poorly to specific types of medications. Security software engineers can work alongside with them to make sure that the database is secure.

​

Healthcare facilities across the nation will have to input patient information into a database. This process will require a lot of paperwork as physicians will have to obtain the permission from their patients to input their personal health information. In this database, doctors will have the option to insert a patient’s medical history, gender, lifestyle, etc. as well as any medication they have been given and what their response was to the drug. Field technicians can speak at different institutions across the nation to encourage doctors to participate and to teach them how to use the database.

​

Once enough data is collected and the computer program is created, statisticians can work with the computer scientists to create statistical tests that determine which correlations among the data are statistically significant. Biologists and Physicians can look at these statistically significant correlations and determine the reasoning behind the correlations. A flow chart can then be created to help physicians follow a logical pathway created by these correlations and decide on the best medication. 

​

The NIH currently has a national project called “All of Us”. (1) For this project, they have already partnered with several hospitals, companies, and universities across the nation to help them recruit volunteers for their data bank. The NIH is also partnered with several associations such as the American Medical Association and the American Academy of Family Physicians.  Therefore, recruiting patients and any necessary personnel (listed below) would start with these institutions that have already partnered with NIH previously and are receptive to projects such as these.

​

Team Members Needed for the Project:

​

  1. Security teams : To make sure that the database is secure and cannot be hacked or reached by unauthorized users

  2. Lawyers : To make sure that all patient privacy rights are being secured and followed

  3. Field teachers : These speakers can talk to healthcare institutions to show doctors how to input information into the database, how to handle the paperwork, and how to explain to their patients what this data will be used for.

  4. Computer Scientists : To help create the database that doctors can input information into as well as create the computer program / algorithm that can sort through all the data

  5. Statisticians : To analyze data and find if there are any statistically significant correlations between any factors

  6. Biologists/Physicians : to look at the factors that seem to have a correlation and find the logical/molecular explanation for why those factors have correlations and help create the “flowchart” for doctors to use to find the best medication for their patient.

  7. Pharmacist: To review the medications recommended by the algorithm and help categorize the medication library

  8. Website designer : A person that works closely with the computer scientists to implement the algorithm results into a user-friendly flow chart or website that doctors can reference and use.

  9. Doctors : To work with the patients and log the information into the database. Doctors will also keep a doctor-patient relationship to ensure the patient is informed.

​

Evaluation Methods:

​

Doctors that create treatment plans derived from this database will follow up and input whether or not this treatment plan was effective or whether the patient had to return due to bad side effects or non-efficacy. A percentage can be calculated for the number of patients that had positive results from a treatment plan using this database.

​

One of the current large-scale patient-derived information projects funded by the NIH is called “All of Us”. This project has been running for about 3 years and, as projected by the agency’s director, expects to reach its goal of having 1 million patients enrolled by 2022. (1) Although the “All of Us” project does not specifically aim to create personalized treatment plans for specific diseases, this project has very similar implantation steps as the plan described above. Therefore, it would be expected that the time frame needed for this project to collect enough data to generate the needed user-friendly treatment plan website would range between 5 to 10 years. Though not possible for every aspect of the plan, many of the team members could work co-linearly to minimize the time needed to have this database fully running.

​

​

Works Cited

​

  1. “All of Us.” National Institutes of Health, U.S. Department of Health and Human Services, <allofus.nih.gov/>.

bottom of page