There are two famous health triangles in the healthcare world. The first triangle comprises diagnosis, treatment and prognosis. The second triangle is called the “Iron Triangle of Healthcare”. Its sides are access, cost, and quality.
We can consider Healthcare Datasets and Big Data Analytics to be the median of every triangle. This is becoming a fact that cannot be ignored or denied.
THE FIRST TRIANGLE
First side: Diagnosis and Testing
The appearance of big data and huge healthcare datasets achieved great improvement in the diagnosis and testing process of health triangle.
The progressive, noticeable and increased rate of migration from traditional paper-based records to Electronic Health Records (EHRs) has spontaneously caused a mutation in the world of big data tools.
The synchronous evolution and development of different sciences that became complimentary enhanced the data analytical process. The rapid improvement in EHRs, Computerized Provider Order Entry (CPOE), Big Data Science, Prognostication and Prediction, in addition to the appearance of the Hadoop Family and Ecosystem led to high-quality Clinical Decision Support Systems (CDSS) and the appearance of reliable Evidence-Based tools. Evidence-Based applications ensure that the clinical decisions are the best and the most updated, whether it is related to diagnosis or treatment.
Clinical cases that share the same symptoms can easily be filtered out and grouped, where different analytical procedures can be done with tools like R, Python or similar tools.
Common diagnostic indicators (like patient demographics, income, habits, or socioeconomic level) can be collected and recorded. This can facilitate the diagnosis process for similar cases in the future. This can also be used to confirm the initial diagnosis.
A clear evidence for the role of Big Data in the diagnostic process is the Million Veteran Program established by the Department of Veterans Affairs. The aim of this research is to determine the variability in treatment effect among Veterans, risk factors for some diseases and the best preventive methods for these diseases. All these questions have to be answered through the analysis of genetic data for more than 150,000 veterans.
Fast and accurate diagnosis can save lives.
Second side: Treatment
By collecting and analyzing different EMRs, we can determine which treatment plan had the best outcomes.
Many medical centers started using big data to build clinical decision support systems that determine the best treatment plan for patients.
Beth Israel Deaconess Medical Center (BIDMC) has very distinct, innovative and important applications depending on the concepts of big data analytics like “BIDMC@Home” and “Screening Sheets”. Moreover, BIDMC has established a clinical query platform. BIDMC clinical query platform is an artificial intelligence system where huge patients’ records are accessed and analyzed to answer big data inquiries. The platform is a self-service application where you can ask simple or complex questions that may lead to a wiser clinical decision like which therapy had the best effect, or which lifestyle can aggravate or resolve the condition.
The Informatics for Integrating Biology and the Bedside (i2b2) center have developed an application that is adopted by 60 medical centers all over the world. Harvard hospitals were among the early adopters of i2b2. The aim of this application is to improve the design of specific therapies for diseases of genetic origin.
Third side: Prognosis
The emergence of the prognostic indicators and IT healthcare companies working in the field of prediction are very clear evidence of the importance of Big Data in the field of prognosis.
Repeated outcomes for specific conditions may help in the estimation of the length of stay (LOS), adverse effects, hospital acquired illness, disability, risk of disease, or possibility of readmission.
Gathering prognostic indicators in huge databases and using different analytical software can now predict the outcome of the clinical case. Soon, by knowing some data about a patient suffering from chronic low back pain for example, we can determine the prognosis (outcome) of the case and whether it could affect his ability to return to work, normal lifestyle, or the possibility of any disability.
THE SECOND TRIANGLE
First side: Quality
Predictive analytics is becoming one of the most important part of quality improvement in healthcare.
By applying artificial intelligent (AI) algorithms to preexisting EMRs, we can determine future patterns and ensures a good quality decision regarding the choice of appropriate lifestyle for the patient, in addition to the selection of the best line of treatment and a high rated service provider on a scientific basis rather than personal judgment.
We can measure the quality of a certain medication, specific line of treatment, service delivered for a predefined clinical condition. Once, we can measure the quality, we can simply figure out the deficiency. Whenever we realize that there is a problem, we can again use big data to analyze the problem and hence we can treat the defects and enhance the service.
Second side: Cost
Using Big Data Analytics and NLP in healthcare process can definitely reduce the overall costs. Making use of Big Data Analytics can save us from conducting unnecessary tests.
In one of his studies, McKinsey reported that U.S Healthcare sector can save $300 billion per year. This can be explained further as follows:
- 8% saving in the national healthcare expenditures, and this could represent two-thirds of the total amount.
- Research and development: $108 billion.
- Clinical operations: $165 billion.
A simple U.S big data analytics research proved that using Aspirin can reduce markedly the risk of coronary heart disease. This can reduce healthcare expenditure up to $30 billion.
It is reported that using big data analytics could reduce the costs of smoking and obesity by over $38 billion. Currently, obesity and smoking cost the world about $2 trillion.
Third side: Access
When emergency rooms are over flooded with patients and wait-time starts to exceed the limits, some physicians prescribe predetermined treatment without any consideration to the full situation of the patient. Using big data analytical tools and evidence-based applications can mitigate the risk of the so-called “Quick Cook Book Predetermined Treatment”, as they depend on reported successful treatment similar clinical cases.
This could definitely reduce the wait-time and improve the accessibility to the Healthcare service without affecting the quality of the clinical decision.