Application of Big Data Analytics in Transfusion Medicine

Abstract

WHAT IS BIG DATA? There is no formal definition of “Big Data.” Big data refers to extremely large and diverse collections of structured, unstructured, and semi-structured data that continues to grow exponentially over time. These datasets are so huge and complex that traditional data management systems cannot store, process, and analyze them. Big data analytics (BDA) may also be called as the science of data management and analysis.[1] For instance, tracking consumer behavior and shopping habits to deliver personalized retail product recommendations tailored to individual customers is done using BDA integrated with artificial intelligence and machine learning (AIML). The term Big data was first coined in 1997 in computer science literature and is best explained by 3 Vs-Volume, Velocity, and Variety.[2] Volume refers to the amount of data generated, Velocity to the speed with which big data are generated and variety refers to the type of data-structured, semi-structured, or unstructured. Today, the world is drowned in data. The common examples of very large data sets or “ Big data” are video subscription libraries such as Netflix, data contained in social media networks such as Linkedin, Facebook, or Instagram that are continuously and rapidly growing every second. The banking and insurance sectors employ large data sets. In healthcare setting, there is a huge amount of data contained in blood donor databases of institutions and nations. There are lots of uninterpreted patient data lying idle in hospital, blood center, and laboratory databases. Businesses understand the value of analyzing huge data pertaining to their interests. Testimony to the fact is the fact that we get bombared by numerous phone calls when we search for a service or product. When you surf the net for vehicle insurance, you are bombarded with phone calls selling them within a minute. Imagine the volume of data (number of customers X variables) that are processed, analyzed simultaneously, and delivered to the right people within seconds for them to call you immediately, a potentially serious customer. This is the beauty of big data analysis integrated with AIML. APPLICATION OF BIG DATA ANALYTICS IN HEALTH CARE Unlike many industries, transfusion medicine is just beginning to understand the value of this untapped data. This is growing and is poised to grow even further in the field of health care, particularly in the fields of diagnostics, preventive medicine, precision medicine, and population health. BIG DATA ANALYTICS IN MOLECULAR BIOLOGY Sequence of nucleotides in human genome is an example for big data. In molecular biology – DNA sequencing and next generation sequencing is used to determine the order of nucleotides in entire genomes or targeted regions of DNA or RNA which generates vast amounts of complex databases which need to be compared and analyzed. When so many fields are analyzing large data sets to make meaningful interpretations for human and business development, can transfusion medicine be far behind? BIG DATA ANALYTICS IN TRANSFUSION MEDICINE In transfusion practices, big data have been used for benchmarking, detection of transfusion-related complications, determining patterns of blood use, and defining blood order schedules for surgery, besides the quality control of red cells.[3] These are some areas where BDA can be employed [Figure 1].Figure 1: Applications of big data in transfusion medicineAnalyzing donor databases Imagine the number of donors worldwide and the data contained in varies national databases. We know that nearly 11 million donations happen in India annually. If we want to compare the demographic profile of Indian blood donors vis a vis the global data comprising of nearly 118.5 million blood donations, it is not easy. Processing, analyzing, and comparing data such as age, sex, weight, blood pressure, hemoglobin levels, and education levels is not an easy task. This is where BDA can come to our aid. EXTRAPOLATION BASED ON SMALL DATA SETS VERSUS BIG DATA ANALYTICS If we want to know the extent of overweight or obesity in a population that is otherwise healthy, but likely to develop lifestyles diseases in future, the population to look at or healthy blood donors or people undergoing routine health checks. These numbers are in millions and are probably the right population to look at. Similarly, if we want to know the status of anemia in a population, anemia in blood donors are indicative of the problem. In the absence of BDA whatever we do is extrapolation based on analysis of small data sets which may not always be correct. A testimony to how extrapolations based on small datasets can go wrong is the exit polls in India in recently conducted National elections, which went completely haywire when the exact poll results were counted. BIG DATA ANALYTICS IN TRANSFUSION PRACTICE If we want to study the rationality of transfusion or its benefits-before and after transfusion, it is a herculean task, analyzing all patients receiving transfusions which will be in hundreds of thousands, if not millions and has to be compared with multiple parameters such as age, department, surgeon, diagnosis, indication for transfusion, hemoglobin, platelet counts, fibrinogen levels, and so on. In the absence of BDA, our ability to process this information manually or using small data sets gives only limited information. This may be accomplished at the institutional level using existing software but if we need to compare this with national or international databases, it would be next to impossible in the absence of BDA. Without BDA which can compare multiple parameters, all these data goes waste. BIG DATA ANALYTICS IN RECIPIENT AND DONOR HEMOVIGILANCE National Hemovigilance Systems contain data pertaining to millions of patients and donors. Processing this data, its analysis and making meaningful interpretations for guiding transfusion services needs the capacity to handle Big Data. Big data analysis was in use by National Hemovigilance Program of India till recently. Transfusion-related immune modulation may lead to an increased risk of patient complications, none of which are currently monitored by post-transfusion surveillance systems.[4] Electronic decision making: It is common knowledge that at the time of monthly transfusions for thalassemia, the previous Laboratory test results such as hemoglobin levels are not easily available and in its absence the quantity of blood transfused is always an approximation. Data analysis at bedside and displaying multiple results with mean and standard deviation will lead to informed decisions. Currently BDA and AI have been employed for quality control of red cells only.[5] CONCLUSION BDA is useful in analyzing donor data bases, transfusion databases, and hemovigilance databases. Combined with artificial intelligence, it can make a huge difference to the way transfusion medicine is practiced today.

Ähnliche Arbeiten