Data Mining in the Pharmaceutical Industry Essay Example

Data Mining in the Pharmaceutical Industry Essay Example Data Mining in the Pharmaceutical Industry Essay Data Mining in the Pharmaceutical Industry Essay A Look at Data Mining in the Pharmaceutical Industry Topics Covered: 1) What is Data Mining and why is it used? 2) How is Data Mining used in the Pharmaceutical Industry? 3) Recent debate in the legality of Data Mining and the Pharmaceutical Industry Pharmaceutical companies are taking advantage of the growing use of technology in the healthcare arena by using data to enhance their marketing efforts and increase the quality of research and development. The process of data mining allows companies to extract useful information from large sets of individual data. This process provides a knowledge that is vital to a pharmaceutical company’s competitive position and organizational decision-making. â€Å"Data Mining enables firms and organizations to make calculated decisions by assembling, accumulating, analyzing and accessing corporate data. It uses variety of tools like query and reporting tools, analytical processing tools, and Decision Support System (DSS) tools† (Rangan, 2007). 1) What is Data Mining and why is it used? Data mining is the practice of automatically searching large stores of data to discover patterns and trends that go beyond simple analysis. Data mining uses sophisticated mathematical algorithms to segment the data and evaluate the probability of future events. Data mining is also known as Knowledge Discovery in Data (KDD)† (Oracle, 2008). As stated, data mining is used to help find patterns and relationships stored within large sets of data, these patterns and relationships are then used to provide know ledge and value to the end user. The data can help prove and support earlier predictions usually based on statistics or aid in uncovering new information about products and customers. It is usually used by business intelligence organizations, and financial analysts, but is increasingly being used in the sciences to extract information from the enormous data sets generated by modern experimental and observational methods. Data mining is being increasingly used in business to help identify trends that would have otherwise gone unnoticed. There are several different opinions on the exact â€Å"steps† of data mining, but they all agree on these basics: planning, modeling and extracting information. Oracle defines 4 steps in the data mining process: 1) problem definition, 2) data gathering and preparation, 3) model building and evaluation, and 4) knowledge deployment. The first step of data mining is to understand the purpose, scope and requirements of the project . Once the project is specified from a business perspective, it can be formulated as a data mining problem and a preliminary implementation plan can be developed† (Oracle, 2008). The data gathering process takes a look at how well the data serves the purpose of the project. In this step many changes can be made to the attributes of the data so that they better serve the objective and requirements of the project. This process can play a large part in the value of the knowledge and information derived from the data. For example, you might transform a DATE_OF_BIRTH column to AGE; you might insert the average income in cases where the INCOME column is null† (Oracle, 2008). The third step of data mining is to build and evaluate the model. The model should be tested and evaluated to make sure that it will answer the question and stay within the requirements of the business objectives stated in the first phase of the process. The final phase includes knowledge deployment which is where actual information and realization comes from the data. Here is where the relationships and patterns are turned into something meaningful that meets the objective of the project. There are several techniques used for data mining, some of them have been used for decades prior to the information technology boom that has changed the system dramatically. According to (Alex Berson, 2000), these â€Å"classic† techniques include Statistics, Neighborhoods and Clustering while the â€Å"next generation† techniques include Trees, Networks and Rules. In the end the purpose of each of these techniques is to explore data (usually large amounts of data typically business or market related) in search of consistent patterns and/or systematic relationships between variables, and then to validate the findings by applying the detected patterns to new subsets of data† (StatSoft, 2011). As stated above, data mining is often used to solve business decision problems, â€Å"it provides ways to quantitatively measure what business users should already know qualitatively† (Linoff, 2004). A growing number of industries are using data mining to become more competitive in their market by primarily focusing on the customers; increasing their customer relationships and increasing customer acquisition. 2) How is Data Mining Used in the Pharmaceutical Industry? The pharmaceutical industry has copious uses for data mining which include increasing the efficiency of research and development, contributing to drug safety information and to increasing the effectiveness of their marketing efforts. Understanding that the benefit of data mining is allowing for the extraction of useful information from large sets of individual data, it is evident that the pharmaceutical industry has a need for this process. The abundance of diseases prevalent in the world, the multitude of drugs available for each disease, and the variety of patients that take the products, produces massive amounts of information available in the industry. Pharmaceutical companies have begun to use this data to benefit patient safety, physician knowledge and their own marketing efforts. Data mining can be used while companies are researching and testing new products. â€Å"Scientists run experiments to determine activity of potential drugs† (Rangan, 2007). They are able to use process that produce results and relationships much faster, they can quickly determine activity on â€Å"relevant genes or to find drug compounds that have desirable characteristics† (Rangan, 2007). â€Å"By relating the chemical structure of different compounds to their pharmacological activity, [data mining] can bringing a degree of predictability to drug screening procedures that, until now, have tended to be a bit hit and miss† (Results, 2009). That should help scientists and pharmaceutical companies identify more effective compounds to treat different diseases, allowing them to find drug leads in a fraction of the time and at a fraction of the cost of current methods. † The earlier methods of experimentation was very time consuming and had to be done over and ov er again each time a new drug was being researched, none of the old information was every used to help with speedier development. Data mining allows the past research to be used when picking compounds as opposed to just randomly choosing and testing. As a drug gets further into the development and into the clinical trial stages, data mining can help predict which diseases and patients will benefit from the drug. Based on past information data mining will provide a correlation between the new molecules, disease states and patients. For example, Pfizer is â€Å"turning to sophisticated data mining techniques to help improve the design of new trials, to better understand possible new uses for existing drugs, and to help examine how drugs are being used after they have been approved† (Salamone). During trial phases they are able to â€Å"understand safety and efficacy profiles within the patient population by tackling the question of patient selection within the framework of demonstrating groups that are most responsive. Data mining framework enables specialists to create customized nodes that can be shared throughout the organization† (Rangan, 2007). Additionally, one of the greatest benefits of data mining in the pharmaceutical industry and the healthcare world is the discovery of adverse events and drug toxicity in patients. It could help determine the adverse reactions associated with a specific drug and still go a step further to show if any specific condition aggravates the adverse reaction for eg age, sex, and obesity (Novartis Business Intelligence report, 2004). Data mining is useful in almost every stage of drug discovery and can aid in toxicity detection, side effect profiles and can work to uncover responsiveness in certain patients. â€Å"The patterns that emerge from data mining this information will not only improve our understanding of this disease, but could give practitioners new insights into prevention and treatment. (Rangan, 2007). As addressed above, a limiting factor in past and current pharmaceutical data is the sheer amount of data and lack of information that exists in the industry. Knowledge and information is being slowed at even a physician and patient level, for example the FDA estimates that only 1% of serious adverse events are actually reported to the companies after they h appen because most practicing healthcare providers do not have the time or means to report the adverse events and have no need for the data at a later time. There is a strong need for data mining techniques within the pharmaceutical industry to understand and detect possible adverse events before they happen to patients. Outside of product research, development and safety, pharmaceutical organizations are using data mining techniques to increase their marketing efforts directly to the consumer as well as to the prescribing physician. They are able to see a better return on the investment of resources based on mining the prescription data released by pharmacies. As discussed earlier, many businesses are using data mining to increase their customer relationships nd encourage product growth. They are able to use the data to gather knowledge and information in order to create more effective and efficient sales strategies. â€Å"Data mining can be used to supplement the pharmaceutical companies marketing efforts by market segmentation, measuring return on investment (ROI) and understanding profitable managed care formulary status. â€Å"A p harmaceutical company can analyze its recent sales force activity and their results to improve targeting of high-value physicians and determine which marketing activities will have the greatest impact in the next few months. The data needs to include competitor market activity as well as information about the local health care systems. The results can be distributed to the sales force via a wide-area network that enables the representatives to review the recommendations from the perspective of the key attributes in the decision process. The ongoing, dynamic analysis of the data warehouse allows best practices from throughout the organization to be applied in specific sales situations. † (Alex Berson, 2000) Market segmentation allows for tailored messaging and information to be given to appropriate customers where their need is specifically met. Prescribing information allows the sales representatives to spend appropriate time and resources on customers that have the most need for individual products based on their patient population and historical prescribing trends. â€Å"Supplemented by survey data, patient and physician interviews, information gleaned from epidemiological studies and managed care organizations, questionnaires on web sites, and other market research, a quite detailed picture of a customer base can be identified, with marketing strategies devised accordingly† (Cohen). This is critical at the launch of a product, in order to determine the â€Å"early adopters† that will drive a product use and share their success with professional peers. A successful product launch to the right market segment can allow a product to surpass its competitors in the field. Especially in the era of â€Å"me-too† products with similar efficacy and slightly lower side effects than earlier competitors, the effectiveness of a product launch is vital to the career of the product. Identifying the early adopters and focusing tailored promotional efforts on this segment (as opposed to broadcasting a general message to all physicians) can be crucial to the success of the product† (Cohen). Measuring the ROI of certain programs and resources the organizations use, can help save time and money by making sure the resources are being put in the most favorable places to produce the most amount of business and patient satisfaction. Data mining allows pharmaceutical companies to get an idea of how their field promotions and direct to consumer promotion programs are driving business results. The promotional efforts of these organizations are tremendous and indlude field promotions: representative sales calls, peer-to-peer dinner meetings, exhibits at conventions, promotional samples, and direct to consumer advertising which include: commercials, websites, patient education materials the companies spend billions of dollars on the promotion of a single product. Data mining can help stream line the customers that are targeted for these promotional events and help make more accurate decisions on where to spend their resources so that they make sense for the physician, patient and the organization. The formulary status of a specific drug is very dependent on the location and area in question. The managed care market dynamics are very critical to effective targeting and marketing of pharmaceutical companies. Physicians are generally unaware of specific prescription coverage on certain health care plans, especially if a pharmacy benefits manager (PBM) is used in lieu of the actual healthcare benefits to manage prescriptions. For this reason it is important that each organization have the data to support the needs of the various customers and plans to help physicians overcome the obstacles they encounter while prescribing certain drugs to their specific environment. There are many facets of the pharmaceutical industry, including patient care and marketing that can widely benefit from utilizing decision support systems and data mining. The process is revolutionizing early drug discovery and increasing the speed and effectiveness that scientists have in uncovering new molecules to treat various disease states. It has a place in patient safety by providing early detection of drug-on-drug interactions, toxicity and adverse reactions. Finally it is widely used to supplement the marketing efforts in the field and increase the business acumen and accuracy of the promotional side of the industry. 3) Recent debate in the legality of Data Mining and the Pharmaceutical Industry In the past 10 years, the pharmaceutical industry has been scrutinized for various activities that have lawmakers on the watch and uncertain about the agenda these organizations have when promoting their products. With the blatant need for healthcare reform and a slow demise of the American healthcare industry this scrutiny of the pharmaceutical industry has steadily increased and the reputation of these organizations has increasingly plummeted. Within this scrutiny, authorities have begun to question the lawfulness of data mining and the use of prescription-drug records used in promotional efforts. Some argue that the data-mining is purely to grow market share for money-hungry companies and has little relation to the care or need of patients and physicians. As recently as 2011 the Supreme Court heard a case assessing the legality of prescription-drug records being used to promote pharmaceutical products. After a patient fills a prescription â€Å"pharmacies can sell the other information in those prescriptions to data-mining companies (they cannot sell patient identification information), who sift through all this information, spot trends and patterns, and then sell that to, as in this case, drug companies, who can then have their sales representatives do targeted marketing of brand-name drugs to doctors† (Coyle, 2011). Drug makers buy prescription records that reveal the prescribing practices of individual doctors from data mining companies and, based on the information, practice a type of marketing called detailing, in which sales representatives, who already know which doctors prescribe certain kinds of medications, pitch information about new drugs they think will be of interest to the doctor† (Lewis, 2011). The discrepancy existed in the State of Vermont where lawmakers made it unlawful to sell this information without the prescribing physicians consent, however this law was ruled unconstitutional in the lower federal appellate court, bringing the decision to the Supreme Court. The following is an excerpt from an interview done after the hearing. Vermonts purpose in enacting the law was to protect the privacy of the doctors information, to encourage prescription of generic drugs, which would help lower health costs in the state, and also to protect the public health, which it felt could be endangered by drug companies sales representatives presenting one-sided information to the doctors. Then, on the other side, you have the drug companies and other businesses concerned that if the court restricts access to this kind of information then, that they wont get the kind of information they say they need to make important business decisions, ot just marketing decisions, research decisions, ot her decisions that they think could be beneficial to consumers† (Coyle, 2011). The Supreme Court ended up over-ruling this decision based on the First Amendment right and gave pharmaceutical industries a big victory in their use of Data Mining. â€Å"The Supreme Court handed down a 6-3 majority decision and ruled that the law interfered with the pharmaceutical industrys First Amendment right to market its products (Lewis, 2011). † Despite the controversy, it is evident that there is a wealth of knowledge and information to be gained by the use of data mining in the pharmaceutical industry. It is a process that allows an organization to streamline the massive amounts of data and make educated research developments and business decisions based on the information. Alex Berson, S. S. (2000). Building Data Application for CRM. McGraw-Hill. Cohen, J. (n. d. ). Data Mining of Market Knowledge in the Pharmaceutical Industry. Data Mining of Market Knowledge in the Pharmaceutical Industry. Coyle, M. (2011, April 26). National Law Journal. (R. Suarez, Interviewer) Lewis, N. (2011, Januray 24). Drug Prescription Data Mining Cleared by the Supreme Court. Retrieved August 09, 2011, from Informtion Week: informationweek. com/news/healthcare/security-privacy/231000397 Linoff, G. (2004). Data Miners. Retrieved July 31, 2011, from Data Miners Inc. : data-miners. com/resources/SUGI29-Survival. pdf Oracle. (2008, May). Data Mining Concepts. Retrieved July 31, 2011, from Oracle: http://download. oracle. com/docs/cd/B28359_01/datamine. 111/b28129/process. htm Rangan, J. (2007). Applications of Data Mining Techniques in the Pharmaceutical Industry. Journal of Theoretical and Implied Information Technology, 7. Results, I. (2009, Feb 3). Data Mining Promises to Dig Up New Drugs. Retrieved August 9, 2011, from Science Daily: sciencedaily. com/releases/2009/02/090202140042. htm Salamone, S. (n. d. ). Pfizer Data Mining Focuses on Clinical Trials. Retrieved August 09, 2011, from Bio. It. Com: bio-itworld. com/newsitems/2006/february/02-23-06-news-pfizer StatSoft. (2011). Statsoft: Data Mining Techniques. Retrieved July 31, 2011, from Statsoft: statsoft. com/textbook/data-mining-techniques/#eda

The History and Domestication of Bananas

The History and Domestication of Bananas Bananas (Musa spp) are a tropical crop, and a staple in the wet tropic areas of Africa, the Americas, mainland and island Southeast Asia, South Asia, Melanesia and the Pacific islands. Perhaps 87% of the total bananas consumed worldwide today are locally consumed; the rest is distributed outside of the wet tropical regions in which they are grown. Today there are hundreds of fully domesticated banana varieties, and an uncertain number are still in various stages of domestication: that is to say, they still are inter-fertile with wild populations. Bananas are basically giant herbs, rather than trees, and there are approximately 50 species in the Musa genus, which includes the edible forms of bananas and plantains. The genus is split into four or five sections, based on the number of chromosomes in the plant, and the region where they are found. Furthermore, over a thousand different types of cultivars of bananas and plantains are recognized today. The different varieties are characterized by wide differences in peel color and thickness, flavor, fruit size, and resistance to disease. The bright yellow one found most frequently in western markets is called the Cavendish. Cultivating Bananas Bananas produce vegetative suckers at the base of the plant which can be removed and planted separately. Bananas are planted at a typical density of between 1500-2500 plants per square hectare. Between 9-14 months after planting, each plant produces some 20-40 kilograms of fruit. After the harvest, the plant is cut down, and one sucker is allowed to grow up to produce the next crop. Banana Phytoliths The evolution of bananas are difficult to study archaeologically, and so the domestication history was unknowable until recently. Banana pollen, seeds, and pseudostem impressions are quite rare or absent at archaeological sites, and much of the recent research has been focused on the relatively new technologies associated with opal phytoliths- basically silicon copies of cells created by the plant itself. Banana phytoliths are uniquely shaped: they are volcaniform, shaped like little volcanoes with a flat crater at the top. There are differences in the phytoliths between varieties of bananas, but variations between wild and domesticated versions are not as yet definitive, so additional forms of research need to be used to fully understand banana domestication. Genetics and Linguistics Genetics and linguistic studies also help in understanding banana history. Diploid and triploid forms of bananas have been identified, and their distribution throughout the world is a key piece of evidence. In addition, linguistic studies of local terms for bananas support the notion of the spread of the banana away from its point of origin: island southeast Asia. Exploitation of early wild forms of bananas has been noted at the Beli-Lena site of Sri Lanka by c 11,500-13,500 BP, Gua Chwawas in Malaysia by 10,700 BP, and Poyang Lake, China by 11,500 BP. Kuk Swamp, in Papua New Guinea, so far the earliest unequivocal evidence for banana cultivation, had wild bananas there throughout the Holocene, and banana phytoliths are associated with the earliest human occupations at Kuk Swamp, between ~10,220-9910 cal BP. Todays Hybridized Bananas Bananas have been cultivated and hybridized a number of times over several thousand years, so well concentrate on the original domestication, and leave the hybridization to botanists. All edible bananas today are hybridized from  Musa acuminata  (diploid) or  M. acuminata  crossed with  M. balbisiana  (triploid). Today,  M. acuminata  is found throughout mainland and island southeast Asia including the eastern half of the Indian subcontinent;  M. balbisiana  is mostly found in mainland Southeast Asia. Genetic changes from  M. acuminata  created by the domestication process include the suppression of seeds and the development of parthenocarpy: the ability of humans to create a new crop without the need for fertilization. Bananas Across the World Archaeological evidence from the  Kuk Swamp  of the highlands of New Guinea indicates that bananas were deliberately planted by at least as long ago as 5000-4490 BC (6950-6440 cal BP). Additional evidence indicates that  Musa acuminata  ssp  banksii  F. Muell was dispersed out of New Guinea and introduced into eastern Africa by ~3000 BC (Munsa and Nkang), and into South Asia (the Harappan site of Kot Diji) by 2500 cal BC, and probably earlier. The earliest banana evidence found in Africa is from  Munsa, a site in Uganda dated to 3220 cal BC, although there are problems with the stratigraphy and chronology. The earliest well-supported evidence is at Nkang, a site located in southern Cameroon, which contained banana phytoliths dated between 2,750 to 2,100 BP. Like  coconuts, bananas were most widely spread as a result of the sea exploration of the Pacific by Lapita peoples ca 3000 BP, of extensive trade voyages throughout the Indian Ocean by Arab traders, and of exploration of the Americas by Europeans. Sources Ball T, Vrydaghs L, Van Den Hauwe I, Manwaring J, and De Langhe E. 2006.  Differentiating banana phytoliths: wild and edible Musa acuminata and Musa Journal of Archaeological Science 33(9):1228-1236.balbisiana.   De Langhe E, Vrydaghs L, de Maret P, Perrier X, and Denham T. 2009. Why Bananas Matter: An introduction to the history of banana domestication.  Ethnobotany Research Applications  7:165-177. Open Access Denham T, Fullagar R, and Head L. 2009.  Plant exploitation on Sahul: From   Quaternary International  202(1-2):29-40.colonisation to the emergence of regional specialisation during the Holocene. Denham TP, Harberle SG, Lentfer C, Fullagar R, Field J, Therin M, Porch N, and Winsborough B. 2003.  Origins of Agriculture at Kuk Swamp in the Highlands of New Guinea.  Science  301(5630):189-193. Donohue M, and Denham T. 2009.  Banana (Musa spp.) Domestication in the Asia-Pacific Region: Linguistic and archaeobotanical perspectives.  Ethnobotany Research Applications  7:293-332. Open Access Heslop-Harrison JS, and Schwarzacher T. 2007.  Domestication, Genomics and the Future for Banana.  Annals of Botany  100(5):1073-1084. Lejju BJ, Robertshaw P, and Taylor D. 2006.  Africas earliest bananas?  Journal of Archaeological Science  33(1):102-113. Pearsall DM. 2008.  Plant . In: Pearsall DM, editor.  Encyclopedia of Archaeology. London: Elsevier Inc. p 1822-1842.domestication Perrier X, De Langhe E, Donohue M, Lentfer C, Vrydaghs L, Bakry F, Carreel F, Hippolyte I, Horry J-P, Jenny C et al. 2011.  Multidisciplinary perspectives on banana (Musa spp.) domestication.  Proceedings of the National Academy of Sciences  Early Edition.