Introduction
Recently, the world has witnessed several large scale natural disasters: the Tsunami that devastated many of the countries around the rim of the Indian Ocean in December 2004, extensive flooding in many parts of Europe in August 2005, hurricane Katrina in September 2005, the outbreak of Severe Acute Respiratory Syndrome (SARS) in many regions of Asia and Canada in 2003, and the earthquake disaster in Pakistan towards the end of 2005 . These emergency and disaster situations (E&DS) serve to underscore the utter chaos that ensues in the aftermath of such events, the many casualties and loss of life, not to mention the devastation and destruction that is left behind. One recurring theme that is apparent in all these situations is that irrespective of the warnings of the imminent threats, countries have not been prepared and ready to exhibit effective and efficient crisis management. This paper examines the application of the tools, techniques, and processes of the knowledge economy to develop a prescriptive model that will support superior decision making in E&DS and thereby enable effective and efficient crisis management.
Background
Changing weather patterns, rapid urbanization, expansion of industry, not to mention development of air and ground transportation networks, population growth and migration, and recently - acts of terrorism , are associated with ever increasing frequency of major disasters involving multiple casualties (von Lubitz, Carrasco, Fausone, Gabbrielli, Kirk, Lary, and Levine, 2005). Emergency Healthcare management is a complex process which has to be tackled on various fronts (Beltrame, Maryni and Orsi, 1998; Kun and Bray, 2002). Such situations require effective crisis management capability, that is, pre-hospital and emergency/trauma, in-hospital medical services, firefighting, disaster-related law enforcement operations, and so forth, and superior decision making capabilities (von Lubitz, et al., 2005; von Lubitz and Wickramasinghe, 2005a; 2005b). Most of these services are governed by different local or national agencies, are subject to different rules and regulations, and develop independent operational plans. This in turn leads to the gathering and storing of data in disparate databases. However, given the interdependent nature of these elements, any decision making based on only one or a few of these data elements will logically provide only a partial picture and, thus, an inferior decision. Hence, it is necessary to collect multi-spectral data, analyze this data in aggregate to develop a complete picture if we are to truly support superior decisions. To do this effectively and efficiently, it is imperative to embrace the tools, techniques and processes of the knowledge economy (Liebowittz, 1999; Maier and Lehner, 2000; Shapiro and Verian, 1999; von Lubitz and Wickramasinghe, 2005b; Wickramasinghe, 2005; Wilcox, 1997; Zack, 1999). Advances in IT, coupled with the advent of Knowledge Management (KM), can facilitate better processes for efficient and effective healthcare (Dwivedi, Bali, James, Naguib, and Johnston; 2002).
Main Focus
The Intelligence Continuum consists of a collection of key tools, techniques, and processes of the knowledge economy; that is, including data mining, business intelligence/analytics and knowledge management which are applied to a generic system of people, process and technology in a systematic and ordered fashion (Wickramasinghe and Schaffer, 2005). Taken together, they represent a very powerful instrument for refining the data raw material stored in data marts and/ or data warehouses and thereby maximizing the value and utility of these data assets. As depicted in Figure 1, the intelligence continuum is applied to the output of the generic information system. Once applied, the results become part of the data set that are reintroduced into the system and combined with the other inputs of people, processes, and technology to develop an improvement continuum. Thus, the intelligence continuum includes the generation of data, the analysis of these data to provide a "diagnosis," and the reintroduction into the cycle as a "prescriptive" solution. In this way, continuous learning is invoked and the future state always builds on the lessons of the current state.
The key capabilities and power of the model are in analyzing large volumes of disparate, multi-spectral data so that superior decision making can ensue. This is achieved through the incorporation of the various intelligence tools and techniques which taken together make it possible to analyze all data elements in aggregate. Currently, most analysis of data is applied to single data sets and uses at most two of these techniques (Newell, Robertson, Scarbrough, and Swan, 2002; Nonaka, 1994; Nonaka and Nishiguchi, 2001; Schultze and Leidner, 2002; von Lubitz and Wickra-masinghe, 2005b; Wickramasinghe, 2005; Wickramasinghe and Schaffer, 2005). Thus, there is neither the power nor the capabilities to analyze large volumes of multi-spectral data (ibid.). Moreover, the interaction with domain experts is typically non-existent in current methods. The benefits of applying the capabilities of the intelligence continuum to E&DS scenarios are profound indeed. E&DS scenarios are concomitant with complex, unstable, and unpredictable environments where the unknown or position of information inferiority prevails. Hence, these scenarios are chaotic and sub-optimal decision making typical results. In contrast, the tools and techniques of the intelligence continuum can serve to transform the situation of information inferiority to one of information superiority in real time through the effective and efficient processing of disparate, diverse, and seemingly unrelated data. This enables decision makers to make superior decisions which in turn lessen the chaos and facilitates the restoring of order. In order to appreciate the power of the intelligence continuum in such scenarios, it is necessary to briefly describe its key elements.
Data Mining
Due to the immense size of the data sets, computerized techniques are essential to help decision makers understand relationships and associations between data elements. Data mining is closely associated with databases and shares some common ground with statistics since both strive toward discovering structure in data. However, while statistical analysis starts with some kind of hypothesis about the data, data mining does not. Furthermore, data mining is much more suited to deal with heterogeneous databases, data sets, and data fields, which are typical of data in E&DS that contain numerous types of text and graphical data sets. Data mining also draws heavily from many other disciplines, most notably, machine learning, artificial intelligence, and database technology.
From a micro perspective, data mining is a vital step in the broader context of the knowledge discovery in databases (KDD) that transforms data into knowledge by identifying valid, novel, potentially useful, and ultimately understand able patterns in data (Adriaans and Zantinge, 1996; Bendoly, 2003; Cabena, Hadjinian, Stadler, Verhees, and Zanasi, 1998; Fayyad, Piatetsky-Shapiro, Smyth, 1996). KDD plays an important role in data-driven decision support systems that include query tools, report generators, statistical analysis tools, data warehousing, and on-line analytic processing (OLAP). Data mining algorithms are used on data sets for model building, or for finding patterns and relationships in data. How to manage such newly discovered knowledge, as well as other organizational knowledge assets, is the realm of knowledge management.
Figure 1
Figure 2 shows an integrated view of the knowledge discovery process, the evolution of knowledge from data to information to knowledge, and the types of data mining (exploratory and predictive) and their interrelationships. In Figure 2, all the major aspects connected with data mining are captured and by so doing the integral role of data mining to knowledge creation is emphasized. This is not normally explicitly articulated in the existing literature although the connection between data, information, and knowledge is often discussed (Becerra-Fernandez and Sabherwal, 2001; Choi and Lee, 2003; Chung and Gray, 1996; Holsapple and Joshi, 2002).
Data mining then, is the non-trivial process of identifying valid, novel, potentially useful, and ultimately understandable patterns from data (Fayyad, et al., 1996). It is essential to emphasize here the importance of the interaction with experts who always play a crucial and indispensable role in any knowledge discovery process in facilitating prediction of key patterns and also identification of new patterns and trends.
Business Intelligence/Analytics
Another technology-driven technique, like data mining connected to knowledge creation, is the area of business intelligence and the now newer term of business analytics. The business intelligence (BI) term has become synonymous with an umbrella description for a wide range of decision-support tools, some of which target specific user audiences (Wickramasinghe, 2005; Wickramasinghe and Schaffer, 2005). At the bottom of the BI hierarchy are extraction and formatting tools which are also known as data-extraction tools. These tools collect data from existing databases for inclusion in data warehouses and data marts. Thus, the next level of the BI hierarchy is known as warehouses and marts. Existing healthcare information systems are not generally designed to cater to new (data) needs (Anderson, 1997). Because the data come from so many different, often incompatible systems in various file formats, the next step in the BI hierarchy is formatting tools. These tools and techniques are used to "cleanse" the data and convert it to formats that can easily be understood in the data warehouse or data mart. Next, tools are needed to support the reporting and analytical techniques. These are known as enterprise reporting and analytical tools. On-line analytic process (OLAP) engines and analytical application-development tools are for professionals who analyze data and do, for example, business forecasting, modeling, and trend analysis. Human intelligence tools form the next level in the hierarchy and involve human expertise, opinions, and observations to be recorded to create a knowledge repository. These tools are at the very top of the BI hierarchy and serve to amalgamate analytical and BI capabilities along with human expertise. Business analytics (BA) is a newer term that tends to be viewed as a sub-set of the broader business intelligence umbrella and concentrates on the analytic aspects within BI by focusing on the simultaneous analysis of patterns and trends in a given context (Wickramasinghe and Schaffer, 2005).
Figure 2
Knowledge Management
Knowledge Management is an emerging management approach that is aimed at solving the current business challenges to increase efficiency and efficacy of core business processes while simultaneously incorporating continuous innovation. Specifically, knowledge management through the use of various tools, processes, and techniques combines germane organizational data, information and knowledge to create business value, and enable an organization to capitalize on its intangible and human assets so that it can effectively achieve its primary business goals as well as maximize its core business competencies (Newell, et al., 2002; Nonaka, 1994; Nonaka and Nishiguchi, 2001; Schultze and Leidner, 2002; von Lubitz and Wickramasinghe, 2005b; Wickramasinghe, 2005; Wickramasinghe and Schaffer, 2005). The importance of knowledge management is confirmed by the increasing attention that the subject has received from both researchers and practitioners (Huang and Newell, 2003).
Broadly speaking, knowledge management involves four key steps of creating/generating knowledge, representing/storing knowledge, accessing/using/re-using knowledge, and disseminating/transferring knowledge (von Lubitz and Wickramasinghe, 2005a; Wickramasinghe, 2005b; Wickra-masinghe and Schaffer, 2005).
Knowledge Management (KM) as a discipline is said not to have a commonly accepted or de facto definition. However, some common ground has been established which covers the following points. KM is a multi-disciplinary paradigm (Gupta, Iyer & Aronson, 2000) which often uses technology to support the acquisition, generation, codification, and transfer ofknowledge in the context of specific organizational processes. Knowledge can either be tacit or explicit (explicit knowledge typically takes the form of company documents and is easily available, whilst tacit knowledge is subjective and cognitive). As tacit knowledge is often stored in the minds of healthcare professionals, the ultimate objective of KM is to transform tacit knowledge into explicit knowledge to allow effective dissemination (Bali, 2005).
KM initiatives should be incorporated in conjunction with the technological revolution that is occurring within healthcare organizations. A balance is required between organizational and technological aspects of the healthcare process (Dwivedi, et al. 2001a).
KM can enable the healthcare sector to successfully overcome the information and knowledge explosion by way of appropriate frameworks customized for healthcare institutions (Dwivedi, et al., 2001b, 2002a).
Knowledge Generation In Dynamic And unpredictable Environments
Hierarchically, gathering of information precedes transformation of information into useable knowledge (Alavi and Leidner, 1999; Massey, Montoya-Weiss, and O'Driscoll, 2002). Hence, the rate of information collection and the quality of the collected information will have a major impact on the quality (usefulness) of the generated knowledge (Chang, et al., 2005). In dynamic and unstable environments, relative to the environment, the decision maker is in a position of tremendous information inferiority. In order to make effective decisions he/she must rapidly process seemingly irrelevant data and information into relevant and useable knowledge (Award and Ghaziri, 2004; Boyd, 1976; Courtney, 2001; Drucker, 1993; Newell, et al., 2002; Schultze and Leidner, 2002; Wickramasinghe, 2005). This necessitates a process perspective to knowledge management (von Lubitz and Wickramasinghe, 2005b; Wickramasinghe and von Lubitz, 2006). The cornerstone of such a perspective is the OODA Loop (Figure 3) which provides formalized analysis of the processes involved in the development of a superior strategy (Boyd, 1976, 1987; von Lubitz and Wickramasinghe, 2006).
The Loop is based on a cycle of four interrelated stages revolving in time and space: Observation, followed by Orientation, then by Determination, and finally Action. At the Observation and Orientation stages, multispectral implicit and explicit inputs are gathered (Observation) and converted into coherent information (Orientation). The latter determines the sequential Determination (knowledge generation) and Action (practical implementation of knowledge) steps. The outcome of the latter affects, in turn, the character of the starting point (Observation) of the next revolution in the forward progression of the rolling loop. The Orientation stage specifies the characteristics and the nature of the "center of thrust" at which the effort is to concentrate during the Determination and Action stages. Hence, the Loop implicitly incorporates the rule of "economy of force," that is, the requirement that only minimum but adequate (containment) effort is applied to insignificant aspects of competitive interaction. The Loop exists as a network of simultaneous and intertwined events that characterize the multidimensional action space (competition space), and both influence and are influenced by the actor (e.g., an organization) at the centre of the network.
Figure 3.
It is the incorporation of the dynamic aspect of the "action space" that makes the Loop particularly useful to environments that are inherently unstable and unpredictable, that is, medicine, business, war and emergency, and disaster scenarios (von Lubitz and Wickramasinghe, 2005a; 2005b; 2006).
future trends
The need for more effective, superior crisis management techniques is clearly becoming apparent as we embark upon post crisis analysis for each of the more recent disasters from 9-11, to the Tsunami in December2004, the floods in Europe, hurricanes Katrina and Rita and the earthquakes in Pakistan. This area is the focus for the emerging discipline of Operations Other than War (OOTW). Commonly considered as military, OOTW now becomes increasingly civilian-driven, and often executed as interventions in potentially unstable environments, or as management activities consequent to destabilizing events. In addition to the most obvious topic of terrorism, where risk assessment and management are prerogative to meaningful counteraction, problems of assessing and managing consequences of natural disasters, epidemic diseases, major industrial or transportation accidents, or humanitarian relief operations, become increasingly relevant. Key issues for OOTW include (Richards, 2004):
• Risk factors and their management
• Preparedness/readiness
• Political factors
• International organizations/national organizations/ NGOs in OOTW
• International cooperation
• Military/civilian interaction
• Law/law enforcement
• Healthcare and medical aspects o telemedical operations
o medical logistics
o medical information networks in disaster operations
• ICT technology-based tools facilitating assessment and management of risk
o data mining/business analytics o simulation/modeling o training in complex synthetic environments
• Field operations and analysis of practical execution
As this nascent field evolves the above areas will form a central research focus for scholars in the near future.
conclusions
Sound emergency management requires the ability to:
1. Focus on solvable problems;
2. Priorities the elements of a problem in terms of how much progress can be achieved with each element in a small amount of time;
3. Delegate responsibility;
4. Manage the "span of control;"
5. Communicate clearly and rationally;
6. Keep a level head in a crisis; and
7. Make sound decisions.
However, when we analyze the recent natural disasters, a common recurring and unfortunate situation is that countries and regions are never as prepared and ready for the eminent disaster as they perhaps could have been. It is too late once the disaster strikes to have an organized and systematic fashion for contending with the aftermath. What is required is to be able to analyze past crises and develop appropriate lessons to apply to future events. In the advent of a health crisis, knowledge management is in a position to improve information sharing and coordination. The intelligence continuum model coupled with a process perspective of knowledge management appears to fill this void as it can be applied to existing and disparate data elements from past disasters in order to build a predictive model that can facilitate in the development of sound procedures and protocols to facilitate preparedness and readiness a priority so that ex-ante operations can, in fact, be more effective and efficient, decision making superior and order replace much of the chaos.
• Preparedness, unless based on broadly-based knowledge is useless. The development of appropriate preparedness is predominantly a strategic task that requires intimate knowledge of several aspects of the environment.
• Readiness, unless based on germane knowledge, is useless. Thus, coping with the sudden and unpredictable event requires the background of germane knowledge that will dictate the nature of the subsequent response. Readiness is, therefore, context dependent.
• Readiness is the most essential tool in response to, and containment of, an unexpected threat. While intuitively obvious, the practical development of readiness is not an easy task. Possession of knowledge is not equivalent to the ability to employ it under the stress of less-than-routine circumstances; that is, E&DS.
Hence in E&DS, what is needed is to be prepared and ready which, in turn, requires not only the possession of pertinent information and germane knowledge, but also the ability to apply it successfully; evoke superior decision making. Efficient flow of information is necessary in managing an outbreak (Kun and Bray, 2002). Hurricane Katrina serves to highlight how vulnerable and insufficient existing crisis management techniques are (CNN 2005a-c) as well as to underscore that developing better techniques through the utilization of critical data sources should be remedied immediately. The Intelligence continuum offers such a possibility.
KEY TERMS
Data Mining and KDD Process: Knowledge discovery in databases (KDD) (and more specifically data mining) approaches knowledge creation from a primarily technology driven perspective. In particular, the KDD process focuses on how data is transformed into knowledge by identifying valid, novel, potentially useful, and ultimately understandable patterns in data (Fayyad, et al., 1996). From an application perspective, data mining and KDD are often used interchangeably.
Explicit Knowledge: Or factual knowledge, that is, "know what", represents knowledge that is well established and documented.
Germane Knowledge: The sum total of all information plus the ability to implement it constructively and purposefully in the dynamic and unstable environment.
Knowledge Spiral: The process of transforming the form of knowledge, and, thus, increasing the extant knowledge base as well as the amount and utilization of the knowledge within the organization.
Pertinent Information: Information structured data, grouped into coherent categories that are easily perceptible and understood.
Preparedness: The availability (prepositioning) of all resources, both human and physical, necessary for the management of, or the consequences of, a specific event or event complex .
Readiness: The instantaneous ability to respond to a suddenly arising major crisis (e.g. sudden slow-down in the manufacturing parts supply chain) that is based on the instantaneously and locally available/ un-prepositioned and un-mobilized countermeasure resources.
Tacit Knowledge: Or experiential knowledge, that is, "know how" represents knowledge that is gained through experience and through doing.