Turning data into information: the cost of making diamonds from coal.
As with any process, there are quantifiable costs associated with taking any raw material (such as data) and refining it into a finished product (useable/actionable information). Diamonds are only converted from coal with the application of intense heat and pressure. So too, does data require the intense heat and pressure of processing to become useable information.
Such "heat and pressure" does not come without cost. Much of the process of turning data into information is hidden from the end-users of the information and is not factored accurately in the determination of whether the information is economically valuable.
The cost associated with processing data into information must be balanced against the benefit gained from the information. To have economic value, the benefit of the information must at least equal, if not exceed, time costs.
Information can be broken clown into four cost centers:
* Data collection
* Data input
* Data analysis
* Information reporting
Each of these can be broken down further into:
* Direct costs
* Labor--Dedicated staff to perform the functions
* Materials--Essentially the "raw" data used to perform the data analysis
* Indirect costs--Materials (or labor) that may not be directly attributable to the data analysis process, but are used to perform analysis functions (such as computers used for data analysis and employees who participate in facets of data analysis but are not dedicated it)
* Overhead costs--Costs (facility, maintenance, etc.) that allow for the data analysis to occur that are not direct materials or direct labor
Before embarking on a process of generating information, you should address each of the cost centers individually. This does not need to be done with accounting precision. It should be an exercise to identify potentially hidden costs in processing data into information.
Data collection can occur in many forms. It may require manual abstraction from review of records or it may be done electronically. For example, if you are abstracting data manually from a medical record, costs may include the direct labor of record abstraction as well as the simple costs of tools to document the data, such as a spreadsheet, database, or pencil and paper.
If data is being abstracted electronically costs may include time data analyst/programmer who needs to create and run the report and materials may include the computer hardware and software necessary to generate data.
While the data input step may at tinges be incorporated as part of the data collection process (or may replace it), many times it is a data processing step in its own right.
Manually collected data has to be entered into a data analysis tool, whether it is a manual tool or electronic database, for further analysis. Electronic data may have to be imported or interfaced from a data repository to your data analysis tool.
This function may also have its own associated costs, such as direct labor to manage and perform this task, as well as material costs if electronic interfaces need to be established between databases.
Ultimately, the goal of data processing is to take raw data and find some meaning in it, whether it is examining trends, testing hypotheses, establishing correlations, determining causes of outcomes or possibly being able to predict outcomes.
To take advantage of data, you must be able to analyze it and this usually requires a direct labor cost of a data analyst and may entail direct costs of a dedicated data analysis tool.
Once information is obtained, it must be communicated appropriately to those who have the ability to use it and make decisions. Communicating data starts with making it comprehensible to the decision makers.
Depending on the complexity of the information, this may require specialized tools/resources to present this information appropriately (direct material cost) as well as staff with specialized skills at data display or facilitation/presentation skills (direct labor costs).
In addition to depicting the information in an understandable format, it must be disseminated to the responsible decision makers. This may range from a minimal cost to prepare a report for a single individual or a more substantial cost to prepare reports/graphics/presentations for large numbers of stakeholders.
The economic value of information
Obviously, significant resources can be consumed generating informarion. Much of this process may remain completely hidden from the end user/decision makers.
There are anecdotes in every organization about reports being produced and distributed (sometimes at significant cost) to managers who admittedly find no value n the information.
For information to be of economic value, it must outweigh the cost associated with generating that information. If weekly length of stay data is generated in a hospital organization and distributed to all managers and there is ultimately a reduction in length of stay that translates to decreased cost to the institution, that information has proven economic value.
That the outcome is achieved is less important than the fact that the cost to generate this data is expected to be less than the potential out come achieved. Moreover, it is important to determine at the start of a data processing project that the information generated has the potential of creating value for the organization.
To knowingly expend resources on an information project that does not have the potential to exceed the costs of the data processing is irresponsible. It is an important task of executives who control information management resources to make their colleagues who request informarion aware of these costs.
It is helpful for managers who rely on information to perform this type of analysis to boost their argument about the need for the information.
Case study: Hidden costs can cripple
Hospital A wanted to implement a risk-adjusted clinical outcomes database to aid in identifying and prioritizing performance improvement projects, generate physician utilization and outcomes profiles and examine resource allocation issues (benefits of implementing information processing system).
A number of technology vendors presented their systems for evaluation to the hospital's senior management and System X was chosen for its ability to generate the correct types of reports.
A contract was signed that included the costs of:
* Software licenses
* Data processing
* Training certain staff to use database (costs of processing data into information)
The IT department, undergoing significant reorganization at the time and only peripherally involved in the evaluation of this software solution, was informed of the need to implement the system.
For System X to process the data and generate reports, data needed to be abstracted in an electronic format from Hospital A's clinical information system, validated for accuracy and sent to System X.
Fortunately, the data required for abstraction resided in the hospital's electronic clinical system. However, the IT department did not have the resources available to create the data abstraction program internally and this function was outsourced to a consultant at significant cost (hidden cost of data extraction).
Additionally, once data were processed by System X, staff at Hospital A were dedicated to generating, interpreting and communicating results from System X to management at Hospital A so that informed decisions could be made regarding performance improvement prioritization and physician and hospital resource utilization (hidden cost of information reporting).
It was also determined that within the terms of the contract, the volume of data being sent to System X for processing exceeded the cap specified in the contract and Hospital A incurred a per case processing fee (unplanned-for cost of data processing).
While use of System X had a number of well-defined benefits, significant costs were hidden from the managers making the decision to implement this system.
Ultimately, once these additional costs were overcome, System X provided significant decision-making information to the organization.
However, it's easy to see how surmounting costs could have easily negated the potential benefits of implementing such a system.
Information Scorecard The Cost of Information The Value of Information Who will collect my data and what Will the information obtained resources will they need to do it? help generate revenue? How will my data get input into an Will the information help analysis tool? contain or reduc cost? What are the costs associated with my Will the information help make data analysis in terms of labor a department/service more and resources? efficient? Who requires this information and in Will the information improve what format will I need to provide it quality? to be valuable in the decision making process? Will the information influence employee or customer satisfaction?
Paul P. Antonecchia, MD, MBA, FACP is medical director, informatics at St. Vincent's Medical Center, Bridgeport, Conn. He. can be reached by phone at 203-576-961 or by e-mail at email@example.com
Teresa Kryspin, MD, MBA, FACP is medical director for the Family Health (.Enter clinic at St. Vincent's Medical Center, Bridgeport, Conn. She can be reached by phone at 203-576-5701 or by e-mail at firstname.lastname@example.org
Mike H. Summerer, MD, MS, FACP, FACC, CPE is corporate senior vice president/chief medical officer at St. Vincent's Medical Center, Bridgeport, Conn. He can he reached by phone at 203-576 5411 or by e-mail at email@example.com.
|Printer friendly Cite/link Email Feedback|
|Date:||Sep 1, 2003|
|Previous Article:||A business case for patient safety.|
|Next Article:||Are you on the right track?|