< back to White Paper Samples

Healthcare Sample
This white paper sample is written in a somewhat informal and conversational style. Usually, white papers have a more formal or businesslike tone, but don't hesitate to use a different approach to reach your audience more effectively.

Tailoring the Data: Using Data Warehousing to Prepare Healthcare-Related Messages — One Patient at a Time


As businesses and organizations acquire more data about the people who buy their products and use their services, they often find themselves wondering how to leverage this newfound wealth of information to reach some desired goal. This may seem simple enough until you realize how often organizations can’t seem to figure that out.

Most of the time, the desired result is usually some change in behavior (or the reinforcement of current behavior, such as continuing to purchase a certain product). Marketing experts think in terms of customer retention, developing brand loyalty, and individualized service, but what is really happening is behavior manipulation.

Of course, in the process of doing this we can often get people to do something that benefits themselves, too. Such is the case with using data warehousing as a tool to help people change their unhealthy behavior patterns into healthier ones.

The health care industry is starting to adopt such tools in their treatment procedures. Using data collected from surveys and established databases, organizations can develop programs to help participants reduce their risk factors for certain health problems.

The concept of tailoring is not new. Tailoring a message to a customer involves taking it from a generic form suitable for a mass audience and individualizing it so it personally relates to an individual or group of individuals. The degree of individuality achieved in the message is a function of both the amount and quality of the data collected and the ability to use this data to trigger messages that are appropriate to the individual.

Of course, it’s not very practical to have one devoted writer tapping away in a cubicle somewhere, laboriously composing a personal message. That’s fine if you have one or two recipients; it’s somewhat inefficient if you have twenty thousand other people waiting for their own “message.” This is where data warehousing becomes indispensable.

Getting Them to Listen

You must relate to each person in your intended audience if you hope to reach them. It’s no secret that people are interested in themselves and those things that are important to them. The data you’ve collected should paint a picture of each person and include a catalog of the things in their lives that matter to them. You can then use this data to craft a message that recognizes all the things that make them unique. This establishes the rapport that will keep them attentive and open to what you have to say.

The other side of this process is to eliminate those points in your message that have no relationship to your intended audience and that will only turn them off. People are bombarded with so much material nowadays that they are especially sensitive to anything that wastes their time. If they start to read material that is obviously sent out using a shotgun approach, they are likely to toss it before they get to your main message.

The issues involved with health care are a bit different because the subject can be so intimate and personal. That’s why it’s especially important to establish rapport early with your audience. Having a well-constructed database can help with this. It can also help make sure that any recommendations you provide are medically correct for that particular person.

The Advantages of Data Warehousing

There are numerous advantages to using data warehousing in the healthcare industry just as there are in other applications. When considering the delivery of tailored messages to healthcare program participants, the two primary advantages are 1) speed of message delivery and 2) timely and accurate database updates.

The information available for an individual, both medical and general, can be comprehensive and come from a variety of sources. For example, in a corporate healthcare program, information on each employee might be kept in a number of separate databases. The company’s Human Resources Department might maintain personal, family, and employment data. General health information might be found in the database run by the Managed Care Organization (MCO) that provides health care coverage for the workers. Individual physicians within the MCO might keep specific medical and test data on their own patients. Various sales and marketing departments involved with the MCO might collect information relating to the type of services requested from the MCO throughout the year. Organizations working with the MCO to administer different programs might keep similar data about their participants. These are just some examples of the types of databases that can maintain data on any one individual.

A tailored message prepared for one participant might require a variety of data from many or all of these disparate sources. Trying to get the databases to talk to each other and still deliver the appropriate message in a timely manner is virtually impossible. Part of the solution is to use a central data warehouse to collect and store the relevant information from the appropriate databases.

However, the “engine” that produces the tailored message doesn’t interact directly with this central data warehouse. Instead, a runtime database (also called a datamart) is generated as a subset of the all the data that is warehoused. This is what supplies the information that generates the message.

The runtime database can be customized according to its purpose. For instance, a company might wish to start a smoking cessation program for its workers who smoke. (Separating the smokers from the non-smokers is the most elementary form of tailoring in this case). The program will include tailored messages designed to reinforce the workers’ efforts to quit. Information about each of the smokers is collected from the warehouse and will describe any related health factors, their position in the company, their interests, marital status, how many children they have, and so on. These pieces of the total picture will then be collated into the runtime database. The tailored messages will be generated using all those bits of information and, if constructed correctly, will speak to each smoker in a personal and intimate way.

Now, say the company also wants to develop a tailored-message program for its workers who suffer from diabetes. The same procedure is put into place, only now the runtime database can contain information regarding the worker’s dietary habits, weight, exercise patterns, and so on. These would be pulled from whatever sources are appropriate.

The next question that might be asked is: “What if you have a smoker who also suffers from diabetes?” This scenario is neither unusual nor difficult to deal with. The key here is to make sure the data flow from warehouse to runtime database travels both ways. In other words, as information in any one database is updated, it will be reflected everywhere that needs to “know” about the update. So, the runtime database that got the diabetes information about the worker from a medical database also recorded information gathered from a participant survey that indicated this particular worker smokes. The medical database is then updated to reflect the smoker status for this patient. Subsequent healthcare related messages would now be tailored to reflect the fact that this person smokes. It is this intercommunication between databases—through the data warehouse—that allows the latest updates to be shared universally. And this is what enables all the tailored messages to have the same level of relevance.

The Data Collection Process

As noted above, much of the information residing in the data warehouse may have been collected from various sources, especially if the program is related to a participant’s employment. However, there will probably come a time when each participant is asked to complete a survey. Collecting data for a healthcare-related survey is often easier than with other intended applications for a couple of reasons. First, people will usually participate in such a survey because they desire the health-related information or advice that is being offered. This is in contrast to surveys that are clearly aimed at collecting data for subsequent sales pitches. They may have elected to take the survey on their own or they may have been given the opportunity through the health care plan they receive from their workplace. The survey can also be part of a voluntary follow-up support program offered in conjunction with the purchase of some health care product such as a patch or gum to help the participant quit smoking.

Second, as mentioned above, this data is not necessarily going to be used for subsequent sales pitches. This is especially true with data acquired as part of a healthcare plan-related program.

The data can be collected many ways. When you set up a large program that involves many participants on a long-term follow-up basis, it is often necessary to gather the bulk of the baseline data using a comprehensive questionnaire. This can be done using a printed form, a call-in center staffed by operators trained to conduct the survey, or a web-based data entry process. Participants could complete the initial survey in stages, picking up where they left off. This way, you would greatly reduce the number of people who would drop out because the survey was “too long” or started to bore them. Subsequent updates could be handled with the ease and simplicity of a toll-free phone call. Of course, other methods such as email, regular mail, fax, points-of-service, smart cards, kiosks, voice response/recognition, etc. could also be employed as seen fit.

Using the Data

Once collected, the data can be wed to a series of “rules.” The rules are logic statements within the computer coding that determine which messages and portions of messages are delivered to each participant. They are determined by such factors as the overall purpose of the program, the type of data collected, the level of personalization desired, the purpose for each message, the number of times you intend to contact the participant, and so on. For example, if you are speaking to a non-smoker about the risk factors for cardiovascular disease, it doesn’t make sense to go on and on about the dangers of smoking. You could, instead, use the data collected to determine if, for example, they are overweight (according to predetermined guidelines). The rules could then be set up to present a message about the health risks of being overweight. You could further refine the message according to how much over the “recommended” ideal weight this participant was. The possibilities and permutations are almost infinite and are ultimately determined by how comprehensive you want your message to be.

So, Why Tailor?

In a nutshell: to produce a more relevant message—one that gets better results than you’d get from something written in a generic, one-size-fits-all format.

In healthcare, using data warehousing to drive a process that produces tailored messages can help someone stay with a program that is designed to help them quit smoking, reduce their weight, get more exercise, and so on. Because the message is tailored to their particular stage in the process, it can act as an incentive to move to the next stage.

The messages can include a congratulatory statement that recognizing the progress the participant has made... suggestions for taking the next step... encouragement to stick with the program... useful information applicable to the participant, and so on. In every case, each subsequent message would be customized for each participant based on the individualized data stored in the system.

Data warehousing is the key to making this work. It has already been proven in real- world applications. As more and more organizations take advantage of this marriage between the psychology of change and the technology of tailoring, we will see an increase in the delivery of tailored messages to the masses—one person at a time.

< back to White Paper Samples