Collecting data, especially Big Data, seems to be a very important trend in healthcare and other businesses today. I have read several articles recently in the Wall Street Journal about the use of big data, especially by the government. Then there is the story of a teen who shopped at Target. Target sent an email to her family providing coupons for new baby goods, such as diapers. The father and mother who read the message did not even know their daughter was pregnant. Target did because of the data that they collected on the teen’s shopping habits; Target used data to predict future purchases of its customers.
These two illustrations provide good examples of the power of data. Both stories are somewhat controversial, as they involve privacy issues. The good news is that data can be collected and analyzed that is both non-invasive and that benefits clients and patients, as well as providers, whether in business or healthcare.
In order to be effective data collection must have a goal(s) that can be coupled with actions. The actions should be based upon the results of the analysis of the data collected. If a provider works with a data specialist who has knowledge of the provider’s service, such as primary care health service, then the data collection and analysis can be very useful. As data collection and analysis is new to many providers, many are not making the best use of the data that they collect. Following are several examples illustrating how useful data can be and how one nonprofit did not make the best use of the data that it collected.
Organization A was conducting a pilot project for an energy company to see if case management of clients who had shut-off notices from the energy company could improve the paying habits of the clients. Could interventions of up to six months by trained case managers using no financial assistance for paying energy bills result in better paying habits of the clients? Up until this time relief for clients who had shut-off notices was in the form of direct payment on the bills with funds from the government or donations for various agencies; the clients live in a state with very cold winters.
The project was well designed with a control group and various sources of data, including regular payment information on the clients, initial assessments and exit assessments. The data yielded a trove of data. From it the analyst on the project was able to determine a number of important conclusions. First, it was determined that clients who were involved in case management with no financial help for their heating bills became better bill payers on their energy bills. The average amount of arrearage for the clients decreased steadily. Up to 60% of the clients became “good bill payers” for the energy company. This was in contrast to the control group who had occasional financial assistance from outside agencies; this group’s arrearage did not decrease and their payment patterns were erratic. The group of clients under case management performed well for 20 months after enrollment in the program even though case management assistance had ceased months earlier. The changes in the habits of the clients were long term.
Besides establishing the success of the program, the analyst of the program was able to determine what were the most effective factors in creating change in the clients. The factors were the number of children in the household, whether the clients completed the two or more goals that they had set for themselves at the beginning of case management, whether the clients had savings accounts and whether the clients were involved in the Dave Ramsey “Financial Peace University”, a program provided by the organization. Knowing these were effective factors, the case managers focused upon them as they interacted with clients to produce ever better outcomes.
Although this program had a very specific goal, improving the payment habits of the energy clients, the processes and techniques are easily transferable to a number of businesses, including healthcare. A similar approach could easily be adapted to improve the outcomes for diabetics at a primary care site.
Researcher nurse B was interested in whether a well-designed educational intervention with a group of perimenopausal women could affect a change in the diet and exercise habits of the subjects. The nurse used three assessments in a pretest and post-test research design to determine whether significant changes occurred. With the help of a data analyst she was able to determine exactly what significant changes occurred in the knowledge base of the patients. Although the results did not determine whether the outcomes were sustainable it did indicate that there were significant changes in attitudes towards diet and exercise of the patients.
Organization C was grant funded in prevention services for teens. It collected data occasionally on some of its projects but did not consistently. When it did collect data it did so because the grant demanded that it do so. Some of the data that it collected was used to drive improvements in outcomes. Other sets of data analysis were not used at all to see if the organization was reaching its goals.
As you can see, the data collected by the first organization and the research nurse was used effectively. Both the organization and the nurse had clear goals that were explicitly written out. From these data assessments and collection techniques were devised. From these analysis yielded not only whether their actions were effective but also how to improve future interactions with clients.
The second organization was not nearly as effective in its use of data. Although some of its work could be proved to be effective, many of the activities in which it engaged could not be proved to be effective.
I believe that the second organization is very typical of businesses and providers who are just beginning to collect data. Such companies collect data without careful setting of goals and actions based upon the data collected. Rather, the data is collected in a rather haphazard fashion without any clear understanding of how it can be effectively used to drive outcomes.
Data collection and analysis can be very effective in improving the outcomes for an organization, whether it be healthcare, a business offering a service such as a media company, or a nonprofit serving a wide variety of clients, including the poor. The examples illustrate that clear goals must be written along with identification of possible actions based upon analysis of the data.