Acceptance of new systems is always difficult, regardless of what industry you are in. Systems require investments of time and money to build, and only pay for themselves if they are used. This is especially true with business intelligence applications because the concepts are not usually a normal part of most users workdays, and yet the final link in the chain between data and business benefits is the decision-maker using the data.
Clinical providers in general and doctors in particular, are especially hard to reach with business intelligence for a number of reasons. First of all, these people are clinical, not business people. Most of the time they have all they can handle just keeping up with the flood of clinical information coming at them daily. To add business concepts and decisions just adds to the wave.
It would be presumptuous of me to tell you that I know how doctors think and work because I am not a doctor. What follows is a composite view of interviews with dozens of doctors and hundreds of healthcare professionals in numerous organizations such as clinics, hospitals, administrative and executive offices, insurance companies, research labs, government agencies and academic settings.
One interview with a Chief Medical Officer (CMO) of an integrated healthcare organization was particularly helpful in developing these five tips to increase doctors’ usage of business intelligence information. This CMO had only 30 minutes for the interview, so I asked him to help me answer the question of what will improve the chances that our information and application will be used. He thought about it a moment and said he would give me his view of how doctors think and work based on his 30 years in medicine, and let me translate that into what our application needs to provide.
Key Concepts: Motion and Focus
Two concepts must be kept clearly in mind when designing a business intelligence application, if you want it to be used. These are motion and focus.
•A Doctor’s World is in Constant Motion. Everything in a doctor’s world moves constantly. Their day is characterized by moving from one patient to the next. The patients themselves are constantly changing in terms of their health status (hopefully for the better) as well as in their thoughts, actions and priorities as their lives change. And the medical information, knowledge, techniques and technology they use is constantly in motion. This year one treatment is recommended for a particular condition, next year it may not be. Diets, drugs and lifestyle guidelines are constantly changing. And of course, there is the relentless march of quality standards, increased paperwork and concerns over how to get paid for the work you do.
•A Doctor’s Work is Focused. As a consequence of all of this motion, doctors need to stay focused on the tasks needed to make sure their patients are getting the best care possible, in the most efficient and cost-effective manner. At times it may seem that doctors, and all healthcare professionals are a little too brusque, but this is a result of attempting to stay focused in an environment of constantly shifting sands.
With these two key concepts in mind, let me share with you two essential ingredients in your business intelligence application, and five ways to ensure that your application will be used to produce benefits for users and the organizations they work for.
Two Musts
Before looking at what will encourage doctors to use the information, it is important to look at two qualities that the data must have. If you ignore either of these, you will lose your audience immediately and probably never get them back.
Data Must Be Pristine. The data absolutely must be accurate from the doctor’s perspective, so data quality is critical. Doctors simply do not have the time or the patience to find your mistakes, wait for you to get it cleaned up or to reconcile the discrepancies. This is not as easy as it sounds, however, even for information that is vital to the organization. Virtually every commercial organization on the planet considers its customer information to be the most critical data it has. But it is likely that these same organizations have several different customer files that don’t match, as well as years of obsolete customer records and garbage data from people working around old systems to get their jobs done. I once worked with a retailer who had 29 major systems that used customer sales data as part of their processing. And they had 29 customer files that did not talk to each other. Therefore, if you asked me what last month’s sales were, I could give you 29 official answers. And rarely did any of them match.
To many users, business intelligence information and concepts are new. Therefore, the data is “on probation” until it proves itself. Make sure it is pristine to begin with, and stays pristine.
Information Must Be Relevant. You may have the most accurate data in the world, but it needs to be relevant. If not, you will lose your readers. To a clinical person, business data comes with a built-in handicap. It is not familiar to them. Dollars, time measures, quick ratios, growth percents and margins are as foreign to a clinician as sensitivity, specificity, incidence and affinity rates are to the average businessperson. This gap can be bridged with information that is tangible and real to the clinical reader such as comparative dollars spent treating a specific condition in one clinic vs. the amount spent in other clinics. This type of information will generate questions in the reader’s mind like what is the best clinic doing and can we copy their method, or what is the worst one doing wrong. But first, it must be relevant and understandable to the user.
Five Tips to Increase Usage of Business Intelligence
Once you have secured the two essential ingredients of accuracy and relevance and have not turned your users off completely, then you can begin to address ways to increase the usage of business intelligence applications. Over the past few years, I have seen and used these five methods repeatedly in designing business applications for use by clinical people.
Clinical Decision Support vs. Business Decision Support. Set aside for a moment the fact that the contents of a clinical decision support system will be different than the contents of a business decision support system, such as a business intelligence application. Not only are the contents different, but the central goal of these two types of systems is different.
The goal of a clinical decision support system is to distill large amounts of information down to one answer. It could be a single diagnosis, a single treatment or a single expected outcome. By contrast, the goal of a business decision support system is to gather information and apply it to many different questions to get many different answers. For instance, customer sales data can be used to drive product development decisions, marketing decisions, investment decisions and staffing decisions, to name a few.
To illustrate this difference, I often use funnels, canisters and pipes as mental pictures. A clinical decision support system can be viewed as a funnel with a number of types of information coming in from studies, research results, lab tests, observation, etc. This data is often stored in a clinical data repository (CDR), which is the canister. From this canister, a pipe is used to deliver the single result to the clinician to answer one of those questions mentioned above.
A business decision support system, on the other hand, has the same type of funnel gathering data from a number of sources such as internal systems, the internet, purchased industry data, etc. This data is funneled into a canister or data repository. Then, instead of a pipe coming out, another funnel, this one flipped over delivers information to many users across the organization, to be used in a number of decisions at various executive, management, analyst and operational levels.
Even in clinical organizations, you still want to develop your business intelligence application with the two funnels and the canister approach. But understanding this difference will help you explain the immense value of business intelligence much better.
Deliver Information Inline. Traditionally, business intelligence applications have been developed to be used by analytical staff like financial analysts, marketing research analysts, informaticians, researchers, etc. For these people, the system is a central part of their work and the interfaces are designed with a stand-alone approach. Even executive dashboards are designed to be a system that an executive or manager “logs into” in order to view the information. This cannot be the case with doctors and clinical professionals. They are unique among professionals in that their central focus is a person and the physical aspects of that person. Attorneys, accountants and educators, for instance, are also professionals who work with people. But in those cases, information either from, about or for that person is the central focus of their work. In clinical healthcare, information needed to support ones work needs to be delivered inline, or along the same stream as other information used in the clinician’s daily work. Doctors are simply accustomed to receiving data this way. Therefore, to improve usage of business intelligence information, it is important to deliver it in the same way.
Use a Familiar Format. Many business intelligence applications developers have an information technical (IT) background, and tend to create interfaces with an IT-centric look and feel. This is jarring to clinical professionals. To increase acceptance of your application, make it look as familiar as possible to them. Make it look like the vehicles through which they currently receive information. In this way, they will easily slip into the interface and quickly begin to use the information. I think the CMO mentioned earlier said it best. For printed documents, the easiest way to get doctors to use them is to make them look like medical journals. Medical journals have a distinctive font, color selection, style of presenting numerical and graphical information, as well as supporting information such as the way citations are done. Doctors are used to consuming medical journals, and they will more readily be used to consuming your information if it looks similar. As far as electronic information, just use online medical journals as your guide.
Line Up Your Numbers. This may seem like a strange suggestion, and one that should be such common sense that it barely warrants mentioning. Line up your numbers so that the entire column has the same unit of measure. For instance, let’s say you present a report of number of patients, average length of stay, and insurance dollars submitted, by hospital department. Respectively, the measures are counts, days and dollars. This may seem simple to present as a matrix where the counts, days and dollars all line up, but for some reason some applications are designed with the information stacked in different ways. This aggravates busy doctors because they have to constantly be checking on what the number is referring to as they read the report. Don’t let this be the last time they read your report.
Imply Action. Finally, and perhaps most importantly, imply action with your reports. Clinical people are action-oriented, not exploratory-oriented. The same information can be presented in an informing way, or as a directive to act upon. If your business intelligence application presents information that implies action, it will be used. Here is an example where the first is informational, and the second is directive: 1. Eighty-seven percent of patients with Condition X come in for focused visits, and 2. Patients requiring focused visits for Condition X is 13 percent. Design your information to imply action.
A Better Way to Reap the Benefits of Business Intelligence
The scope of your business intelligence application can include any of a number of subject areas for analysis, such as patient analysis, outcomes analysis, pay-for-performance analysis, efficiency analysis, patient safety analysis, etc. Each of these can be very valuable for a clinical healthcare organization to measure, monitor and analyze. But that is the key. The information must be used, or the potentially significant investment in business intelligence applications will be wasted. The keys to getting a return on this investment by getting your application used may be simple, or may require an intuitive leap. The goal is simple. Make it usable, and it will pay for itself.
Thanks for reading. I look forward to your comments.