What's the Big Deal About Data?

In my previous positions, I often worked closely with data. The data would often reveal to those who could analyze, understand, and interpret the information what was truly going on in the hospital, throughout the organization as a whole, and even in the community. Working with those while reviewing data also showcased their levels of understanding about data. Their perception or grasp of the concepts or errors and willingness to share those concepts or errors became apparent very quickly.

One of the data issues that I dealt with regularly was integrity of the data. Because so many fingers touched the data, it was not always pristine.  Imagine a patient presents to the emergency department. As soon as the patient enters, data is collected on that patient. Each time someone enters a patient's demographic information into a hospital system, for example, the information could be entered differently.  Next, imagine an intake specialist selecting the wrong Dr. Smith or Dr. Jones. The information then travels with the patient. Dr. Smith or Dr. Jones now has been attached to this patient. The patient goes through the hospital process. Several nurses and techs on different shifts chart in the patient's record, and some may even use a different system. Some nurses may write on paper, even though the computerized record is available. All of this information comprises a patient's record, which is full of data. The patient's bill is calculated by coders, and the patient is contacted by internal and external sources to determine patient satisfaction. The data is entered into multiple systems by multiple users to hopefully create reports and demonstrate outcomes on which to base practice, create programs, and improve processes.

Photo courtesy of Pixabay.com

Unfortunately, by the time the data reaches the end user or the report creator...which used to be me, there were errors in the data. Information would be entered into different legacy systems which maybe did not talk to the new system. Drilling down into the data might have revealed that an incorrect diagnostic code was entered, transfer times were not entered appropriately, or the patient was assigned to a different Dr. Smith or Dr. Jones than the one that treated them. Identifying these issues may have seemed to some intuitive, but there is nothing intuitive about data. Rather, understanding and ensuring that data is pristine and accurate requires examining the integrity of the data and ensuring transparency of those reporting.

As a data person, a lot of times I would hear the phrase that "We need to find the single source of truth" for the data. However, there really was no single source of truth because there was no single source of data. Each element was entered by a separate someone creating a separate and distinct pattern of how they entered data or how the data was interpreted. Each time a patient presented, someone different may have recorded vital signs or fall assessments or pneumonia vaccine history or length of time between readmission. Rather, instead of finding a single source of truth, data experts needed to create a source of truth of data. That is, the data had to be analyzed to determine, based on best estimates and reviews, what the true data demonstrated. Clearly, an otolaryngologist did not perform a hysterectomy and a pediatrician more than likely did not treat geriatric patients at an adult only facility. To truly marry the data, legacy systems would have to be reviewed and analyzed. True systems thinking (not just information technology systems) would have to occur. Determining where each piece of data originated and how that data was utilized in the environment was essential.  

Knowing these pieces of information was crucial because, as many know, reimbursement rates are based on quality data. Accreditation, reviews, payments and even certificate of need determinations are made based on data. Grant funding, program development, and research uses data to fund, create, and innovate. Each time a flaw in the data is ignored, a flaw in the program has a chance to develop. Although these flaws may seem small and insignificant to some, they could cost the hospital money, time, energy, talent, and possibly lives. Data is the proof in the pudding. It shows that someone did something. It shows they documented it. It shows how well they performed their duties, how good the outcomes were, how closely clinicians are meeting regulatory requirements.  Thus, each time the data is analyzed without closer scrutiny, it becomes a big deal.