Special Focus: Medical Negligence Analyzing and Visualizing Health Care Metrics Obtaining and presenting defendant’s health data as a compelling storytelling tool
By Alex Ackel
This article was reposted from the May 2024 issue of Trial News, the monthly newspaper of the Washington State Association for Justice.
In the simplest terms, data visualization is the graphic representation and presentation of data. Effective data visualization can be a potent storytelling tool. Our brains rely on visual context for processing and storing information. When done well, data visualization can distill complex—and sometimes monotonous—information so that it can be easily understood and remembered.
For various reasons, in the context of medical malpractice cases, it can often be difficult to obtain usable or relevant data to present to the jury. First, obstacles such as concerns for patient privacy and quality improvement privileges can sometimes be legitimate objections to requests for relevant health care data. Additionally, because of the inherent information imbalance we face, it can be impossible to know what health data is discoverable and obtainable. The goal of this article is to provide a brief introduction to data visualization and provide some considerations for obtaining and presenting data in your next medical malpractice case.
Considerations for making effective data visualizations
In creating an effective data visual, your main goal is to make it easy for the viewer to understand the point you are trying to make. Complexity is your enemy. The whole purpose of data visualization is to convey complicated information in a manner that is easy to digest and remember. The challenge is balancing the need to present enough information so that the visual is impactful, but without too much detail to overwhelm or distract the viewer. Your audience should know exactly what they are looking at within five seconds of seeing your visual and then understand the conclusion your visual is making within the next five seconds, regardless of the complexity of the subject matter. As a general rule, your visual should only include data and elements that are strictly necessary for making your point. Simplicity and clarity should be your two driving considerations.
Generally, an effective visual has four key elements: (1) the data/information; (2) your story; (3) your conclusion; and (4) the visual form itself. The story is what makes the data relevant to your viewer and your conclusion or goal is what makes the visual useful to the jury in determining the issues in the case.
Obtaining useful data
In the world of business, data is essential for understanding and measuring performance. The healthcare industry is no different. In fact, the proliferation of electronic medical records systems has made health care providers especially capable of generating a vast array of different types of health care statistics. Overall, analysis of this data has led to higher quality health care and improved the accessibility of patient information. Health care providers regularly collect and analyze this data to evaluate their overall performance. This data can be helpful, for instance, in proving a pattern or practice of delivering substandard care.
For example, in Ryan v. Staten Island Univ. Hops.,1 the plaintiff claimed that her terminally ill husband was fraudulently lured into unnecessary treatment by defendants’ aggressive false advertising. The plaintiff moved to compel the production of a database which listed all of the patients who were treated with the surgery her husband received. The database included a treasure trove of data fields such as diagnoses, year of treatment, type and location of cancer, date of birth, diagnosis date, date of death, dates of follow-up examinations, treatment start and end date, date of recurrence, and treatment completion.
In rejecting the defendant’s argument that HIPAA protected information like diagnosis and dates of treatment, the court held that as long as all personal identifying information was redacted, information that might otherwise by protected by HIPAA was discoverable.
Hospitals also collect and analyze data to evaluate the performance of its contractors. Also, hospitals often set targets for various metrics in their service agreements with their contractors. I have personally seen hospitals offer payment incentives to their contracted emergency room providers for meeting certain goals related to patient satisfaction survey results, wait times, and adherence to documentation requirements. All of that is potentially valuable data depending on the nature of your case.
For example, you might consider a request for production for any and all data collected and compiled from patient satisfaction surveys. Or even broader, a request for any and all data collected and compiled for the purpose of evaluating contractor performance. The latter may be more likely to yield an objection, but the raw data itself should not be protected. Remember, one of the marks of a good database is the ability to export discrete portions of data for analysis. Any burdensome objections are likely to be overstated.
One way of circumventing a potential objection is by issuing an interrogatory asking for a detailed description of the means by which the hospital evaluates the performance of its contractors, including any and all categories of data that it collects or analyzes. That may allow you to narrow your request for production to relevant data fields and may also give you valuable information for contesting the defendant’s objection.
But what if the data being tracked isn’t useful or specific enough to be meaningful in your case? Then you may have to collect and analyze your own dataset through discovery. For instance, if you are anticipating the defense arguing that the hospital was overburdened by a particular high volume of patients, you might consider the following:
INTERROGATORY NO. 1: For each and every patient who was admitted on the day of the Incident between 6:00am and 6:00pm, please specify the:
(a) time of arrival;
(b) time of triage;
(c) time roomed;
(d) time seen by a provider;
(e) chief complaint;
(f) acuity designation;
(g) provider name;
(h) discharge diagnosis; and
(i) exit time.
INTERROGATORY NO. 2: What was the average and median Arrival to Room time (time from when the patient arrived to when the patient was roomed) at the Defendant’s Emergency Department on the day of the Incident?
INTERROGATORY NO. 3: What was the average and median Room to Provider time (time from when the patient was roomed to when the patient was seen by a provider) at the Defendant’s Emergency Department on the day of the Incident?
We used the above in a failure to timely triage case in which our client presented with obviously life-threatening symptoms but was left waiting for 6 hours before being seen. That information, by itself, wasn’t enough to prove our liability case. The defense could argue that there were other patients whose symptoms were either more or just as critical as our client’s. Or, they could simply argue that they were a lot busier than normal that day. We needed to collect information about what was actually happening in the emergency department that day before we could know the best strategy for attacking those defenses. We propounded the above discovery and then converted the data into the visualization below.
Scatter plot visuals like the one at the bottom of this page are particularly good at showing the relationship between two variables from one set of data; in this case, patient acuity and wait times. They are most powerful when used to demonstrate contrast between the general population and an outlier, as seen in the illustration.
This visualization also demonstrates the importance of clarity and simplicity. Within only a few seconds, the viewer knows exactly what is being shown. The data is clear and easily read. The visual fits seamlessly into the story of a delay in treatment case and the conclusion almost goes without saying: the plaintiff arrived as one of the most critically ill patients at the hospital and was forced to wait the longest.
Overall, data visualization is a powerful tool for conveying large amounts of information into a small space. By thinking of visuals as an extension of your theme and theory, your visuals can add meaning to the data you obtain and make it useful and usable to the jury.
1 No. 04-CV-2666, 2006 WL 3497875 (E.D.N.Y. Dec. 5, 2006).
Alex Ackel is an EAGLE member whose practice focuses on medical malpractice, civil rights, products liability, and personal injury at Friedman Rubin PLLC in Seattle.
