Subscribe to The Podcast by KevinMD. Watch on YouTube. Catch up on old episodes!

Internist and psychiatrist Muhamad Aly Rifai discusses his article, “How data accuracy failures are costing lives and money in health care.” He explores the consequences of inaccurate data in health care, from billing errors and wrongful prosecutions to the broader impact on medical decision-making and patient care. Muhamad highlights real-world cases, including his own legal battle over Medicare claims and the flawed prosecution of an interventional cardiologist, to illustrate how poor data fidelity can lead to catastrophic errors. He also explains how artificial intelligence may amplify these risks if data integrity issues remain unresolved. Tune in for insights into the urgent need for accurate, reliable health care data and solutions to improve data quality in medicine.

Our presenting sponsor is Microsoft Dragon Copilot.

Microsoft Dragon Copilot, your AI assistant for clinical workflow, is transforming how clinicians work. Now you can streamline and customize documentation, surface information right at the point of care, and automate tasks with just a click.

Part of Microsoft Cloud for Healthcare, Dragon Copilot offers an extensible AI workspace and a single, integrated platform to help unlock new levels of efficiency. Plus, it’s backed by a proven track record and decades of clinical expertise—and it’s built on a foundation of trust.

It’s time to ease your administrative burdens and stay focused on what matters most with Dragon Copilot, your AI assistant for clinical workflow.

VISIT SPONSOR → https://aka.ms/kevinmd

SUBSCRIBE TO THE PODCAST → https://www.kevinmd.com/podcast

RECOMMENDED BY KEVINMD → https://www.kevinmd.com/recommended

Transcript

Kevin Pho: Hi, and welcome to the show. Subscribe at KevinMD.com/podcast. Today, we welcome back Muhamad Aly Rifai. He is an internal medicine physician and psychiatrist. Today’s KevinMD article is “How data accuracy failures are costing lives and money in health care.” Muhammad, welcome back to the show.

Muhamad Aly Rifai: Thank you very much for having me to talk about this important and timely topic, especially in medicine.

Kevin Pho: All right. For those of you who get a chance to read it, what is this article about?

Muhamad Aly Rifai: Sure. So in, in the article, I talk a little bit about the importance of data in health care. The field of medicine now has become data driven. We have a large amount of data that is available, and that produces results for us in terms of our decisions with payments, our decisions with outcomes, and our decisions with treatment modalities. We have seen how a significant amount of data can affect different decisions, as well as the fact that if we do not have data accuracy, that actually impacts our decisions. We have seen recent events in government of how data accuracy has caused a significant amount of, um, ridicule in the media about some data accuracy issues with the Social Security Administration and Medicare.

Kevin Pho: All right, so you are in the world of psychiatry; I am in the world of internal medicine and primary care, and absolutely I am seeing it on my end about how everything is data driven, how everything is metrics based—from how hospitals and physicians are reimbursed to patient care in terms of the metrics that a lot of doctors have to meet to, you know, ensure that patients are receiving the best care. Everything is becoming more data-based. So how about in your world of psychiatry? How does that intersect with what you do in terms of data-driven medicine?

Muhamad Aly Rifai: It is very important, in terms of data-driven practice of psychiatry, for us to be able to identify truly what is going on and whether the data we have truly represents what is going on. For example, treatment modalities sometimes produce results, and some of those results may not be very meaningful. Over the last few years, we have come to dispute some of the notions that we had in the field of psychiatry, such as the efficacy of antidepressant medications. We have seen that placebo effects in some of these studies have risen significantly, and we have seen that the efficacy of antidepressant medications may not be as we would like it to be. Now there is a whole movement distrusting the efficacy of antidepressant medications based on re-analysis of some of this data, so having very accurate data would be very important.

Also, we have seen that not having the correct data in terms of payment or in terms of assessment can be problematic. I can tell you my personal experience with not having very accurate data in terms of analyzing things—it has gotten me into some, in, in, in trouble. We have heard recently in the news that Elon Musk and the president talked about their analysis of the master file of the Social Security Administration, where they discovered some individuals above the age of one hundred years and above the age of one hundred fifty years, and even an individual who is three hundred sixty years old on the rolls receiving Social Security benefits. Maybe those were anomalies, maybe those were errors in data entry. But having a just, reliable data set is very important because it gives legitimacy to the field of medicine, as well as giving people confidence that the data they are receiving is reliable and meaningful.

Kevin Pho: So you are talking about what sounds like two different sources of potential confounders in data. The first is regarding reassessment of previous clinical trials that could change the data we are using. The second is simply inaccurate data that we may be entering about patients in our electronic medical records—data that could lead to mistakes and potential fraudulent billing going forward. How common is that in medicine, from what you see, in terms of the second one—just simple mistakes in the patient record that can lead to problems with accuracy in patient records and in billing?

Muhamad Aly Rifai: Sure, sure. The error rate in our data in health care records actually stems from the Institute of Medicine report in the early 2000s. They titled it “To err is human.” We are human, and so data entry errors are pretty rampant in health care. Some of the estimates are around 40 to 60 percent; there is some data entry error somewhere in the data set in the health care system, and sometimes that can have a significant impact in health care, medication errors, billing, and so forth. It is significant. In my case, I had staff who made data entry errors, resulting in submitting bills on patients who were deceased. We corrected that, but that got me into significant trouble with the government—with a prosecution. But when it was revealed that these were data entry errors of entering data on one patient versus another with the same first and last name but different date of birth (one was deceased and one was alive), that was corrected. However, sometimes that can lead to significant bad outcomes. For example, entering orders on patients with a different date of birth and giving medications that are not needed to another patient could be significantly detrimental. Errors exist, but being able to identify that they happen and being able to fix them is very significant.

Kevin Pho: And I think you have told your story on prior episodes, but just to reiterate, you had staff members enter the wrong data, and that led you into an experience with federal prosecution. How long did that experience last? You do not have to rehash your whole story, but how long was this ordeal for you?

Muhamad Aly Rifai: It lasted for a few years, and one of the catalysts for that was billing on deceased patients. When the data was unhashed and we reviewed it with the government and the jury, we found out that they claimed I billed on deceased patients who had died before I was born. Even the government, when they presented these allegations, had inaccurate data, because the artificial intelligence collected made-up data on deceased patients who died before I was born. Errors go both ways, but acknowledging that this was an error and was being corrected was crucial in convincing the jury that this was not fraudulent intent but just a simple data entry error. However, sometimes this is not just a fraud prosecution issue. Errors in data entry on different patients can lead to giving the wrong medication to a patient, which could be fatal. We could have bad outcomes with that, too.

Kevin Pho: In your article, you also talk about the case of Dr. Richard Paulus and how flawed data contributed to his prosecution. For those who are not familiar with that case, please tell us about that connection between his prosecution and flawed data.

Muhamad Aly Rifai: Sure, sure. Dr. Paulus was an interventional cardiologist in Kentucky who came under the attention of the federal government because he had a large interventional cardiology practice, which led the government to focus on the rates at which he inserted stents into occluded or narrowed coronary arteries. They found what they thought was a significant anomaly in his billing. He went to trial twice and was incarcerated for almost one year, enduring a prosecution of about twelve years, while ultimately the federal government dropped all charges against him.

However, his prosecution stemmed from the government identifying a large universe of cases in which they believed he did something wrong but only presenting a small subset of those cases to the jury. The government looked at around one thousand or eleven hundred cases but showed only about seventy to the jury, stating that in all those seventy, he inserted stents that other cardiologists felt should not have been inserted. That made it seem like one hundred percent of his stent insertions were inappropriate. But when you look at the full set of around one thousand cases, only seventy were disputed, which is closer to five or six percent—a much more acceptable margin of difference in interpreting whether a narrowing is fifty percent or seventy percent. Once those facts were shown to the jury, it significantly affected his case. But that was not until after twelve years of prosecution, two trials, and one year in jail, after which the appeals court reversed his conviction and set him free. Finally, the government decided to drop all charges against him.

Kevin Pho: So tell us about the path forward. What are some ways that physicians, hospitals, and medical institutions can prevent these data errors and reduce fraudulent billing and data inaccuracies going forward? What are some potential solutions?

Muhamad Aly Rifai: I think it starts with human training and adding additional layers of review to try to catch errors and correct them before they have a meaningful impact. We need to identify that errors happen, and most of the time these are not fraudulent. For example, in the case of the person who is three hundred sixty years old receiving Social Security benefits—if that is a data entry error, it could be easily investigated and corrected. But because such errors get publicized, it makes people lose confidence in these data sets. A lot of work needs to be done. For instance, the Social Security Administration decided that investing time and money in a team of staff who comb through the master file to clean up inaccuracies is worthwhile. Hospitals should similarly have teams that review data and correct errors, and the government should admit that errors can happen and are not always fraudulent or criminal.

Kevin Pho: We are talking to Muhamad Aly Rifai. His KevinMD article today is “How data accuracy failures are costing lives and money in health care.” Muhamad, let us end with some take-home messages that you want to leave with the KevinMD audience.

Muhamad Aly Rifai: To err is human. We have to admit our limitations, but we also have to redouble our efforts to provide our patients with accurate, appropriate, and safe care.

Kevin Pho: Thank you again for sharing your perspective and insight. Thanks again for coming back on the show.

Muhamad Aly Rifai: Thank you.