Serum uric acid-to-creatinine ratio and preeclampsia: a retrospective cross-sectional analysis with machine learning discrimination using the MIMIC-IV database
This retrospective cross-sectional study using the MIMIC-IV database found that a higher serum uric acid-to-creatinine ratio at admission is significantly associated with preeclampsia and serves as a key feature in machine learning models with moderate discriminative ability, though prospective validation is required before clinical application.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Preeclampsia is a serious condition that can strike during pregnancy, turning a routine pregnancy into a medical emergency. It involves high blood pressure and signs of damage to other organ systems, most often the kidneys, and it remains a leading cause of illness and death for both mothers and their babies worldwide. For decades, doctors have known that when this condition develops, the levels of uric acid in a woman's blood tend to rise. Uric acid is a waste product created when the body breaks down certain chemicals found in food and cells. However, looking at uric acid alone can be misleading because the kidneys play a major role in filtering it out of the blood. If the kidneys are struggling, uric acid builds up, but if the kidneys are working fine, the level might stay normal even if the body is producing too much. To get a clearer picture, researchers often look at the ratio of uric acid to creatinine, another waste product that serves as a reliable marker for how well the kidneys are filtering. By comparing these two numbers, doctors can better understand the true load of uric acid in the body, independent of how fast the kidneys are working.
A team of researchers set out to see if this specific ratio, measured when a pregnant woman first arrives at the hospital, could help identify those at risk for preeclampsia. They also wanted to test whether modern computer programs, known as machine learning, could use this ratio along with other routine blood test results to spot the condition more accurately than a human doctor might on their own. To do this, they turned to a massive, anonymous collection of medical records from a major hospital in the United States. They focused on nearly five thousand pregnant women who were at least twenty weeks along in their pregnancy. After filtering out records that were incomplete, they were left with a group of just over five hundred women to study. Within this group, about seventy-eight women had been diagnosed with preeclampsia, while the rest had not.
The researchers began by comparing the blood work of the women with preeclampsia against those without the condition. They found a clear pattern: the women who developed preeclampsia had higher ratios of uric acid to creatinine upon admission than the healthy women. The relationship appeared to be straightforward and linear, meaning that as the ratio went up, the risk of the condition increased steadily. When they looked at the women with the highest ratios, they were nearly three times more likely to have preeclampsia compared to those with the lowest ratios. This connection held true even after the researchers accounted for factors like the woman's age and language spoken.
To see if this finding could be turned into a practical tool, the team trained six different computer algorithms to learn from the data. These programs were fed the admission blood test results, including the uric acid ratio, along with other standard measurements like liver enzyme levels and blood cell counts. The goal was to see if the computers could distinguish between the women who had preeclampsia and those who did not. The results showed that the computers could indeed make this distinction with moderate accuracy. One of the algorithms performed the best, correctly identifying the condition in a significant portion of cases. When the researchers asked the computer to explain which factors mattered most, the uric acid ratio consistently ranked as one of the top clues, second only to a measure of red blood cell concentration.
Despite these promising results, the researchers were careful not to overstate what their study could do. Because they looked at a snapshot of data taken all at once, rather than following women over time, they could not prove that the high ratio caused the disease or that it could predict the disease before it started. The study was also limited by the fact that many records were missing, which meant the final group of women was smaller than the original pool, and the number of actual cases of preeclampsia was relatively low. The authors concluded that while the uric acid ratio is a useful piece of information and can help computer models make better guesses, it is not yet a standalone crystal ball. Before this method can be used to guide treatment decisions in real hospitals, it needs to be tested in larger, forward-looking studies to confirm that it works reliably for all pregnant women. For now, it stands as a valuable, low-cost indicator that, when combined with other data, helps paint a clearer picture of risk for those already in the hospital.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.