Towards Electronic Health Records-Based Paediatric Growth References: Results from the SwissPedGrowth Project

Every child grows at their own pace, but doctors need a common map to know if that pace is healthy. For decades, pediatricians have relied on growth charts—curves that show the average height, weight, and head size of children at every age—to spot potential health issues early. These charts act as a standard ruler, helping clinicians decide if a child is thriving or if something is wrong. However, creating these maps is difficult. Traditionally, researchers have had to recruit thousands of families, measure their children by hand, and track them over many years. This process is slow, expensive, and often misses the full diversity of the population. As a result, many countries have struggled to keep their growth charts up to date with the changing reality of their children.

In Switzerland, a team of researchers decided to try a different approach. Instead of starting from scratch, they asked a simple question: could they use the millions of measurements already sitting in hospital computer systems to build a better picture of how Swiss children grow? They turned to electronic health records, the digital files that hospitals keep for every patient visit. These records contain routine measurements taken by doctors and nurses, such as height and weight, collected over many years. The challenge was to sift through this massive amount of data, which includes children with serious illnesses, to find the patterns of healthy growth hidden within.

The researchers, working through a project called SwissPedGrowth, gathered data from seven major children's hospitals across the country. They looked at records from 2017 to 2023, covering hundreds of thousands of children. In total, they analyzed over 448,000 weight records, more than 213,000 height records, and nearly 210,000 body mass index calculations. To make sense of this, they first cleaned the data, removing impossible numbers and fixing errors, much like a librarian organizing a chaotic stack of books. Then, they separated the children into two groups. The first group included every child with a measurement, regardless of their health. The second group was a carefully selected subset of children who did not have diseases known to affect growth, such as severe hormonal disorders or chronic digestive issues.

To ensure this second group truly represented the children of Switzerland, the researchers adjusted the data to match the country's demographics. They balanced the sample so that the mix of ages, sexes, nationalities, and neighborhood wealth levels in their study matched the actual population of the country. This step was crucial because hospital patients often differ from the general public; for instance, a hospital might see more children from certain backgrounds or with specific health needs than a random street corner would. By weighting the data, they created a virtual snapshot of a healthy Swiss child.

When they compared these hospital-based measurements to the official growth charts currently used in Switzerland, the results were striking. The average height, weight, and body mass index of the healthy children in the hospital records matched the new national standards almost perfectly. The curves that traced the growth of these children from birth to eighteen years old followed the same path as the official charts. This alignment held true even though the hospital data came from sick children's hospitals, which might seem like an unlikely source for healthy growth data. The researchers found that the middle of the distribution, representing the typical child, was nearly identical to the national average.

However, the study also revealed subtle differences at the edges of the charts. While the average child fit the standard, the extremes—the very shortest and tallest, or the lightest and heaviest children—showed slightly more variation in the hospital data than in the official charts. This suggests that while hospital records are excellent for tracking the average child, they might capture a wider range of extremes, perhaps because hospitals see more children with complex conditions even after filtering for health. Despite this, the overall picture was clear: the data from routine hospital visits can successfully reflect the growth patterns of the general population.

This finding suggests a new way forward for medical research. If hospitals can provide accurate growth data, the need for expensive, time-consuming studies that measure children manually might decrease. The researchers noted that using existing electronic records could allow for growth charts to be updated more frequently, perhaps every few years instead of every decade. This would help doctors stay ahead of changes in how children grow, whether due to shifts in nutrition, environment, or other factors. The study confirmed that the new Swiss growth references, introduced in 2026, are a good fit for the country's children, and it demonstrated that the digital footprints left in hospital systems are a powerful, underused resource for understanding child health.

The work does not mean that hospital data is perfect. The researchers acknowledged that some measurements in hospitals might be less precise than those taken in a dedicated research setting, and that the data is currently limited to the French and German-speaking parts of Switzerland. Yet, the core message is one of possibility. By turning to the data that is already being collected every day, scientists can build a clearer, more current map of childhood growth. This approach offers a practical path to keeping medical standards aligned with the real lives of children, ensuring that the tools doctors use to monitor health are as accurate and up-to-date as possible.