Nurses, especially highly experienced nurses, underrate moderate and severe patient pain associated with pediatric burn dressing changes, according to new research published in the journal Burns.

In that age group and injury type, procedural pain is measured through the FLACC (Face, Legs, Activity, Cry, Consolability) system, which relies more on the observed presentation of pain and less on self-reporting and physiological signs of pain. And according to the research, the pain associated with burns and burn dressings is high among the more than 70,000 children under the age of 6 who need burn care in a given year. But inaccurately assessing patient pain could be a barrier to properly treating it.

The study examined the FLACC rating practices of 24 nurses within one U.S. pediatric burn center who performed a little more than one and a half burn dressing changes per week. They watched videos of four different dressing changes rated at different points along the 10-point FLACC scale.

Among three different watching sessions of the same four videos, the nurses consistently under-rated the pain of dressing changes previously determined to be a 1 and those determined to be a 10. Among those burn dressing changes determined to be rate at 4 and 6, the nurses slightly over-rated the pain on the FLACC scale.

For example, upon the first watching of the 1-point pain video, the nurses gave it a mean score of 0.42 on the FLACC scale. On the first watching of the 10-point pain video, the nurses gave it a mean score of 8.75.

The researchers also found that as the pain intensity increased, the more the nurses seemed to agree with the previous designation (and each other) when assigning a FLACC score.

The study authors also noted that the more experience a nurse had, the less accurate he or she became when determining FLACC scores—maybe because more experienced nurses had necessarily become desensitized to patient pain the longer they worked in a pediatric burn unit.

It’s possible the nurses’ inaccuracy stems from the 10-point limit of the system, according to the study. Nurses might be biased in distributing pain scores, knowing that it can never reach a point higher than 10. It’s also possible that moderate pain is more difficult to recognize than extremely high or extremely low pain, which would be rated at either end of the FLACC score spectrum. It’s also possible the nurses were relying on their own experience to rate the patients’ pain instead of systemically calculating the FLACC score based on observable factors, therefore underestimating pain.

The concern is that if caregivers are unable to accurately measure those pain levels, they are unable to accurately treat them, the researchers claimed.