Predictive validity of a pressure injury risk assessment tool at different time-points in patients admitted to the intensive care unit.

Background: Multiple risk assessment scales are available for predicting the development of pressure injuries (PIs) in patients in the intensive care unit (ICU). Most PI risk assessment tools have been validated at the time of admission; however, another time point during treatment could better reflect clinical changes and therefore, the risk of PIs.

Aim: The study aimed to examine the predictive validity of PI risk assessment scale designed for ICU patients, the conscious level, mobility, haemodynamic, oxygenation and nutrition (COMHON) index, at several time points or intervals during ICU stay.

Study Design: This was an observational prospective study undertaken over a period of 1 year (July 2021-June 2022). Patients admitted to ICU for >3 days were included. The number, location and degree of the PIs were recorded. The level of risk for developing PIs during the stay was determined by calculating the COMHON scores at admission, and 72 h, as well as the highest and mean score. Predictive validity was studied using accuracy parameters and areas under the receiver operating characteristic curve (AUC). The best cutoff point was also determined and used to compare risk between categories.

Results: Of the 286 patients included in the study, 160 (59%) were male. The level of severity evaluated using the APACHE II scale was 18.4 ± 5.8 points. Invasive mechanical ventilation was used in 32.1% (n = 92) of the patients and 20.6% (n = 59) received high flow oxygen therapy. The incidence of PI was 15.4% (n = 44), with sacral location in 47.7% (n = 21) and grade II in 75% (n = 33) of the patients. The AUC was 0.907 (0.872-0.942); 0.881 (0.842-0.920); 0.877 (0.835-0.920) and 0.749 (0.667-0.831) at the mean, the highest, 72 h and ICU admission scores, respectively. The best cutoff point was 13 in all patients. The risk of developing a PI was 6.4 times higher in the high-risk group (>13 points).

Conclusions: The best predictive capacity for the COMHON index risk assessment was the mean and highest scores. The predictive accuracy was higher on the third day of the patient's stay than on admission, and this was attributed to the clinical changes observed in some patients over the course of their critical illness.

Relevance For Clinical Practice: Patients in ICU are at high risk of developing PIs, therefore, preventive measures should be maximized. Risk assessment should be carried out sequentially owing to the changes that patients present throughout their ICU stay and preventive measures should be used according to the risk level.