Pattern analysis of hospital nurses' knowledge, attitudes, and practices regarding incontinence-associated dermatitis

To analyze hospital nurses' knowledge, attitudes, and practices regarding incontinence-associated dermatitis (KAP-IAD).

This study utilized responses from hospital nurses to the Knowledge, Attitudes, and Practices of Incontinence-Associated Dermatitis Questionnaire (KAP-IAD-Q). Three clustering methods, Hierarchical Clustering on Principal Components (HCPC), K-means, and Partitioning Around Medoids (PAM), were applied to analyze the correlations of KAP-IAD. A classification method was used to explain the underlying behavioral patterns behind these correlations.

Two clusters were found to be most appropriate. Decision attributes (D) were generated for the KAP-IAD data using the three clustering methods: HCPC, K-means, and PAM. Three datasets with categorical labels were generated, and predictive models and decision rules were established for each dataset using the Rough Set (RS) method. The PAM method demonstrated the highest accuracy among the three datasets. After five rounds of stochastic modeling, 57 decision rules were generated. Additionally, patterns or rules with a support threshold of 50 or more, as discussed by domain experts, were considered the primary behaviors or rules.

Our study suggests clear decision rules for KAP-IAD nursing practice, which have been absent in previous research. The key variables and rules identified can serve as a guide for KAP-IAD nursing practice, as well as for recognizing the etiology, risk factors, and key influences of dermatitis associated with KAP-IAD in nursing practice. This study provides an important management approach for the prevention and treatment of incontinence-associated dermatitis.