Explainable AI for Prospective Pressure Injury Staging

NCT07891689 — Not yet recruiting, 415 participants, hatice akaltun

ClinicalTrials Daily · Areas · Conditions · Drugs · Sponsors ·

Staging of Pressure Injuries Using Explainable Artificial Intelligence: A Prospective Clinical Decision Support Study With Expert Panel-Labeled Images

Registry	NCT07891689
Status	Not yet recruiting
Study type	Observational
Enrollment	415 (Estimated)
Sponsor	hatice akaltun (Other); collaborators: Yuzuncu Yil University
Conditions	Pressure Injury
Dates	start 2026-11-01 (estimated), primary completion 2027-06-01 (estimated), completion 2027-10-01 (estimated), first posted 2026-10-23, last updated 2026-10-23
Locations	1 site in Turkey (Türkiye)

-Summary

This prospective observational mixed-methods study aims to develop and evaluate an explainable artificial intelligence model for staging pressure injuries using standardized clinical photographs and expert panel-labeled images from adult patients. The study will also explore nurses' views on the clinical acceptability and usability of the decision-support system through semi-structured interviews.

Expand all sections · Collapse all

+Arms and interventions

Arm	Type	Interventions	Description
Adult Patients with Pressure Injuries			Adult patients with pressure injuries whose standardized clinical photographs and clinical data will be collected prospectively. Pressure injuries will be independently staged by an expert panel, and the images will be used to develop and evaluate an explainable artificial intelligence model. No intervention will be administered.
Nurses Participating in Interviews			Approximately 10-15 nurses working in intensive care and surgical units who provide pressure injury care will participate in individual semi-structured interviews. The interviews will explore their views on the clinical acceptability, usability, interpretability, and potential use of the explainable artificial intelligence decision-support system. No intervention will be administered.

+Outcome measures

Primary outcomes

Accuracy of the Explainable Artificial Intelligence Model in Pressure Injury Staging [From completion of image collection through completion of data analysis, approximately 8 months]: The proportion of pressure injury images correctly staged by the explainable artificial intelligence model compared with the reference staging determined by the expert panel. Accuracy will be calculated as the number of correctly classified images divided by the total number of evaluable images.

Secondary outcomes

Nurses' Views on the Clinical Acceptability and Usability of the Explainable Artificial Intelligence Decision-Support System [From completion of nurse interviews through completion of qualitative data analysis, approximately 4 months]: Nurses' views will be explored through individual semi-structured interviews. The interviews will be analyzed using thematic analysis to identify perceptions of clinical acceptability, usability, interpretability, potential benefits, concerns, and barriers to using the system.

+Eligibility

Sex: All, ages: 18 Years to any

Inclusion Criteria:

1. Adults aged 18 years or older.
2. Patients with at least one pressure injury.
3. Patients whose pressure injury can be photographed according to the study protocol.
4. Patients or legally authorized representatives who provide informed consent.

   For the qualitative component:
5. Nurses working in intensive care or surgical units.
6. Nurses involved in pressure injury care.
7. Nurses who provide informed consent and agree to participate in an individual semi-structured interview.

Exclusion Criteria:

1. Patients who do not provide informed consent.
2. Patients whose clinical condition does not allow the photography procedure.
3. Images that are technically inadequate or do not meet the study quality criteria.
4. Nurses who do not provide informed consent.
5. Nurses who are unable to complete the interview.

+Locations (1)

Facility	City	Country	Status
Sbu. Van Education and Research Hospital	Van	Turkey (Türkiye)

My Notes

Public

To Read Reading Read