Hidden knowledge gaps in mechanical ventilation care: a multicenter study of ICU nurses’ competence and patient safety risks in Iraq

Abstract Background and aim Mechanical ventilation is a life-saving but high-risk intervention whose safety depends largely on the bedside knowledge and competence of intensive-care nurses. Knowledge deficits in initiating ventilation, managing alarms, preventing ventilator-associated complications, and conducting weaning have been linked to increased patient harm in low- and middle-income settings. This study aimed to assess nurses’ knowledge regarding initiating mechanical ventilation and nursing procedures in the intensive care units of four major hospitals in Amara City, Iraq, and to identify the demographic and training-related determinants of competence. Method A descriptive cross-sectional study was conducted from November 20, 2024, to July 15, 2025, across four governmental hospitals in Amara City: Al-Shaheed Al-Sadder Teaching Hospital, Al-Hakeem Teaching Hospital, Al-Zahrawi Surgical Hospital, and Child & Maternal Hospital. A non-probability purposive sample of nurses with at least one year of ICU experience and direct involvement in the care of mechanically ventilated patients was recruited. Data were collected using a structured, self-administered questionnaire covering demographic characteristics and a 20-item knowledge assessment, with each correct answer scored as 1 and each incorrect answer as 0. Statistical analyses included descriptive statistics, Chi-square tests, Pearson/Spearman correlations, hierarchical multiple linear regression as the primary inferential model, and an exploratory binary logistic regression. Exact p-values were reported to three decimal places, and effect sizes were reported alongside every significance test (Cramer’s V for Chi-square associations, Fisher z-transformed 95% confidence intervals for correlations, and R² change with Cohen’s f² for the regression blocks). Results A total of 152 ICU nurses participated in the study. Findings revealed a generally moderate level of knowledge with a mean score of 0.40 ± 0.117, with 65.8% classified as moderate, 26.3% as poor, and only 7.9% as good. Critical knowledge deficits were identified in items related to forced artificial respiration (90.8% incorrect), high-pressure-alarm interpretation (89.5% incorrect), and ventilator-associated pneumonia prevention (81.6% incorrect). The primary inferential model, a hierarchical multiple linear regression using the continuous knowledge score and all 152 participants, demonstrated that personal, experience-related and training-related variables explained 48.2% of the variance in knowledge (Adjusted R² = 0.456; ΔR² for the training block = 0.183; p < 0.001). An exploratory binary logistic regression, in which the adequate-knowledge threshold was reached by only 12 nurses (7.9%), giving approximately three events per predictor variable, associated ICU experience (OR = 1.21 per year) and ≥ 5 training sessions (OR = 4.40) with adequate knowledge. Because that model was neither adequately powered nor internally validated, its correct-classification rate (82.9%) and discrimination (AUC = 0.871) are likely to be optimistic and are reported as exploratory only. Conclusions ICU nurses in Amara City demonstrated hidden knowledge gaps in domains carrying a high potential for patient harm. The FMEA-derived Patient-Safety Risk Index is a severity- and detectability-weighted transformation of the same item-level error rates from which the knowledge score is calculated; its inverse association with knowledge is therefore an internal re-expression of those data and is not independent evidence of patient harm. Structured, simulation-based mechanical ventilation training, alarm-management and ventilator-associated pneumonia bundle education, and a tier-based clinical decision-support framework merit prospective evaluation as strategies for reducing preventable harm in mechanically ventilated patients. Clinical trial registration This study did not constitute a clinical trial and therefore did not require registration.