A Behavior-Based Quality Improvement Initiative for Intensive Care Unit Calmness: A Single-Center Before-After...

Introduction Intensive care units (ICUs) are acoustically and cognitively demanding environments in which modifiable behavioral factors, unnecessary verbal communication, non-actionable alarms, disorganized workflow, and unstructured handovers contribute substantially to noise burden and cognitive overload for nursing staff. Published Indian ICU literature on this problem remains predominantly measurement-focused. We developed the CALMER bundle, a six-component, no-capital-expenditure behavioral intervention (communication discipline, alarm discipline, leadership modeling, mindful workflow and equipment handling, emotional reset, and reinforcement). We evaluated its feasibility and perceived impact in an Indian tertiary ICU. Materials and methods This was a single-center, uncontrolled, before-after quality improvement study conducted in a 20-bed mixed medical-surgical ICU at GEM Hospital, Coimbatore, India, reported per the Standards for Quality Improvement Reporting Excellence (SQUIRE) 2.0 guidelines. The primary outcome, composite perceived ICU calmness, was assessed using a retrospective paired then-now 11-item questionnaire (five-point Likert scale; Cronbach's alpha 0.831), administered to all 40 nursing staff with complete-case analysis in 28 nurses (28/40, 70%). Secondary outcomes included a six-item work environment score (same 28 nurses) and a post-intervention adapted NASA Task Load Index (NASA-TLX) workload assessment (n = 40). We assessed patient/surrogate perception cross-sectionally post-intervention (n = 100). We analyzed paired ordinal data using the Wilcoxon signed-rank test, with effect size r and 95% confidence intervals (Fisher's z-transformation). Results Composite ICU calmness improved from a pre-intervention median of 2.0 (IQR: 1.75-2.0) to 4.0 (IQR: 4.0-4.0) (p < 0.001; r = 0.910; 95% CI: 0.813-0.958). All four calmness domains improved significantly (r = 0.842-0.874). Twenty-six of 28 nurses (92.9%) met a prespecified ≥2-point responder threshold; none declined. The work environment score improved from 2.25 to 4.0 (p < 0.001; r = 0.874), with all six secondary items improving significantly (r = 0.751-0.872). Post-intervention workload was at the scale midpoint (NASA-TLX median: 3.0/5). Patient/surrogate perceptions were favorable across all five domains (median: 5.0/5); as a post-intervention-only assessment, these are reported descriptively without a baseline comparator. A later-phase implementation-fidelity audit (141 shift-level audits, May 7 to June 30, 2026) showed 82.9% overall compliance (highest: leadership modeling 91.8%; lowest: emotional reset 76.2%). No alarm-related sentinel events occurred. Conclusions A zero-capital-expenditure, behavior-based bundle was associated with large, statistically significant improvements in nurse-perceived ICU calmness and work environment in an Indian tertiary ICU, alongside a post-intervention workload rating at the scale midpoint. Unblinded, investigator-administered assessment and the absence of objective acoustic data are important limitations. Multicenter validation with objective noise measurement and independent data collection is needed before generalizing these findings.