Author(s): Jodi Mullen, MS, RN-BC, CCRN, CCNS, ACCNS-P, FCCM, CPHQ
Senior Quality Improvement Specialist
UF Health

Competency Domain: Population Health & Care Transitions

Description: The Agency for Healthcare Research and Quality identifies Patient Safety Indicators (PSIs) that track preventable in-hospital complications following elective surgeries and procedures. Since PSI data are nationally reported for hospital comparisons, ensuring accuracy is critical. This session describes how a multidisciplinary team identified inconsistencies in surgical admission classification, leading to inaccurate reporting of PSI events. By revising outdated criteria and aligning classifications with clinical need, PSI-10, PSI-11, and PSI-13 rates significantly decreased. Quality tools, including control charts, process mapping, and patient-level case reviews, drove data-informed quality improvements on key safety metrics. These results have been sustained over time. Problem:Patient safety indicators (PSIs) are preventable complications that occur following elective surgeries and procedures. We noted increasing rates of PSIs related to postoperative acute kidney injury requiring dialysis (PSI-10), respiratory failure (PSI-11), and sepsis (PSI-13). This increase contributed to poor performance on publicly-reported safety measures. Measurement:Control chart graphs were used to visualize trends over time for PSI 10, 11, and 13 rates. A process flow map was created to better visualize the workflow of the coding quality, clinical documentation integrity, and revenue cycle departments as it related to coding and billing for inpatient surgical admissions. Analysis:A multidisciplinary team performed patient-level case review for PSI 10, 11, and 13 events that occurred in the previous two years, looking for high-level opportunities for improvement. Baseline mean rate per 1000 elective surgical discharges was measured for each of the three PSIs during the same time period.Implementation:Inconsistences in classifying a surgical admission as elective, urgent, or emergent were identified. Criteria varied by department and were outdated. Admission criteria were revised based on clinical need rather than time. Some surgeries, such as cardiothoracic, defaulted to urgent. Cases with potential misclassifications underwent further review and admission classification revision.Results/DiscussionPSI rates 9-months post-implementation were compared to baseline utilizing a two-tailed t-test. PSI-10 decreased from 0.481 to 0.166 (P=0.015), PSI-11 decreased from 1.743 to 0.180 (p=0.000), and PSI-13 decreased from 1.427 to 0.234 (p=0.003), and variation was reduced by roughly 50%. Results were sustained during the subsequent control time period.