P2A086: Implementation of a Short Stay Protocol Reduces Post-operative Length of Stay After Pediatric Heart Surgery

Background: Prioritization of value-based care has led to greater emphasis on process and quality improvement (QI). Clinical practice pathways and protocols are common strategies to achieve this, though evidence-based practices are often lacking. Additionally, there is variability in medical management between providers and for patients with differing anatomy and physiology. We hypothesized that implementation of a protocol with clearly defined clinical targets would significantly decrease post-operative length of stay (LOS) for common and low-risk STAT1 category surgeries over approximately 1 year.

Methods: This was a prospective QI study at a single institution. Baseline pre-implementation data was collected retrospectively over the prior 18 months, and post-implementation data was collected prospectively for consecutive patients that had surgery for an ASD, VSD >6 months of age, vascular ring, coarctation off-bypass >6 months, pulmonary valve replacement, anomalous coronary artery, pacemaker implantation, sub-aortic stenosis, or partial/transitional AV septal defect. Patients with a significant comorbidity (coagulopathy or bleeding disorder, immunodeficiency, pulmonary hypertension, failure to thrive, obesity) or ≥2 prior sternotomies were excluded.

The short stay protocol (SSP) clinical targets were limited to extubation on post-operative day (POD) 0, chest tube (CT) and central venous line (CVL) removal on POD1, ICU transfer on POD1, and a primary goal of hospital discharge on POD3. Plan-Do-Study-Act (PDSA) cycle #1 was implementation of SSP. PDSA #2 corresponds to an effort for more efficient communication, including weekly messaging for upcoming patients and monthly feedback on progress, to all participating team members. Variables are reported as mean ± standard deviation. Statistical analysis includes t-test and Chi-square where appropriate. A run chart was utilized to display mean percentage of patients discharged by POD3.

Results: There were 96 patients in the baseline, pre-implementation cohort (Table). Mean LOS was 3.9 ± 1.5 days, 52% patients achieved discharge by POD3, and 30-day readmission rate was 5.0%. The SSP cohort (n = 55) generally achieved significant improvement in all clinical targets compared to baseline. LOS improved significantly to 3.2 ± 1.0 days (p = 0.003) and 80% achieved discharge by POD3 (p = 0.001) (Figure). This translated to 41.6 potential hospital days saved for the SSP cohort when compared to baseline, without an increase in 30-day readmission rate (4.0%, p = 0.4).

Conclusion: Implementation of SSP led to significant improvement in average LOS without corresponding increase in re-admission rate. Protocols utilizing simplified clinical targets have potential to improve efficiency of care delivery after pediatric cardiothoracic surgery.