Background

Quality improvement is frequently assessed by benchmarking, often one-dimensionally using Standardised Mortality Ratios (SMR).

Objectives

Development of a multidimensional score to compare PICUs.

Methods

Virtual Pediatric Systems, LLC (VPS) data from 147 PICUs (82,000 admissions) was analysed. Performance metrics included (weights in parenthesis): SMR (2), standardized length of stay ratio (1), discharge delay (1) while other metrics (1) assessed deviation from cohort average (standard) performance: 24-h readmissions, extubation failure, use of mechanical ventilation, and discharge directly home (0.5). For metrics, differences between individual ICU metric and standard care were calculated. Percentile ranges 0-10 (best), 10-20, 20-50, 50-80, 89-90, and 90-100 (worst) were converted to metric-scores, 6, 5, 4, 3, 2, and 1. The unit global score was the weighted sum of metric-scores scaled to 100. The heatmap (figure) shows metrics (rows) and 16 ICUs (columns). Performance above or below the standard is indicated (+/-).

Results

ICU performance is ranked by the multidimensional global score. The figure shows a highlighted hospital (score 73) compared to 15 reference ICUs. The figure demonstrates individual relative metric performance, highlighting ICU strengths and weaknesses.

Conclusion

This multidimensional score permits ICU benchmarking and also allows numerical and visual comparisons.

Global score rankings could be used to rate ICUs, for example, 3 stars for 90-100%, 2 for 50-89%, one for 10-49% and <10% not rated.

Background

Background: Improving quality often relies on benchmarking and research. Virtual Pediatric Systems, LLC (“VPS”) has supported the North American PICU community since 2002 with high quality data aggregation from over 160 hospitals collecting severity adjusted data with over 1.4 million admissions (140,000 cardiac). VPS is the largest pediatric collaborative quality improvement and research registry and provides comprehensive benchmarking.

Objectives

Objective: To report the VPS experience providing analytics, reporting, research and quality improvement for PICUs.

Methods

Methods and Results: VPS records were reviewed to provide information. All VPS comparative reports are adjusted for severity of illness. VPS provides data validation, reporting and statistical support. All programing is done in house. VPS data security is certified by HiTrust. In the last ten years over 5,000 annual and quarterly reports have been provided to participating ICUs. Data has been provided for 237 research papers (see figure). Hundreds of quality improvement projects have been undertaken. VPS has a members’ medical advisory committee, supports quality improvement collaboratives, provides near real time dashboards and ad hoc reporting for its members. There are periodic user group meetings. Several states either require or are considering mandatory VPS participation.

Results

Results Data:

Conclusion

Conclusions: VPS provides innovative, comprehensive, detailed and actionable quality reports for practitioners and administrators and supports improved quality of critical care and has become the largest PICU research collaborative.

Background

Quality improvement is frequently assessed by benchmarking, often one-dimensionally using Standardised Mortality Ratios (SMR).

Objectives

Development of a multidimensional score to compare PICUs.

Methods

Virtual Pediatric Systems, LLC (VPS) data from 147 PICUs (82,000 admissions) was analysed. Performance metrics included (weights in parenthesis): SMR (2), standardized length of stay ratio (1), discharge delay (1) while other metrics (1) assessed deviation from cohort average (standard) performance: 24-h readmissions, extubation failure, use of mechanical ventilation, and discharge directly home (0.5). For metrics, differences between individual ICU metric and standard care were calculated. Percentile ranges 0-10 (best), 10-20, 20-50, 50-80, 89-90, and 90-100 (worst) were converted to metric-scores, 6, 5, 4, 3, 2, and 1. The unit global score was the weighted sum of metric-scores scaled to 100. The heatmap (figure) shows metrics (rows) and 16 ICUs (columns). Performance above or below the standard is indicated (+/-).

Results

ICU performance is ranked by the multidimensional global score. The figure shows a highlighted hospital (score 73) compared to 15 reference ICUs. The figure demonstrates individual relative metric performance, highlighting ICU strengths and weaknesses.

Conclusion

This multidimensional score permits ICU benchmarking and also allows numerical and visual comparisons.

Global score rankings could be used to rate ICUs, for example, 3 stars for 90-100%, 2 for 50-89%, one for 10-49% and <10% not rated.

Background

Background: Improving quality often relies on benchmarking and research. Virtual Pediatric Systems, LLC (“VPS”) has supported the North American PICU community since 2002 with high quality data aggregation from over 160 hospitals collecting severity adjusted data with over 1.4 million admissions (140,000 cardiac). VPS is the largest pediatric collaborative quality improvement and research registry and provides comprehensive benchmarking.

Objectives

Objective: To report the VPS experience providing analytics, reporting, research and quality improvement for PICUs.

Methods

Methods and Results: VPS records were reviewed to provide information. All VPS comparative reports are adjusted for severity of illness. VPS provides data validation, reporting and statistical support. All programing is done in house. VPS data security is certified by HiTrust. In the last ten years over 5,000 annual and quarterly reports have been provided to participating ICUs. Data has been provided for 237 research papers (see figure). Hundreds of quality improvement projects have been undertaken. VPS has a members’ medical advisory committee, supports quality improvement collaboratives, provides near real time dashboards and ad hoc reporting for its members. There are periodic user group meetings. Several states either require or are considering mandatory VPS participation.

Results

Results Data:

Conclusion

Conclusions: VPS provides innovative, comprehensive, detailed and actionable quality reports for practitioners and administrators and supports improved quality of critical care and has become the largest PICU research collaborative.