National registries: Lessons learnt from quality improvement initiatives in intensive care

National clinical quality registries (CQRs) are effective tools for improving the outcomes of patients admitted to the intensive care unit (ICU), and are increasingly important as healthcare needs evolve. A high-quality ICU CQR is built from a foundation of common requirements and challenges. First, performance indicators of the structure, process, or outcomes of patient care should measure what is important. Second, high data quality is essential and can be collected and curated through standardized processes. Third, standardized mortality ratio (SMR) is a cornerstone for benchmarking ICU performance, but application requires a comprehensive understanding of its context and potential pitfalls. Fourth, data collection alone is insufficient. Quality improvement comes from closing the feedback loop by identifying and managing unwarranted practice variation. Fifth, the process of improving healthcare is fundamentally a human enterprise, subject to behavioural change, including those that modify performance. Sixth, ICU CQRs must be dynamic to meet the needs of an evolving healthcare system and stakeholders. Finally, these lessons are far from comprehensive. Sharing perspectives on the development of ICU CQRs can help maximise their value as a powerful platform for informing policy development and improving the outcomes of patients admitted to the ICU.

Predicting Acute Kidney Injury After Cardiac Surgery Using a Simpler Model

OBJECTIVE

To develop a simple model for the prediction of acute kidney injury (AKI) and renal replacement therapy (RRT) that could be used in clinical or research risk stratification.

DESIGN

Retrospective analysis.

SETTING

Multi-institutional.

PARTICIPANTS

All cardiac surgery patients from September 2016 to December 2018.

INTERVENTIONS

Observational.

MEASUREMENTS AND MAIN RESULTS

The study cohort was divided into a development set (75%) and validation set (25%). The following 2 data epochs were used: preoperative data and immediate postoperative data (within 4 h of intensive care unit admission). Univariate statistics were used to identify variables associated with AKI or RRT. Stepwise logistic regression was used to develop a parsimonious model. Model discrimination and calibration were evaluated in the test set. Models were compared with previously published models and with a more comprehensive model developed using the least absolute shrinkage and selection operator. The study included 22,731 patients at 33 hospitals. The incidences of AKI (any stage) and RRT for the present analysis were 5,829 patients (25.6%) and 488 patients (2.1%), respectively. Models were developed for AKI, with an area under the receiver operating curve (AU-ROC) of 0.67 and 0.69 preoperatively and postoperatively, respectively. Models for RRT had an AU-ROC of 0.77 and 0.80 preoperatively and postoperatively, respectively. These models contained between 3 and 5 variables. Comparatively, comprehensive least absolute shrinkage and selection operator models contained between 21 and 26 variables, with an AU-ROC of 0.71 and 0.72 for AKI and 0.84 and 0.87 for RRT respectively.

CONCLUSION

In the present study, simple, clinically applicable models for predicting AKI and RRT preoperatively and immediate postoperatively were developed. Even though AKI prediction remained poor, RRT prediction was good with a parsimonious model.

Continuous Magnesium Infusion to Prevent Atrial Fibrillation After Cardiac Surgery: A Sequential Matched Case-Controlled Pilot Study

OBJECTIVE

The authors aimed to test whether a bolus of magnesium followed by continuous intravenous infusion might prevent the development of atrial fibrillation (AF) after cardiac surgery.

DESIGN

Sequential, matched, case-controlled pilot study.

SETTING

Tertiary university hospital.

PARTICIPANTS

Matched cohort of 99 patients before and intervention cohort of 99 consecutive patients after the introduction of a continuous magnesium infusion protocol.

INTERVENTIONS

The magnesium infusion protocol consisted of a 10 mmol loading dose of magnesium sulphate followed by a continuous infusion of 3 mmol/h over a maximum duration of 96 hours or until intensive care unit discharge.

MEASUREMENTS AND MAIN RESULTS

The study groups were balanced except for a lower cardiac index in the intervention cohort. The mean duration of magnesium infusion was 27.93 hours (95% confidence interval [CI]: 24.10-31.76 hours). The intervention group had greater serum peak magnesium levels: 1.72 mmol/L ± 0.34 on day 1, 1.32 ± 0.36 on day 2 versus 1.01 ± 1.14 and 0.97 ± 0.13, respectively, in the control group (p < 0.01). Atrial fibrillation occurred in 25 patients (25.3%) in the intervention group and 40 patients (40.4%) in the control group (odds ratio 0.49, 95% CI, 0.27-0.92; p = 0.023). On a multivariate Cox proportional hazards model, the hazard ratio for the development of AF was significantly less in the intervention group (hazard ratio 0.45, 95% CI, 0.26-0.77; p = 0.004).

CONCLUSION

The magnesium delivery strategy was associated with a decreased incidence of postoperative AF in cardiac surgery patients. These findings provide a rationale and preliminary data for the design of future randomized controlled trials.

Mortality Prediction After Cardiac Surgery: Higgins' Intensive Care Unit Admission Score Revisited

BACKGROUND

This study was performed to develop and validate a cardiac surgical intensive care risk adjustment model for mixed cardiac surgery based on a few preoperative laboratory tests, extracorporeal circulation time, and measurements at arrival to the intensive care unit.

METHODS

This was a retrospective study of admissions to 5 cardiac surgical intensive care units in Sweden that submitted data to the Swedish Intensive Care Registry. Admissions from 2008 to 2014 (n = 21,450) were used for model development, whereas admissions from 2015 to 2016 (n = 6463) were used for validation. Models were built using logistic regression with transformation of raw values or categorization into groups.

RESULTS

The final model showed good performance, with an area under the receiver operating characteristics curve of 0.86 (95% confidence interval, 0.83-0.89), a Cox calibration intercept of -0.16 (95% confidence interval, -0.47 to 0.19), and a slope of 1.01 (95% confidence interval, 0.89-1.13) in the validation cohort.

CONCLUSIONS

Eleven variables available on admission to the intensive care unit can be used to predict 30-day mortality after cardiac surgery. The model performance was better than those of general intensive care risk adjustment models used in cardiac surgical intensive care and also avoided the subjective assessment of the cause of admission. The standardized mortality ratio improves over time in Swedish cardiac surgical intensive care.

Variation between hospitals in outcomes following cardiac surgery in the UK

INTRODUCTION

We examine the influence of variations in provision of cardiac surgery in the UK at hospital level on patient outcomes and also to assess whether there is an inequality of access and delivery of healthcare. Cardiothoracic surgery has pioneered the reporting of surgeon-specific outcomes, which other specialties have followed. We set out to identify factors other than the individual surgeon, which can affect outcomes and enable other surgical specialties to adopt a similar model.

MATERIALS AND METHODS

A retrospective analysis of prospectively collected data of patient and hospital level factors between 2013 and 2016 from 16 cardiac surgical units in the UK were analysed through the Society for Cardiothoracic Surgery of Great Britain and Ireland and the Royal College of Surgeons Research Collaborative. Patient demographic data, risks factors, postoperative complications and in-hospital mortality, as well as hospital-level factors such as number of beds and operating theatres, were collected. Correlation between outcome measures was assessed using Pearson's correlation coefficient. Associations between hospital-level factors and outcomes were assessed using univariable and multivariable regression models.

RESULTS

Of 50,871 patients (60.5% of UK caseload), 25% were older than 75 years and 29% were female. There was considerable variation between units in patient comorbidities, bed distribution and staffing. All hospitals had dedicated cardiothoracic intensive care beds and consultants. Median survival was 97.9% (range 96.3-98.6%). Postoperative complications included re-sternotomy for bleeding (median 4.8%; range 3.5-6.9%) and mediastinitis (0.4%; 0.1-1.0%), transient ischaemic attack/cerebrovascular accident (1.7%; range 0.3-3.0%), haemofiltration (3.7%; range 0.8-6.8%), intra-aortic balloon pump use (3.3%; range 0.4-7.4%), tracheostomy (1.6%; range 1.3-2.6%) and laparotomy (0.3%; range 0.2-0.6%). There was variation in outcomes between hospitals. Univariable analysis showed a small number of positive associations between hospital-level factors and outcomes but none remained significant in multivariable models.

CONCLUSIONS

Variations among hospital level factors exists in both delivery of, and outcomes, following cardiac surgery in the UK. However, there was no clear association between these factors and patient outcomes. This negative finding could be explained by differences in outcome definition, differences in risk factors between centres that are not captured by standard risk stratification scores or individual surgeon/team performance.

Quality metrics in coronary artery bypass grafting

Studies on the association between care quality, case volume, and outcomes in coronary artery bypass grafting (CABG) have concluded that consistent adherence to quality measures improves mortality rates and outcomes. However, the quality metrics are not well-defined, and their significance to surgeons and healthcare providers remains uncertain. We review the concept of "quality and quality metrics" and discuss their importance in the context of CABG.

Measuring the quality of perioperative care in cardiac surgery

Quality of care is of increasing importance in health and surgical care. In order to maintain and improve quality, we must be able to measure it and identify variation. In this narrative review, we aim to identify measures used in the assessment of quality of care in cardiac surgery and to evaluate their utility. The electronic databases Pubmed/MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Database of Systematic Reviews, and CINAHL were searched for original published studies using the terms 'cardiac surgery' and 'quality or outcome or process or structure' as either keywords in the title or text or MeSH terms. Secondary searches and identification of references from original articles were carried out. We found a total of 54 original articles evaluating measurements of quality. While structure, process, and outcome indicators remain the mainstay of quality measurement, new and innovative methods of risk assessment have improved reliability and discrimination. Continuous assessment provides a promising method of both maintaining and improving quality of care. Future studies should focus on long-term and patient-centred outcomes, such as quality-of-life measures.

Acute Risk Change: An Innovative Measure of Operative Adverse Events and Perioperative Team Performance

OBJECTIVES

Cardiac surgical risk models predict mortality preoperatively, whereas intensive care unit (ICU) models predict mortality postoperatively. Finding a large difference between the 2 (an acute risk change [ARC]) may reflect an alteration in the status of the patient related to the surgery. An adverse ARC was associated with morbidity and mortality in an Australian population. The aims of this study were to validate ARC in a UK population and to investigate the possible mechanisms behind ARC.

DESIGN

This was a retrospective case-control study.

SETTING

Single, high-volume cardiothoracic hospital.

PARTICIPANTS

Data from 4,842 cardiac surgical patients were collected between 2013 and 2015.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

EuroSCORE was recalibrated to each preceding year's data. ARC was defined as postoperative minus preoperative percentage mortality risk. Association among ARC, morbidity, and mortality was tested. Cases with large adverse ARC (greater than +15%) were compared with cases with large favorable ARC (less than -10%) with regard to intraoperative adverse events, unmeasured patient risk factors, and postoperative events. Adverse ARC was associated with hospital mortality, ICU stay, ICU readmission, renal support, prolonged intubation and return to the operating room (p < 0.001). Intraoperative adverse events occurred in 23 of 33 patients with adverse ARC; however, only 2 of 17 patients with favorable ARC reported adverse events (p < 0.001). Unmeasured risk factors were present in 48% of patients in the adverse ARC group.

CONCLUSION

ARC is a readily available and sensitive marker that correlates strongly with morbidity and mortality. The use of ARC in local and national quality monitoring could identify areas for improvement of the quality of cardiac surgical care.

Assessing contemporary intensive care unit outcome: development and validation of the Australian and New Zealand Risk of Death admission model

The Australian and New Zealand Risk of Death (ANZROD) model currently used for benchmarking intensive care units (ICUs) in Australia and New Zealand utilises physiological data collected up to 24 hours after ICU admission to estimate the risk of hospital mortality. This study aimed to develop the Australian and New Zealand Risk of Death admission (ANZROD₀) model to predict hospital mortality using data available at presentation to ICU and compare its performance with the ANZROD in Australian and New Zealand hospitals. Data pertaining to all ICU admissions between 1 January 2006 and 31 December 2015 were extracted from the Australian and New Zealand Intensive Care Society Adult Patient Database. Hospital mortality was modelled using logistic regression with development (two-thirds) and validation (one-third) datasets. All predictor variables available at ICU admission were considered for inclusion in the ANZROD₀ model. Model performance was assessed using Brier score, standardised mortality ratio and area under the receiver operating characteristic curve. The relationship between ANZROD₀ and ANZROD predicted risk of death was assessed using linear regression. After standard exclusions, 1,097,416 patients were available for model development and validation. Observed mortality was 9.5%. Model performance measures (Brier score, standardised mortality ratio and area under the receiver operating characteristic curve) for the ANZROD₀ and ANZROD in the validation dataset were 0.069, 1.0 and 0.853; 0.057, 1.0 and 0.909, respectively. There was a strong positive correlation between the mortality predictions with an overall R² of 0.73. We found that the ANZROD₀ model had acceptable calibration and discrimination. Predictions from the models had high correlations in all major diagnostic groups, with the exception of cardiac surgery and possibly trauma and sepsis.