Special Session 7 - Clinical Outcomes Monitoring Programmes in Hospital Authority

Data-driven care in surgery refers to the integration of data analytics, advanced technologies, and computational methods in surgical practice to enhance decision-making, optimize procedures and improve patient outcomes and satisfaction. Back in early 2000s, the Hospital Authority (HA) Clinical Outcomes Monitoring Programme (COMP) was begun with the ongoing comparative audits piloted by frontline surgical units. Subsequently, with the support from the HAHO, a frontline surgeons-initiated Surgical Outcomes Monitoring and Improvement Programme (SOMIP) was launched in 2008.
Through clinical audit and risk-adjusted analysis, clinical specialties can identify factors that may impact the outcomes of patients. In turn, evidence-based improvement practices can be implemented, leading to better outcomes. Availability of these data can assist stakeholders in inidivudal hospitals to address deficiencies in both the software and the hardware. Those units with the best outcomes could also share their practice with their peers through various forms of collaboration.
COMP has demonstrated its impact on reducing mortality rate and enhancing care quality. Not only does COMP reflect actionable insights for evidence-based decision-making and quality improvement within the HA, but it also fostered a culture of patient safety and quality improvement. Recognizing the importance of outcomes monitoring, other initiatives, namely Intensive Care Unit Outcomes Monitoring Programme (ICUOMP) and Neurosurgical Outcomes Monitoring and Improvement Programme (NOMIP), have joined under the umbrella of COMP since 2015 and 2020, respectively.

The Surgical Outcomes Monitoring and Improvement Programme (SOMIP) is an outcome-based, risk-adjusted, hospital-focused programme implemented across 17 public hospitals in Hospital Authority (HA) with surgical unit since 2008. Its primary objective is to enable risk-adjusted comparison among surgical units and identify improvement areas in clinical services provided by public hospitals. 
The data generated by SOMIP has enabled hospitals and the specialty to address issues that may not have been identified during their own in-house surgical audits or quality assurance exercises. Through provision of risk-adjusted outcome data, providing feedback to surgical units and organizing discussion forum, SOMIP provides a unique platform for sharing of surgical knowledge and skills among different specialties and hospitals. 
With the concerted effort from the stakeholders, a sustainable low mortality rate has been observed in surgical units in HA. Looking ahead, SOMIP will put more emphasis on reduction of post-operative complications.

Intensive care is a vital service in the public health care system, providing critical care to patients with life-threatening illnesses. The Intensive Care Unit Outcomes Monitoring and Improvement Programme (ICUOMP) was commissioned by the Coordinating Committee in Intensive Care (COC(ICU)) in 2015. Using Acute Physiology and Chronic Health Evaluation (APACHE) IV system as a reference, ICUOMP established refined and contemporary risk adjusted models to effectively benchmark the ICU performance in Hospital Authority (HA).
The ICUOMP aims to develop a reliable local contemporary clinical audit related to critically ill patients in intensive care units (ICUs), measure and enhance the quality of critical care services, and advise strategic planning of ICU services within HA.
ICUOMP provides reliable and high-quality information on several aspects of ICUs. It assists COC(ICU) to improve professional care by increasing understanding of patient characteristics, disease severity, outcomes, resource utilization, and capacity planning in the short and long term.
Over the years, ICU utilization has been increasing, while the crude hospital mortality rate has remained relatively stable. Demographic characteristics of patients and multi-level analysis have identified statistically significant variables related to mortality and long-stayers.

The Neurosurgical Outcomes Monitoring and Improvement Programme (NOMIP) aims to monitor and enhance the outcomes of neurosurgical procedures in public hospitals in Hong Kong. NOMIP utilizes a non-punitive and transparent process to encourage hospitals to improve their performance. The programme focuses on evaluating postoperative mortality and complications within 30 days, taking into account patient risk factors and specific indicators for neurosurgery.
Through NOMIP, this programme aims to monitor risk-adjusted mortality and morbidities within 30 days after neurosurgical operations and assess the stability of a mortality prediction model. Additionally, the feasibility of establishing a neurosurgical risk calculator will be examined.
Data on major and ultra-major neurosurgical operations were retrieved from the Clinical Data Analysis and Reporting System (CDARS). Risk-adjusted models for mortality and morbidities were developed using logistic regression, and model performance was assessed through cross-validation. 
Before regularization of NOMIP, during April 2020 to March 2023, risk-adjusted mortality models were built and yielded satisfactory results for each year. A cumulative three-year risk-adjusted model on 30-day mortality even achieved excellent discrimination power. The prediction model for 30-day mortality included significant risk factors and demonstrated fabulous predictive performance.
The next steps involve refining the risk prediction model of mortality and morbidity so as to support clinical decision-making before neurosurgical procedures. This advancement will provide valuable guidance for improving patient outcomes and optimizing the delivery of neurosurgical care.