The use of quality indicators in trauma care is an evolving process. Ideal quality indicators should have high reliability, sensitivity, and specificity, while process measures should have empirical links with patient outcomes [11]. With respect to trauma, ideal quality indicators should apply to a significant number of cases, rely on current and best practice, consider a specific/appropriate population, include a risk adjustment strategy, and reflect outcomes other than mortality [12]. Most health care providers have experienced cases where good patient outcomes have occurred despite poor care (i.e., faulty process or inadequate structure) and cases when poor outcomes have occurred despite optimal care in well-equipped facilities. Thus, any measure that assesses quality must be designed such that it truly captures the aspect of the healthcare system/patient care that it is thought to capture.

The American College of Surgeons Committee on Trauma (ACS-COT) used expert consensus to create an initial set of quality indicators (audit filters) in 1987 to facilitate quality improvement from the peer review process [6]. Since then, a plethora of indicators have been implemented in a variety of jurisdictions, which allows for standards to be set and for other organizations to work toward in order to improve trauma patient care. In a recent survey of trauma centers in Canada, the USA, and Australasia, 10,587 quality indicators were identified from 242 institutions, including 1,102 unique indicators. The ten most common quality indicators identified are listed in Table 11.2 [13].

Several limitations in the reported quality indicators were noted. Most quality indicators were not well specified; a descriptive statement was included, but lacked detail regarding data elements or construction of the measure, which impacts reliability and validity of the indicator. Furthermore, not all aspects of trauma care were captured with these measures. While many indicators reflected hospital process and outcomes for trauma care, few measured prehospital care and even fewer measured posthospital care or secondary prevention. In all phases of trauma care, structure-based quality indicators were rarely used. With respect to the Institute of Medicine’s six aims for improvement (described earlier in this chapter), patient-centered care and equitable care were measured by less than 1 % of the 10,587 quality indicators reported [13]. Process measures put forth by ACS-COT have also been questioned in their ability to reflect patient outcomes [14]. A recent review of current ACS-COT quality indicators found those that are strongly associated with clinical outcomes may lack face validity to identify poor-quality care for complex multi-trauma patients [7]. It is clear that gaps exist in the current assessments of trauma quality.

Despite these gaps, several initiatives in North America show promise to advancing the measure of quality in trauma. In 2006, the American College of Surgeons created the Trauma Quality Improvement Program (TQIP) to move forward from ACS-COT audit filters and provide reliable, high-quality, and risk-adjusted data for mortality rates and ten common in-patient complications, e.g., deep venous thrombosis, across participating trauma centers [15]. Rigorous standardization of the National Trauma Data Standard (the means by which hospital data is collected) has allowed TQIP to provide benchmarking measures such that individual trauma hospitals can compare themselves to other centers and identify areas of strength and weakness in prespecified areas [15, 16]. As well, there is the ability for institutions to network in order to facilitate and offer mentorship around some of the QI initiatives. This allows institutions to focus QI efforts in addressing areas of substandard performance. It also highlights best practices of care, which can be shared among institutions, and can influence funding in pay-for-performance models. Currently, over 200 institutions in North America are using TQIP [17].

At this time, TQIP relies on ordinary logistic regression to compare institutions. Ongoing work in Quebec, Canada, to assess interhospital mortality suggests that hierarchical logistic regression may be a better method to assess hospitals given the relationship among patients treated at the same institution [18]. However, this method is more complex than ordinary logistic regression and the implications of widespread implementation are still being evaluated. Related work suggests that risk-adjusted models for length of stay [19] and unplanned readmissions [20] are valid quality indicators to reflect acute trauma care. These models are based on routinely collected registry and administrative data and can be used to drive performance improvement. Another method to increase reliability of trauma quality indicators has been the development of composite measures for predicting mortality. Computation of composite scores, based on multiple indicators, takes interactions between processes into account and may better reflect the complexity of trauma care [21]. Composite measures have been used to identify top hospital performers in management of medical issues such as congestive heart failure, pneumonia, and acute myocardial infarction [22]. In the setting of trauma care, performance on nine process measures, e.g., head CT within 2 h of injury, combined as a composite measure successfully predicted mortality rates at the individual hospital level (Table 11.3) [12].