Less is more: chest pain pathways in clinical care

The significant benefits for patients and hospitals of reduced testing can be achieved without compromising safety

Public awareness of the importance of early treatment for chest pain has contributed to improved outcomes for patients with acute coronary syndromes (ACS), but also to the large and growing number of people with chest pain attending emergency departments (EDs). Some have threatening pathology, but in most cases the underlying cause is benign. Clinicians’ desire for diagnostic accuracy, minimal risk, and low personal liability has led to high rates of low value admission and investigation, draining resources from already strained services; in the United States, 10% of ED visits are motivated by chest pain, costing US$12 billion annually.1 The rapid growth in the numbers of clinical pathways and risk scores for chest pain is one response to this unsustainable demand, each aiming to integrate clinical information, assess risk, uphold consistency and quality, and minimise diagnostic error. Successful trials of a range of pathways have been reported, but questions remain about their content and local applicability, impact on outcomes, implementation, and financial benefits.1 Two articles published in this edition of the MJA investigate these questions.

The Improved Assessment of Chest pain Trial (IMPACT) applied a rapid risk stratification tool for identifying patients at low short term (30-day) risk of an ACS who could be safely discharged home without additional cardiac testing.2 The strengths of IMPACT included its large cohort of patients, the intuitiveness of the assessment tool, and its inclusion of broader outcome measures important to clinicians and patients. IMPACT could be implemented in a diverse range of EDs, with meaningful reductions in hospital length of stay and resource consumption.

Minimising the number of missed cases of ACS and other serious pathology is critical, and has been explored by several studies of accelerated diagnostic protocols. The 2-hour protocol of the Asia–Pacific ASPECT trial categorised about 10% of patients as low risk, with 99.3% sensitivity;3 the ADAPT study in Australia and New Zealand found one ACS event among 392 low risk patients (99.7% sensitivity).4 IMPACT similarly categorised 17.9% of patients as low risk, none of whom had experienced an ACS event by 30 days. A clinical pathway would ideally expedite care for a greater proportion of patients than the 10–20% achieved by ADAPT, ASPECT and IMPACT, but extending the approach to patients at greater risk might result in a higher miss rate than is clinically acceptable. When the IMPACT risk tool was combined with judicious inpatient testing, the rate of undiagnosed ACS in an intermediate risk cohort of patients was 0.12%.2 Although promising, this finding requires further validation before the approach is more widely adopted.

IMPACT used a sensitive, but not highly sensitive, troponin assay. Given the recent emphasis on high sensitivity assays, this may seem contentious,5 but is consistent with other reports.6 As many Australian and New Zealand EDs do not have access to high sensitivity assays, the safety of sensitive assays is reassuring. Further, IMPACT does not emphasise chest pain characteristics; comparable scores, including the Emergency Department Assessment of Chest Pain Score (EDACS), incorporate subjective features with low diagnostic value and possibly cause ascertainment bias.1,7 Patients at low risk can be identified without pain descriptors.8

Additional diagnostic testing of patients with low risk chest pain is controversial. The patients categorised by IMPACT as low risk were young (mean age, 33.8 years), but 24% underwent objective testing at the request of a clinician, usually exercise electrocardiography (ECG); the authors concluded that “discharging patients with normal findings from this low risk, inexpensive and widely available investigation is safe.”2 The utility of exercise ECG for risk stratification of low risk patients, however, is limited, and testing does not influence revisit rates for similar symptoms.9 Several guidelines recommend against testing patients at low risk;10,11 false positive results can lead to a cascade of inappropriate intervention and the risk of harm.

A clinician’s appetite for risk is rarely discussed. One study found that the most risk-averse ED clinicians admitted 14% more low risk patients than the least risk-averse; the availability of a chest pain unit did not alter this behaviour.12 The impact of clinician behaviour on the consistency of decision making needs to be taken into account when implementing clinical pathways.

The Accelerated Chest pain Risk Evaluation (ACRE) project, on the other hand, illustrates the system-level benefits of a chest pain pathway.13 This commendable initiative implemented the ADAPT assessment protocol in 16 public hospitals across Queensland, and then evaluated pre- and post-implementation process measures and costs; it did not assess the outcomes for individual patients, which the authors concede was a weakness of the study. The statewide project achieved an estimated released capacity of $11.2 million through reduced admission rates and $2.3 million through shorter hospital stays, impressive gains in view of the fact that only 20% of patients with chest pain were managed on the accelerated pathway. Further savings could be realised should expanding the approach to groups with higher risk prove possible, as in the IMPACT trial.

The ability of ACRE to drive rapid system change is also encouraging. Changing models of care requires a whole-of-system approach, but modifying clinician behaviour is often the greatest challenge. Ian Scott has noted that for doctors with an entrenched personal belief in an established clinical practice, “even multiple studies involving patients representative of everyday practice may not prove persuasive”.14 Traditional methods for achieving best practice, including education, audit feedback, academic detailing, and professional leadership, are ineffective.14

Unwelcome “forced function” manoeuvres, including financial disincentives, are receiving increased attention from funders and policy makers;14 but the system-wide implementation of ACRE was successfully achieved without disincentives. A related article described the engagement of senior clinicians, collaboration between departments, and the incorporation of ADAPT into local pathways.15 Some EDs implemented the ACRE pathway in 3 months, while others needed more than 12. The order in which sites were approached, starting with the most receptive, was an important factor in the success of the project. Whether all ED physicians accepted the reform is unclear; survey results were positive, but the response rate was only 36%.15 Interdepartmental relationships, the busy nature of the ED, the accessibility of pathway documents, and clinicians having time to engage with the project all proved challenging.

The longer term sustainability of clinical pathways needs more research. Adherence to pathways may decline after 12 months,16,17 the time point at which the ACRE report ends. Reasons for regression include suboptimal clinician engagement, organisational support systems, feedback mechanisms, and re-education. The ACRE group note it is easier for clinicians faced with competing priorities to revert to standard practice,15 but responding to these problems can sustain improvements.18

Finally, any clinical intervention must keep the patient at the centre of decisions about their health care. In a randomised trial, a patient-focused decision aid for patients with low risk chest pain resulted in greater patient knowledge, as well as reduced observation unit admission and stress testing rates. The decision aid was acceptable to patients and physicians, and increased patient satisfaction.19 Combined with other improvements in applying chest pain pathways, shared decision making is vitally important for avoiding low value care and improving resource use.