Early identification of patients in the emergency department (ED) who would benefit from palliative care is crucial for improving patient outcomes and aligning treatment with patient goals. A recent study highlighted the effectiveness of the Palliative Care and Rapid Emergency Screening (P-CaRES) tool in predicting mortality among older adults in the ED setting. This article delves into the utility of the P-CaRES screening tool and its potential to transform early intervention strategies in emergency medicine.
The study, consistent with previous research, underscores the P-CaRES tool’s ability to identify patients at higher risk of mortality. Notably, it found that patients screening positive on the P-CaRES tool had a significantly higher mortality rate within three months of hospital discharge. This aligns with findings from Paske et al., who reported over 50% mortality within six months for P-CaRES positive patients. However, while Paske’s study assessed patients hours after admission, this recent research focused on ED assessment, emphasizing the importance of early intervention at the point of initial contact. By utilizing the P-CaRES screening tool in the ED, clinicians can proactively establish care objectives, potentially preventing unnecessary aggressive treatments for patients nearing the end of life.
Furthermore, the study validated the P-CaRES tool’s capacity to identify patients with pre-existing conditions. Its predictive power is amplified when combined with readily available triage vital signs such as elevated body temperature (BT > 37.5 °C), rapid pulse rate (PR > 100 bpm), and increased respiratory rate (RR > 25 bpm). This combination offers a robust method for identifying older adults who would significantly benefit from serious illness conversations and the initiation of palliative care services, either within the ED or shortly thereafter. The P-CaRES tool incorporates potential terminal illnesses, which, when coupled with abnormal vital signs indicative of exacerbated conditions, heighten the predicted risk of mortality.
Interestingly, the study also examined the “Surprise Question” (SQ) – asking clinicians “Would you be surprised if this patient died within [ timeframe ]?” – alongside the P-CaRES tool. The findings regarding SQ (“no surprise” indicating higher mortality risk) were consistent with prior research, confirming its accuracy in mortality prediction. However, systematic reviews have pointed out the varying sensitivity of SQ over different timeframes, suggesting that while highly specific in the short-term, its sensitivity increases over longer periods. The study’s results indicated that combining SQ with triage data and electronic medical records, including vital signs and pre-existing palliative care plans, enhanced its predictive accuracy for 3-month mortality. This combined approach using SQ alongside readily available patient data can be a valuable asset in identifying older adults in the ED who require serious illness conversations and potential palliative care interventions.
Extending beyond previous work that often focused on 12-month mortality prediction with SQ, this study specifically evaluated 3-month survival. The magnitude of association between SQ and mortality observed in this ED-focused study was notably higher than in prior research conducted in different settings. This difference could be attributed to the unique context of the ED, where emergency physicians’ prognostic perceptions are shaped by a multitude of factors, including immediate clinical presentations, underlying conditions, and the acute nature of medical emergencies. Experienced clinicians, as noted in other studies, tend to demonstrate greater accuracy in such predictions, likely due to their accumulated clinical judgment and pattern recognition. The ED environment, with its specific patient population and resource availability for life-sustaining treatments, further influences these prognostic assessments.
While broad and narrow criteria have been explored for identifying patients with pre-existing conditions in the ED who might benefit from palliative care, this study did not find that these criteria directly decreased 3-month survival. The study suggests that the inclusion of broad criteria, particularly those employing the Charlson Comorbidity Index (CCI), which was originally designed for long-term (10-year) survival prediction, might explain this finding. The CCI, while valuable in many contexts, may not be as directly applicable to short-term mortality prediction in the acute ED setting.
Clinical Implications
This study reinforces the clinical utility of the P-CaRES screening tool and the Surprise Question in the fast-paced environment of the emergency department. The findings demonstrate that both P-CaRES positivity and a “no surprise” answer to the Surprise Question are significant predictors of short-term mortality. When integrated with vital signs and readily accessible patient information, these tools can empower ED physicians to engage in timely and sensitive discussions about care objectives with patients and their families. Future research should explore the synergistic potential of combining vital signs and disease-specific prognostic tools for conditions like dementia, COPD, or decompensated heart failure to further refine and enhance prognostic accuracy in the ED.
Limitations
It is important to acknowledge the limitations of this single-center study, which utilized convenience sampling and focused on weekday data collection during specific hours. This methodology introduces potential selection bias. Furthermore, the study period coincided with the COVID-19 pandemic, and patients under investigation for COVID-19 were excluded, potentially impacting the generalizability of the findings during pandemic conditions. Data collection relied on research assistants and PGY-3 residents, and inter-rater reliability for data collection was not specifically evaluated, although prior studies have indicated the P-CaRES tool exhibits high inter-rater reliability and is less subjective. Finally, the relatively small sample size limited the ability to perform stratified analysis for specific diagnoses. Larger, multi-center studies are warranted to validate these findings and further explore the P-CaRES tool’s application across diverse ED populations and clinical scenarios.
