The Association Between Abnormal Vital Signs and Mortality in the Emergency Department

Background The emergency department (ED) receives patients from all over the world every day. Hence, using various triage scales to detect sick patients and the need for early admission are essential. Triage is a process used in the ED to prioritize patients requiring the most urgent care over those with minor injuries based on medical urgency and medical needs. These decisions may be based on patients’ chief complaints at the time of their ED visit and their vital signs. Vital signs, including blood pressure (BP), respiratory rate (RR), heart rate (HR), and body temperature, are necessary tools that are traditionally used in the ED during procedures such as triage and recognizing high-risk hospital inpatients. This study aimed to determine the relationship between abnormal vital signs and mortality in the ED. Method and Material This retrospective record review study was performed at the ED of King Abdulaziz University Hospital (KAUH). Altogether, 641 patients fulfilled our inclusion criteria. Data including patients’ demographics, vital signs, in-hospital mortality, triage level, and precipitating factors were collected. Results The mean age of the patients was 45.66 ± 18.43 years (69.3% females), and the majority of them had Canadian Triage and Acuity Scale (CTAS) level 3 (71.1%). The total number of in-hospital mortalities was 32 (5%). Lower systolic blood pressure (SBP) and diastolic blood pressure (DBP), high respiratory rates, and low oxygen saturation (O2SAT) were significantly associated with high mortality rates. Conclusion Abnormal vital signs play a major role in determining patient prognosis and outcomes. Triage score systems should be adjusted and carefully studied in each center according to its population.


Introduction
The emergency department (ED) has a highly stressful environment with a large number of sick patients [1]. Due to a large number of patient visits, quick evaluation of each patient by an emergency doctor is nearly impossible [2]. Hence, the use of various triage scales to detect medically ill patients who require quick and early evaluation and admission is mandatory in the ED [3]. Triage is a process used in the ED to prioritize patients requiring the most urgent care over those with minor injuries based on their medical urgency and medical needs [3][4][5][6]. These decisions are usually based on patients' vital signs and chief complaints at the time of their ED visit [7].
Vital signs such as heart rate (HR), blood pressure (BP), respiratory rate (RR), and body temperature are important parameters that are widely used in the ED during procedures such as triage [7,8] and in the identification of high-risk hospital inpatients [9]. They are also recorded upon admission to the medical wards and throughout the patients' hospital stay during almost all procedures [10]. Vital signs also reflect the patients' current health status [11]. Thus, changing trends in patients' vital signs may indicate clinical deterioration, which may result in adverse effects or death if not identified and treated promptly [12]. Furthermore, they can be used to determine the urgency for intensive care unit transfer [13,14].
A systemic review published in 2011 examined the evidence regarding the association of abnormal vital signs and presenting symptoms with increased mortality in the ED [7]. In addition, a large number of

Study participants
The medical records of 105,837 men and women who were aged above 17 years and who visited the ED between 2018 and 2020 were analyzed. Among these, 641 patients were randomly selected using a computed system and included in this study. We excluded the patients who are below 17 years old from our study.
Vitals including HR, O 2 SAT, BP, and temperature was measured using Philips Medical Systems model SureSigns VS3 (Andover, MA, USA).

Ethical considerations
Ethical approval for the study was obtained from the institutional review board of the Ethics Research Committee of KAUH, Jeddah, Saudi Arabia (reference number: 287-20).

Statistical analysis
Data were analyzed using IBM SPSS Statistics version 26 (IBM Corp., Armonk, NY, USA). Categorical variables were presented as numbers and percentages, and the chi-squared (χ2) test was used to test the relationship between variables. Continuous variables were expressed as mean ± standard deviation (SD). The Mann-Whitney test was used for nonparametric variables. Multivariate logistic regression analysis was performed to assess the risk factors (independent predictors) of death among the studied patients. The odds ratio was determined at a confidence interval of 95%. Statistical significance was set at p < 0.05. Table 1 shows the descriptive data of the included patients. The mean age was 45.66 ± 18.43 years, 69.3% were females, and 44.6% had normal weight. The mean body mass index (BMI) was 26.71 ± 6.19 kg/m 2 . Altogether, 71.1% of the patients had CTAS level 3. More than half of the patients (55.7%) had an ER stay of less than six hours, and 76.8% of the patients had a history of ER visits in the last six months. Altogether, 40.95% of the patients had a past medical history, and the most common medical conditions were hypertension (HTN) (31.3%) and diabetes mellitus (DM) (29.8%).

Variable
No. (%)      Patients with CTAS level 1 or low SBP/DBP had a significantly higher mortality rate (p ≤ 0.05) (Figures 1, 2). Moreover, patients who had urgent RR or less urgent O 2 SAT had a significantly higher mortality rate (p ≤ 0.05) (Figure 3). 2021    The multivariate logistic regression analysis to assess the risk factors for death revealed that CKD and less urgent O 2 SAT were significant risk factors for death (95% confidence interval: p ≤ 0.05) (

Discussion
In this retrospective record review study of ED patients, we aimed to analyze the association between abnormal vital signs and mortality rate in the ED. We also aimed to identify the value of abnormal vital signs as prognostic tools for identifying patients at an increased risk of death in the hospital and enhancing the triage system.
We demonstrated that the categories of vital signs and the triage system used in our study were valid tools for predicting in-hospital mortality. Impaired RR and O 2 SAT were strongly associated with adverse outcomes. Moreover, we observed that increased BMI and a history of CKD, DM, and HTN were strongly associated with higher odds of mortality. However, the CTAS did not seem to impact mortality but helped prioritize patients with the most urgent needs.
Surprisingly, the results indicated that among the vital signs significantly associated with a higher mortality rate, greater deviation of the vital signs from their normal range was associated with lower odds of mortality. In contrast, a previous study involving a larger population showed that greater deviation of the vital signs from their normal range was associated with higher odds of mortality [21].

Respiratory rate
Low O 2 SAT [17,18,21,22] and low or high RR [3,17,21] have been identified as significant independent predictors of a high mortality rate. Our findings are consistent with the findings of these previous studies ( Figure 3).

Heart rate
We observed that HR categorization was associated with decreased odds of in-hospital mortality (Figure 2). A similar conclusion was reported by other studies [3,18,21]. However, two other studies demonstrated that HR was negatively correlated with adverse outcomes [6,17,22].
A popular explanation is that HR could be affected by many environmental or internal factors, such as anxiety, noise, stress, and emotions, in addition to serious illnesses. Thus, it might contribute to inaccurate measurements and results.

Blood pressure
Interestingly, we observed that low SBP was a significant risk factor for in-hospital mortality, but high BP was not associated with a higher mortality rate. The explanation for this finding is discussed in previous studies in Switzerland [22], Sweden [21], and South Africa [3]. Other studies have failed to reveal a significant association between low SBP and higher in-hospital mortality [17,18].

Temperature
Another novel finding in the present study was that low and high body temperatures were not significant predictors of in-hospital mortality. In-hospital mortality was negatively correlated with temperature in sub-Saharan Africa [23], South Africa, and Sweden [3,17]. However, two of the seldom published studies that investigated temperature (low or high) and its relationship with mortality in the ED found contrasting outcomes [21,24,25].
It is important to highlight that the findings in these previous studies may simply reflect the abnormal temperature value distribution in the study populations (one population may be dominated by hyperpyretic patients, while the other may be dominated by hypothermic patients). A significant correlation between temperature and the in-hospital mortality rate has hardly been defined in previous studies [26].
We observed that the length of stay in the ED did not have a significant impact on the mortality rate. This finding is supported by two other studies that failed to identify the length of stay in the ED as an independent predictor of in-hospital mortality [26,27].
The present study revealed that age was not significantly associated with higher odds of one-day mortality. Similar findings were also observed in other studies in South Africa and sub-Saharan Africa [3,23]. However, previous studies in the United Kingdom and Sweden have shown that age was associated with increased mortality [6,14,17,18,28]. In these studies, the mean age of the study population was 60 years or above, while the mean age of our study population was approximately 45 years. The study populations from sub-Saharan Africa and South Africa had mean ages of 36 and 43 years, respectively.
Thus, the differences between the previous studies and our study are possibly due to the differences in the mean age of the population, since the older population is likely to exhibit higher mortality than the younger population.

Limitations
Some of the limitations of the present study include incomplete charts in the hospital's data system and the possible inaccuracies in the chart data. Moreover, we were unable to include all patients during the study period. We selected a random sample that represented the study population.