The effect of the shock index and scoring systems for predicting mortality among geriatric patients with upper gastrointestinal bleeding: a prospective cohort study

ABSTRACT BACKGROUND: Gastrointestinal (GI) bleeding is an important cause of mortality and morbidity among geriatric patients. OBJECTIVE: To investigate whether the shock index and other scoring systems are effective predictors of mortality and prognosis among geriatric patients presenting to the emergency department with complaints of upper GI bleeding. DESIGN AND SETTING: Prospective cohort study in an emergency department in Bursa, Turkey. METHODS: Patients over 65 years admitted to a single-center, tertiary emergency service between May 8, 2019, and April 30, 2020, and diagnosed with upper GI bleeding were analyzed. 30, 180 and 360-day mortality prediction performances of the shock index and the Rockall, Glasgow-Blatchford and AIMS-65 scores were evaluated. RESULTS: A total of 111 patients who met the criteria were included in the study. The shock index (P < 0.001) and AIMS-65 score (P < 0.05) of the patients who died within the 30-day period were found to be significantly different, while the shock index (P < 0.001), Rockall score (P < 0.001) and AIMS-65 score (P < 0.05) of patients who died within the 180-day and 360-day periods were statistically different. In the receiver operating characteristic (ROC) analysis for predicting 360-day mortality, the area under the curve (AUC) value was found to be 0.988 (95% confidence interval, CI, 0.971-1.000; P < 0.001). CONCLUSION: The shock index measured among geriatric patients with upper GI bleeding at admission seems to be a more effective predictor of prognosis than other scoring systems.

Various scoring systems are used to predict prognosis and mortality among patients with upper GI bleeding. The most frequently used scoring systems for this purpose are Rockall, Glasgow-Blatchford and AIMS-65. These use clinical information and results from laboratory tests and endoscopy. 10 The Rockall score, which has the aim of helping to discharge low-risk patients and reduce costs, was created based on criteria such as age, comorbidity, shock status, endoscopic diagnosis and findings of new bleeding, in order to predict rebleeding in patients with upper GI bleeding. 11 The aim of the Glasgow-Blatchford scoring system is to predict the need for intervention to control bleeding. Endoscopic findings are not included in the evaluation. Scoring is between 0 and 23, and as the score increases, the need for endoscopy also increases. 12  , which is an easy-to-remember and simple scoring system, provides a risk score for predicting in-hospital mortality, length of stay and cost, for patients with acute upper GI bleeding. It is based on the patient's age, systolic blood pressure, mental status and laboratory data. 13

OBJECTIVE
The aim of this study was to investigate whether the shock index and other scoring systems measured at admission to the emergency department are effective predictors of mortality and prognosis among geriatric patients with upper GI bleeding.

Patient selection and location
This study was carried out in the Department of Emergency In this study, patients over 65 years of age with a diagnosis of upper GI bleeding who presented to the Department of Emergency

Methods and measurements
A total of 128 patients were included in the study. Two of the patients were excluded because endoscopy could not be performed; five were excluded because no focus of bleeding could be detected through endoscopy; and ten patients were excluded because they could not be reached. Thus, a total of 111 patients over the age of 65 years who met the criteria and were diagnosed with upper GI bleeding through the tests and examinations were included in the study.
The patients' vital signs and laboratory findings, and especially their demographic information, pulse rate and systolic and diastolic blood pressures, were recorded at admission. The "shock index" was calculated by dividing the patients' pulse rate at the time of first admission by the systolic blood pressure. The patients' existing comorbidities, current medications and endoscopy results were followed up and recorded on the case report forms.  Levene's test.
The significance of differences between the groups, in terms of continuous numerical variables in which the statistical assumptions of parametric tests were met, was evaluated using Student's t test.
The significance of differences, regarding continuous numerical variables in which the statistical assumptions of parametric tests were not met, was investigated using the Mann-Whitney U test.
Spearman's correlation analysis was used to evaluate the relationship between variables with nonparametric distribution.
A receiver operating characteristic (ROC) curve was drawn to investigate the 30, 180 and 360-day mortality prediction performances of the shock index and Rockall, Glasgow-Blatchford and AIMS-65 scores. Logistic regression analysis was performed to determine the factors affecting mortality. The results were reported with the 95% confidence interval (CI), and P < 0.05 was considered statistically significant.
The Mann-Whitney U test was performed to investigate whether there were any differences in the patients' median shock index or Rockall, Glasgow-Blatchford and AIMS-65 scores, with regard to 30, 180 and 360-day mortality. The results showed that the shock index and AIMS-65 score were significantly different among patients who died within 30 days (P < 0.001 and P < 0.05).
Additionally, the shock index, Rockall score and AIMS-65 score were found to be significantly different in patients with 180-day mortality (P < 0.001, P < 0.001 and P < 0.05). Lastly, the shock index, Rockall score and AIMS-65 score were found to be significantly different among patients with 360-day mortality (P < 0.001, P = 0.001 and P < 0.05) ( Table 3).
The diagnostic value of the patients' shock index and Rockall, Glasgow-Blatchford and AIMS-65 scores for 30, 180 and 360-day mortality were analyzed using ROC. For 30-day mortality, the area under the curve (AUC) for the shock index was 0.911 (P < 0.001) while the AUC for the AIMS-65 score was 0.662 (P < 0.05). For 180-day mortality, the AUC for the shock index was found to be 0.960 (P < 0.001), the AUC for the Rockall score was 0.714 (P < 0.001) and the AUC for the AIMS-65 score was 0.657 (P < 0.05).
For 360-day mortality, the AUC for the shock index was 0.988 (P < 0.001), the AUC for the Rockall score was 0.690 (P < 0.05) and the AUC for the AIMS-65 score was 0.641 (P < 0.05) (Figure 1).
When the cutoff value of the shock index for 30-day mortality was 1.240, the sensitivity was found to be 82.4% and the specificity was 81.9%. When the cutoff value of the AIMS-65 score was 1.5, the sensitivity was found to be 76.5% and the specificity was 50.0%.
When the cutoff value of the shock index for 180-day mortality was 1.205, the sensitivity was 91.2% and the specificity was 92.2%.
When the cutoff value of the Rockall score was 5.5, the sensitivity was found to be 38.2% and specificity was 61.8%. When the cutoff value of the shock index for 360-day mortality was 1.06, the sensitivity was 95.3% and the specificity was 94.1%. Accordingly, it can be seen that the performance of the shock index was significantly better than that of the other scoring systems ( Table 4).
Logistic regression analysis was performed using variables of gender, comorbidities and drug use history, which were thought to have an effect on 360-day mortality. The history of drug use was found to be an effective factor for diagnosing 360-day mortality (Exp beta = 6.489; 95% CI, 1.607-26.208; P = 0.009) ( Table 5). The shock index is obtained by dividing the heart rate by the systolic blood pressure, and its normal value is between 0.5 and 0.7. When the shock index is greater than 0.9, presence of  18 Most of the studies in the literature investigated in-hospital or 30-day mortality. In our study, in addition to 30-day mortality, we also examined 180 and 360-day mortality rates. We think that these results will contribute to the literature. In addition, the mortality rates in our study were generally higher, contrary to the data in the literature. We think that this was because the population that we examined consisted of elderly patients and because they had high numbers of comorbidities.    These are obtained using the clinical information, laboratory data and endoscopy results of the patients. 10 found that the performance of the shock index was weak with regard to predicting the 30-day mortality rate in their study. 26 We think that the fact that the patients included in their study were younger and had less comorbidity may have caused that result.

DISCUSSION
We also believe that the shock index could not be evaluated as a significant predictor of mortality in the study by Saffouri et al.
due to the hemodynamical instability of their patients at the time of admission, relatively young age of the patients and presence of cardiovascular compensation.
In our study, we found that the shock index was much more effective for predicting 30, 180 and 360-day mortality, which were the endpoints of the study, than the other scoring systems. In particular, both the AUC and the cutoff values of the shock index were statistically more significant than other scoring systems. When the cutoff value for the shock index regarding 30-day mortality was 1.240, the sensitivity was found to be 82.4% and the specificity was 81.9%. When the 180-day mortality cutoff value was 1.205, the sensitivity was found to be 91.2% and the specificity was 92.2%.
Lastly, when the 360-day mortality cutoff value was 1.06, the sensitivity was found to be 95.3% and the specificity was 94.1%. We believe that these cutoff values determined for the shock index will have a major role in the treatment and follow-up of geriatric patients with upper GI bleeding.

Limitations
The most important limitation of our study was that it was a single-center study. In addition, we think that the relatively small number of patients was another important limitation. In order to obtain results with greater accuracy and reliability, further multicenter studies with larger populations would be required.

CONCLUSION
The shock index, which is a simple, inexpensive and noninvasive parameter that can be obtained only from vital signs, is a more effective predictor for prognosis than other scoring systems, among geriatric patients with upper GI bleeding presenting to emergency departments.