Dataset for evaluation of threescoring systems for forecasting the clinical outcomes of patients with upper gastrointestinal bleeding (UGIB) – Ahvaz, Iran

Upper gastrointestinal bleeding (UGIB) which occurs proximal to the Treitz ligament is one of the most common cases is emergency medical conditions. The aim of this data article is to evaluation of Rockall and Blatchford scoring systems for predicting the clinical outcomes of patients with upper gastrointestinal bleeding in Imam-Khomeini Hospital, Ahvaz, Iran. This dataset was collected by retrospective descriptive epidemiologic survey which 350 non-cirrhotic patients with UGIB who referred to Ahwaz Imam-Khomeini Hospital for six months. According to the obtained data, in both clinical Rockall and complete Rockall systems, the need for re-endoscopy and the risk of re-bleeding in patients with high scores was more compared to patients with low scores. While, the obtained data about Blatchford score for re-endoscopy and re-bleeding risk was showed which no significant difference. Based on to present dataset, the Rockall systems was superior to Blatchford systems in predicting the re-bleeding as well as the need for re-endoscopy while, none of the systems were efficient in terms of predicting the need for urgent endoscopy and surgery.


Value of the data
To conduct further research on the topic under discussion, the obtained data from this dataset can be the basis for performance of the future similar studies.
The acquired data from present dataset can be useful to decide about selection of the most effective scoring systems for predicting the clinical outcomes of patients with upper gastrointestinal bleeding.
The obtained data from this dataset can be useful to increase patient satisfaction as well as reduce the cost of the health system.

Data
According to the demographic data, the patients were 18-94 years old, with the standard deviation of 20.63. Out of 350 patients studied, 230 patients were males (65.7%) and 120 patients were female (34.3%).
Among the patients, 126 (36%) patients suffered from hemostasis, 124 (35.42%) from melena, and 88 patients (25.14%) from both hemostasis and melena. Moreover, 4 patients (1.4%) complained about syncope, and 9 patients (2.57%) complained both about syncope and melena which were the most common causes of the referral of patients with gastrointestinal hematemesis. The frequency of clinical findings related to selected patients were shown in Table 1. The classification of the patients into low-risk and high-risk groups based on mean value (cut-off) for the three scoring systems were shown in Table 2. In addition, the characteristic curve for the Blatchford, clinical Rockall and complete Rockall scoring systems were provided in Figs. 1-3, respectively. Table 3 shows the relationship between clinical parameters and scoring systems. In Table 4, the clinical outcomes of patients was divided into low-risk and high-risk groups for three scoring systems.

Experimental design, materials and methods
The aim of the data article is to evaluate the three scoring systems including clinical Rockall, Complete Rockall and Blatchford for forecasting the clinical outcomes of patients with upper gastrointestinal bleeding (UGIB) This dataset was collected by retrospective descriptive epidemiologic survey which 350 non-cirrhotic patients with UGIB who referred to Ahwaz Imam-Khomeini Hospital for six months. The medical records and required data were extracted through a questionnaire. The questionnaire included demographic data, clinical symptoms, endoscopic findings, and clinical outcomes [1,2]. In addition, the outcomes included re-bleeding, the need for re-endoscopy, surgery and patient mortality. The need for blood transfusion or endoscopy was recorded as an emergency in Table 2 Classification of the patients into low-risk and high-risk groups based on mean value (cut-off). the questionnaire. After Blatchford score were inserted, the clinical Rockall and complete Rockall scored were determined for each patient. The age of over 18 years was regarded as inclusion criterion, while the patients with cirrhosis were excluded from the study. According to Blatchford, clinical Rockall, and complete Rockall scoring systems, patients were evaluated and scored. In order to classify the patients into two low-risk and high-risk groups based on the scoring systems mentioned above, the statistical characteristic curve (ROC) was used which was defined as a cut-off with specific sensitivity and specificity. In order to determine the relationship between the variables and the clinical outcomes, the Chi Square test was used by SPSS.21 software (SPSS Inc. Chicago, IL, USA).