INTRODUCTION
Gastrointestinal system (GIS) bleeding has an important place among patients who applied to the emergency department (1,2). Its frequency is between 36 and 172 per 100,000 people (3). GIS hemorrhages are pathologies that develop suddenly and result in death if proper intervention is not performed. The mortality rate can go up to 30-40% (4).
Bleeding that originates above the ligament of Treitz is defined as upper GIS bleeding, and 80% of bleeding occurs in this area. In 80% of GIS bleeding cases, bleeding stops spontaneously and 20% require intervention. 20% of all GIS bleeding patients, apply to the hospital with recurrent attacks and surgical intervention is planned in 15-30% of these recurrent bleeding (4-7).
Lactate is a biomarker of hypoperfusion that can be measured from arterial blood as well as from venous blood. It has focused on the relationship of morbidity as well as mortality in emergency and intensive care patients and has been a useful biomarker in predicting the risk of death (8,9).
Foreseeing the prognosis in these patients will contribute to clinicians in terms of treatment efficacy. In our study, we determined the diagnostic and prognostic value of lactate level in patients with GIS bleeding.
METHODS
The study was conducted by retrospectively scanning the information of patients in the tertiary education and research hospital emergency department for 2 years, after the approval of the local ethics committee (B.10.1.TKH.4.34.H.GP.0.01/66). Patient data were obtained from the hospital automation system.
Patients’ sex, age, application complaint, known heart and liver diseases, known malignancy, antiaggregant, anticoagulant, and nonsteroidal anti-inflammatory drugs used, arrival arterial blood pressure and pulse rate per minute, pre-existing GIS bleeding presence, blood gas lactate value, the amount of hemoglobin (Hgb) and hematocrit (Hct) in the first hemogram taken, urea and creatinine values, endoscopy and colonoscopy findings, the amount of erythrocyte suspension (ES) need during the patient’s stay in hospital, and mortality were evaluated.
According to the results of the screening of the hospital data processing system, patients who were over 18 years of age who had melena, hematemesis, or hematoquesia on physical examination were determined to have no bleeding other than GI bleeding or were diagnosed with gastrointestinal bleeding by endoscopy and colonoscopy. Patients under the age of 18 years, pregnant patients, and patients with missing physical examination, history, and laboratory data were excluded from the study.
Statistical Analysis
After all the data was collected, the program called SPSS for Windows® 16.0 was used. The conformity of the data to normal distribution was tested with the Kolmogorov-Smirnov test. Continuous data that did not conform to the normal distribution were expressed as median and percentile slices, and those that did not conform were expressed as mean and standard deviation. Student’s t-test was used to compare the data that fit the normal distribution, and the Mann-Whitney U test was used to compare those that did not. The chi-square test was used to compare categorical data, and Fisher’s Exact test was performed where appropriate. The diagnostic value of lactate level for predicting mortality was evaluated by ROC analysis. In the analyses, p<0.05 value was accepted as statistically significant.
RESULTS
In the study 866 patients were evaluated as all GIS bleeding cases without discriminating between upper and lower GIS bleeding. Of these, 294 were excluded because of the lack of blood lactate level at the time of application, and 66 were excluded because of insufficiency in various other data, especially endoscopy and colonoscopy. As a result, 506 patients were included in the study.
In our study, the median age of the patients was 67 (IQR: 53-80) years and 310 (61.3%) were male. Of the patients, 443 (87.5%) had upper GIS bleeding and 63 (12.5%) had lower GIS bleeding. Other demographic data of the patients are shown in Table 1.
The median lactate level in our study was 1.8 mmol/L (IQR: 1.3-2.7); the lactate level of 311 (61.5%) of the cases was above the reference (minimum: 0.5-maximum: 1.6) values. The median lactate level in upper GIS bleeding was 1.8 mmol/L (1.3-2.7), and the median lactate level in lower GIS bleeding was 1.8 mmol/L (1.3-2.4). There was no statistical difference between the groups (p=0.759) (Figure 1).
Three hundred eighty six (76.3%) of the patients needed ES. The median lactate level in patients who required ES was 1.9 mmol/L (IQR: 1.3-2.88), and the median lactate level in patients without ES requirement was 1.6 mmol/L (IQR: 1.1-2.2). The lactate level of patients who were given ES was found to be significantly higher (p<0.001).
Nine of the patients (1.8%) were discharged from the emergency service, while 475 (93.8%) were hospitalized. Twenty-two patients (4.4%) were admitted to the intensive care unit or died in the emergency service. The lactate median level of patients discharged was 1.3 mmol/L (IQR: 1.1-1.7), the lactate median level of patients hospitalized in the service was 1.8 mmol/L (IQR: 1.3-2.7), and the lactate median level of patients hospitalized in intensive care was 2.55 mmol/L (1.7-4.77). There was a statistically significant difference between these 3 groups in terms of lactate levels (Kruskal-Wallis, p=0.008), and in the post hoc evaluation, the number of patients who were admitted to intensive care were significantly higher than that in the other groups.
There was no correlation between lactate level and creatinine, platelet count and forrest classification (p=0.094, p=0.243, p=0.148, respectively). Lactate level was found to have a weak negative correlation with systolic blood pressure, diastolic blood pressure, Hgb, and Htc value (r=-0.174, r=-0.114, r=-0.123, r=-0.124, respectively) ; and a weak positive correlation with pulse rate, urea, amount of ES given, and duration of hospitalization (r=0.185, r=0.172, r=0.201, r=0.093, respectively) (Table 2).
It was determined that 62 (0.3) of the patients died. The lactate median value of patients who died was 2.7 mmol/L (IQR: 1.7-4.2), and the lactate median value of those living was 1.7 mmol/L (IQR: 1.2-2.5). The lactate level of the patients who died was found to be significantly high (p<0.001).
In our study, the area under the curve was found to be 0.714 cm2 (Figure 2). When the cut-off value of lactate is taken as 4 mmol /L, the test performance results in predicting mortality are as follows: sensitivity, 29%; specificity, 91%; positive predictive value, 31%; negative predictive value, 90%; positive likelihood ratio, 3.22; and negative likelihood ratio, 0.78. The sensitivity and specificity values of lactate for different cut-off values are given in Table 3.
DISCUSSION
In our study, the median value of the lactate level was 1.8 mmol/L in accordance with the literature, and in 61.5% of the cases, the lactate level exceeded the reference level. In studies in the literature, similar to our study, it was stated that lactate level increased in patients with GIS bleeding, and this situation developed due to hypoperfusion and hypoxia (10-12). Shrestha et al. (13) stated that a single venous lactate level provides clinically useful information in patients with acute GIS bleeding and is an important parameter in the planning of treatment.
In our study, the mortality rate was found to be 12.3% in line with the literature, and it was shown that this rate increased to 23% according to some study results (14,15). The lactate levels of patients who died were found to be higher, and when the lactate threshold value of 4 mmol/L for mortality was taken, it was observed that they had similar specificity and sensitivity with other studies (15-18). In other studies, they concluded that serum lactate levels are better than prognostic indicators, such as the Rockall or Glasgow-Blatchford systems, physiological triage criteria, such as heart rate, blood pressure, Glasgow coma scale, and respiratory rate (19,20). Although it was determined in our study that high lactate levels contributed significantly to predicting mortality, we believe that it is not reliable to estimate prognosis with lactate level alone because the test performance results are not strong enough. However, when used with other prognosis indicators, we believe that lactate levels is a very useful marker in the follow-up of patients with GI bleeding.
In our study, no difference was found between the upper and lower GIS bleeding lactate levels. Similarly, Kollef et al. (10) also stated that lactate levels increased in all GIS bleeding cases,and lactate levels were similar in upper and lower GIS bleeding. We believe that the lactate level in GIS bleeding is related to hypoxia caused by the amount of bleeding rather than the bleeding focus.
Some studies in the literature have stated that the lactate level is correlated with the severity of the lesion and may indicate rebleeding (11,12). However, in our study, no significant relationship was found between Forrest classification and lactate level. The data to explain the reason of this condition have not been determined in our study, and we believe that factors such as mild and chronic losses in the cases do not give symptoms unless they reach serious rates, and rapid intervention to serious lesions may have led to this result.
Shah et al. (16) reported that there was no relationship between lactate, pulse rate, and Htc in their study in patients with GIS bleeding. Ko et al. (21) stated that high serum lactate level can predict hypotension in patients with GIS bleeding. In their study in GIS bleeding cases, Shrestha et al. (13) stated that patients with high and normal lactate levels had similar heart rate and blood pressures, and patients with high lactate levels had low Hgb. In our study, it was determined that lactate level had negative and weak correlations with blood pressure, Hgb, and Hct, and positive and weak correlations with pulse rate and urea.
Previous studies have shown a relationship between blood transfusion and mortality and length of hospital stay (13,19,22,23). Although Kollef et al. (10) reported in their study that there is a similarity between the lactate level and the need for ES, Ayık et al. (10,24) reported that they did not find a significant relationship between the two in their studies. In our study, it was determined that 76.3% of the patients needed ES, the median value of ES requirement was 3 units, the lactate level of patients with ES requirement was high, and there was a positive correlation between the amount of ES delivered and the lactate level. This may be related to the amount of blood lost from the body, leading to a decrease in the amount of blood flowing to the tissues and thus to hypoxia.
Study Limitations
The main limitation of this study is that it was designed retrospectively. The second important limitation is the deficiencies in patient anamnesis and the fact that the lactate intake time of the patients is unknown. Again, we believe that the fact that the study was single-centered reduces the generalizability, while the relatively high number of patients increases the reliability of the study results.
CONCLUSION
As a result, lactate level is a parameter that can be used in conjunction with other clinical and laboratory markers and is effective in determining prognosis in patients with GIS bleeding. In the cases with high lactate value, more hospitalization time, more ES supplement requirement, and increased risk of hospitalization and death were determined.
Ethics
Ethics Committee Approval: The study was conducted by retrospectively scanning the information of patients in the tertiary education and research hospital emergency department for 2 years, after the approval of the local ethics committee (B.10.1.TKH.4.34.H.GP.0.01/66). Patient data were obtained from the hospital automation system.
Informed Consent: A retrospective study.
Peer-review: Externally peer-reviewed.
Authorship Contributions
Surgical and Medical Practices: H.A.K., G.A., Ö.B., S.E.E., Concept: H.A.K., G.A., S.E.E., Design: H.A.K., G.A., S.E.E., Data Collection or Processing: H.A.K., Ö.B., Analysis or Interpretation: G.A., Literature Search: H.A.K., Ö.B., Writing: H.A.K., G.A., Ö.B., S.E.E.
Conflict of Interest: No conflict of interest was declared by the authors.
Financial Disclosure: The authors declared that this study received no financial support.