Estudio de las escalas pronósticas de gravedad al ingreso en UCI y su impacto en la sepsis y la mortalidad

Abstract

Infection is one of the most frequent reasons for admission to the ICU and mortality secondary to septic shock is high. In the last years, there is a high interest to get diagnostic tools to predict sepsis and mortality and to optimize healthcare resources. Objective: To carry out a descriptive study of the epidemiological characteristics of patients admitted to the Intensive Care Unit (ICU) at the San Pedro Hospital ( Logroño, Spain) with a diagnosis of infection over a year, and to evaluate the prognostic performance of sepsis and mortality on the qSOFA and SOFA severity scales. In addition, to develop a new scale, called qSOFA PLUS, to provide improved results. Material and Methods: A descriptive, observational and retrospective cohort study. All patients admitted to the ICU during 2013 with a diagnosis of infection were included. A value ≥ 2 points was considered positive for the SOFA and qSOFA scales. For the elaboration of the qSOFA PLUS scale, the cut-off value for procalcitonine (PCT) > 2ng/ml and for lactate > 2 mmol/L was used. A cut-off value for lactate > 4 mmol/L was also considered, as it is a threshold value of extreme severity. Results: 151 patients were identified, of which 107 (9% of the total admissions that year) presented sepsis and 41 died (27.15%). A clear dominant sex was not observed. The overall mean age was 67 years. The 65 years and over age group was the most prevalent and showed the highest rate of mortality. The most frequent septic focus was respiratory (40%). 58.27% of the patients presented one or two risk factors. A total of 8 risk factors were detected that could negatively influence the prognosis. Incidence according to prevalence of these factors were: cardiovascular disease, immunosuppression and diabetes mellitus. Only 4.87% of the patients presented obesity. Cardiovascular disease was the one with the highest rate of mortality (51.21%), followed by immunodeficiency and diabetes mellitus. For the diagnosis of sepsis, the positive predictive value (PPV) of the qSOFA was 92.7% compared to 73.8% in the case of the SOFA scale. To predict mortality, the PPV for qSOFA was 36.5% versus 29.1% for SOFA. 73% of the deceased had serum lactate levels > 2 mmol/L and almost half of them (48.78%) had serum levels > 4 mmol/L. There are no statistically significant differences that justify the inclusion of lactate, in any of its values, in the multivariate qSOFA PLUS model. There are statistically significant differences in the proportion of sepsis in each classification group (chi-square p <0.001) that would justify the use of PCT in the multivariate qSOFA PLUS model. The qSOFA PLUS model, based on the original qSOFA scale and the qualitative PCT value, has an Area under the curve (AUC) value of 0.879 (p <0.001) and a PPV of 92%. The diagnostic accuracy of the qSOFA PLUS scale is 87.8% compared to 83.2% of the qSOFA scale, with a sensitivity rate of 92% and 83%, respectively. Conclusion: 1. The best performing scale for predicting sepsis and mortality outside the ICU was qSOFA. 2. The lactate biomarker was useful as a prognosis of death. However, it was not used in the qSOFA PLUS model, since no statistically significant differences were found to justify its inclusion. 3. The qSOFA PLUS model improved the prognostic performance of the qSOFA scale, but not significantly. However, it could be of practical relevance as a screening method.