A prospective longitudinal study similar to the one performed by Claushuis and colleagues (2016) will be performed in order to further understand the epidemiology and clinical relationship between platelet levels and mortality secondary to septic shock in a different population. The primary objective is to compare the mortality due to septic shock between patients with thrombocytopenia and patients with normal platelet levels in the ICU of the General Hospital of León, Gto. The secondary objectives are to identify the association between mortality due to septic shock and mild, moderate and severe thrombocytopenia in patients admitted to the ICU at 30, 60 and 90 days.

Research questions

Is there an association between thrombocytopenia and mortality due to septic shock in patients admitted to the critical medicine service? Our hypotheses are that:

1. Mortality from septic shock and thrombocytopenia at 30, 60 and 90 days will be higher in patients with thrombocytopenia than in patients normal platelet counts.

Is there an association between the degree of thrombocytopenia and mortality from septic shock in patients admitted to the critical medicine service? Our hypotheses are that:

1. Mortality from septic shock and thrombocytopenia at 30, 60 and 90 days will be higher in patients with mild thrombocytopenia than in patients without thrombocytopenia.
2. Mortality from septic shock and thrombocytopenia at 30, 60 and 90 days will be higher in patients with moderate thrombocytopenia than in patients without thrombocytopenia.
3. Mortality from septic shock and thrombocytopenia at 30, 60 and 90 days will be higher in patients with severe thrombocytopenia than in patients without thrombocytopenia.

Background

Sepsis has a high morbidity and mortality, hence is a challenging medical problem throughout the world. Three therapeutic principles must be taken into account to improve organ dysfunction and survival in sepsis: 1) early and adequate antimicrobial therapy; 2) reestablishment of adequate tissue perfusion; and 3) timely identification and control of the septic foci. Moreover, survival in sepsis depends on the timely identification and appropriate treatment. Progressive advances in sepsis management over the past decades have led to reduced mortality indices. Mortality rates have decline progressively from 43% in 1993 to 37% in 2003, to 29% in 2007; furthermore, reaching a low in 2012 of 18.4%. All in all, in order to establish a global outlook and adequate therapeutic algorithms for sepsis, the factors associated with high mortality rates in septic shock must be identified.

Severe sepsis and septic shock are two clinical entities that have a great impact on intra-hospital morbidity and mortality, as well as on the cost of health care systems; accounting for a large proportion among the causes of admission to the intensive care unit (ICU). In the United States in 2013, septicemia was the hospital condition that accounted for the highest cost, reaching a total cost of $23.7 billion dollars (i.e. 6.2% of total hospitalization costs); followed by osteoarthritis (i.e. $16.5 billion dollars, 4.3% of total hospitalization costs), births (i.e. $13.3 billion, 3.5%), complications arising from medical devices, implants or grafts (i.e. $12.4 billion, 3.3%), and acute myocardial infarction (i.e. $12.1 billion, 3.2%).

The prevalence of sepsis at the ICU across multiple centers in Spain ranges between 6 and 30%. Furthermore, more than 50% of patients with sepsis develop severe sepsis and 25% septic shock. Carrillo-Esper and colleagues (2009) report that in Mexican public and private institutions, 27.3% of 40,957 hospitalizations (i.e. 11,183 cases) developed sepsis. This study which included data from ICUs of 24 out of 32 federal entities reports a mortality secondary to sepsis of 30.4%. In Mexico, the hospitalization costs of patients with sepsis is also a cause of concern, with a total estimated cost of $835 million dollars and an average patient cost of $73,000 dollars.

Sepsis definition

The systemic inflammatory response syndrome or SIRS response is the complex pathophysiological reaction to an insult (e.g. infection, trauma, burns). The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3), defines sepsis as a life-threatening organ dysfunction caused by a dysregulated host response to an infection. Severe sepsis is the sepsis complicated by organ dysfunction, which could progress to septic shock (i.e. sepsis-induced hypotension persisting despite adequate fluid resuscitation). The definitions for sepsis, septic shock, and organ dysfunction have essentially not changed in the past two decades. Despite its global importance, the recognition of sepsis is inadequate. Furthermore, the various manifestations of sepsis make diagnosis difficult, even for experienced physicians. Therefore, a comprehensive definition and clinical guidelines for sepsis management are needed; this would facilitate the clinical identification of sepsis by health professionals, thus improving the diagnostic accuracy and quantification of sepsis.

Thrombocytopenia

Platelets are the smallest blood cells, being only fragments of the megakaryocyte cytoplasm; however, they have a critical role in normal hemostasis and contribute to thrombotic disorders. The normal range of platelets in humans is 150,000-400,000 platelets per microliter of blood. The production of platelets is critically dependent on thrombopoietin, which acts on the differentiation and proliferation of megakaryocyte progenitors and in the maturation of megakaryocytes. Platelets have a short life of up to 10 days. Thrombocytopenia can be classified as: 1) mild, from 100 x 10^9 a 149 x 10^9/L; 2) moderate, from 50 x 10^9 to 99 x 10^9/L; and 3) severe, <50 x 10^9/L. It has been reported that 20% to 50% of patients in the ICU have thrombocytopenia. Aside from hemostatic functions and playing a central role in thrombosis, platelets participate in the immune response. Platelets react to an infectious agent by altering tissue integrity; contributing to the inflammatory cascade, as well as in the elimination process of pathogens and tissue repair. Platelets are activated in patients with SIRS and sepsis. Activation is followed by platelet isolation within the microcirculation and consequently giving rise to a thrombocytopenia. Platelets regulate inflammation and sepsis through multiple mechanisms. Among these mechanisms is innate immunity, where platelets express a receptor for lipopolysaccharide (LPS) Toll-like receptor-4, which contributes to thrombocytopenia through recruitment of neutrophils to the pulmonary system in a systemic response to LPS. Platelets also interact with other leukocytes, including monocytes. The interaction of activated platelets with monocytes induces the nuclear translocation of nuclear factor-kB (NF-kB) and expression of NF-kB dependent inflammatory genes. In addition to direct interactions with leukocytes, platelets contribute to inflammation and immune activation by releasing cytokines and cellular mediators stored in dense and alpha granules. The gold standard for the evaluation of thrombopoiesis is through bone marrow aspiration; however, since bone marrow aspiration is an invasive procedure, it is not frequently performed in clinical conditions such as sepsis.

The association between level of thrombocytopenia (i.e. mild, moderate, and severe) and mortality has been previously studied in sepsis. Claushuis and colleagues (2016), reported the mortality rates of 1,483 consecutive patients admitted to the intensive care unit with sepsis and degrees of thrombocytopenia (i.e. mild, moderate and severe, depending on platelet counts, very low <50 × 10^9/L, intermediate-low 50 × 10^9 to 99 × 10^9/L, low 100 × 10^9 to 149 × 10^9/L) against patients with normal platelet count (i.e. 150 × 10^9 to 399 × 10^9/L). The mortality reported at 30 days was 25.1%, 37.2%, and 54.1% for patients with mild, moderate and severe thrombocytopenia respectively. More studies are needed to confirm the relationship between the platelet levels and septic shock. We decided to conduct a prospective longitudinal study similar to the one performed by Claushuis and colleagues (2016) in our hospital in order to further understand the epidemiology and clinical relationship between platelet levels and mortality secondary to septic shock in a different population.

Complications: septic shock and multiple organ dysfunction

Sepsis can progress to septic shock, multiple organ failure and death if not recognized early. More than 50% of patients with sepsis develop severe sepsis and 25% septic shock; these figures represent 15% of all admissions to the ICU. Septic shock is defined as a septic process associated with circulatory, cellular and metabolic abnormalities; therefore, there is a higher risk of mortality compared to the unique presence of sepsis. Clinically, it includes patients who meet sepsis criteria and who, despite adequate fluid resuscitation, require vasopressors to maintain a mean arterial blood pressure (MAP) above 65mmHg and serum lactate concentration levels above 2 mmol/L. The factors that contribute to septic shock are 1) vasodilation; 2) endothelial dysfunction (i.e. increased permeability due to loss of vascular smooth muscle reactivity secondary to cellular and humoral mediators); 3) hypovolemia; and 4) bilateral ventricular dysfunction.

Multiple organ dysfunction (MOD) refers to progressive organic dysfunction in a severely ill patient; where homeostasis cannot be maintained without intervention. Because of its severity, the organic dysfunction is at one end of the spectrum of severity of the disease. Both infectious conditions (e.g. sepsis, septic shock), and non-infectious conditions (e.g. SIRS for pancreatitis) can present in MOD. MOD can manifest with affection to different systems such as: cardiovascular; pulmonary; hepatic; renal; gastrointestinal; and hematologic, among others. Multiple organ dysfunction can be defined as an increase of two or more points in the Sequential (Sepsis-Related) Organ Function Assessment score (SOFA). It is important to note that the SOFA score is a mark of organic dysfunction; consequently, it does not determine the individual treatment strategies, nor does it predict mortality according to demographic data (e.g. age) or the underlying conditions (e.g. stem cell transplant recipient, postoperative patient). However, the SOFA score helps identify those patients who potentially have a high risk of dying from an infection.

Singer and colleagues (2016) proposed the quickSOFA (qSOFA) scale in order to facilitate the identification of patients at risk of dying from sepsis. The qSOFA scale is a modified version of the SOFA scale, where a score greater than two points is associated with sepsis; consequently, having a poor prognosis. The qSOFA score can be easily calculated as it only has three components. Each component is assigned a point: 1) respiratory rate ≥22 ventilations per minute; 2) alteration of the mental state; and 3) systolic blood pressure ≤100mmHg). Hypoperfusion duration is directly associated with MOD, accounting for a mortality of 70%. According to SOFA score predictions, patients who meet these criteria for septic shock have a higher mortality than patients who do not meet the criteria (i.e. ≥40% versus ≥10%). Raith and colleagues (2017) report that in 182 ICUs in Australia and New Zeland, between 2000 and 2015, 184,875 patients were admitted with a primary admission diagnosis associated with an infectious process. The objective of the aforementioned study was to validate and evaluate the discriminatory capabilities of a two or more points increase in the SOFA score, a two or more increase in SIRS criteria, or a two or more points increase in the qSOFA score among critically ill patients with an infection. The retrospective study reported a greater capacity to predict the intra-hospital mortality of the SOFA scale over the SIRS criteria and qSOFA criteria.

Number of required patients and power calculation

The different sample sizes were calculated to detect statistically significant differences taking as parameters an α = 0.05 and with a statistical power of 0.8 (i.e. 1-β). The work carried out by Claushuis and colleagues (2016) was used to determine the minimum sample size needed to detect statistically significant changes using the categorical variables of mild, moderate and severe thrombocytopenia as previously defined in this protocol at 30, 60 and 90 days of stay in the ICU. The sample size calculator based on proportions of two samples considering the equality of the two extremes (i.e. tails of a Gaussian distribution), was used, available on the web page http://powerandsamplesize.com/. Table 8 summarizes the parameters used to perform these calculations. The parameters are the following: nA, number of patients with normal platelet concentration (i.e. higher than 150,000/μL and lower than 399,000/μL); nB, number of patients with thrombocytopenia; pA, mortality, in percentage, of patients with normal platelet concentration; pB, mortality, in percentage, of patients with thrombocytopenia; k (nA/nB), sampling proportion; N, sample size needed. Taking into account the results, we consider that a sample of more than 30 patients with severe thrombocytopenia is sufficient, while a sample of 105 patients with moderate thrombocytopenia is sufficient. A similar number of patients without thrombocytopenia is needed to make the comparative analysis of mortality.

Inclusion Criteria:

• Patients diagnosed with sepsis (white blood cells >14,000μL) and thrombocytopenia (<150,000/μL).
• Patients who receive care in the ICU of the General Hospital of León, Gto.
• Patients with a complete file with 90-day evolution notes.

Exclusion Criteria:

• Patients diagnosed with cardiogenic, hypovolemic, anaphylactic and neurogenic shock
• Patients older than 80 years or younger than 18 years of age
• Patients referred from another ICU
• Patients with a diagnosis of neoplasia (i.e. hematologic and/or solid tumor)
• Pregnant women
• Patients with liver cirrhosis
• Patients with splenectomy
• Patients with thrombocytosis (i.e.> 400,000/μL) upon admission to ICU
• Patients with previous use of drugs that could produce thrombocytopenia ( e.g. calcium carbonate, acetylsalicylic acid, clopidogrel, dipyridamole, tamoxifen, cisplatin, bexarotene, doxorubicin, and lovastatin.

Elimination Criteria:

• Patients who do not have a follow-up at 30, 60 and 90 days.

Patients will be recruited from the intensive care unit at the General Hospital of León, Guanajuato, Mexico. The sampling technique will be a convenience sampling as this is a regional urban hospital that serves a total population of approximately 1.5 million people.

Abiding by the Declaration of Helsinki, patient anonymity will be guaranteed.