Determinants of fatal outcome in septic shock patients with euthyroid sick syndrome

Abstract

BACKGROUND

Septic shock is a leading cause of mortality in intensive care, and euthyroid sick syndrome (ESS) may influence outcomes. Identifying predictors of fatal outcome in this population is crucial for guiding management.

AIM

To identify clinical and laboratory predictors of 28-day mortality in patients with septic shock and ESS, and to evaluate the prognostic value of markers such as neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), prognostic index (PI), and modified Glasgow prognostic score (mGPS), the vasoactive drug-dependent index (VDI), and shock index (SI).

METHODS

In this prospective observational study, 95 patients with septic shock and ESS admitted between May 2024 and August 2025 were analyzed. Demographic, clinical, and laboratory data were collected at admission and on days 1, 3, and 7. Prognostic markers - NLR, PLR, PI, and mGPS, VDI, and SI were analyzed. Associations with 28-day mortality were evaluated using standard statistical tests.

RESULTS

Of the 95 patients (mean age 61 ± 14.6 years; female/male ratio 52/43), 53 (52%) died. Duration of mechanical ventilation (P = 0.013) and intensive care unit (ICU) length of stay (P = 0.005) were significantly associated with mortality. Acute Physiology and Chronic Health Evaluation II (P = 0.014) and Simplified Acute Physiology Score II (P = 0.001) scores correlated positively with fatal outcome. Predictive laboratory parameters included base excess and free thyroxine (FT4) at admission (P = 0.013, P < 0.001); free triiodothyronine and FT4 on day 1 (P = 0.007, P < 0.001); red blood cells (RBC) and lymphocytes on day 2 (P = 0.027, P = 0.049); and white blood cells, pO₂, thyroid-stimulating hormone, and FT4 on day 3 (all P < 0.05). Variables positively correlated with mortality included neutrophils, lactate, C-reactive protein, and albumin, while pH, pO₂, bicarbonate, RBC, platelets, and thyroid hormones were negatively correlated. SI on day 3 (P = 0.027) and mGPS on day 1 (P = 0.030) were significant prognostic indices.

CONCLUSION

The 28-day mortality among patients with septic shock and ESS was 52%. Duration of mechanical ventilation, ICU stay, severity scores, laboratory parameters, and indices such as SI and mGPS were significantly associated with outcome. The presence of ESS may contribute to adverse prognosis, and combined evaluation of clinical and laboratory factors can improve risk stratification in this high-risk population.

Key Words: Septic shock; Euthyroid sick syndrome; Vasoactive drug index; Inflammatory prognostic scores; Mortality

Core Tip: This prospective observational study identifies clinical and laboratory determinants of fatal outcome in septic shock patients with euthyroid sick syndrome. Recognizing these predictors may help clinicians improve risk stratification and optimize management strategies in critically ill patients.

INTRODUCTION

Septic shock remains one of the most critical and challenging conditions encountered in intensive care units, characterized by persistent hypotension despite adequate fluid resuscitation, leading to tissue hypoperfusion and multi-organ failure[1]. It is a major contributor to mortality among critically ill patients, with studies reporting intensive care unit (ICU) mortality rates ranging from 37.9% in sepsis patients to 63.4% in those with septic shock[2]. Despite advancements in early recognition and management strategies, the fatality rates associated with septic shock remain unacceptably high. The pathophysiology of septic shock involves a complex interplay between the host's immune response and invading pathogens, leading to widespread inflammation, endothelial dysfunction, and microcirculatory failure. These processes result in impaired oxygen delivery to tissues, cellular metabolic disturbances, and ultimately, organ dysfunction[3]. The Sequential Organ Failure Assessment (SOFA) score, which quantifies the degree of organ failure, has been identified as a significant predictor of mortality in septic shock patients. A study found that a SOFA score greater than 12 was associated with a sixfold increased risk of death[4]. In addition to organ dysfunction, other factors such as age, comorbidities, and delayed administration of appropriate antimicrobial therapy have been identified as independent risk factors for mortality in septic shock patients. Early recognition and timely administration of appropriate antimicrobial therapy are crucial in improving patient outcomes[5]. A prospective observational study highlighted that failure to early recognize sepsis and inappropriate antimicrobial treatment were associated with significantly higher mortality rates in pediatric patients[6]. Recent studies have also explored the role of biomarkers and endocrine alterations in predicting outcomes in septic shock. Euthyroid sick syndrome (ESS), characterized by low circulating triiodothyronine (T3) with normal or low thyroxine (T4) and normal thyroid-stimulating hormone (TSH), is frequently observed in critically ill patients, including those with septic shock[7]. ESS has been associated with disease severity and may contribute to adverse outcomes, highlighting the importance of assessing thyroid function in this population[8]. Systemic inflammation and cytokine release suppress deiodinase activity, reducing the peripheral conversion of T4 to T3 and leading to low T3 levels despite normal or low TSH. This hormonal alteration impairs cellular metabolism and reflects the severity of critical illness, thereby linking ESS to worse clinical outcomes in sepsis[9]. Understanding the determinants of fatal outcomes, including both biochemical and endocrine factors such as ESS, is essential for developing targeted interventions aimed at improving patient survival. This study aims to identify and analyze the clinical, laboratory, and therapeutic factors, with particular attention to ESS, associated with fatal outcomes in patients with septic shock admitted to the ICU. By examining these determinants, we seek to provide evidence-based insights that can guide clinical decision-making and enhance patient outcomes in this critical condition.

MATERIALS AND METHODS

We carried out a prospective observational study after receiving approval from the institutional ethics committee (No. 00-03-35-38-14/22). Written informed consent was obtained either directly from the patients or, when this was not possible, from their closest relatives. Recruitment took place consecutively from May 2024 to August 2025 and included all septic shock patients with ESS. Excluding criteria were patients younger than 18 years, those with previous pituitary and thyroid disease, patients taking corticoids and those who declined participation in the study. Definitions According to the surviving sepsis campaign definition, septic shock represents a severe form of sepsis characterized by persistent hypotension that requires vasopressor support to keep the mean arterial pressure (MAP) at 65 mmHg or higher, despite adequate fluid resuscitation, together with a serum lactate concentration above 2 mmol/L[10]. ESS is defined as an alteration in thyroid hormone levels occurring in the absence of any preexisting hypothalamic–pituitary or thyroid gland disorder. The most common pattern involves isolated low T3 levels, followed by combined reductions in T3 and T4 with normal TSH concentrations[11]. All patients were treated following the 2016 Surviving Sepsis Campaign guidelines[10].

Data collection within the first 24 hours of ICU admission, demographic variables (age, sex, and body mass index) as well as severity scores [Acute Physiology and Chronic Health Evaluation II (APACHE-II), Simplified Acute Physiology Score II (SAPS II), and SOFA] were recorded. Mechanical ventilation was documented as present or absent, ICU length of stay was measured in days, and 28-day outcome was categorized as survival (discharge/transfer) or death. All laboratory parameters were measured using the same assay and at consistent time points for all patients, on admission (T0) and subsequently on day 1 (T1), day 3 (T2), and day 7 (T3) of the ICU stay. Tests included a complete blood count with differential, C-reactive protein (CRP), procalcitonin, arterial blood gas analysis with lactate, albumin, TSH, free triiodothyronine (FT3) and free thyroxine (FT4). At the same time points, inflammatory prognostic markers were calculated: Neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, prognostic index, and modified Glasgow prognostic score (mGPS). Each inflammation score was calculated according to Wang et al[12]. The effect of vasoactive therapy was assessed by calculating the vasoactive drug-dependent index and the shock index (SI), which integrate vasopressor dosage with MAP and heart rate.

Statistical analysis

All statistical analyses were performed using SPSS software, version 23.0 (IBM Corp., Armonk, NY, United States). Categorical variables were summarized as counts and percentages and compared using the Pearson χ² test. Continuous variables were presented as mean ± SD or median with range, depending on their distribution; comparisons between continuous variables were performed using the Mann–Whitney U test when data were not normally distributed. Linear correlations between continuous variables and mortality were assessed using the Pearson correlation coefficient (r). A P value < 0.05 was considered statistically significant.

RESULTS

During the sixteen-month study period, 135 patients were admitted to the ICU with septic shock. Of the 95 patients included in the analysis, 53 (52%) died, representing the overall fatality rate. The mean age of the patients was 61.5 years. Duration of mechanical ventilation (P = 0.013) and length of stay (P = 0.005) were significantly associated with fatal outcome, whereas other variables-including demographic factors, mortality scores, and endotracheal intubation-showed no significant association. Positive correlations with fatal outcome were observed for APACHE and SAPS scores, as well as for the duration of mechanical ventilation (Table 1). Among the laboratory variables, the following showed statistical significance: Base excess (BE) (P = 0.013) and FT4 (P = 0.00) at T0; FT3 and FT4 at T1 (P = 0.007 and P = 0.000), red blood cells (RBC), lymphocytes, FT3 and FT4 at T2 (P = 0.027, P = 0.049, P = 0.028 and P = 0.021, respectively); and white blood cells (WBC), RBC, pO₂, TSH and FT4 at T3 (P = 0.005, P = 0.045, P = 0.027, P = 0.026 and P = 0.011, respectively). Positive correlations with fatal outcome were observed for neutrophils at T0; pH and lactate at T1; neutrophils, lymphocytes, BE and lactate at T3; and WBC, pCO₂, and CRP at T3. Negative correlations were found for bicarbonate and BE at T1; pH, pO₂, bicarbonate, FT3 and FT4 at T2; and RBC, platelets, pH, pO₂, albumin, FT3 and FT4 at T3 (Table 2). SI was significantly associated with fatal outcome (P = 0.027) at T3, while the mGPS was significant at T1 (P = 0.030). Overall, SI and inflammatory prognostic scores mostly showed positive correlations with fatal outcome (Table 3).

Table 1 Demographic parameters, mortality scores and duration parameters.

Variable		n (%)/mean ± SD	P value1	r	P value2
Age		61.5 ± 14.6	0.594	-0.186	0.071
Gender	Male/female	52/43 (54.73/40.85)	0.464	0.100	0.333
BMI		26.2 ± 5.4	0.107	0.035	0.736
Mortality scores
APACHE		21.7 ± 7.6	0.345	0.252	0.014
SAPS II		57.07 ± 15.19	0.338	0.328	0.001
SOFA		10.93 ± 3.63	0.700	-0.165	0.109
Duration indicators
ETI, yes/no		84/11 (79.80/10.45)	0.407	0.146	0.159
MV		9.21 ± 15.12	0.013	0.280	0.006
LOS		13.63 ± 16.05	0.005	-0.098	0.342

¹χ² test.

²Pearson’s correlation coefficient test.

r: Correlation coefficient; BMI: Body mass index; APACHE II: Acute Physiology and Chronic Health Evaluation II; SAPS II: Simplified Acute Physiology Score II; SOFA: Sequential Organ Failure Assessment; ETI: Endotracheal intubation; MV: Mechanical ventilation; LOS: Length of stay.

Table 2 Correlation of laboratory findings with fatal outcome.

Variable	T0, day 0			T1, day 1			T2, day 3			T3, day 7
	P value1	r	P value2	P value1	r	P value2	P value1	r	P value2	P value1	r	P value2
WBC	0.399	-0.070	0.499	0.483	-0.116	0.264	0.195	-0.094	0.367	0.005	0.208	0.044
RBC	0.161	0.117	0.258	0.095	0.628	0.050	0.027	0.062	0.553	0.045	-0.249	0.015
PLT	0.216	0.106	0.309	0.152	0.107	0.301	0.444	0.147	0.155	0.409	-0.270	0.008
NEU	0.458	0.244	0.017	0.626	-0.077	0.459	0.111	0.300	0.003	0.120	0.304	0.003
LYM	0.758	0.150	0.148	0.417	0.112	0.279	0.049	0.280	0.006	0.052	0.121	0.242
pH	0.467	-0.109	0.295	0.123	0.401	0.000	0.122	-0.424	0.000	0.063	-0.451	0.000
pO2	0.575	0.096	0.353	0.380	-0.036	0.739	0.249	-0.335	0.001	0.027	-0.249	0.015
pCO2	0.265	-0.171	0.097	0.365	-0.044	0.671	0.413	-0.179	0.082	0.916	0.232	0.024
HCO3	0.069	0.158	0.127	0.090	-0.251	0.014	0.567	-0.235	0.002	0.150	0.113	0.276
BE	0.013	0.257	0.012	0.144	-0.285	0.005	0.663	0.289	0.004	0.491	0.194	0.060
Lactate	0.541	-0.198	0.055	0.796	0.261	0.011	0.058	0.219	0.033	0.217	-0195	0.059
Albumin	0.556	-0.027	0.798	0.113	0.007	0.943	0.074	-0.236	0.021	0.051	-0.337	0.001
CRP	0.600	-0.006	0.953	0.932	-0.152	0.141	0.291	-0.150	0.146	0.364	0.372	0.000
PCT	0.372	-0.044	0.671	0.400	-0.022	0.831	0.106	-0.145	0.161	0.493	-0.097	0.348
TSH	0.587	-0.069	0.508	0.248	0.111	0.165	0.079	0.095	0.359	0.026	-0.067	0.548
FT3	0.089	0.039	0.706	0.007	-0.102	0.326	0.028	-0.236	0.021	0.198	-0.249	0.015
FT4	0.000	-0.084	0.420	0.000	0.006	0.951	0.021	-0.334	0.001	0.011	-0.331	0.001

¹Mann–Whitney U test.

²Pearson’s correlation coefficient test.

r: Correlation coefficient; WBC: White blood cells; RBC: Red blood cells; PLT: Platelets; NEU: Neutrophils; LYM: Lymphocytes; pCO₂: Partial pressure of carbon dioxide; pO₂: Partial pressure of oxygen; HCO₃: Bicarbonate; BE: Base excess; CRP: C-reactive protein; PCT: Procalcitonin; TSH: Thyroid-stimulating hormone; FT3: Free triiodothyronine; FT4: Free thyroxine.

Table 3 Correlation of the effects of vasoactive agents and inflammatory prognostic scores with fatal outcome.

Variable	T0, day 0			T1, day 1			T2, day 3			T3, day 7
	P value1	r	P value2	P value1	r	P value2	P value1	r	P value2	P value1	r	P value2
Effects of vasoactive agents
VDI	0.389	-0.125	0.229	0.074	-0.122	0.229	0.814	-0.128	0.217	0.842	-0.122	0.237
SI	0.206	-0.155	0.134	0.080	-0.197	0.134	0.027	-0.182	0.078	0.000	0.268	0.009
Inflammatory prognostic scores
NLR	0.712	0.264	0.010	0.689	0.259	0.010	0.236	0.281	0.006	0.176	0.337	0.001
PLR	0.351	-0.160	0.120	0.367	-0.193	0.120	0.200	0.214	0.037	0.659	-0.145	0.162
mGPS	0.496	0.060	0.566	0.030	-0.224	0.566	0.128	-0.106	0.306	0.308	-0.108	0.298
PI	0.328	0.220	0.032	0.305	0.286	0.005	0.100	0.239	0.019	0.327	0.283	0.006

¹Mann–Whitney U test and χ² test for modified Glasgow prognostic score.

²Pearson’s correlation coefficient test.

r: Correlation coefficient; VDI: Vasoactive drug-dependent index; SI: Shock index; NLR: Neutrophil-lymphocyte ratio; PLR: Platelets-lymphocyte ratio; mGPS: Modified Glasgow prognostic score; PI: Prognostic index.

DISCUSSION

This prospective observational study of 95 patients with septic shock demonstrated a high case fatality rate of 52%, which is comparable to previously reported mortality rates in similar cohorts[13]. Our analysis identified several clinical, laboratory, and prognostic markers associated with outcome, highlighting the multifactorial nature of mortality in septic shock. The duration of mechanical ventilation and length of ICU stay were significantly longer among non-survivors, suggesting that prolonged organ support reflects both disease severity and limited recovery potential. These findings are consistent with prior studies indicating that mechanical ventilation not only serves as a therapeutic measure but also functions as a marker of critical illness and poor prognosis when extended[14,15]. Longer ICU stays among non-survivors may reflect refractory shock, complications, or multi-organ dysfunction. Severity scoring systems such as APACHE II and SAPS II showed strong positive correlations with outcome. Their prognostic value has been extensively validated, and our results confirm their role as reliable indicators of illness severity in septic shock[16]. However, while these scoring systems provide valuable risk stratification, their complexity and need for multiple variables may limit real-time use at the bedside. This underscores the potential value of simpler indices, such as SI and mGPS, which also emerged as significant predictors in our cohort. Laboratory markers demonstrated notable associations with mortality. Abnormalities in acid–base balance (low pH, reduced bicarbonate, and base excess) and gas exchange (low pO₂) were associated with adverse outcomes, consistent with the pathophysiology of septic shock where hypoperfusion, anaerobic metabolism, and impaired oxygen delivery are central features[17]. Elevated lactate further supported the presence of tissue hypoxia and metabolic derangements in non-survivors, in line with its well-established role as a prognostic marker[18]. Markers of systemic inflammation also correlated with poor outcome. Increased neutrophil counts and elevated CRP levels in non-survivors were associated with mortality, reflecting the intensity of the inflammatory response and potential dysregulation of host defense mechanisms[19]. In contrast, sepsis-induced lymphopenia has been associated with worse outcomes, including secondary infections, multi-organ failure, and mortality, suggesting that immune suppression plays a key role in septic shock progression[20]. Platelet reduction was also associated with fatal outcome, which may reflect disseminated intravascular coagulation and microvascular dysfunction frequently seen in septic patients[21]. Interestingly, our results identified thyroid hormone alterations as prognostic indicators. Non-survivors more frequently exhibited low FT3 and FT4 and altered TSH values during follow-up, consistent with the pattern of ESS[22]. Previous studies have shown that reductions in thyroid hormones are common in critically ill patients and correlate with illness severity and outcomes[23,24]. In this context, thyroid dysfunction in sepsis may contribute to adverse clinical progression by decreasing T3-dependent mitochondrial energy production, impairing cardiovascular performance and modulating immune function, thereby exacerbating cellular metabolic failure and multiorgan dysfunction[25]. Moreover, dynamic changes in thyroid hormones over the course of critical illness, rather than single measurements alone, have been shown to carry prognostic significance[26]. In our cohort, the predictive role of thyroid hormones persisted across several time points, reinforcing their potential as biomarkers of systemic dysfunction in septic shock[27]. Among the composite prognostic indices analyzed, SI on day 3 and mGPS on day 1 were significantly associated with mortality. Elevated SI likely reflects persistent hemodynamic instability despite resuscitation efforts[28], while the early prognostic value of mGPS underscores the importance of systemic inflammation and nutritional status in the initial phase of septic shock[29]. Nevertheless, interpretation of thyroid hormone alterations in sepsis should account for potential confounders, such as variations in nutritional status, exposure to systemic corticosteroids, and vasopressor administration, all of which can independently affect circulating hormone levels and peripheral metabolism[30].

Our study has several strengths. The prospective design allowed systematic data collection at multiple time points, enabling dynamic assessment of prognostic factors rather than relying solely on admission values. By integrating demographic, clinical, laboratory, and prognostic indices, we provided a comprehensive analysis of determinants of outcome in septic shock. Importantly, all patients were managed according to standardized ICU protocols, reducing variability in care. Nevertheless, limitations must be acknowledged. This was a single-center study with a modest sample size and lack of hormonal replacement intervention analysis which may restrict the generalizability of our findings and limit the statistical power to detect weaker associations. Finally, as with all observational studies, causality cannot be established, and associations should be interpreted cautiously. Despite these limitations, the clinical implications of our results are important.

CONCLUSION

The identification of both traditional (severity scores, lactate, acid–base disturbances) and novel (thyroid hormones, mGPS, SI) predictors of outcome suggests that a combined evaluation of multiple domains may improve risk stratification in septic shock. In summary, our study demonstrates that mortality in septic shock is driven by a complex interplay of organ dysfunction, systemic inflammation, hemodynamic instability, and endocrine alterations. Duration of organ support, severity scores, acid–base imbalance, inflammatory markers, and thyroid hormone abnormalities were all associated with adverse outcomes. Notably, simple indices such as SI and mGPS demonstrated clear prognostic value and could be leveraged within clinical risk assessment tools to facilitate early identification and targeted management of high-risk patients.

ACKNOWLEDGEMENTS

The authors would like to thank the staff of the Department of Anesthesia, Resuscitation and Critical Care for their support in patient management and data collection.