1. Background
Sepsis, defined as a systemic response to severe infection, is the third leading cause of death from infectious diseases after pulmonary infections and AIDS (1). It is a major cause of morbidity and mortality among hospitalized patients and places a substantial burden on healthcare systems (2). Severe sepsis may involve organ dysfunction, impaired organ perfusion, or hypotension and is often accompanied by lactic acidosis, oliguria, or acute alterations in mental status (3).
Septic shock, a severe form of sepsis, occurs when hypotension persists despite adequate fluid resuscitation and is associated with impaired tissue perfusion (4, 5). A variety of microorganisms, including gram-negative and gram-positive bacteria and fungi, can trigger sepsis. Positive culture results are observed in approximately half of patients with sepsis; therefore, early administration of broad-spectrum antibiotics is recommended to improve survival (6).
Early signs of sepsis are often nonspecific and may include fever or hypothermia, chills, tachycardia, tachypnea, nausea and vomiting, hyperglycemia, myalgia, fatigue, proteinuria, tissue hypoxia, leukocytosis, and hyperbilirubinemia. Clinical presentation may vary according to underlying conditions and comorbidities.
2. Objectives
Despite the frequent occurrence of culture-negative sepsis, evidence regarding outcomes in these patients remains limited. Given the critical need for accurate diagnosis and management of sepsis in hospital settings, particularly the role of culture results in guiding treatment, this study aimed to compare mortality rates between patients with culture-negative and culture-positive sepsis admitted to the ICU at Imam Reza Hospital.
3. Methods
3.1. Study Design
This cross-sectional study was conducted in the general ICUs of Imam Reza Hospital, Tabriz, Iran, over a 6-month period from January to June 2024. The study aimed to compare hospital mortality and clinical outcomes between patients with culture-positive and culture-negative sepsis. The minimum required sample size was calculated as 118 participants based on the effect size, 80% statistical power, and a 95% confidence interval. To ensure sufficient power and account for a potential 10% attrition rate, the final number of participants included in the study was 150. The sample size calculation was performed using G*Power software version 3.1, ensuring adequate statistical reliability for the planned analyses.
3.2. Selection of Study Groups
Eligibility criteria were age 18 years or older, provision of informed consent, willingness to cooperate, a specialist-confirmed diagnosis of sepsis, and accessible culture data. Patients who had received antibiotics before culture sampling, those who did not provide consent or were unable to cooperate, and those diagnosed with microbiologically verified fungal, viral, or parasitic infections or tuberculosis were excluded.
3.3. Data Collection
Data were collected from 150 patients with sepsis admitted to General ICUs 7 and 8 at Imam Reza Hospital, Tabriz, Iran, over a 6-month period from January to June 2024. A structured checklist was used to systematically extract data from patients' electronic medical records and ICU charts. Baseline data recorded at ICU admission included demographic characteristics (age and sex), the reason for and timing of hospitalization, comorbidities such as diabetes mellitus, hypertension, and cardiovascular disease, vital signs (blood pressure, heart rate, respiratory rate, and body temperature), and initial laboratory tests, including white blood cell count and C-reactive protein (CRP) levels.
Data from the first day of ICU admission were also collected, including APACHE II scores, administered treatments, interventions such as mechanical ventilation and central venous catheterization, prescribed antibiotics, and microbiological culture results from blood, urine, or other clinically relevant specimens. Culture samples obtained from 2 days before to 2 days after ICU admission were reviewed to identify the causative pathogens associated with sepsis. Patients were followed until hospital discharge or death. The primary outcome was hospital mortality, and secondary outcomes included ICU mortality, duration of mechanical ventilation, and lengths of ICU and hospital stay. Based on culture results, patients were categorized into culture-positive and culture-negative groups, and outcomes were subsequently compared between these groups.
3.4. Statistical Analysis
The Kolmogorov-Smirnov test was used to assess the normality of data distribution. Normally distributed quantitative variables were presented as mean ± SD, whereas nonnormally distributed variables were expressed as median (25th - 75th percentile). Categorical variables were reported as frequency and percentage. Baseline characteristics were compared between the culture-positive and culture-negative groups using the independent t-test or Mann-Whitney U test, as appropriate. The Fisher exact test was used for categorical comparisons. Statistical analyses were performed using SPSS version 26 (SPSS Inc., Chicago, IL, USA), and P < 0.05 was considered statistically significant.
To identify factors associated with hospital mortality, univariate logistic regression analyses were first performed for clinically relevant variables. Variables with P < 0.10 in the univariate analysis, along with variables considered clinically important (age and sex), were subsequently entered into a multivariable logistic regression model. Results were reported as odds ratios (ORs) with 95% confidence intervals (CIs).
3.5. Ethical Considerations
All participants or their legal guardians provided written informed consent after receiving a comprehensive explanation of the study aims and procedures. Participants could withdraw at any time. The study protocol was approved by the Ethics Committee of Tabriz University of Medical Sciences (Ethics Code: IR.TBZMED.REC.1402.643).
4. Results
4.1. Descriptive Results
The mean age of the patients was 49.1 ± 19.7 years, and 112 patients (74.7%) were male. Hypertension was the most common comorbidity, reported in 21 patients (14%), followed by diabetes mellitus in 6 patients (4%). Diffuse axonal injury was the most frequent primary diagnosis, observed in 35 patients (23.3%), followed by intracranial hemorrhage (19.3%) and subdural hematoma (8.0%).
Overall, 86 patients (57.3%) had positive culture results from blood, urine, tracheal secretions, or bronchoalveolar lavage, whereas 64 patients (42.7%) had negative cultures (Figure 1). Among patients with positive cultures, Acinetobacter baumannii was the most frequently isolated organism in tracheal secretion samples (22%), whereas Klebsiella species were the most common pathogens identified in blood and urine cultures. Colistin (51.1%) and meropenem (29.1%) were the most frequently administered antibiotics in culture-positive patients, whereas cefazolin was the most commonly prescribed antibiotic in the culture-negative group (25.0%).
Figure 1.
The occurrence of positive and negative cultures among the patients.
4.2. Demographic Characteristics
As shown in Table 1, the mean age of patients with positive culture results was 47.4 ± 19.9 years, compared with 42.1 ± 19.2 years in the culture-negative group; this difference was not statistically significant (P = 0.102). The proportion of male patients was similar between the groups (74.4% vs. 75.0%; P = 0.545).
Table 1.Clinical Outcomes of Patients with Culture-Positive and Culture-Negative Sepsis a
| Variables | Culture-Negative (n = 64) | Culture-Positive (n = 86) | P-Value b |
|---|---|---|---|
| Demographic characteristics | |||
| Age (y) | 42.1 ± 19.2 | 47.4 ± 19.9 | 0.102 |
| Male sex (%) | 75.0 | 74.4 | 0.545 |
| Vital signs at ICU admission | |||
| Systolic blood pressure (mm Hg) | 115.8 ± 21.6 | 116.4 ± 65.9 | 0.956 |
| Diastolic blood pressure (mm Hg) | 71.3 ± 14.2 | 65.1 ± 15.1 | 0.015 |
| Heart rate (beats/min) | 89.1 ± 17.3 | 88.8 ± 21.3 | 0.925 |
| Respiratory rate (breaths/min) | 18.5 ± 11.3 | 18.6 ± 4.7 | 0.953 |
| Oxygen saturation (%) | 93.1 ± 5.1 | 91.1 ± 9.6 | 0.140 |
| Severity of illness | |||
| APACHE II score | 21.6 ± 6.6 | 24.9 ± 9.6 | 0.030 |
| Laboratory findings at ICU admission | |||
| CRP (mg/dL) | 58.1 ± 44.6 | 83.1 ± 42.1 | 0.002 |
| White blood cell count (× 103/μL) | 10.7 ± 4.6 | 11.8 ± 6.7 | 0.445 |
| Hemoglobin (g/dL) | 12.9 ± 12.4 | 12.9 ± 11.2 | 0.978 |
| Platelet count (× 103/μL) | 210.4 ± 107.1 | 184.1 ± 96.8 | 0.116 |
| Laboratory findings at sepsis diagnosis | |||
| Platelet count (× 103/μL) | 187.5 ± 103.6 | 189.5 ± 103.0 | 0.583 |
| Hemoglobin (g/dL) | 10.9 ± 2.5 | 11.9 ± 2.3 | 0.934 |
| White blood cell count (× 103/μL) | 16.9 ± 9.2 | 13.9 ± 4.5 | 0.137 |
| Patient outcomes | |||
| Discharge; No. (%) | 58 (92.1) | 49 (59) | < 0.001 |
| Mortality; No. (%) | 5 (9.7) | 34 (41) |
a Values are expressed as mean ± SD unless otherwise indicated.
b P-values were calculated using the Mann-Whitney U test for continuous variables and Fisher exact test for categorical variables.
4.3. Vital Signs and Clinical Status
Patients with positive culture results had a significantly lower mean diastolic blood pressure than those in the culture-negative group (P = 0.015). In addition, the mean APACHE II score was significantly higher in the culture-positive group (P = 0.030). No significant differences were observed between the groups in systolic blood pressure, heart rate, respiratory rate, or oxygen saturation (Table 1).
4.4. Laboratory Variables
Laboratory parameters measured at the time of sepsis diagnosis are presented in Table 1. The mean serum CRP level was significantly higher in patients with positive culture results than in those with negative cultures (P = 0.002). No other laboratory variables differed significantly between the two groups.
4.5. Association Between Oxygen Delivery and Positive- or Negative-Culture Sepsis
Modes of oxygen delivery are summarized in Table 2. Endotracheal intubation was more frequent in the culture-positive group (93.0%) than in the culture-negative group (84.4%); however, this difference was not statistically significant (P = 0.194).
Table 2.Methods of Supplemental Oxygen Delivery Across Study Groups a
| Variables | Positive Culture Group (n = 86) | Negative Culture Group (n = 64) | P-Value b |
|---|---|---|---|
| Receiving supplemental oxygen | 0.194 | ||
| Nasal | 1 (1.2) | 0 (0) | |
| Mask | 3 (3.5) | 7 (10.9) | |
| Mask with reserve bag | 2 (2.3) | 3 (4.7) | |
| Intubation | 80 (93.0) | 54 (84.4) |
a Values are expressed as No. (%).
b The P-value was calculated using the chi-square test.
4.6. Length of Hospital Stay
Patients with positive culture results had significantly longer hospital stays (30.1 ± 17.7 days vs. 20.2 ± 9.9 days; P = 0.001) and ICU stays (26.9 ± 17.3 days vs. 15.5 ± 9.7 days; P = 0.001) than patients with negative cultures. Among intubated patients, the duration of mechanical ventilation was also significantly longer in the culture-positive group (23.4 ± 16.1 days vs. 11.1 ± 7.8 days; P = 0.001) (Table 3).
Table 3.Lengths of Hospital Stay, ICU Stay, and Mechanical Ventilation by Study Group a
| Variables | Positive Culture Group (n = 86) | Negative Culture Group (n = 64) | P-Value b |
|---|---|---|---|
| Duration of hospitalization (d) | 30.1 ± 17.7 | 20.2 ± 9.9 | 0.001 |
| ICU admission duration (d) | 26.9 ± 17.3 | 15.5 ± 9.7 | 0.001 |
| Duration of intubation (d) | 23.4 ± 16.1 | 11.1 ± 7.8 | 0.001 |
a Values are expressed as mean ± SD.
b P-values were calculated using the independent samples t-test.
4.7. Patient Outcomes
As shown in Table 1, hospital mortality was significantly higher among patients with positive culture results (41%) than among those with negative cultures (9.7%) (P < 0.001).
4.8. Predictors of Hospital Mortality: Univariate and Multivariable Logistic Regression Analysis
In the univariate logistic regression analysis, culture-positive sepsis was significantly associated with increased hospital mortality (OR = 7.72; 95% CI, 2.81 - 21.19; P < 0.001). Increasing age, a higher APACHE II score, and elevated CRP levels were also significantly associated with mortality in unadjusted analyses.
In the multivariable logistic regression model adjusted for age, sex, APACHE II score, CRP level, and diastolic blood pressure, culture-positive sepsis was no longer independently associated with hospital mortality (adjusted OR = 3.30; 95% CI, 0.93 - 11.69; P = 0.064). Age remained an independent predictor of hospital mortality (adjusted OR = 1.04 per year; 95% CI, 1.01 - 1.07; P = 0.027), whereas the associations of APACHE II score, CRP, sex, and diastolic blood pressure were attenuated after adjustment (Table 4).
| Variables | Unadjusted B | Unadjusted OR | Unadjusted 95% CI | Unadjusted P-Value | Adjusted B | Adjusted OR | Adjusted 95% CI | Adjusted P-Value |
|---|---|---|---|---|---|---|---|---|
| Culture-positive sepsis | 2.04 | 7.72 | 2.81 - 21.19 | < 0.001 | 1.2 | 3.30 | 0.93 - 11.69 | 0.064 |
| Age (per year) | 0.030 | 1.03 | 1.01 - 1.05 | 0.004 | 0.036 | 1.04 | 1.01 - 1.07 | 0.027 |
| Male sex | 0.57 | 1.77 | 0.71 - 4.43 | 0.222 | 0.53 | 4.62 | 0.87 - 25.54 | 0.073 |
| APACHE II score | 0.081 | 1.08 | 1.03 - 1.15 | 0.004 | 0.049 | 1.05 | 0.99 - 1.11 | 0.232 |
| CRP | 0.009 | 1.01 | 1.00 - 1.02 | 0.041 | 0.010 | 1.01 | 0.99 - 1.02 | 0.278 |
| Diastolic blood pressure | -0.013 | 0.99 | 0.96 - 1.01 | 0.272 | -0.008 | 0.99 | 0.96 - 1.03 | 0.655 |
a Results are presented as regression coefficients (B), odds ratios (ORs), 95% confidence intervals (CIs), and P values. Unadjusted estimates represent univariate logistic regression analyses. Adjusted estimates were obtained from a multivariable logistic regression model including culture status, age, sex, APACHE II score, CRP, and diastolic blood pressure.
b P < 0.05 was considered statistically significant.
5. Discussion
Globally, the incidence of sepsis and septic shock has increased over the past decade. The associated morbidity and mortality remain alarmingly high, ranging from 20% to 40% depending on disease severity (1). Over the past 2 decades, various guidelines and protocols have been developed and updated for the management of sepsis and septic shock. These guidelines commonly recommend the early use of broad-spectrum antibiotics, as numerous previous studies have consistently demonstrated that reducing the delay in initiating effective antibiotic treatment can lower mortality in patients (2, 3).
Sepsis is a highly heterogeneous syndrome. Given differences in pathophysiological mechanisms, epidemiological patterns, and therapeutic outcomes, patients with culture-negative sepsis likely have distinct characteristics compared with those with culture-positive sepsis. In this study, 150 patients with sepsis admitted to the ICU were included, and several findings supported these differences.
The present study showed that 42.7% of patients with sepsis had negative culture results. Similar findings were reported in the single-center study by Kim et al., in which approximately 41% of patients with septic shock had negative culture outcomes (4). In the study by Phua et al., 41.5% of patients with sepsis had negative culture results, which is consistent with the present findings (5). Multicenter studies have shown that the percentage of culture-negative severe sepsis among ICU patients varies by region, with reported rates of 28% in North America, 35% in Spain, and 48% in Canada (8, 9). In the United States, nearly half of patients hospitalized for sepsis had negative culture results (10). Taken together, the present study and similar research indicate that patients with septic shock often have negative culture results. A recent worldwide retrospective analysis spanning 10 years corroborated this finding and showed that the proportion of patients with severe sepsis and negative culture results increased by approximately 28% each year (11).
Multiple factors contribute to the increasing frequency of culture-negative results. One important factor is the administration of antibiotics within the 48 hours before testing (12). The widespread accessibility of antibiotics, which are commonly used through self-medication or empirical prescription in clinical practice, frequently results in exposure to these agents before sepsis develops. In addition, sepsis caused by nontraditional pathogens, including viruses and fungi, has increased (13). Infections caused by these organisms may not be accurately detected using conventional culture methods with blood, urine, stool, and pus specimens. For example, sputum cultures often have low detection rates, whereas bronchial lavage increases the likelihood of identifying pathogens (14). Recent studies using polymerase chain reaction (PCR) techniques have shown that DNA from various organisms can be rapidly identified in the blood of patients with sepsis. This method enables the detection of clinically important fungi and bacteria that often remain undetected in traditional blood cultures (15). Patients diagnosed with sepsis who do not have detectable pathogens may nevertheless develop the syndrome due to noninfectious causes, such as metabolic disturbances, inflammatory disorders, adverse drug reactions, or malignancies (16).
Although culture-negative sepsis is frequently observed, studies addressing outcomes in this patient population remain limited. This study evaluated several outcomes, including clinical status, hospital and ICU lengths of stay, duration of mechanical ventilation, and mortality, in the 2 study groups. The findings showed that, although the need for supplemental oxygen was comparable between the groups, patients with culture-positive sepsis had a significantly longer duration of intubation than those with culture-negative sepsis. Consequently, patients with culture-negative sepsis had shorter ICU and hospital stays. Furthermore, hospital mortality was substantially higher among patients with culture-positive sepsis than among those with culture-negative sepsis (41% vs. 9.7%). In the study by Phua et al., findings related to hospital stay duration and mortality were similar to those of the present study. However, after adjustment for confounders such as medical history and patient age, Phua et al. found no significant difference in mortality between the groups (5).
In the present study, after controlling for age, sex, APACHE II score, CRP concentration, and diastolic blood pressure in the multivariable logistic regression analysis, culture-positive sepsis did not remain independently associated with in-hospital mortality. However, age remained an independent predictor of in-hospital mortality. Li et al. conducted a comprehensive analysis with results that differed from those of the present study, particularly regarding hospital mortality and ICU length of stay, as these outcomes were similar between the 2 groups. However, the duration of mechanical ventilation and overall hospital stay were longer in patients with positive cultures, corroborating the outcomes observed in the present study (17). In the study by Kim et al., both hospital mortality and 90-day mortality were analyzed, and no significant differences in mortality rates were observed between the 2 groups (4). In a previous multicenter study by Brun-Buisson et al., which contrasts with the findings of the present study, no significant difference in mortality was observed between 310 patients with culture-negative severe sepsis and 742 patients with culture-positive severe sepsis (19). In a European ICU study, ICU mortality rates were comparable between the 2 groups, at 40% and 39%, respectively (18). In contrast to the present study and most investigations in this area, Gupta et al. reported increased mortality among patients with culture-negative severe sepsis (11). One possible explanation for this finding is the absence of adequate protocols for administering antibiotic treatment to patients with negative cultures.
Variation in findings across studies can generally be attributed to differences in patient demographics, sources of infection, and levels of bacterial resistance to antibiotics. For instance, the present study focused on a younger population (mean age, 41 years compared with 63 years) and found a lower percentage of comorbid conditions (14% vs. 45%) than the retrospective study conducted by Phua et al. Furthermore, infection sites differed notably between the studies. Given that sepsis is a heterogeneous syndrome, infection sites associated with specific pathogens may have distinct clinical characteristics (5). A multicenter prospective study conducted in Japan provided evidence that the clinical characteristics and outcomes of patients with severe sepsis and septic shock vary according to the source of infection (20). A prior retrospective analysis indicated that patients with pulmonary and intra-abdominal infections who had positive cultures had worse outcomes (21).
Furthermore, the specific organism implicated in sepsis can significantly influence the clinical trajectory and prognosis. For example, in the study by Phua et al., among patients with culture-positive results, the presence of Pseudomonas aeruginosa was independently associated with an increased risk of mortality (5). The spectrum of pathogenic organisms varies substantially across hospital settings. In the present study, Acinetobacter was more prevalent than other pathogens. In addition, antibiotic resistance patterns also vary depending on the specific setting analyzed.
5.1. Study Limitations
This study has several limitations. First, sepsis was categorized into 2 main groups based on whether pathogenic microorganisms were identified; however, these categories may encompass multiple diagnostic groups. As discussed, the group without detectable pathogens likely included patients with nonbacterial forms of sepsis and possibly some patients without sepsis. Second, the absence of microbial growth in some cultures may have been attributable to inadequate sample collection. Third, because this study was conducted in a single medical ICU, the generalizability of the findings to other ICU settings may be limited. Finally, given the observational study design, potential unadjusted and unmeasured variables may have influenced the outcomes and interpretation of the findings.
5.2. Conclusions
The present study identified notable differences between patients with culture-negative and culture-positive sepsis. At admission, patients in the culture-negative group had better vital signs, including higher diastolic blood pressure, as well as lower APACHE II scores. In addition, this group had shorter ICU stays, shorter overall hospital stays, and shorter durations of mechanical ventilation. Mortality rates were also lower in the culture-negative group than in the culture-positive group.
