Early peripheral perfusion index predicts 28-day outcome in patients with septic shock

Abstract

BACKGROUND:

To investigate the prognostic value of the peripheral perfusion index (PPI) in patients with septic shock.

METHODS:

This prospective cohort study, conducted at the emergency intensive care unit of Peking University People’s Hospital, recruited 200 patients with septic shock between January 2023 and August 2023. These patients were divided into survival (n=84) and death (n=116) groups based on 28-day outcomes. Clinical evaluations included laboratory tests and clinical scores, with lactate and PPI values assessed upon admission to the emergency room and at 6 h and 12 h after admission. Risk factors associated with mortality were analyzed using univariate and multivariate Cox regression analyses. Receiver operator characteristic (ROC) curve was used to assess predictive performance. Mortality rates were compared, and Kaplan-Meier survival plots were created.

RESULTS:

Compared to the survival group, patients in the death group were older and had more severe liver damage and coagulation dysfunction, necessitating higher norepinephrine doses and increased fluid replacement. Higher lactate levels and lower PPI levels at 0 h, 6 h, and 12 h were observed in the death group. Multivariate Cox regression identified prolonged prothrombin time (PT), decreased 6-h PPI and 12-h PPI as independent risk factors for death. The area under the curves for 6-h PPI and 12-h PPI were 0.802 (95% CI 0.742–0.863, P<0.001) and 0.945 (95% CI 0.915–0.974, P<0.001), respectively, which were superior to Glasgow Coma Scale (GCS), Sequential Organ Failure Assessment (SOFA) scores (0.864 and 0.928). Cumulative mortality in the low PPI groups at 6 h and 12 h was significantly higher than in the high PPI groups (6-h PPI: 77.52% vs. 22.54%; 12-h PPI: 92.04% vs. 13.79%, P<0.001).

CONCLUSION:

PPI may have value in predicting 28-day mortality in patients with septic shock.

INTRODUCTION

Septic shock refers to sepsis complicated by severe circulatory and cellular metabolic disorders, and has a significant higher mortality than sepsis alone.[1,2] Early identification and intervention are critical to improve overall outcomes. Currently, systemic circulatory indices, such as blood pressure and heart rate, are used as the evaluation criteria for risk stratification and treatment effects. However, the regulation and compensation mechanisms maintaining hemodynamic consistency are seriously impaired in septic shock, resulting in a mismatch between systemic circulation and microcirculation.[3] At present, clinical indicators for monitoring microcirculation are relatively limited, and noninvasive monitoring indicators are even fewer.[4]

Since adequate microcirculatory perfusion is crucial for improving oxygenation,[3,4] monitoring microcirculatory indices and enhancing the microcirculation of patients should be the primary goal in the management of septic shock. The peripheral perfusion index (PPI), a variable derived from pulse oximeter readings, has been validated to reflect patient microcirculation.[5] It has the advantages of being noninvasive, simple, and providing continuous monitoring,[6] making it increasingly used in clinical practice. However, few studies have investigated the feasibility of using PPI in the assessment of sepsis and septic shock. Therefore, this study aimed to explore the predictive value of PPI in the prognosis of patients with septic shock.

METHODS

Study design and setting

This prospective cohort study, conducted at the emergency intensive care unit of Peking University People’s Hospital, recruited 200 patients with septic shock between January 2023 and August 2023. These patients were divided into survival and death groups based on their 28-day outcomes. Demographic and clinical data were collected. Blood samples were drawn within 24 h of septic shock onset to evaluate complete blood count, bilirubin, creatinine, procalcitonin, and interleukin-6 levels. The Glasgow Coma Scale (GCS), Sequential Organ Failure Assessment (SOFA), and Acute Physiology and Chronic Health Evaluation II (APACHE II) assessments were evaluated. The duration of hospitalization and patient outcomes were recorded for analysis.

Additionally, arterial lactate levels and PPI were measured upon admission to the emergency room and at 6 and 12 h after admission. The rates of lactate clearance and PPI change from 0 to 6 h and 0 to 12 h were calculated. Arterial lactate measurements were uniformly performed using the GEM3500 blood gas analyzer (American Experimental Instruments Company, USA). Lactate clearance rate was calculated using the formula: (initial lactate value – retested lactate value) ÷ initial lactate value × 100%. Similarly, the PPI change rate was determined by: (retested PPI value – initial PPI value) ÷ initial PPI value × 100%.

The study further analyzed individual risk factors associated with mortality using both univariate and multivariate Cox regression analyses. To assess the predictive performance of these factors, receiver operating characteristic (ROC) curves were plotted, and the area under the curves (AUCs) was calculated. Kaplan-Meier survival analysis was then conducted to evaluate survival probabilities, and corresponding survival curves were generated to comprehensively illustrate the findings.

Study participants

This study recruited 200 patients with septic shock, diagnosed and treated according to the International Guidelines for the Management of Sepsis and Septic Shock. Inclusion criteria were: age ≥18 years, admission within 24 h of onset, ICU stay >24 h and complete clinical data. Exclusion criteria were: peripheral vascular disease, Raynaud’s syndrome, extensive pulmonary embolism, septic shock onset ≥24 h prior to admission, prior treatment elsewhere before admission, conditions affecting PPI detection (e.g., finger injuries, hypothermia), acute and chronic heart failure, advanced malignant tumors, and cases with missing or abandoned treatment. All patients received timely and adequate treatment based on the progression of septic shock. This study was approved by the Ethics Review Committee (K202310-12) and conducted in accordance with relevant guidelines and regulations. Written informed consent to participate was obtained from all patients. Initial fluid therapy consisted of lactated Ringer’s solution, normal saline, and other crystalloids, administered at a rate of 5–10 mL/kg per hour. Hemodynamic indices were continuously monitored, with treatment targets set at a mean arterial pressure (MAP) >65 mmHg (1 mmHg=0.133 kPa) and a central venous pressure (CVP) of 8–12 mmHg.

The measurement method of PPI

PPI was measured using a Philips Medical Systems IntelliVue MP5 pulse monitor (Boblingen, Germany) placed on the index finger. The sensor was positioned on the palm, with the limb kept level with the heart. PPI readings were recorded every 30 s for 10 min, resulting in an average of 20 values. Measurements were taken during a hemodynamically stable state, defined as changes in MAP of less than 10% over 5 min and a systolic arterial pressure below 180 mmHg.

Statistical analysis

Data were analyzed using SPSS 27.0 software. Quantitative data are presented as mean ± SD or median with interquartile range (IQR), while categorical data are shown as frequencies and percentages. Differences betwwen groups were assessed using Student’s t test, Mann-Whitney test, χ² test, or Fisher’s exact test as appropriate. Univariate and multivariate Cox regression analyses identified risk factors for septic shock, with significant univariate indicators entered into multivariate analysis. ROC curves were plotted, and AUCs were calculated. Kaplan-Meier survival analysis was conducted, and survival curves were generated. A P-value < 0.05 was considered as statistical significance.

RESULTS

The study included 200 patients, 84 patients in the survival group and 116 patients in the death group. Patient characteristics are detailed in Table 1. The 28-day mortality rate was 58.00%. Patients in the death group were older and severe illness, with lower GCS scores and higher SOFA and APACHE II scores. They also developed liver dysfunction and coagulopathy, requiring higher doses of norepinephrine and larger volumes of fluid replacement to maintain blood pressure. A higher proportion of patients in the death group required mechanical ventilation.

Table 1.

The demographic and baseline clinical characteristics of patients with septic shock

Furthermore, lactate levels in the death group were significantly elevated at admission, 6 h, and 12 h compared to the survival group, with a significant drop in lactate clearance rate (all P<0.001). Additionally, PPI levels were significantly decreased at admission, 6 h, and 12 h in the death group compared to the survival group (all P<0.001), and the improvement in PPI over time was also inferior in the death group. These details are summarized in Table 1.

To investigate the risk factors that affect the 28-day prognosis of patients with septic shock, univariate Cox regression analysis identified several indicators correlating with 28-day mortality. These indicators included age, GCS, SOFA, APHCHE II score, lactate levels at various time points, PPI at different time intervals, clotting parameters (PT, APTT), mechanical ventilation, 24-h norepinephrine dose, and 24-h fluid infusion volume (Table 2). Subsequent multivariate Cox regression analysis highlighted prolonged PT (hazard ratio [HR] 1.028, 95% CI 1.003–1.054, P=0.031), decreased 6-h PPI (HR 0.647, 95% CI 0.418–0.985, P=0.032), and decreased 12-h PPI (HR 0.218, 95% CI 0.140–0.339, P<001) as independent risk factors for death within 28 d in patients with septic shock (Table 3).

Table 2.

Univariate Cox regression analysis of risk factors of 28-day mortality in septic shock patients

Table 3.

Multivariate Cox analysis of risk factors of 28-day mortality in septic shock patients

ROC curve analysis revealed a strong predictive value for both 6-h PPI (AUC=0.802, 95% CI 0.742–0.863, P<0.001) and 12-h PPI (AUC=0.945, 95% CI 0.915–0.974, P<0.001) regarding 28-day mortality in patients with septic shock (Figure 1). Comparisons with GCS and SOFA scores showed the superior predictive ability of 12-h PPI (0.945 vs. 0.864 vs. 0.928) (Figure 1, supplementary Table 1). To verify these findings, optimal cut-off values for 6-h PPI (1.68) and 12-h PPI (1.86) were used for statistical analysis. Kaplan-Meier survival curves indicated significantly higher cumulative mortality in the low PPI patients compared to the high PPI patients (6-h PPI, 77.52% vs. 22.54%; 12-h PPI, 92.04%% vs. 13.79%, P<0.001) (supplementary Table 2, Figure 2).

Figure 1.

The ROC curve of 6-h PPI, 12-h PPI, along with GCS, SOFA and APACHE II scores for predicting the 28-day prognosis in patients with septic shock. GCS: Glasgow Coma Scale; SOFA: Sequential Organ Failure Assessment; APACHE II: Acute Physiology and Chronic Health Evaluation II; PPI: peripheral perfusion index.

Figure 2.

Survival curve of 6-h PPI (cut-off at 1.68) and 12-h PPI (cut-off at 1.86) in low and high PPI patients. PPI: peripheral perfusion index; HR: hazard ratio; CI: confidence interval.

DISCUSSION

Septic shock is a common and severe acute condition. Despite recent advancements in monitoring and treatment methods, the mortality rate remains high.[7-9] During septic shock, the body undergoes an excessive inflammatory response, resulting in damage to vascular endothelial cells,[10] activation of the coagulation system, capillary leakage, and disturbance of systemic and microcirculation. This ultimately leads to inadequate tissue perfusion and hypoxia.[11] The primary function of the microcirculation is to maintain tissue oxygenation, therefore, adequate microcirculatory perfusion is critical for improving oxygenation.[3] Thus, enhancing microcirculation should be the primary goal of anti-shock therapy.[12]

However, the regulatory and compensatory mechanisms for maintaining hemodynamic consistency are severely disrupted during the progression of septic shock, leading to a mismatch between systemic and microcirculatory hemodynamics.[3] Consequently, using systemic improvement to assess dynamic status of the microcirculation has not been justified. A study has shown that using norepinephrine to improve the patient’s systemic circulation does not significantly improve microcirculation.[13] In cases of microcirculatory shock, systemic circulatory indicators cannot effectively predict or replace microcirculatory status.[14,15] The microcirculatory perfusion indicators may be superior to systemic circulation parameters in predicting outcomes of patients with shock.[5] Monitoring peripheral tissue microcirculation not only aids in forecasting the prognosis of shock patients but also signals the potential onset of multiple organ failure in the later stages of the disease. The study by Merdji et al[16] found that patients with peripheral microcirculatory failure experienced significantly longer hospital stay and worse prognosis. Similarly, a multicenter prospective study reported that early tissue hypoperfusion and microcirculatory failure were strongly associated with 90-day all-cause mortality in shock patients.[17] Furthermore, using microcirculatory indicators to guide fluid resuscitation has been shown to significantly reduce fluid volume requirements. These indicators provide a more accurate and dynamic reflection of a patient’s volume status, thereby minimizing the risk of fluid overload and associated mortality.[6,7] Therefore, monitoring microcirculation holds substantial clinical significance and prognostic value.

Several methods and measurements have been employed clinically to monitor dynamic changes in microcirculatory status, including blood lactate levels, mixed central venous oxygen saturation, CO₂ gap, capillary refill time, and skin temperature gradient changes.[18,19] However, these methods often lack objectivity or require advanced monitoring equipment and technology, which may compromise the accuracy, continuity, and universality of the evaluation. Consequently, clinicians may struggle to detect early microcirculatory failure effectively.[15] In contrast, the PPI offers the advantages of being noninvasive, easily obtainable, and capable of continuous monitoring, addressing many of the limitations associated with other methods.[6]

A previous study has identified a PPI value <1.4 as indicative of inadequate tissue perfusion,[20] which may predict the early onset of organ failure and shock. In our study, the PPI values for the death group at 0, 6, and 12 h were consistently below 1.4 (median 0.44, 0.76, and 0.92, respectively), indicating persistent tissue hypoperfusion and microcirculatory dysfunction despite comprehensive resuscitation efforts. This contributed to their poor outcomes. Rasmussen et al[21] demonstrated that PPI values in patients with septic shock effectively predict prognosis, with PPI values <0.5 correlating with increased mortality. Thus, improving PPI values is associated with better outcomes.[22]

Our study showed that the PPI improvement rate was significantly lower in the death group compared to the survival group, indicating that changes in PPI may predict poor prognosis. The Cox multivariate analysis identified prolonged PT, decreased 6-h PPI, and decreased 12-h PPI as independent prognostic factors. The AUCs for 6-h and 12-h PPI in predicting outcomes were both over 0.8, with the 12-h PPI showing superior predictive value compared to GCS and SOFA scores, and comparable to the APACHE II score. Given its convenience and real-time capability, PPI offers faster and more practical prognostic information than traditional scoring systems, making it an effective tool for identifying high-risk septic shock patients and guiding clinical interventions.

Moreover, the present study validated the prognostic value of 6-h and 12-h PPI to predict 28-day outcomes in septic shock patients. By applying cut-off values of 1.68 for 6-h PPI and 1.86 for 12-h PPI, we stratified patients and generated survival curves. The findings revealed that patients with 6-h PPI <1.68 and 12-h PPI <1.86 experienced significantly higher mortality rates at 28 d.

Additionally, the Cox risk model established in this study confirmed that prolonged PT is an independent risk factor for 28-day mortality in septic shock patients. This finding suggests that septic shock patients are more susceptible to early development of sepsis-induced coagulopathy (SIC). The combined impact of coagulopathy and microcirculatory failure contributes significantly to poor prognosis, aligning with conclusions from previous research.

This study has several limitations. It was a single-center investigation with a small sample size, necessitating validation in larger, multi-center cohorts. Additionally, the PPI is influenced by factors such as vasopressors, body temperature, pH, and vascular disease. Despite efforts to ensure baseline homogeneity and control confounding variables through multivariable analysis, the complexities of treating severely ill patients may still introduce biases in PPI measurements.

CONCLUSION

In conclusion, PPI may have value in predicting 28-day mortality in patients with septic shock. It serves as an effective tool for early identification of high-risk individuals and could enhance clinical management strategies. By integrating PPI into routine assessments, healthcare providers may better stratify patient risk and tailor interventions more precisely, ultimately improving patient outcomes in septic shock.