Article Text

Download PDFPDF

Moderate to severe hyperphosphataemia as an independent prognostic factor for 28-day mortality in adult patients with sepsis
  1. Dong-Hyun Jang1,
  2. You Hwan Jo1,2,
  3. Jae Hyuk Lee1,
  4. Joonghee Kim1,
  5. Seung Min Park1,
  6. Ji Eun Hwang1,
  7. Dong Keon Lee1,
  8. Inwon Park1,
  9. Che Uk Lee1,
  10. Sang-Min Lee1
  1. 1 Emergency Medicine, Seoul National University Bundang Hospital, Seongnam, The Republic of Korea
  2. 2 Emergency Medicine, Seoul National University College of Medicine, Seoul, The Republic of Korea
  1. Correspondence to Dr You Hwan Jo, Emergency medicine, Seoul National University Bundang Hospital, Seongnam 13620, the Republic of Korea; emdrjyh{at}gmail.com

Abstract

Background Ischaemic tissue injury caused by tissue hypoperfusion is one of the major consequences of sepsis. Phosphate concentrations are elevated in ischaemic tissue injury. This study was performed to investigate the association of phosphate concentrations with mortality in patients with sepsis.

Methods This was a retrospective cohort study of patients with sepsis conducted at an urban, tertiary care emergency department (ED) in Korea. Patients with sepsis arriving between March 2010 and April 2017 were stratified into four groups according to the initial phosphate concentration at presentation to the ED: group I (hypophosphataemia, phosphate <2 mg/dL), group II (normophosphataemia, phosphate 2–4 mg/dL), group III (mild hyperphosphataemia, phosphate 4–6 mg/dL), group IV (moderate to severe hyperphosphataemia, phosphate ≥6 mg/dL). Multivariable Cox proportional hazard regression analyses were performed to evaluate the independent association of initial phosphate concentration with 28-day mortality.

Results Of the 3034 participants in the study, the overall mortality rate was 21.9%. The 28-day mortality rates were group I (hypophosphataemia) 14.6%, group II 17.4% (normophosphataemia), group III (mild hyperphosphataemia) 29.2% and group IV (moderate to severe hyperphosphataemia) 51.4%, respectively (p<0.001). In the multivariable analyses, patients with severe hyperphosphataemia had a significantly higher risk of death than those with normal phosphate levels (HR 1.59; 95% CI 1.23 to 2.05). Mortality in the other groups was not significantly different from mortality in patients with normophosphataemia.

Conclusions Moderate to severe hyperphosphataemia was associated with 28-day mortality in patients with sepsis. Phosphate level could be used as a prognostic indicator in sepsis.

  • infection
  • death/mortality
  • emergency department
  • intensive care

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What is already known on this subject?

  • Ischaemic injury caused by tissue hypoperfusion is one of the major consequences of sepsis.

  • Phosphate concentrations are elevated in ischaemic tissue injury. Elevated phosphate might be predictive of sepsis mortality

What this study adds?

  • In this retrospective, single centre study of patients with sepsis seen in the emergency department, hyperphosphataemia was associated with significantly greater 28-day mortality compared with those without hyperphosphataemia, suggesting that phosphate could be used as a prognostic marker.

Introduction

Sepsis may result in life-threatening organ dysfunction, which is caused by the infectious organism and dysregulated immune responses of the host.1 2 Although many advances have been made in the treatment of sepsis, sepsis still has high morbidity and mortality.3 4 Identifying factors predictive of mortality is the first step to recognising patients whose condition may worsen and who may need appropriate management to reduce mortality.

Phosphorus is involved intracellular energy exchange and mineral metabolism. Phosphorus binds to oxygen in the form of phosphate.5 Approximately 80%–85% of phosphate is stored in the bone and teeth, and the remainder is primarily in the intracellular space. Phosphate is excreted into urine through the kidneys; thus, deterioration of renal function could lead to an increase of phosphate concentration in the body. Phosphate concentrations may also increase when intracellular phosphate is excreted into the extracellular space as a result of ischaemic tissue damage.6 Several studies have reported that phosphate concentrations are associated with prognosis in diseases that cause ischaemic injury of tissues, such as intestinal ischaemia, coronary heart disease, critical limb ischaemia and cardiac arrest.6–10 In addition, the higher the phosphate concentration, the worse the observed outcome, indicating that the phosphate concentration may reflect the severity of the disease. These results also suggest that the elevation of the phosphate concentration could reflect the extent of ischaemic tissue injury.

In sepsis, circulatory tissue hypoperfusion induces cytopathic hypoxia at the cellular level and results in ischaemic tissue injury and organ dysfunction.11 12 Therefore, we hypothesised that the phosphate concentration, which is a marker of ischaemic tissue injury, could be used to predict mortality in patients with sepsis. We performed this study to investigate the association between initial phosphate concentration obtained at emergency department (ED) presentation and 28-day mortality in patients with sepsis.

Methods

Study population

This study was a retrospective analysis of adult patients (18 years old or older) with sepsis or septic shock who received medical care at an urban, tertiary care ED between March 2010 and April 2017 and had a phosphate level drawn. Patients for whom phosphate concentrations were not measured on admission were excluded. Patients or legal representatives who refused treatment for sepsis were also excluded.

Because the definition of sepsis was revised in 2016, inclusion criteria were changed during the study. From March 2010 to February 2016, we used the Second International Sepsis Definition Consensus to include patients with severe sepsis and septic shock; severe sepsis was defined as sepsis with organ dysfunction, and septic shock was defined as sepsis with hypotension despite adequate fluid resuscitation.1 For patients arriving from March 2016 to April 2017, patients with sepsis and septic shock were included based on the sepsis-3 definition.2 In the sepsis-3 definition, patients with the sequential organ failure assessment (SOFA) score of 2 or greater than 2 were diagnosed as sepsis.

The ED has an annual patient census of approximately 90 000. Patients received care according to the international guidelines for management of sepsis.4 13 In the absence of any contraindications, fluid resuscitation with 30 mL/kg of crystalloids was performed, cultures were drawn and antibiotics administered as soon as possible after arrival at the ED. Medications such as vasoactive agents, insulin and corticosteroids were administered according to the discretion of the treating physician, and transfusions were also performed if necessary. Serum phosphate concentration was included in the routine serum chemistry panel and measured at the central laboratory of our institution. The results could usually be obtained within 1 hour.

Data collection

Baseline characteristics, including age, sex, comorbidities and presumed or confirmed site of infection, were obtained with standardised data collection forms.14 Haemodynamic variables and laboratory data, such as arterial blood gas analysis, complete blood count, serum chemistry and coagulation, were collected. From this data, the Acute Physiology and Chronic Health Evaluation II (APACHE II) score was calculated.

The primary outcome of this study was 28-day mortality after admission to the ED. If the patients were discharged before 28 days, a structured telephone follow-up questionnaire was administered by a research nurse to determine if death had occurred. Patients who were lost to follow-up were excluded from the cohort.

Statistical analysis

Continuous variables were examined with the Shapiro-Wilk test for the normality of the distribution and are expressed as the mean and SD or the median (IQR). Student’s t-test or the Mann-Whitney U test was performed for comparisons of continuous variables between survivors and non-survivors. Categorical variables were described as a number with percentage, and were compared using the χ2-test or Fisher exact test, as appropriate.

Patients were divided into 28-day survivors and non-survivors. A restricted cubic spline curve was used to determine the relationship between the phosphate concentration and the 28-day mortality (figure 1). The adjusted OR for mortality was lowest at approximately 2.5 mg/dL. The mortality rate was not significantly different at lower phosphate concentrations. However, mortality rate increased with phosphate concentrations above 2.5 mg/dL. To evaluate the non-linear relationship in the lower phosphate concentration as well as the positive correlation in the higher phosphate concentration, we divided patients into four groups according to the phosphate concentration: group I (hypophosphataemia, phosphate <2 mg/dL), group II (normophosphataemia, phosphate 2–4 mg/dL), group III (mild hyperphosphataemia, phosphate 4–6 mg/dL), group IV (moderate to severe hyperphosphataemia, phosphate ≥6 mg/dL). Patients with hyperphosphataemia were divided into two groups to evaluate the differences as the phosphate concentration increased. Group III were patients with mild hyperphosphataemia, and group IV were patients with moderate to severe hyperphosphataemia.

Figure 1

Restricted cubic spline plot of mortality according to the phosphate concentration. The model fitted with five knots and adjusted for age, sex, serum calcium, serum creatinine, APACHE II score and the site of infection. The dark line indicates the adjusted ORs for 28-day mortality, and the grey shade area represents the 95% CI. APACHE II, Acute Physiology and Chronic Health Evaluation II.

One-way analysis of variance or the Kruskal-Wallis test was performed for comparisons of continuous variables between four groups. Categorical variables were compared using the χ2 test or Fisher’s exact test, as appropriate. The Bonferroni correction method was used for post hoc analysis, and statistical significance was determined using the adjusted p value.

The survival curves of the four phosphate level groups were plotted using the Kaplan-Meier method, and the log-rank test was performed. Multivariable Cox proportional hazard regression analysis was performed to evaluate the independent association of phosphate concentrations with 28-day mortality, and the results were expressed as HRs and 95% CIs. Age, sex, APACHE II score, the site of infection and variables considered to be associated with phosphate homeostasis, such as respiratory rate, serum creatinine and calcium, were included in the analysis.

The lactate concentration is widely used as a marker of ischaemic tissue injury in many clinical situations.15–17 Therefore, we also assessed the association of the phosphate concentration with the lactate concentration to investigate whether the phosphate concentration could reflect the degree of ischaemic injury caused by tissue hypoperfusion. Pearson correlation coefficients were assessed to investigate the correlation between lactate and phosphate concentrations.

All data processing and statistical analyses were performed using R-package, V.3.5.1 (R Foundation for Statistical Computing, Vienna, Austria). A two-tailed p<value less than 0.05 was considered statistically significant.

Results

Characteristics of the study population

During the study period, a total of 3173 patients met the definition of sepsis. Of these patients, 117 were excluded because they refused treatment, and 22 were excluded because the initial phosphate concentration was not measured. Therefore, 3034 patients were included in the final analysis (figure 2).

Figure 2

Flow chart of the study population group I (hypophosphataemia), phosphate <2 mg/dL; group II (normophosphataemia), phosphate 2–4 mg/dL; group III (mild hyperphosphataemia), phosphate 4–6 mg/dL; group IV (moderate to severe hyperphosphataemia), phosphate ≥6 mg/dL.

The median age of study participants was 74 (IQR, 65–81) years, 1745 (57.5%) were male, and the overall mortality rate was 21.9% (table 1). The non-survivors were older and included more males than the survivors. The non-survivors had higher frequencies of chronic liver disease and underlying malignancy than the survivors. The frequency of lung infection was higher in the non-survivors than in the survivors, whereas the frequency of urogenital infection was higher in the survivors than in the non-survivors. The median phosphate concentrations were higher in the non-survivors than the survivors, 3.7 (IQR, 2.8–5.1) mg/dL vs 3.1 (IQR, 2.4–3.9) mg/dL (p<0.001).

Table 1

Baseline characteristics of patients according to 28-day mortality

Patients in group I (hypophosphataemia) and IV (moderate to severe hyperphosphataemia) were younger than those in group II and III, and sex was not different between the four groups (table 2). The frequency of comorbidities was significantly different between the groups and the rates of chronic kidney disease and renal replacement therapy were highest in the group IV. The mean arterial pressure and body temperature were lower and the respiratory rate was higher in the group IV than the other groups. Blood urea nitrogen and creatinine concentrations were higher in group III and group IV than in the other two groups, and they were highest in the group IV. The APACHE II score was also higher in the group III and group IV than in the other groups.

Table 2

Baseline characteristics of patients according to the serum phosphate concentration

Phosphate concentrations and mortality

The 28-day mortality rates of group I (hypophosphataemia), group II (normophosphataemia), group III (mild hyperphosphataemia) and group IV (moderate to severe hyperphosphataemia) were 14.6%, 17.4%, 29.2% and 51.4%, respectively (p<0.001), and the mortality rate increased with increasing the phosphate concentrations (table 2). The Kaplan-Meier curve with the log-rank test showed higher mortality in the group III (mild hyperphosphataemia) and group IV (moderate to severe hyperphosphataemia) than in the other groups (figure 3, p<0.001). The mortality rate was highest in group IV, followed by group III, and there was no significant difference between group I and group II.

Figure 3

Kaplan-Meier curve of the three groups divided by phosphate concentration. Group I (hypophosphataemia), phosphate <2 mg/dL; group II (normophosphataemia), phosphate 2–4 mg/dL; group III (mild hyperphosphataemia), phosphate 4–6 mg/dL; group IV (moderate to severe hyperphosphataemia), phosphate ≥6 mg/dL.

In the multivariable Cox proportional hazards regression analysis, age, serum calcium, APACHE II score and urogenital infection were independently associated with 28-day mortality (table 3). Group IV (moderate to severe hyperphosphataemia) had an approximately 60% higher risk of death than group II (normophosphataemia) during the 28-day period (HR 1.59; 95% CI 1.23 to 2.05). Group I (hypophosphataemia) and the group III (mild hyperphosphataemia) did not have a greater risk of death as compared with group II.

Table 3

Multivariable COX regression analysis for 28-day mortality

Lactate and phosphate concentrations

The median lactate concentration among all patients was 2.8 (1.6–4.9) mmol/L. The lactate concentrations in group III (mild hyperphosphataemia) and group IV (moderate to severe hyperphosphataemia) were 3.2 (1.7–5.7) mmol/L and 7.0 (3.2–12.8) mmol/L, respectively, and they were significantly higher than that in the other groups (p<0.001, table 2). The lactate concentration had a positive correlation with the phosphate concentration; Pearson correlation coefficient was 0.35 (figure 4).

Figure 4

Correlation between serum phosphate and lactate concentrations.

Discussion

In this study of ED patients with sepsis, the initial serum phosphate concentration was higher in those who did not survive within 28 days after ED presentation than those who did survive. Patients with mild hyperphosphataemia and moderate to severe hyperphosphataemia had a higher mortality rate than those with hypophosphataemia or normophosphataemia in univariable analysis and in Kaplan-Meier survival analysis. In multivariate analysis, moderate to severe hyperphosphataemia (group IV), but not mild hypophosphataemia was an independent prognostic factor of mortality. Thus, hyperphosphataemia appears to be associated with mortality in a dose-dependent manner.

Previous studies have reported that the phosphate concentration is elevated in tissue ischaemia-induced diseases. Experimental studies on intestinal ischaemia reported that the serum phosphate concentration increased as ischaemia progressed, and the degree of elevation of the phosphate concentration was related to the degree of intestinal ischaemic injury.6 8 Zettervall et al 10 reported that a higher phosphate concentration was associated with higher mortality in patients with critical limb ischaemia, and hyperphosphataemia also was associated with amputation-free survival. In patients with cardiac arrest, a high phosphate concentration after successful resuscitation is associated with poor outcomes.9 In patients with sepsis, Miller et al 18 reported that a high phosphate concentration was associated with a poor prognosis. However, this study was performed in an intensive care unit and included patients with mechanical ventilation. In addition, the time-weighted phosphate concentration using all phosphate measurements obtained during the intensive care unit stay was used for the analysis. In contrast, we included patients with sepsis or septic shock in an ED and used the initial phosphate concentration for the analysis, so the characteristics of the studies are quite different.

Phosphate is essential for cellular activity, including signal transduction, metabolism and energy exchange. Several studies have reported the effects of hypophosphatemia on poor outcomes in patients with sepsis and critically ill patients.19 20 The results of these studies suggest that severe phosphate deficiency may be associated with poor prognosis, especially in critically ill patients. However, other studies have reported that hypophosphatemia was not associated with mortality.18 21 In our study, mortality rates were not different between patients in group I (hypophosphataemia) and group II (normophosphataemia) in either the univariable or multivariable model. These different results for hypophosphataemia might result from the use of different cut-off levels and differences in the patient groups and their conditions included in these studies.

The mechanism for the association of hyperphosphataemia and mortality in sepsis remains unclear. Potential mechanisms have been suggested in preclinical studies, including vascular endothelial dysfunction, and the release of mitochondrial reactive oxygen species induced by phosphate.22 23 In addition, it has been reported that phosphate concentrations increase in ischaemic tissue injury. In a rabbit mesenteric ischaemia model, phosphate concentrations increased as the duration of ischaemic injury increased.8 In a human study, phosphate and lactate concentrations increased during cardioplegic arrest, and the changes in lactate and phosphate concentrations exhibited similar patterns.24 Ischaemic tissue injury is one of the major mechanisms of organ damage in patients with sepsis, and lactate has been used as an indicator of the extent of ischaemic injury.16 25 In our study, lactate concentration was higher in group III (mild hyperphosphataemia) and group IV (moderate to severe hyperphosphataemia) than in the other groups, and lactate concentrations had a positive correlation with phosphate concentrations. Therefore, these findings suggested that phosphate was released from the damaged tissues because of ischaemic injury in sepsis.

In our study, patients with mild hyperphosphataemia and those with moderate to severe hyperphosphataemia demonstrated a higher APACHE II score, and were more likely to have unstable haemodynamic variables and abnormal laboratory results. These findings suggest that hyperphsophataemia is associated with the severity of the sepsis these patients had and consequently resulted in high mortality. However, even when the APACHE II score and laboratory results including creatinine were adjusted, patients with moderate to severe hyperphosphataemia had significantly higher mortality. Therefore, moderate to severe hyperphosphataemia was an independent prognostic factor for mortality in sepsis or septic shock.

There are some limitations of this study. First, this study is a retrospective analysis of data collected at a single institution, so we cannot say if we can generalise these results directly to other institutions. Therefore, further studies performed at other institutions are required to determine the generalisability of the findings. Second, the definitions of sepsis and septic shock changed during the study period, so differences in the composition of study sample is possible. Third, parathyroid hormone, vitamin D and calcitriol could influence the phosphate concentration, but we did not measure their concentrations.

Conclusion

Moderate to severe hyperphosphataemia was associated with greater 28-day mortality in patients with sepsis or septic shock, and could be used as a predictor of mortality in sepsis or septic shock.

Abstract translation

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

References

Footnotes

  • Contributors YHJ, JHL and JK conceived and design the study. D-HJ, DKL and CUL conduct data analysis. JK and JEH provide statistical advice. D-HJ, IP and S-ML drafted the manuscript. YHJ and SMP contributed to its revision.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Patient consent for publication Not required.

  • Ethics approval This study was approved by institutional review board of Seoul National University Bundang Hospital. (Approval number: B-1908/556-103).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Data are available on reasonable request. Only deidentified participant data are available on reasonable request to the corresponding author. drakejo@snubh.org.