Association between elevated central venous pressure and outcomes in critically ill patients
Li et al. Ann. Intensive Care
Association between elevated central venous pressure and outcomes in critically ill patients
Dong‑kai Li 0
Xiao‑ting Wang 0
Daw‑ei Liu 0
0 share first authorship Department of Critical Care Medicine, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Science , 1 Shuaifuyuan, Dongcheng District, Beijing 100730 , China
Background: Some prior studies have shown that elevated mean central venous pressure in certain patient populations and disease processes may lead to poor prognosis. However, these studies failed to generalize the concept of elevated central venous pressure (ECVP) load to all patients in critical care settings because of the limited cases and exclusive cohorts. The aim of the study was to investigate the association between elevated central venous pressure and outcomes in critical care. Methods: We performed a retrospective analysis on a single‑ center public database (MIMIC) of more than 9000 patients and more than 500,000 records of central venous pressure measurement. We evaluated the association between mean central venous pressure level and 28‑ day mortality after intensive care unit admission. The secondary outcomes were duration of mechanical ventilation, vasoactive drug use, laboratory results related to organ dysfunction and length of intensive care unit hospitalization. Accordingly, we proposed the concept of ECVP10 (the time sum of CVP above 10 mmHg) and investigated its association with outcome. Results: There were 1645 deaths at 28 days after admission. Compared with the lowest quartile of mean central venous pressure [mean (SD) 7.4 (1.9) mmHg], the highest quartile [17.4 (4.1) mmHg] was associated with a 33.6% (95% CI 1.117-1.599) higher adjusted risk of death. Poor secondary outcomes were also associated with higher quartiles of elevated mean central venous pressure. After stratification by mean central venous pressure, elevated duration of central venous pressure above 10 mmHg was significantly higher in the non‑ survival group than in the survival group. Conclusions: Elevated central venous pressure level correlated with poor outcomes and prolonged treatment in critical care settings. Level and duration of elevated central venous pressure should be both evaluated to establish its cause and apply appropriate treatment.
Central venous pressure; Survival; Critical care
Elevated right atrial pressure or central venous
pressure (CVP) occurs frequently in critical care settings
] and may be caused by several conditions, such as
congestive heart failure syndrome, constrictive
pericardial disease, tension pneumothorax, and resuscitation/
evacuation phases of septic shock. Although CVP has
been utilized as a surrogate of intravascular volume in
critically ill patients, recent studies have challenged the
validity of elevated CVP (ECVP) in critical care settings
]. Based on rationale provided by the Starling curves
and Guyton model on cardiac function, CVP is
determined by the interaction of cardiac function and the
return of blood to the heart. An ECVP might indicate
an impediment to the venous return and
microcirculatory blood flow as well as accompanying lung edema and
splanchnic congestion [
], which may further worsen
the potential organ failure in critical patients.
Some prior studies have shown that increased mean
CVP in certain patient populations and disease processes
may lead to poor prognosis. In septic patients, increased
mean CVP was associated with worse outcome, including
increased 28-day mortality as well as the development and
progression of acute kidney injury [
6, 7, 11
]. For patients
with cardiovascular diseases, ECVP might be associated
with impaired hepatic and renal function and may be
independently related to all-cause mortality [
studies, however, could not generalize the concept of elevated
central venous pressure (ECVP) load to all patients in
critical care settings because of the limited cases and exclusive
cohorts. Besides, the importance of ECVP duration was
underestimated in estimating the ECVP load and
analyzing its relationship with outcome, which is only accessible
with the advantage of “big data” technology .
By using the large, public, de-identified clinical
database, Multi-parameter Intelligent Monitoring in
Intensive Care (MIMIC-III) [
], we evaluated the effect of
ECVP on all patients in critical care settings, for whom
CVP data were carefully recorded. We aimed to further
characterize the association between mean CVP level
and outcome as well as treatment received in the
intensive care unit (ICU). Specifically, we investigated the
potential effect of ECVP duration on outcome when
mean CVP were stratified at different levels.
We conducted a large-scale, single-center, retrospective
cohort study using data collected from the MIMIC-III
open source clinical database (version 1.3, released on
December 10, 2015), which was developed and
maintained by the Massachusetts Institute of Technology
(MIT), Philips Healthcare, and Beth Israel Deaconess
Medical Center (BIDMC) [
]. Information derived from
the electronic medical records of 46,476 unique critical
care patients admitted to the ICUs at BIDMC between
2001 and 2012 was included in this free accessible
database. Use of the MIMIC-III database has been approved
by the Institutional Review Boards of BIDMC and MIT,
and waiver of informed consent was granted.
All patients in the database were screened. The inclusion
criteria in this study were as follows: (1) adults (≥18 years
of age) at ICU admission, with complete medical records
including available CVP measurement records; (2) ICU
stay ≥72 h; and (3) consecutive CVP monitoring ≥12 h.
For patients with multiple ICU stays, only data related
the first ICU admission were considered.
All available CVP measurements recorded during ICU
stay were extracted. Other variables including ICU type,
demographic data, age, sex, Elixhauser comorbid
] and admission illness severity scores [including
the Simplified Acute Physiology Score (SAPS) [
Sequential Organ Failure Assessment (SOFA) [
extracted from the database. Additionally, data on use of
vasopressors, mechanical ventilation, laboratory results
related to organ dysfunction and length of ICU stay and
hospitalization were extracted from the database.
The primary exposure was the mean CVP during the first
72 h after ICU admission. We also calculated the duration
of ECVP10 as the time of CVP above the level of 10 mmHg,
which was considered as a relatively higher CVP level that
might by unbeneficial for patients in critical care settings
], and used it as an alternative indicator of the mean
CVP level to estimate the association between ECVP load
and outcome in the critical care settings.
The primary outcome was 28-day mortality after ICU
admission. The secondary outcomes included duration
of mechanical ventilation and vasoactive drug use
(epinephrine, norepinephrine, vasopressin, dopamine,
dobutamine, milrinone and phenylephrine), laboratory results
related to organ dysfunction and the length of ICU
admission and hospitalization. For some patients, whose
death occurred outside the hospital, the Social Security
Death Index was linked to the database for investigations
related to mortality.
Baseline characteristics were stratified by quartiles of
mean CVP during the first 72 h after ICU admission.
Cox regression analyses were undertaken to compare
the 28-day mortality among different mean CVP
quartiles. All other covariates, which comprised demographic
characteristics, ICU type, SAPS II score at admission,
comorbidities and mean duration of CVP measurement,
were included into the multivariable regression model.
We also compared the survivors and non-survivors for
their ECVP10 duration with the outcome independently
through the stratification on mean CVP.
The results are expressed as mean ± SD (standard
deviation) for normally distributed data or median
[interquartile range (IQR)] for non-normally distributed data.
Continuous variables were compared using one-way
analysis of variance for normally distributed quantitative data,
and Mann–Whitney U test for non-normally distributed
quantitative data to determine differences between groups.
All statistical analyses were performed by using the IBM®
SPSS® Statistics version 22 (SPSS Inc., Chicago, IL, USA).
Any p value <0.05 was considered statistically significant.
Subset and sensitivity analyses
Since the mean CVP level can only describe the overall
level of CVP, in sensitivity analyses, we replaced the mean
CVP level with the duration of ECVP10 as an alternative
indicator to estimate the association between ECVP load
and outcome in the critical care settings.
We examined whether elevated mean CVP level in the
highest quartile was associated with poor outcome
compared to the other quartiles in patients with and without
sepsis, which was defined as suspected or documented
infection and an acute increase of ≥2 SOFA points
], and AKI, as defined by either a greater than or
equal to 0.3 mg/dL increase within 48 h or a greater than
or equal to 50% increase within 7 days of ICU
admission, or acute dialysis, in keeping with the Kidney Disease
Outcome Quality Initiative guidelines [
individually tested association between elevated mean CVP and
these variables in adjusted analysis and provided
graphical representation of the stratified risks.
Among the 46,476 ICU patients and 61,532 ICU
admissions in the MMIC-III v1.3, 17,324 patients underwent
CVP measurement during the first 3 days of the ICU
hospitalization and 791,606 CVP records were available
in the database. Sequentially, we excluded 911 patients
whose age at admission was below 18 years, 7323 patients
with ICU stay less than 3 days and CVP measurement
duration of less than 12 h, as illustrated in Fig. 1. The
final cohort comprised 9090 patients with their first ICU
admission and the corresponding 505,317 records of
CVP measurement. The mean (±SD) interval between
two consecutive CVP measurements was 0.9 ± 1.2 h, and
the mean (±SD) duration of CVP measurement during
the first 72 h in ICU hospitalization was 49.5 ± 18.8 h.
In the 9090 critically ill patients included in the study,
the mean (± SD) CVP level was 11.8 ± 4.0 mmHg and
median (IQR) ECVP10 duration of 23.0 (9.0, 43.8) h.
As shown in Table 1, patients in the highest quartile of
mean CVP level presented greater SOFA score and more
comorbidities. The mean duration of CVP measurement
and ECVP10 duration were also increased in higher
quartiles of mean CVP level.
Mean CVP level and outcome
During the first 28 days after ICU admission, 1645 patients
died. Higher quartiles of mean CVP level during the first
3 days after ICU admission were associated with higher
unadjusted mortality. After adjusting for sex, age, gender,
ethnicity, congestive heart failure, cardiac arrhythmias,
hypertension, valvular disease, pulmonary circulation
disease, renal failure and other 24 Elixhauser comorbidities,
ICU type, SAPS II score at admission and duration of
CVP measurement, the mean CVP level remained a
significant predictor of 28-day mortality, as shown in Table 2.
The other potential determinants are shown in Additional
file 1: Supplement Table 1. The survival curve of 28-day
mortality by the quartiles of mean CVP level is shown in
Fig. 2, and the detailed comparison of 28-day mortality
between the deciles of mean CVP level is shown in
Additional file 1: Supplement Table 2 .
In addition to the 28-day mortality, we investigated the
association between quartiles of mean CVP level and
secondary ICU outcomes, which included length of
hospitalization, treatment received in ICU and laboratory
results related to organ dysfunction, as shown in Table 1.
The result showed that prolonged ICU stay and
hospitalization were required for patients with higher mean CVP
level during the first 3 days from ICU admission, and the
duration of vasopressor or mechanical ventilation
treatment also was prolonged. Furthermore, leukocytosis,
higher serum bilirubin, creatinine, and lactate were all
more commonly seen in the higher mean CVP quartiles.
Subset and sensitivity analyses
Duration of ECVP10 was investigated between 28-day
survival and non-survival patients. After stratification by
different levels of mean CVP, as in Additional file 1:
Supplement Figure 2, the results showed that the ECVP10
duration in non-survival group was significantly higher
than in the survival group. Replacement of mean CVP
with ECVP10 duration as indicator of ECVP load did not
meaningfully change the association between ECVP load
and 28-day mortality (odds ratio in the highest quartile,
1.354, 95% CI 1.151–1.594, p < 0.001).
To examine whether the higher mean CVP level
remained associated with increased 28-day mortality
Hazard ratios (95% CI) provided. *Adjusted for sex, age, gender, ethnicity,
congestive heart failure, cardiac arrhythmias, hypertension, valvular
disease, pulmonary circulation disease, renal failure and other 24 Elixhauser
comorbidities, ICU type, SAPS II score at admission and duration of CVP
across subset of patients, we explored multiplicative
interaction terms with and without sepsis and AKI. No
significant effect of ECVP on mortality was observed
in patients with sepsis [n = 5121, odds ratio (OR)
1.090, 95% CI 0.941–1.263, p = 0.252] and without
AKI [n = 6131, odds ratio (OR) 1.097, 95% CI 0.945–
1.274, p = 0.224], while there was a significant trend
toward increased mortality from higher ECVP load
in patients without sepsis [n = 3969, odds ratio (OR)
1.36, 95% CI 1.087–1.702, p = 0.007], as illustrated in
This study investigated the association between ECVP
load and outcomes of patients in critical care setting.
Based on comparison among patients with different
levels of mean CVP, which is a measurement of ECVP load,
we found that the higher mean CVP level was associated
with increased 28-day mortality. Further investigation on
length of hospitalization, duration of vasopressor
treatment and mechanical ventilation, and laboratory results
related to organ dysfunction also demonstrated that
higher mean CVP level was associated with poor ICU
outcome for patients in critical care settings.
In contrast to the common misleading statement of
increasing CVP to a higher level to increase cardiac
output, Starling curves, with the contribution of Guyton’s
work, did not suggest any causal relationship between
CVP and cardiac output, but emphasized the
comprehensive role in the interaction between cardiac function
and venous return of CVP. Recent studies have showed
that not only the CVP failed as a useful measure for the
assessment of preload and fluid responsiveness [
that the ECVP is independently associated with a higher
mortality and increased risk of AKI in patients with
sepsis and heart failure [
7, 9, 11
]. These results indicated that
the CVP should be considered part of the evaluation in
patients with hemodynamic instability. Furthermore, an
ECVP may signify an impediment to venous return [
Based on the study for the consensus of congestive
heart failure treatment, the pathophysiology of chronic
systemic venous congestion should include decreased
cardiac output from increased resistance to venous
return, with the subsequent splanchnic damage and
tissue perfusion insufficiency [
]. Integrated with the
Starling–Guyton cardiac curve, the direct cause of ECVP
may be defined as “the impediment to venous return by
a relatively fatigued heart” [
]. Therefore, the potential
causes of ECVP may include elevated venous return,
cardiac dysfunction, increased vascular resistance, increased
pericardial or intrapleural pressure, among other
conditions. Our study showed that a higher ECVP10 load was
associated with prolonged duration of treatment and
length of ICU hospitalization. In sensitivity analysis, the
result showed that impact of ECVP on mortality was
limited in some subgroup patients, especially in the sepsis
group, which was inconsistent with previous result [
Considering the difference in enrolling criteria, our result
indicated that for the sepsis patients who survived the
resuscitation phase, the benefit from successful
resuscitation, which was always presented as higher CVP level,
may attenuate the impact of ECVP on outcome and did
not lead to statistically significant result. The fluid
balance, mean CVP level and duration of ECVP10 result in
different subgroups, and outcome (survival and
non-survival) is shown in Additional file 1: Supplement Table 3.
In contrast to the long-term influence of systemic
venous hypertension caused by cardiac dysfunction
such as right heart failure, pericardial effusion and other
conditions, the pathophysiological changes that occur
with ECVP, as manifestation of systemic venous
congestion, generally have acute consequences in patients
being treated in critical care settings. Based upon the
significance of association between ECVP load and ICU
outcome, we also showed that the ECVP load should be
considered and corrected in two independent aspects,
level and duration, because both played important
pathophysiological roles. According to the present results,
we suggest that ECVP, with its association with worse
outcome in critical care settings, should be considered
seriously and further actions should be undertaken to
discover potential causes and treatment. Our study was
based on data extracted from electronic medical records
in the MIMIC-III v1.3 [
], a large, open clinical
database, which allowed precise research on ECVP load. To
our knowledge, this study is the first to evaluate ECVP
on a large population of patients managed in critical
settings. More than 500,000 CVP records and 9000 patients
were involved in our study. Furthermore, the use of
database technologies and statistics played a critical role in
reaching a meaningful conclusion in the present study.
Given the extreme complexity and diversity of critically
ill patients and their treatment, the traditional approach
of knowledge discovery may not satisfy the need to
address clinical concerns, and further efforts should be
made to organize, share and analyze the “big data” in
critical care [
This study has several limitations. First, our study is
limited by its retrospective nature and the source of
data used. For this reason, no causal relationships could
be established. Additionally, CVP records could only
be assessed in critical care patients with central venous
catheter and we only extracted data from the first 3 days
of ICU admission. Thus, our conclusions cannot be
generalized to patients treated in a setting other than a
critical care setting or without opportunities of continuous
treatment during the first 3 days in ICU, for the latter
of which extremely low or high CVP values may occur
during the rescue/emergency phase of certain disease
or pathophysiology process. Second, unlike the other
studies on elevated mean CVP in certain patient
populations and disease processes [
], we focused on a
relatively long phase (72 h from ICU admission) during the
general management of patients in critical care settings.
Besides, it should be noted that we do not oppose to the
significance of CVP targeting in some emergency events
of critical care, such as the initial resuscitation phase of
septic shock [
]. Third, although some predictors of
illness severity were included in our study and adjusted
analysis confirmed the association between ECVP10 load
and poor outcome, our results may be affected by other
confounding variables associated with organ damage or
mortality. Additional prospective studies should be
performed to investigate these parameters and the potential
causes of ECVP load.
In conclusion, this study showed that for patients in
critical care settings, higher ECVP load was associated with
poor outcome as well as prolonged treatment in ICU.
Level and duration of ECVP should be evaluated, and
more effort should be made to establish the cause and
appropriate treatment of ECVP.
Additional file 1. Supplementary result for study on association between
CVP and outcome
CVP: central venous pressure; ECVP: elevated CVP; ICU: intensive care unit; MIT:
Massachusetts Institute of Technology; BIDMC: Beth Israel Deaconess Medical
Center; MIMIC: Multi‑parameter Intelligent Monitoring in Intensive Care; SAPS:
Simplified Acute Physiology Score; SOFA: Sequential Organ Failure Assess‑
ment; IQR: interquartile range; SD: standard deviation.
LDK and WXT contributed equally to the study, and they designed the study
and performed data analysis and manuscript preparation. LDW conceived the
study and helped in the result interpretation. All authors read and approved
the final manuscript.
The authors acknowledge all of those involved in the design, construction
and maintenance of the open access MIMIC‑III database at the MIT Laboratory
of Computational Physiology as well as the online community at Open Data
Stack Exchange (http://opendata.stackexchange.com) and GitHub (https://
The authors declare that they have no competing interests.
Availability of data and materials
The data set supporting the results of this article is included within the article.
Ethics, consent and permissions
The Institutional Review Boards of the MIT and BIDMC approved the establish‑
ment of the database. De‑identification was performed to ensure patients’
confidentiality. Our access to the database was approved after completion
of the National Institute of Health(NIH) Web‑based training course named
“Protecting Human Research Participants” by the author LDK (Certification
The authors declare that they have no relative funding support.
Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations.
1. Marik PE . Iatrogenic salt water drowning and the hazards of a high central venous pressure . Ann Intensive Care . 2014 ; 4 ( 1 ): 1 - 9 .
2. Dellinger RP , Levy MM , Rhodes A , Annane D , Gerlach H , Opal SM , et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock , 2012. Intensive Care Med . 2013 ; 39 ( 2 ): 165 - 228 .
3. Berlin DA , Bakker J . Starling curves and central venous pressure . Crit Care . 2015 ; 19 ( 1 ): 1 - 8 .
4. Malbrain ML , Marik PE , Witters I , Cordemans C , Kirkpatrick AW , Roberts DJ , et al. Fluid overload, de‑resuscitation, and outcomes in critically ill or injured patients: a systematic review with suggestions for clinical practice . Anaesthesiol Intensive Ther . 2014 ; 46 ( 5 ): 361 - 80 .
5. Semler MW , Wheeler AP , Thompson BT , Bernard GR , Wiedemann HP , Rice TW . Impact of initial central venous pressure on outcomes of conservative versus liberal fluid management in acute respiratory distress syndrome . Crit Care Med . 2016 ; 44 ( 4 ): 782 - 9 .
6. Wang XT , Yao B , Liu DW , Zhang HM . Central venous pressure dropped early is associated with organ function and prognosis in septic shock patients: a retrospective observational study . Shock . 2015 ; 44 : 426 - 30 .
7. Legrand M , Dupuis C , Simon C , Gayat E , Mateo J , Lukaszewicz AC , et al. Association between systemic hemodynamics and septic acute kidney injury in critically ill patients: a retrospective observational study . Crit Care . 2013 ; 17 ( 6 ): 1 - 8 .
8. Damman K , van Deursen VM , Navis G , Voors AA , van Veldhuisen DJ , Hillege HL . Increased central venous pressure is associated with impaired renal function and mortality in a broad spectrum of patients with cardiovascular disease . J Am Coll Cardiol . 2009 ; 53 : 582 - 8 .
9. Mullens W , Abrahams Z , Francis GS , Sokos G , Taylor DO , Starling RC . Importance of venous congestion for worsening of renal function in advanced decompensated heart failure . J Am Coll Cardiol . 2009 ; 53 : 589 - 96 .
10. Vellinga NAR , Ince C , Boerma EC . Elevated central venous pressure is associated with impairment of microcirculatory blood flow in sepsis: a hypothesis generating post hoc analysis . BMC Anesthesiol . 2013 ; 13 ( 1 ): 1 - 7 .
11. Boyd JH , Forbes J , Nakada TA , Walley KR , Russell JA . Fluid resuscitation in septic shock: a positive fluid balance and elevated central venous pressure are associated with increased mortality . Crit Care Med . 2011 ; 39 ( 2 ): 259 - 65 .
12. Pinsky MR , Dubrawski A . Gleaning knowledge from data in the intensive care unit . Am J Respir Crit Care Med . 2014 ; 190 : 606 - 10 .
13. Johnson AE , Pollard TJ , Shen L , Lehman LWH , Feng M , Ghassemi M , et al. MIMIC‑III, a freely accessible critical care database . Sci Data . 2016 ; 3 : 160035 .
14. Elixhauser A , Steiner C , Harris DR , Coffey RM . Comorbidity measures for use with administrative data . Med Care . 1998 ; 36 : 8 - 27 .
15. Le Gall JR , Loirat P , Alperovitch A , Glaser P , Granthil C , Mathieu D , et al. A simplified acute physiology score for ICU patients . Crit Care Med . 1984 ; 12 : 975 - 7 .
16. Vincent JL , Moreno R , Takala J , Willatts S , De Mendonça A , Bruining H , et al. The SOFA (Sepsis‑related Organ Failure Assessment) score to describe organ dysfunction/failure . Intensive Care Med . 1996 ; 22 : 707 - 10 .
17. Magder S , Bafaqeeh F. The clinical role of central venous pressure measurements . J Intensive Care Med . 2007 ; 22 : 44 - 51 .
18. Singer M , Deutschman CS , Seymour CW , Shankar‑Hari M , Annane D , Bauer M , et al. The third international consensus definitions for sepsis and septic shock (sepsis‑3) . JAMA. 2016 ; 315 ( 8 ): 801 - 10 .
19. Palevsky PM , Liu KD , Brophy PD , Chawla LS , Parikh CR , Thakar CV , et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for acute kidney injury . Am J Kidney Dis . 2013 ; 61 : 649 - 72 .
20. Marik PE , Cavallazzi R . Does the central venous pressure predict fluid responsiveness? An updated meta‑analysis and a plea for some common sense . Crit Care Med . 2013 ; 41 : 1774 - 81 .
21. Ghassemi M , Celi LA , Stone DJ . State of the art review: the data revolution in critical care . Crit Care . 2015 ; 19 : 118 .