Emergency. 2017; 5 (1): e1 OR I G I N A L RE S E A RC H Effect of Positive End-Expiratory Pressure on Central Ve- nous Pressure in Patients under Mechanical Ventilation Majid Shojaee1, Anita Sabzghabaei2, Hossein Alimohammadi1, Hojjat Derakhshanfar1, AfshinAmini1, Bahareh Esmailzadeh1∗ 1. Department of Emergency Medicine, Imam Hossein Hospital, Shahid Beheshti University of Medical Sciences, Tehran, Iran. 2. Department of Emergency Medicine, Loghmane Hakim Hospital, Shahid Beheshti University of Medical Sciences, Tehran, Iran. Received: November 2015; Accepted: January 2016; Published online: 8 January 2017 Abstract: Introduction: Finding the probable governing pattern of PEEP and CVP changes is an area of interest for in- charge physicians and researchers. Therefore, the present study was designed with the aim of evaluating the relationship between the mentioned pressures. Methods: In this quasi-experimental study, patients under me- chanical ventilation were evaluated with the aim of assessing the effect of PEEP change on CVP. Non-trauma patients, over 18 years of age, who were under mechanical ventilation and had stable hemodynamics, with in- serted CV line were entered. After gathering demographic data, patients underwent 0, 5, and 10 cmH2O PEEPs and the respective CVPs of the mentioned points were recorded. The relationship of CVP and PEEP in different cut points were measured using SPSS 21.0 statistical software. Results: 60 patients with the mean age of 73.95 ± 11.58 years were evaluated (68.3% male). The most frequent cause of ICU admission was sepsis with 45.0%. 5 cmH2O increase in PEEP led to 2.47 ± 1.53 mean difference in CVP level. If the PEEP baseline is 0 at the time of 5 cmH2O increase, it leads to a higher raise in CVP compared to when the baseline is 5 cmH2O (2.47 ± 1.53 vs. 1.57 ± 1.07; p = 0.039). The relationship between CVP and 5 cmH2O (p = 0.279), and 10 cmH2O (p = 0.292) PEEP changes were not dependent on the baseline level of CVP. Conclusion: The findings of this study revealed the direct relationship between PEEP and CVP. Approximately, a 5 cmH2O increase in PEEP will be associated with about 2.5 cmH2O raise in CVP. When applying a 5 cmH2O PEEP increase, if the baseline PEEP is 0, it leads to a significantly higher raise in CVP compared to when it is 5 cmH2O (2.5 vs. 1.6). It seems that sex, history of cardiac failure, baseline CVP level, and hypertension do not have a significant effect in this regard. Keywords: Positive-pressure respiration; central venous pressure; ventilators, mechanical; catheterization, central venous; emergency service, hospital © Copyright (2017) Shahid Beheshti University of Medical Sciences Cite this article as: Shojaee M, Sabzghabaei A, Alimohammadi H, Derakhshanfar H, Amini A, Esmailzadeh B. Effect of Positive End-Expiratory Pressure on Central Venous Pressure in Patients under Mechanical Ventilation. Emergency.2017;5(1): e1. 1. Introduction A vast number of patients visiting emergency de- partment (ED) need intubation due to their level of consciousness and clinical status, and are kept under mechanical ventilation. These patients may also require central venous (CV ) line insertion for different reasons such as inability to access peripheral vein, blood transfusion, or administration of vasoactive agents (1). Central venous pressure (CVP) is a good indicator of circulatory volume ∗Corresponding Author: Bahareh Esmailzadeh; Emergency Department, Imam Hossein Hospital, Shahid Madani Avenue, Imam Hossein Square, Tehran, Iran. Tel:00989124034956 , Email: bahareesmailzadeh@gmail.com. and cardiac function, which may be influenced by various factors such as function of right atrium and ventricle, ve- nous tone, and intra-thoracic pressure(1). Considering the afore-mentioned utilities, CVP monitoring is routinely used in operating rooms, intensive care units (ICU), and EDs. The normal range of CVP is 8-12cmH2O, which increases to 12-16 cmH2O in patients under mechanical ventilation (1, 2). Factors such as incorrect adjustment of the ruler’s zero point, patient’s poor condition, inappropriate placement of the catheter, and using vasopressor may interfere with accu- rate CVP measurement (1). Positive end-expiratory pressure (PEEP) in patients under mechanical ventilation can affect CVP via increasing intra-thoracic pressure. Various reports exist on the direct relationship between the 2 pressures This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com bahareesmailzadeh@gmail.com M. Shojaee et al. 2 (3). Yet, there is no accurate formula or solution based on PEEP level for adjusting CVP in patients under mechanical ventilation. Yang et al. showed that a 0.38 cmH2O increase in PEEP leads to 1 cmH2O raise in CVP (4). A study on 70 cardiac surgery patients in 2007 showed that mean CVP of the patients in 0, 5, and 10 cmH2O PEEPs are 11, 12, and 14 cmH2O, respectively (5). Finding theprobable governing pattern of PEEP and CVP changes is an area of interest for in-charge physicians and researchers. Therefore, the present study was designed with the aim of evaluating the relationship between the mentioned pressures. 2. Methods 2.1. Study design and setting In this quasi-experimental study, patients under mechanical ventilation who were hospitalized in the ICU of Imam Hos- sein Hospital, Tehran, Iran, were evaluated with the aim of assessing the effect of PEEP change on CVP. The study was approved by the Ethics Committee of Shahid Beheshti Uni- versity of Medical Sciences in their 147t h meeting in Febru- ary 2013. Since the patients were not able to give written informed consent, it was obtained from their relatives. The researchers were committed to protecting patient rights and confidentiality in line with the principles of Helsinki Decla- ration. 2.2. Participants Non-trauma patients, over 18 years of age, who were under mechanical ventilation and had stable hemodynamics, with inserted CV line were entered. To eliminate other confound- ing factors of CVP, all the patients were deeply sedated and under the same ventilator settings including tidal volume = 8 - 10 cc/kg, pressure support = 12 - 14 cmH2O, FiO2 = 40 - 70%, flow = 10 - 12 l/minute, and PEEP = 0, 5, and 10 cmH2O. Intake volume per hour was in a similar range for all the pa- tients and the patient was excluded if more than 200 cc/hour intravenous isotonic fluid was needed to maintain hemody- namic stability. Patients with auto PEEP (more than 2 cmH2O difference between the PEEP reported by the device and the one set for the patient), what ever the reason, were excluded. In addition, patients who were in need of > 10 cmH2O PEEP, for example patients withacute respiratory distress syndrome (ARDS), were not evaluated. The ventilator devices used for the patients were all the same model and from the same manufacturer. 2.3. Data gathering After gathering demographic data of the patients using a checklist designed for this study, patients underwent 0, 5, and 10 cmH2O PEEPs and the respective CVPs of the men- tioned points were measured and recorded. The time con- sidered for adjustment of CVP with any of the PEEP cut-off points was considered 10 minutes (3). To accurately measure CVP, all measurements were carried out by the same person, in supine position, and by setting the zero point of the CV line ruler at sternal notch level. To minimize errors, mea- surement for every patient was done twice for each PEEP cut point with 30minutes intervals and their mean was consid- ered the reference CVP measure. In cases of wide difference between the 2 measurements, a third measurement was done and the mean of the 2 closest measures was considered to be the reference. If aparticipant showed hypoxia and hemody- namic instability at any time of the study, he/she was elimi- nated from the study and necessary interventions were car- ried out to stabilize his/her status. All the calculated mea- sures for CVP in each PEEP cut point was recorded in the pre- pared checklist and used for analysis. 2.4. Statistical analyses The sample size needed for the present study was calculated based on a pilot study and using a standard deviation to es- timate the minimum sample size needed.Therefore, by con- sidering Zα = 0.5%, p = 95%, minimum clinically consider- able CVP difference of 1 cmH2O (d = 1), and the difference between the standard deviation of CVP in PEEP 10 and 0 of 1.4 (Sd i f f = 1.4), the sample size needed was calculated to be 26 cases. Patient data were analyzed using SPSS 21.0. Quan- titative data were reported as mean and standard deviation and qualitative ones as frequency and percentage. To com- pare CVP before and after applying various PEEPs, paired t- test or non-parametric Wilcoxon test were used. In all tests, p < 0.05 was considered as significance level. 3. Results 60 patients with the mean age of 73.95 ± 11.58 years (46 - 93) were evaluated (68.3% male). The most frequent cause of ICU admission was sepsis with 45.0%. Table 1 shows the baseline characteristics of the studied patients. Table 2 shows the relationship between various PEEP measures and CVP. 5 cmH2O increase in PEEP led to 2.47 ± 1.53 mean difference in CVP level. If the PEEP baseline is 0 at the time of 5 cmH2O in- crease, it leads to a higher raise in CVP compared to when the baseline is 5 cmH2O (2.47 ± 1.53 vs. 1.57 ± 1.07; p = 0.039). Adjusting the analyses done in table 2 based on sex, pres- ence of cardiac failure and history of hypertension did not show any significant differences in the mentioned relations (table3). Evaluation of the relationship between changes in PEEP and CVP measures based on different levels of CVP are summarized in table 4. The relationship between CVP and 5 cmH2O (p = 0.279), and 10 cmH2O (p = 0.292) PEEP changes were not dependent on the baseline level of CVP. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com 3 Emergency. 2017; 5 (1): e1 Table 1: Baseline characteristics of the studied population Studied variables Mean ± SD (range) Systolic BP (mmHg) 125.533 ± 20.55 (100 - 170) Diastolic BP (mmHg) 74.583 ± 11.21 (60 - 100) Heart rate/minute 90.96 ± 15.34 (60 - 120) Age (year) Frequency (%) 18 - 39.9 0 (0) 40 - 59.9 7 (11.7) 60 - 79.9 29 (48.3) ≥ 80 24 (40) Hx of cardiac failure Yes 12 (20) No 48 (80) Hx of hypertension Yes 43 (71.7) No 17 (28.3) Cause of admission Pneumosepsis 14 (23.3) Urosepsis 13 (21.7) Cerebrovascular accident 8 (13.3) Metastatic cancer 4 (6.7) Other 21 (35) BP: blood pressure; SD: standard deviation. Table 2: Correlation between different positive end-expiratory pressures (PEEP) and central venous pressures (CVP) in studied pa- tients PEEP (cmH2 O) CVP (mean ± SD) (cmH2 O) P value Paired 1 0 7.81 ± 5.99 < 0.001 5 10.29 ± 5.67 Paired 2 0 7.81 ± 5.99 < 0.001 10 11.86 ± 5.72 Paired 3 ∆ 0-5 2.47 ± 1.53 < 0.001 ∆ 0-10 4.05 ± 2.09 Paired 4 ∆ 0-5 2.47 ± 1.53 < 0.003 ∆ 5-10 1.57 ± 1.07 SD: standard deviation. 4. Discussion The findings of this study showed that an increase in PEEP has a direct relationship with CVP increase. Approximately, a 5 cmH2O increase in PEEP will be associated with about 2.5 cmH2O raise in CVP. When applying a 5 cmH2O PEEP increase, if the baseline PEEP is 0, it leads to a significantly higher raise in CVP compared to when it is 5 cmH2O (2.5 vs. 1.6). It seems that sex, history of cardiac failure, hyper- tension, and baseline CVP do not significantly affect CVP in- crease rate. In a study by Yang et al. 1 cmH2O increase in PEEP, led to 0.38 cmH2O increase in CVP, which is approxi- mately in line with the present study (4). A study on the effect of PEEP in patients under mechanical ventilation showed a significant direct relation ship between 0, 5, and 10 cmH2O PEEPs with CVP. The CVP increase was related to mean PEEP during mechanical ventilation when PEEP was set 10 or less, in a study by Cao et al. which is in line with this study (6). Evaluating the effect of PEEP on CVP and stroke volume in 20 patients with cardiac diseases, revealed that PEEP signif- icantly affects CVP, while no significant relationship was de- tected between heart rate and mean arterial pressure (7). In the present study, 5 cmH2O increase in PEEP, led to 1.5 - 2.5 cmH2O increase in CVP. Currently, CVP is used as a guide for fluid therapy efficiency and monitoring the effects of intake volume on cardiovascular system. Many current treatment protocols, especially regarding septic shock patient manage- ment, define the aim of the treatment as achieving a CVP of 8 -12 cmH2O in patients without ventilator and 12 -16 cmH2O in those under mechanical ventilation. However, the study by Cao et al. showed that in patients under ventilator and PEEP, CVP alone is not a good reference for estimation of circula- tory volume and required fluid volume for resuscitation (8). Considering the existing controversies in this regard, it seems that we should seek more accurate scales for determining the efficacy of fluid therapy in patients under mechanical venti- lation. Until then, the best way might be using modified CVP based on PEEP rate. 5. Limitations Since the present study was carried out on patients hospi- talized in the ICU, some limitations should be noted: First, patients with a variety of underlying illnesses were included, which can affect the results. Second, the patients have been in different phases of hospitalization, therefore the rates and efficiencies of treatments received (fluid, vasoactive drugs) were different among them. In addition, CVP measurement using a ruler has some limitations in its nature, such as ad- justing the zero point. Naturally, there are some limitations for applying long-term PEEPsin these patients, which can af- fect the conclusion. It is suggested to eliminate the afore- mentioned limitations to accurately evaluate the effect of PEEP on CVP in future studies. 6. Conclusion The findings of this study revealed the direct relationship be- tween PEEP and CVP. Approximately, a 5 cmH2O increase in PEEP will be associated with about 2.5 cmH2O raise in CVP. When applying a 5 cmH2O PEEP increase, if the base- line PEEP is 0, it leads to a significantly higher raise in CVP compared to when it is 5 cmH2O (2.5 vs. 1.6). It seems that sex, history of cardiac failure, baseline CVP level, and hyper- tensiondo not have a significant effectin this regard. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com M. Shojaee et al. 4 Table 3: Correlation between different positive end-expiratory pressures (PEEP) and sex, and history of heart failure and hypertension Sex∗ Heart failure∗∗ Hypertension∗∗∗ PEEP (cmH2 O) Mean ± SD P value Mean ± SD P value Mean ± SD P value PEEP 0 0 7.23±5.00 0.27 7.22±5.79 0.13 8.02±1.14 0.86 1 9.07±7.73 10.16±6.46 7.73±4.81 PEEP 5 0 9.74±4.73 0.27 9.94±5.48 0.35 10.61±7.92 0.78 1 11.47±7.32 11.66±6.47 10.16±4.61 PEEP 10 0 11.54±4.97 0.53 11.60±5.52 0.48 11.91±7.72 0.97 1 12.55±7.20 12.91±6.64 11.84±4.83 ∆0-5 0 2.51±1.66 0.78 2.71±1.55 0.01 2.58±1.20 0.72 1 2.39±1.23 1.50±0.97 2.43±1.65 ∆5-10 0 1.80±1.15 0.01 1.65±1.06 0.24 1.29±0.83 0.20 1 1.07± 0.65 1.25±1.07 1.68±1.14 ∆0-10 0 4.31±2.24 0.14 4.37±2.12 0.01 3.88±1.63 0.70 1 3.47±1.61 2.75±1.37 4.11±2.26 Sex∗(0= Male, 1= Female), Heart failure ∗∗(0= No, 1= Yes), Hypertension∗∗∗ (0= No, 1= Yes). Table 4: of correlation between positive end-expiratory pressures (PEEP) and central venous pressure (CVP) in different baseline CVP levels PEEP (cmH2 O) ∆ CVP (cmH2 O) P value ∆ PEEP 5 CVP < 8 2.7±1.7 0.279 CVP = 8-12 2.4±1.2 CVP > 12 1.8±1.2 ∆ PEEP 10 CVP < 8 4.3±2.3 0.279 CVP = 8-12 4.1±1.3 CVP > 12 3.1±1.9 7. Appendix 7.1. Acknowledgements The authors appreciate the insightful cooperation of Emer- gency Department staff. 7.2. Author contribution All authors passed four criteria for authorship contribution based on recommendations of the International Committee of Medical Journal Editors. 7.3. Funding None. 7.4. Conflict of interest None. References 1. Marino PL, Sutin KM. The ICU book: Williams , Wilkins Baltimore:; 1998. 2. Wai A. Roberts and Hedges: Clinical Procedures in Emer- gency Medicine. LWW; 2010. 3. Roberts JR, Hedges JR. Clinical procedures in emergency medicine: Elsevier Health Sciences; 2009. 4. Yang Z, Zhou J, Sun B, Qian Z, Zhao H, Liu W. [The in- fluence of positive end-expiratory pressure on central ve- nous pressure in patients with severe craniocerebral in- jury]. Zhongguo wei zhong bing ji jiu yi xue= Chinese crit- This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com 5 Emergency. 2017; 5 (1): e1 ical care medicine= Zhongguo weizhongbing jijiuyixue. 2012;24(5):283-5. 5. Marx J, Walls R, Hockberger R. Rosen’s Emergency Medicine-Concepts and Clinical Practice: Elsevier Health Sciences; 2013. 6. Cao F, Liu X, Chen R, Wang X. [Effect of positive end- expiratory pressure on central venous pressure and com- mon iliac venous pressure in mechanically ventilated pa- tients]. Zhongguo wei zhong bing ji jiu yi xue= Chinese critical care medicine= Zhongguo weizhongbing jijiuy- ixue. 2008;20(6):341-4. 7. Geerts B, Aarts L, Groeneveld A, Jansen J. Predicting cardiac output responses to passive leg raising by a PEEP-induced increase in central venous pressure, in cardiac surgery patients. British journal of anaesthesia. 2011;107(2):150-6. 8. Cao F, Chen R, Liu X, He R. [Effect of positive end- expiratory pressure on the pressure gradient of venous re- turn in hypovolemic patients under mechanical ventila- tion]. Zhongguo wei zhong bing ji jiu yi xue= Chinese crit- ical care medicine= Zhongguo weizhongbing jijiuyixue. 2009;21(10):583-6. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com Introduction Methods Results Discussion Limitations Conclusion Appendix References