Fluid Therapy in the Septic Critically Ill Patient

Introduction

Sepsis and severe sepsis is a condition which, in my role as an Advanced Clinical Practitioner, I will encounter many times in my career in the Emergency Medicine Department. There is a high incidence (up to 300 cases per 100,000) reported in both the United Kingdom (P. Mouncey et al. 2015) and Australia (Bailey et al. 2014), with a similar incidence in the United States (The ProCESS Investigators 2014).

The Surviving Sepsis Campaign (Dellinger et al. 2013) has aimed to ensure that this kind of patient is identified and then treated quickly, and its approach has been to break this down into the Sepsis Six, a bundle of investigations and treatment plans that will ensure the patients care gets off to the best possible start.

Part of this care bundle is the infusion of intravenous fluids which, for a number of physiological reasons, becomes necessary in ensuring that the patient’s vital organs remain perfused with the vital oxygen they need to work effectively.

The amount of fluid given to this type of patient, and the parameters that are the ultimate goal in assessing the effectiveness of the fluid has been a matter for some debate over the last 10-15 years. Three important trials were published in the last twelve months which may mean a new approach is needed (P. Mouncey 2015; Bailey et al. 2014; The ProCESS Investigators 2014).

This post aims to critically evaluate some of the evidence behind the current guidelines issued by the Surviving Sepsis group in 2013 (Dellinger et al. 2013), which have been controversial, and then to also look at some of the more recent studies which will help drive the new treatment protocols for the patient with sepsis.

 

Case Study

Mr Smith was a 78 year old gentleman whose condition is summarised in this figure .Capture_opt (1)

A decision was made to give Mr Smith an intravenous fluid infusion via cannula that had been placed in his veins. During critical illness the endothelial layer of the circulatory system can become compromised which can result in the fluids in the body being in the wrong spaces. Dehydration may also be a part of the illness as well may blood loss. These processes together can lead to circulatory collapse if intravenous fluid support is not given (Edwards & Mythen 2014).

The main question to be answered at the time was how much fluid should we give him and what parameters would help us decide when to stop giving fluids and, if necessary, try something else.

The evidence tells us that giving too much fluid can cause microvascular problems and encourage fluid to collect in the wrong places in the body, making the patients problems worse (Cordemans et al. 2012), but due to the nature of the sepsis processes taking place the patient does need some support.

 

Evidence Behind the Guidelines- Early Goal Directed Therapy (EGDT)

One of the biggest influencers in the early management of the septic patient was the paper published by Dr Emanuel Rivers (Rivers & Nguyen 2001).

The principle of Early Goal Directed Therapy (EGDT) is based on much of the findings of the Rivers trial and indeed this is reflected by the Surviving Sepsis group (Dellinger et al. 2013).

The Rivers trial was a single centre, parallel group, prospective, randomised study. The aim of the study was to assess the efficacy of EGDT in reducing the risk of mortality in the septic patient.

The two arms of the study comprised a group treated with the EGDT protocol  and the other with ‘standard care’. The primary outcome was in-hospital mortality with two secondary outcomes of 28 day and 60 day mortality.

EGDT

In the primary outcome there was 16% absolute risk reduction[1] in mortality between the two groups (EGDT 30.5%, standard therapy 46.5%, P= 0.009), and similar findings in the secondary outcomes (28 day mortality- 33.3% vs 49.2%, P= 0.01; 60 day mortality- 44.3% vs 56.9%, P= 0.03).

The 16% absolute risk reduction correlates to a number needed to treat of 6.[2]

There was more fluid used in the EGDT group over the first 6 hours (4.9L vs 3.5L P<0.001). The reduced use of vasopressors over the first six hours was not significant, but was over the next 7-72 hours (29.1% vs 42.9%, P= 0.03).

This reduction on mortality in the septic patient using EGDT has since been replicated (Jones et al. 2007; Crowe et al. 2010; Xue 2010), however these studies had different population mixes from the Rivers trial and certainly in the Crowe study the results were not so clear between the two groups.

There have been many criticisms of the Rivers trial since its publication for a variety of reasons (Marik 2014; Patocka et al. 2010), some methodological and some related to the parameters being targeted.

Firstly the Emergency Department staff were not blinded, and indeed those in the EGDT arm had a physician devoted to their care. The study was also a single centre study. It was hard to tell which of the interventions had actually made a difference and the use of specialist equipment to measure the ScvO2 was controversial. The consequence of this is that lactate clearance has been suggested as a better parameter to use as it is more easily measured and some studies have shown that it is equivalent to ScvO2 in its efficacy (Jones & Shapiro 2010). It has also been shown to be associated with improved outcomes (Nguyen et al. 2004).

 

How Much Fluid?

Boyd found that a more positive fluid balance in the early stages of resuscitation and after four days was associated with an increased mortality in septic shock (Boyd et al. 2011). This was a multi-centre randomised controlled trial which was based on the Vasopressin in Septic Shock Trial (VASST) (Russell 2008).

The question to be answered is clear in the Boyd study which was an analysis of whether a positive fluid balance after the first twelve hours of resuscitation and over the next four days was associated with an increase in 28 day mortality.

The hypothesis in the study was further developed into the question that both too much and too little fluid would be harmful. For this reason, the patients were further divided into four quartiles depending on how much fluid they had received in the early stages. Correlation between fluid balance and CVP and 28 day mortality were measured using survival analysis[3] and regression analysis[4]. Strength of correlation were given using r values[5]

The population in this study is very relevant to the issue in question in that they all had septic shock.

The study certainly demonstrated that this kind of patient is given a lot of intravenous fluids in the first 12 hours (6.3litres) and over the four days with a final fluid balance (difference between input and output) of 11 litres.

There was a correlation between those in the first quartile (i.e. those who received the least fluid in the first 12 hours) versus those in the 3-4th quartiles (i.e. those who received the greater amount of fluids in the first 12 hours). From the survival analysis it was demonstrated that those in quartile 1 and 2 had survival advantages over those in quartile 4. So it would appear that too much fluid is associated with an increased mortality.

The equally important question of whether the CVP correlates with the patient’s mortality is divided into two conclusions. There is evidence that those patients with the lower CVP had an improved survival but then the CVP does not then correlate with mortality on subsequent days. It may seem that there is a correlation at the early stage of treatment but not in the later stages.

So the Boyd paper possibly leads us to ask as many questions as it answers. Possibly the most important one is how much fluid do we give the septic patient. The Surviving Sepsis guidelines suggest targeting the CVP between 8-12mmHg (Dellinger et al. 2013) but we can demonstrate that this is possibly a flawed measurement. The VASST study found that a more positive fluid balance at 12 hours and a more positive cumulative fluid balance at day 4 increased mortality (Russell 2011) suggesting that lower levels of fluid given to the patient would be better. But Boyd’s study showed that 62% of the patients had a CVP >12mmHg indicating that they had been given much more fluid than needed (Boyd et al. 2011).

Other studies have also found that a lower fluid balance significantly lowered patient mortality (Rosenberg et al; Murphy et al. 2009; Sirvent et al. 2015).

 

Central Venous Pressure as a Parameter

Another area of concern with the Rivers trial was the use of the CVP to determine how much fluid the patient received during their treatment. Using the CVP in such a way assumes that it is a good surrogate for the patient’s fluid volume.

There are some that would suggest it is not a good surrogate for the patients fluid volume (Marik & Cavallazzi 2013; Marik et al. 2011; Cordemans et al. 2012; Osman et al. 2007) and should not, therefore, be one of the targets we aim for.

Marik performed a meta-analysis looking at trials that investigated the ability of the CVP to predict fluid responsiveness (Marik & Cavallazzi 2013). The question was a clearly focused one with both population and interventions being well specified. It is not made explicit in the paper whether the trials involved were all randomised controlled trials, which are considered the gold standard where interventions are being compared. A multi method approach was used with two authors independently searching several databases and bibliographies of all selected articles. There is no indication of how the quality of the included studies were assessed but the results from the statistical analysis were made clear and kept to only two measures- area under the curve and correlation coefficients, which both seemed appropriate in this case. The range of confidence intervals were quite narrow for the area under the curve results, but seem broader for the correlation coefficients, a point the author will come back to later.

Perhaps one of the main issues with this meta-analysis is that the studies included had a mixed group of patients- they were heterogeneous. There was a mix of patients from cardiac surgery, neurosurgery and others described as various. To apply Marik’s findings to those patients with sepsis, i.e. those we are discussing in this paper, therefore becomes harder. He does include some papers that looked at the septic patient, but many were not that specific (CASP 2013).

The definition of fluid responsiveness in these trials was an increase in the patient’s stroke volume[6] which, in most of the studies, was reassessed after the delivery of a 500ml fluid bolus. The increase in the CVP was then compared with the rise in the stroke volume. The search generated 43 studies, twenty two of which were on Intensive Care Unit (ICU) patients and twenty on cardiac surgery patients in the operating room. Area under the curve data (AUC)[7] were available for 33 of the studies and for twenty there was correlation[8] data.

The AUC in the meta-analysis of the studies strongly indicated that there was very little relationship between the change in the CVP and the rise in the stroke volumes of the patients. For those studies done in the ICU the AUC was 0.56 (95% Confidence Interval (CI), 0.52- 0.58)[9]. For those studies done in the operating room the AUC was also 0.56 (95% CI, 0.54- 0.58). The correlation coefficients between CVP and stroke volume was 0.28 (95% CI, 0.16-0.40) in the ICU patients and 0.11% (95% CI, 0.02- 0.21) in those patients in the operating room.

These statistics would both indicate then that there is little relationship between the patients increased stroke volume and a rise in their CVP. There is, however, a large range in the 95% confidence intervals for the correlation coefficients in both sets of patients, perhaps indicating a need for some larger studies in the future.

It is because of the use of this parameter that some would say that all the studies that use it are essentially flawed studies (Marik & Cavallazzi 2013). Others go onto to speculate that using the central venous pressure as a target can also contribute to other problems such as acute kidney injury (Legrand et al. 2013)

 

Current Ongoing Research

When considering, therefore, how much fluid to give to a patient in the Emergency Department (once they have had their initial fluid bolus as suggested by the Surviving Sepsis Campaign (Dellinger et al. 2013)) and, bearing in mind some of the difficulties with some of the measurement tools that we use as discussed above, there have been several important pieces of research over the last 2 to 3 years which will help guide this.

The ProCESS trial (The ProCESS Investigators 2014) was a multi-centre randomised controlled trial which aimed to compare protocol based resuscitation (EGDT) with usual care asking if the protocol which included haemodynamic monitoring, blood transfusions, vasopressors and dobutamine was better than a protocol which did not include these.

Patients included in the study were those with hypotension (systolic blood pressure less than 90 mmHg or required vasopressors after an intravenous fluid challenge) or who had a serum lactate greater than 4 mmol per litre.

The trial broke the patient’s down into three groups: protocol-based EGDT, protocol-based standard therapy, or usual care. The primary outcome was in-hospital mortality at 60 days.

This was a large study which had a final cohort of 1341 patients.

One of the criticisms of this study, which is in the Surviving Sepsis Campaigns response (Surviving Sepsis Campaign 2014), was that the 18% mortality rate in the “usual care” arm is a dramatically different one from that found in the River’s trial in which the mortality from septic shock was 46.5%. These different mortality rates however are potentially a reflection of the improvements in the overall care and detection rates in sepsis over the 14 years between the two studies.

The final conclusion of the ProCESS trial was that protocol-based resuscitation of patients with septic shock did not improve outcomes when compared with usual care.

The ARISE trial (Bailey et al. 2014) was also a multicentre randomised controlled trial which recruited 1600 patients. Unlike the ProCESS trial this was a more straightforward comparison between EGDT and usual care.

The hypothesis being tested in this study was that EGDT would decrease 90 day all-cause mortality among patients presenting to the Emergency Department with early septic shock as compared to usual care. Again, those patients in the EGDT arm of the study were treated using the algorithm in the River’s trial. Those receiving the usual resuscitation care were not permitted, amongst other things, to have a ScVO2 monitor.

In the primary outcome there was no significant difference between the two groups; 18.6% EGDT vs. 18.8% usual care group (RR[10] 0.98, p, 0.9), however the volume of fluid delivered during the first six hours in the EGDT group was just over 2 L more than in the usual care group.

The final conclusion was that EGDT did not reduce mortality at 90 days in those critically ill patients presenting to the Emergency Department.

The ProMISe trial had a similar design to the ARISE trial in that it was also a multicentre randomised controlled trial which recruited 1260 patients into two groups, comparing EGDT and usual care. Again the primary outcome was 90 day mortality.

Like the other two trials this was once again a test of the algorithm used in the River’s trial and its outcomes were similar. The 90 day mortality in the EGDT was 29.5% and in the usual care group was 29.2%. The relative risk was 1.01 which would infer that there was little difference to be had between the two types of treatment.

Perhaps the best way to conclude the discussion is to look at one of the most recent meta-analysis conducted on the subject by a group of Chinese physicians (Zhang et al. 2015). The CASP checklist helps to evaluate the value of this meta-analysis. The population for the analysis was clearly identified, being those patients with severe sepsis or septic shock who were treated using EGDT. The EGDT was either standard as defined by Rivers (Rivers & Nguyen 2001), or slightly modified and also looked at papers that compared EGDT with lactate clearance as in the Jones study (Jones & Shapiro 2010).

They looked at a total of ten randomised controlled trials. This does seem like a small number but it does include all the trials which are considered to have influenced practice in the last ten years and the decision taken which to include was taken by two independent reviewers. Bias in all the trials was assessed, and the main point here is that none of the trails were double-blinded. The nature of the problem would make this an ethical issue, so it was a weakness in all of the trials that was accepted.

It is an interesting point that overall mortality in the ten studies was 30.4%, with 29.1% mortality in the EGDT group versus 32% in the control group. This perhaps adds weight to the argument that the cohort of patients in the Rivers trial was much sicker than those sampled in later studies, as the mortality in this study for those patients receiving standard care was 46.5%. The single centre in the Rivers trial was in a low income, inner city area and it has been speculated that, because of the American health insurance system, patients were leaving it longer to come to hospital before receiving treatment. These patients were consequently sicker as a result. This would make the 16% difference between the two groups in the Rivers trial a difficult one to reproduce.

When referring to the amount of fluid given, then there was significantly more fluid given in the EGDT group in the first six hours compared to the usual care group. Bearing in mind the comments already made about the dangers of excess fluids (Annane 2013; Raghunathan et al. 2014; McDermid et al. 2014), this is an important finding when deciding which method to use when delivering fluids.

The overall conclusion of the meta-analysis probably still leaves us with some questions. They show that the available trials do not show a significant difference in mortality between the control group and the EGDT group, but does show that there is a significant difference in the mortality rate when comparing EGDT with lactate clearance.

 

Conclusions

Was the patient in this case given the treatment which current guidelines might dictate? Fluid boluses were given, but it was unclear exactly how much was the actual target. The patient’s weight was an estimated one only, which is often the case in the Emergency Department. Looking back at the fluid balance chart it would seem that the patient was given much more than the initial fluid bolus of 30ml/kg. It was also unclear what parameter was being used to assess the patient’s responsiveness to the treatment. Several arterial blood gases were taken to assess lactate clearance (Jones & Shapiro 2010), which did show signs of responding but it was several hours before the patient had a urinary catheter inserted to assess their kidney function (Dellinger et al. 2013). From review of the records the patient was possibly overloaded with fluids before being started on other drugs to improve the blood pressure (Polderman & Varon 2015).

However it would be fair to say that many of the patients that do present in septic shock in the author’s experience are usually treated with many elements from the EGDT algorithm. Fluids are often given above and beyond the normal bolus recommended as long as the patient is responsive. This responsiveness to the fluid is often measured by very simple measures such as whether there systolic blood pressure improves. Arterial lines are very often placed along with a central line which is used, not necessarily to monitor the patient’s response, but in order to be able to give the volumes of fluids required as expediently as necessary. The use of vasopressors is common in this group of patients and it would be difficult to say that this drug is given when a specific volume has been delivered or a specific parameter has not been achieved.

The only element of the Rivers algorithm that the author has never seen in use is the ScVO2 catheter. This would be used to measure the mixed venous oxygen saturations but, in his study, Jones has inferred that lactate clearance is just as good a parameter to aim for. (Jones & Shapiro 2010).

Perhaps then, current evidence would seem to indicate that fluids should be given carefully, using lactate clearance as a marker for improvement. The practitioner should always be aware of giving too much fluid, where possible using measurements of cardiac output to help decide when vasopressors may be necessary.

The author has many years’ experience in the critical care environment, and whilst in that role was often required to intervene in the care of the very sick patient in the Emergency Medicine Department. I am now a practitioner in the Emergency Medicine Department where care of the critically ill is only a part of my workload. I feel that this gives me some insight into the way we care for this type of patient, which can be beneficial.

Audit of current practice in the resuscitation room may allow us to focus on perhaps one of the areas for improvement in this group of patients which would be a stricter monitoring of their response to the treatments being put into place (ensuring that the clinicians prescribing the treatments are also make clear the goals to be achieved) and an earlier introduction of vasopressor drugs to elevate the blood pressure.

Beyond the initial sepsis six guidelines, which deal with the initial resuscitation, there is no trust guideline or policy about how much fluid to give to the sick patient. With an audit of current practice within the department and the interaction between the various specialities when treating this type of patient, it may become possible to assess the way forward using the copious amount of good research on the subject.

[1] Absolute risk reduction is the change in risk of a given activity or treatment in relation to a control activity or treatment.

[2] The number needed to treat is the average number of people who need to be treated to prevent one additional bad outcome. The lower the number the better. 1 would be best as this would be where everyone improves with the treatment and no one improves with the control.

[3] Survival analysis is used on data with time to events. This is often used with time to death but need not always be this variable.

[4] Regression analysis allows us to model the relationship between two or more variables.

[5] r values range from -1 (very strong negative relationship) to +1 (very strong positive relationship).

[6] Stroke volume is how much blood is ejected from the ventricles with each contraction. This can be affected by how much fluid is in the heart at the time of the contraction (preload), the strength of that contraction (contractility), and the relative resistance of the circulation opposing that contraction (afterload).

[7] The area under the curve (AUC) is a measure of how well a parameter can distinguish between two groups. The closer the value is to one, the stronger the relationship between the two variables. 0.5 would indicate that the change in the relationship is most likely only due to chance.

[8] Correlation is a measure that shows to what extent two or more variables change together. A positive correlation would indicate that those variables change at the same rate and in the same direction. A negative correlation would indicate that there was little relationship in the way they changed.

[9] A confidence Interval is a reflection of the fact that the data being sampled is from a larger overall population, and that the mean we have obtained is within the range we would expect if we could sample the whole population from which our sample is obtained. A smaller confidence interval shows greater statistical evidence that we have obtained a mean reflective of our population- this is affected by our sample size as a larger sample size should lead to less sampling error.

[10] RR- relative risk. When a treatment has an RR greater than one, the risk of a bad outcome is increased by the treatment; when the RR is less than one, the risk of a bad outcome is decreased, meaning that the treatment is likely to do good. When the RR is exactly or close to 1, the risk is unchanged.

 

Al, R. et, 2009. Review of a large clinical series: association of cumulative fluid balance on outcome in acute lung injury: a retrospective review of the ARDSnet tidal volume study cohort. Journal of Intensive Care Medicine, 24, pp.35–46.

Annane, D., 2013. Effects of Fluid Resuscitation With Colloids vs Crystalloids on Mortality in Critically Ill Patients Presenting With Hypovolemic Shock. Jama, pp.1–9.

Bailey, M. et al., 2014. Goal-Directed Resuscitation for Patients with Early Septic Shock. New England Journal of Medicine, (371), pp.1496–1506.

Boyd, J.H. et al., 2011. Fluid resuscitation in septic shock: a positive fluid balance and elevated central venous pressure are associated with increased mortality. Critical care medicine, 39(2), pp.259–65.

Campaign, S.S., 2014. Surviving Sepsis Campaign Responds to ProCESS Trial. The New England journal of medicine, (May), pp.1–2.

CASP, 2013. Systematic Reviews in the Social Sciences.

Cordemans, C. et al., 2012. Fluid management in critically ill patients: the role of extravascular lung water, abdominal hypertension, capillary leak, and fluid balance. Annals of intensive care, 2(Suppl 1 Diagnosis and management of intra-abdominal hyperten), p.S1.

Crowe, C. a. et al., 2010. Evaluation of a modified early goal-directed therapy protocol. American Journal of Emergency Medicine, 28(6), pp.689–693.

Dellinger, R.P. et al., 2013. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012. Intensive care medicine, 39(2), pp.165–228.

Dellinger, R.P. et al., 2008. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008.,

Edwards, M.R. & Mythen, M.G., 2014. Fluid therapy in critical illness. , pp.1–9.

Jones, A. & Shapiro, N., 2010. Lactate clearance vs central venous oxygen saturation as goals of early sepsis therapy: a randomized clinical trial. JAMA : the journal of the American Medical Association, 303(8), pp.739–746.

Jones, A.E. et al., 2007. Prospective external validation of the clinical effectiveness of an emergency department-based early goal-directed therapy protocol for severe sepsis and septic shock. Chest, 132(2), pp.425–432.

Legrand, M. et al., 2013. Association between systemic hemodynamics and septic acute kidney injury in critically ill patients: a retrospective observational study. Critical care (London, England), 17(6), p.R278.

Marik, P. & Bellomo, R., 2013. Lactate clearance as a target of therapy in sepsis: A flawed paradigm. OA Critical Care, 1(1), pp.1–6.

Marik, P.E., 2014. Iatrogenic salt water drowning and the hazards of a high central venous pressure. Annals of intensive care, 4, p.21.

Marik, P.E. & Cavallazzi, R., 2013. Does the central venous pressure predict fluid responsiveness? An updated meta-analysis and a plea for some common sense. Critical care medicine, 41(7), pp.1774–81.

Marik, P.E., Monnet, X. & Teboul, J.-L., 2011. Hemodynamic parameters to guide fluid therapy. Annals of intensive care, 1(1), p.1.

McDermid, R.C. et al., 2014. Controversies in fluid therapy: Type, dose and toxicity. World journal of critical care medicine, 3(1), pp.24–33.

Murphy, C. V et al., 2009. The importance of fluid management in acute lung injury secondary to septic shock. Chest, 136(1), pp.102–9.

Nguyen, H.B. et al., 2004. Early lactate clearance is associated with improved outcome in severe sepsis and septic shock*. Critical Care Medicine, 32(8), pp.1637–1642.

NICE, 2014. Sepsis | Guidance and guidelines | NICE. Available at: http://www.nice.org.uk/guidance/indevelopment/gid-cgwave0686 [Accessed June 12, 2015].

Osman, D. et al., 2007. Cardiac filling pressures are not appropriate to predict hemodynamic response to volume challenge. Critical care medicine, 35(1), pp.64–68.

Mouncey, et al, 2015. Trial of Early, Goal-Directed Resuscitation for Septic Shock. New England Journal of Medicine, (372), pp.1301–1311.

Patocka, C., Turner, J. & Lang, E., 2010. Quantitative resuscitation in sepsis. Canadian Journal of Emergency Medicine, 12(2), pp.150–153.

Polderman, K.H. & Varon, J., 2015. Do not drown the patient: appropriate fluid management in critical illness. The American Journal of Emergency Medicine, 33(3), pp.9–11.

Raghunathan, K. et al., 2014. Choice of fluid in acute illness: what should be given? An international consensus. British Journal of Anaesthesia, 113(5), pp.772–783.

Rivers, E.P. & Nguyen, B., 2001. Early goal directed therapy in the treatment of severe sepsis and septic shock. The New England Journal of Medicine, 345(19), pp.1368–1377.

Russell, J., 2008. Vasopressin versus Norepinephrine Infusion in Patients with Septic Shock. New England Journal of Medicine, 358, pp.877–887.

Russell, J. a, 2011. Bench-to-bedside review: Vasopressin in the management of septic shock. Critical care (London, England), 15(4), p.226.

Sirvent, J.-M. et al., 2015. Fluid balance in sepsis and septic shock as a determining factor of mortality. The American journal of emergency medicine, 33(2), pp.186–9.

The ProCESS Investigators, 2014. A Randomized Trial of Protocol-Based Care for Early Septic Shock. New England Journal of Medicine, (370), pp.1683–1693.

Xue, Z.W.Z.B.J.J.Y., 2010. The effect of early goal directed therapy in the treatment of critical patients with severe sepsis/septic shock. Chinese Critical Care Medicine, 6, pp.331–334.

Zhang, L. et al., 2015. Early goal-directed therapy in the management of severe sepsis or septic shock in adults: a meta-analysis of randomized controlled trials. BMC Medicine, 13(1), pp.1–12.

 

Leave a comment

Your email address will not be published. Required fields are marked *