Comorbidities of a Patient When Monitoring the Colloid Osmotic Pressure During Cardiothoracic Surgery

Is there a benefit of understanding the comorbidities of a patient when monitoring the colloid osmotic pressure during having cardiothoracic surgery with cardiopulmonary bypass?

Introduction

Colloid osmotic pressure and hydrostatic pressure are the primary forces that control the exchange of fluid between the aqueous plasma solutions and the interstitial fluids. Protein concentrations in plasma generate a pressure that draws water out of the interstitium in to the plasma, this is known as colloid osmotic pressure.  The capillary wall acts in this process as the filtration barrier (Guth, et al., 2015). The colloid osmotic pressure of plasma and interstitial fluid play important roles in transvascular fluid exchange (Golab, 2011). Colloid osmotic pressure affects the movement of fluid from the blood vessels into the interstitial space (fluid-filled spaces surrounding cells), organs and cells.  A low colloid osmotic pressure indicates that the blood is less viscous and therefore more fluid is available to move through the membrane out of the blood vessels. Under normal physiological conditions, fluid gain and loss from the plasma are closely balanced and there is little or no change in plasma and interstitial volumes (Boldt, 2009).

All organs are affected by the movement of fluid, including the lungs which can become stiff, difficult to ventilate and result in a lower oxygen transfer. The heart is also affected by myocardial water which can result in ventricular stiffness resulting in systolic and diastolic being impaired.

Excess fluid and large protein molecules which have accumulated in the interstitial spaces are normally removed by the lymphatic drainage system. The main driver of fluid movement through the lymphatic system is the lymphatic pump. Whilst this has a small amount of contractility driven by muscles in the walls of the vessels, the majority of the pumping is driven by compression of the vessels by external factors.  These are in order of importance: contraction of the surrounding muscles, movements of the body, arterial pulsation and compression of the body from external sources. In the majority of cardiopulmonary bypass cases all of these drivers are either severely attenuated or not present.  This can result in the lymphatic drainage system being unable to remove fluid accumulated during low colloid osmotic pressure found in cardiopulmonary bypass (Guyton, 1982).

The amount of water in the blood vessels is controlled by the kidney and is under regulation by the hormone ADH. High ADH levels cause the kidney to reabsorb an increased amount of water back into the blood plasma. If the kidneys are damaged this could result in a large amount of fluid being reabsorbed into the circulation and a decreased usual colloid osmotic pressure. Some external factors can change the working of the kidneys and the amount of fluid removed from circulation including; temperature, exercise, pain, stress, fluid and salt intake. Some drugs including diuretics and antidiuretics  among; alcohol and nicotine can also have an effect.  If excessive amounts of fluid move into the cells then there will be problems with the function of that cell and in some cases lysis can occur.

Colloid osmotic pressure is measured using an Onkometer. The Onkometer has a semi permeable membrane usually made of (cellulose triacetate) and will take between one and two minutes to give a result. The sample is placed onto the membrane, and is then pulled through the membrane and a further sample added around 15 seconds later. The machine sounds an alert when the final result has been given. A higher value indicates an increased colloid osmotic pressure; so fluid will stay within the blood vessels and not move into the interstitial space. A lower value indicates a diminished colloid osmotic pressure and fluid will move out of the blood vessels (London, 1990). The reading is given in mmHg.

The osmotic pressure can be affected by the osmolarity. Osmolarity is a measurement of the amount of solute (usually sodium, potassium, urea and glucose) that has been concentrated into the fluid. At times this measurement can be unreliable in patients with specific illnesses affecting the kidney. Hydrostatic pressure is the opposing force to oncotic pressure. This is the pressure blood has on the blood vessels. As the fluid moves out of the blood vessels inside the capillaries the hydrostatic pressure will drop. This will then allow fluid to move back into the vessels within the capillary bed.

Cardiopulmonary bypass is the procedure by which the function of the heart and lungs are performed by a machine in order to allow the heart and lungs to be stopped to permit the surgeon a still and bloodless operating field in order to perform the open heart surgery. The cardiopulmonary bypass machine is composed components  connected by tubing through which blood is pumped. Before the machine can be combined with a patients’ circulation fluid must displace all air to provide volume for the extra-corporeal circulation and prevent air from reaching a patient. The types of fluids that are used to prime the circuit can vary between hospital departments and surgeons or perfusionists’ might have specific preferences for fluids. These liquids will combine with the patients’ blood and so disturb the balance between the regulatory forces of plasma and the interstitial fluid exchange (Guthe, 2015).

In neonatal and infant cardiac surgery with cardiopulmonary bypass, the inflammatory response to cardiopulmonary bypass (Tassani, 2007), haemodilution with reduction of plasma albumin concentration and low colloid osmotic pressure (Farstad, 2005; Jonas, 2007) are the main factors associated with oedema and post-operative weight gain. Any dilution is usually more extreme for paediatric patients as the amount of fluid, and surface area of the bypass tubing is greater when compared to the patients weight and circulating volume.

The priming fluids used through cardiopulmonary bypass can be divided into two groups depending on their properties; crystalloid and colloid. Crystalloids contain minerals that are all water soluble, and colloids contain large molecules that are not water soluble, for example gelatine. The more common types of crystalloid and colloid have a lower colloid osmotic pressure than a patients’ baseline, so placing a patient on cardiopulmonary bypass (and so mixing these with their circulation) is expected to lower the colloid osmotic pressure.

Research has been carried out into the safety of using both colloids and crystalloids, and the difference seen in the patients’ outcomes when using either or both in the cardiopulmonary bypass machine. The way this can affect the colloid osmotic pressure through and after the surgery has been monitored and presented. The different types of fluids have been compared during cardiopulmonary bypass below. Colloids are more expensive than crystalloids, but in terms of expenditures associated with cardiac surgery it is a small cost. The associated drop in colloid osmotic pressure has been linked to various health conditions including oedema, respiratory problems and death.

In our institution we have published data on colloid osmotic pressure and cardiopulmonary bypass which show a significantly lower baseline colloid osmotic pressure in our patient group than in other studies of congenital heart disease patients but comparable with other groups of sick neonates. The prime solution had the colloid osmotic pressure raised through the addition of gelatine, to approximately match the patients’ baseline level. It was then increased during and post bypass, by filtration. Our results showed no correlation between the colloid osmotic pressure during or post cardiopulmonary bypass with length of ventilation or hospital stay, but did demonstrate a link between pre-operative albumin levels and post-operative ventilation time. They also showed a correlation between disease severity and baseline colloid osmotic pressure, with a lower pressure being linked to increased severity (Crook, et al. 2017).

I am looking at the effect that a change in colloid osmotic pressure has on a patient during cardiopulmonary bypass and which fluids can cause this change and risks associated with this drop. Research has been carried out on patients of differing ages and with various co-morbidities. I would like to understand if it will be beneficial to target the management of colloid osmotic pressure during surgery and tailor it to the patient. I will look to understand what might cause colloid osmotic pressure changes to be more dangerous for a patient and present my suggestions at the end, including which patients who should be more closely monitored or those who should have a particular management strategy.

Methods

Search methods

I used the PRISMA guidelines to search for papers that had relevant information that I can analyse. I used PubMed and the MeSH term that I searched was:

“Colloid osmotic pressure” or “COP” and “Cardiopulmonary bypass” or “CPB”

Which would give papers that were specific to the topic that I was investigating, no filters were added. No language or date restrictions were added. Published papers were included in adult and paediatric settings, for patients undergoing cardiac surgery and all others.

Results

Search results

A total of 92 papers were returned from PubMed, and 52 papers were accessed from the Great Ormond Street Hospital Perfusion departmental library on colloid osmotic pressure. 16 duplicates were removed. Abstracts were screened for eligibility and 71 papers were found to be relevant for the review. The full version of these papers were accessed and a further 21 papers were excluded as not relevant. This resulted in 50 papers which have been included. The reference lists of these were also investigated for relevant papers that might provide further information or research. This is presented as a PRISMA diagram (Figure 1).

Figure 1: The PRISMA flow diagram showing article selection.

I excluded papers that were not relevant to the topic that I am searching, including a paper that investigated fluids given only in the post-operative period, one that was retracted from the journal and another that investigated the safety of the priming fluids.

Table 1. Studies that were included in my review, with the group that was studied and important findings summarised.

Author	Patient group studied	Outcomes recorded	Key results	Comments
Sola et al. (1981), Crit Care Med.	199 new-born term infants;   Group 1: 99 natural birth Group 2: 40 born via caesarean section Group 3: 60 prematurely born infants	Colloid osmotic pressure	There is a significant difference between colloid osmotic pressure depending on the birth method (19.4 mmHg ± 2.2 for natural birth, 16.0 ±2.1 for caesarean section). Colloid osmotic pressure was lowest for premature patients (12.5 mmHg ± 2.5).   Surgery (not cardiopulmonary bypass) decreases colloid osmotic pressure by 32%. Crystalloid transfusion decreases colloid osmotic pressure by 22%.	No separation provided of preterm patients by method of delivery to understand if the same patterns are seen. No analysis of gestation time for natural and caesarean births is given.
Zimmermann et al. (1997), Arch Pediatr.	84 newborn   Group 1: Intubated and ventilated Group 2: without respiratory distress	Colloid osmotic pressure	Low birth colloid osmotic pressure correlates with respiratory distress and low patient blood pressure.
Bhat et al. (1981), Crit Care Med.	94 infant patients.   Group 1: 18 healthy term Group 2: 21 sick term Group 3: 5 healthy preterm Group 4: 50 sick preterm	Colloid osmotic pressure	In critically ill preterm patients, survivors had an increase in colloid osmotic pressure. Patients that did not survive saw colloid osmotic pressure drop to <10mmHg.	The group size was not consistent. These were not cardiac patients.
Guthe et al. (2015), PLos ONE.	99 patients, 2-10 years old undergoing tonsillectomy, adenotomy, and tympanic paracentesis.	Colloid osmotic pressure	Colloid osmotic pressure increases with age and is comparable to recorded adult levels.	All patients had been fasted and were in hospital for surgery and measurements were taken after anaesthesia. This is not a projection of a healthy child.
Ohqvist et al. (1981), Cand J Thorac Cardiocasc Surg.	14 adults undergoing aortic valve replacement. Group 1: 2 litres of Ringerdex prime.   Group 2: 1.8 litres of Ringerdex and 40g albumin prime.	Colloid osmotic pressure, albumin in plasma, blood erythrocyte, volume fraction, PaO2	Crystalloid priming fluid causes a drop in colloid osmotic pressure by 50%. With added albumin the colloid osmotic pressure remained unchanged.	Small study group of 14 patients. It would be useful to see the values in regards of a patients body surface are as dilution would be more extreme for a patient with lower body surface are with regard to the same circuit size and fluid dilution.
Sanchez et al. (1982), Br j Anaesth.	16 patients undergoing cardiopulmonary bypass	Colloid osmotic pressure, right to left shunt, alveolar-arterial PO2 difference.	Colloid osmotic pressure returned to normal after 6 hours. Drop in colloid osmotic pressure can be tolerated without pulmonary oedema.	Small study group, only 16 patients. Many variables seen in patient groups and length of cardiopulmonary bypass.
Yeoman et al. (1981), Resuscitation.	20 adults undergoing cardiopulmonary bypass, all with a crystalloid prime.	Colloid osmotic pressure	Using crystalloid prime results in a 45% drop in colloid osmotic pressure.   Colloid osmotic pressure returns to normal after 24 hours.	No comparison group. Wide range of cardiopulmonary bypass length and surgery types. Three patients died.
London et al. (1990), Journal of Cardiothoracic Anesthesia.			On bypass colloid osmotic pressure is between 8-15 mmHg. Low colloid osmotic pressure prime causes oedema and drops colloid osmotic pressure by 50%. Cardioplegia reduces this further. Membrane permeability is increased by compliment activation and tissue ischemia.	Some information does not have references attached.
Jansen et al. (1996), British Journal of Anaestesia.	20 adults undergoing coronary artery bypass grafting surgery.   Group 1: Crystalloid prime with no colloids during cardiopulmonary bypass Group 2: Crystalloid and colloid prime with colloid infusions during cardiopulmonary bypass.	Fluid balance, blood loss, means of temperature during surgery.   Relative change in post-operative PO2/FtO2 ratio and intubation time.	Colloid osmotic pressure was lower in group 1 during and in the post-operative period. The colloid osmotic pressure for group 2 remained comparable to baseline during bypass.   Group 1 needed more fluid infusions during surgery and had a more positive fluid balance. Respiratory function was not significantly different. Colloid prime took 5 days to leave hospital, and the crystalloid group 7.	Small patient group studied.
Sussmane et al. (2001), Critical care.	37 healthy children aged 1-11 months old.	Plasma colloid osmotic pressure every month.	Plasma colloid osmotic pressure increases with age from 1-9 months and is comparable to adult colloid osmotic pressure.	Patients were separated into ethnic groups, and most investigated were Hispanic patients. One black female was recorded, with statistically significant colloid osmotic pressure results.
Mota et al. (2008), Brazilian Journal of Cardiovascular Surgery.	Review paper		Many factors will affect the outcomes for cardiopulmonary bypass; elderly, impregnated circuit, pulsatile flow, inflammatory response, diabetes. How to improve lung function.
Morisette. (1977), CMA Journal.	Review paper		Albumin infusions increase the colloid osmotic pressure. No correlation between colloid osmotic pressure and pH or haematocrit.   Colloid osmotic pressure is linked to survival.
Kirov. Measurement of EVLW at the bedside: Why and how?	Information paper		How to measure Extravascular lung water. Why, what it can be used for. Allows the understanding of lung oedema formation.
Lumb. (1987), Ann Surg.	20 adults having coronary artery bypass grafting surgery.   Group 1: 150ml Albumin added to prime. Group 2: 500ml Hespan added to prime	Serum oncotic pressure before, during and after surgery. Extravascular lung water measurements.   Colloid osmotic pressure of the prime.	Increased extravascular lung water measurements are linked to lower serum oncotic pressure.	Group 2 has 350ml extra priming volume.   Does not show the outcomes of the extravascular lung water measurements, how this relates to length of stay. Some baseline patient values are not consistent.
Wendt-Hornickle et al. (2011), Vet anaesth Analg.	14 horses.   Group 1: Ringers solution 5-10ml /kg/hour during anaesthesia. Group 2: 2.5mL/kg of 6% Hetastarch over one hour of anaesthesia.	Changes in colloid osmotic pressure, total protein and osmolality during anaesthesia.	No benefit to the colloid osmotic pressure when Hetastarch is given alongside the Ringers solution.	Colloid osmotic pressure was higher in the Ringers group before anaesthetic when compared to Hetastarch group before and after, so the two groups do not have comparable baselines. This is an animal model studied.
Stanton et al. (2005), Chest.	Group 1: 97 patients having coronary artery bypass grafting carried out using cardiopulmonary bypass. Group 2: 100 patients having off pump coronary artery bypass grafting.	Fluid balance, fluid intake, respiratory, compliance, haemodynamics, chest radiograph scoring, gas exchange, time to extubation, spirometry, pulmonary complications.	Off pump surgery resulted in a better gas exchange and earlier extubation. No difference seen in radiographs, spirometry, rates of death, pneumonia, pleural effusion or pulmonary oedema. Off pump patients were extubated earlier.	Good methodology. Single surgeon carried out all operations.
Elliot. (1993), Ann Thorac Surg.	Review of infants being filtered after cardiopulmonary bypass surgery.		Modified ultrafiltration increases haematocrit by 35-40%, removes accumulation of water, reduces blood loss and need for blood transfusion. Also increases myocardial contractility.
Tonnesen et al. (1977) Crit Care Med.	84 consecutive intensive care patients (not post cardiac surgery).	Colloid osmotic pressure	Finds a relationship between colloid osmotic pressure and survival, and identifies a figure for survival (15.0 mmHg = 50% survival rate).	Patients are not separated depending on their primary diagnosis.
Morissette et al. (1975), Crit Care Med.	99 consecutive patients in cardiopulmonary failure (not post cardiac surgery).	Colloid osmotic pressure	Patients with colloid osmotic pressure under 10.5 mmHg died; 21 % patients. Survival increased with colloid osmotic pressures between 10-19 mmHg. Colloid osmotic pressure over 17 mmHg is ‘safe level’.	Does not link cause of cardiopulmonary failure with colloid osmotic pressure.
Riegger et al. (2002), Crit Care Med	86 patients under 14kg, aged 3 days- 4 years.   Group 1: 44 patients, albumin prime Group 2:  44 patients crystalloid prime.	Fluid balance, colloid osmotic pressure, haematocrit.	After cardiopulmonary bypass Group 1 had a net negative fluid balance. Group 2 had a net positive fluid balance.   24 hours post-surgery there was no difference in colloid osmotic pressure and haematocrit and both groups returned to baseline after 48 hours.
Ziyaeifard et al. (2014), Res Crdiovasc Med.		Haemodynamics, pulmonary function, decreases blood transfusion, post-operative blood loss, total body water and morbidity.	Modified ultrafultration improves haemodynamics, pulmonary function and decreases blood transfusion, post-operative blood loss, total body water and morbidity.
Tassani et al. (2007), J Cardiothorac Vasc Anesth.	11 neonatal patients undergoing surgery to repair transposition of the great arteries.	Colloid osmotic pressure, plasma interleukin-6, fractional escape rate	Low colloid osmotic pressure is association with oedema. No link found between capillary leak after cardiopulmonary bypass.	Small patient group.
Loeffelbein et al. (2008), Eur J Cardiothorac Surg.	20 neonates or infants.   Group 1: 10 with a blood prime including fresh frozen plasma. Group 2: 10 patients had a blood prime, fresh frozen plasma and albumin.	Weights, fluid balance, transfusion, colloid osmotic pressure, inflammatory parameters, renal function.	Albumin added to the priming fluid results in less weight gain (caused by less capillary leakage).	Small patient group.
Golab et al. (2010), Eur J Cardiothorac Surg	70 patients under 10kg   Group 1: 0.5g/kg of human albumin in the prime and albumin added to maintain the colloid osmotic pressure over 15.0mmHg Group 2: 50% of the prime was human albumin and the colloid osmotic pressure was maintained over 18.0mmHg	Weight gain, colloid osmotic pressure, albumin concentration, fluid transfusion, blood loss, urine production, haemodynamic results.	Group two (higher colloid osmotic pressure sustained) results in lower ventilation needs and lower plasma lactate readings.	By adding albumin to the prime and aiming for a higher colloid osmotic pressure will result in higher albumin levels and a higher colloid osmotic pressure.
Mehlhorn et al. (1998), Cardiovasc Surg.	6 dogs were put onto cardiopulmonary bypass using cardioplegia that has a higher colloid osmotic pressure.		Cardioplegia causes oedema. Increasing the colloid osmotic pressure of cardioplegia will result in less myocardial oedema and prevents cardiac dysfunction.	Small sample size and dogs used in the study.   This was not compared with cardioplegia that has a normal colloid osmotic pressure.
Darling et al. (2000), Perfusion	20 patients   Group 1: 10 with the departmental standard circuit and protocols. Group 2: 10 patients, after modification of protocol changes.	Osmolality, oncotic pressure, total protein, haematocrit, glucose and electrolytes.	A colloid osmotic pressure below 10 results in a low prognosis. Cardiopulmonary bypass priming fluids do not affect the colloid osmotic pressure.
Miao et al. (2014b), Perfusion	80 patients;   Group 1: 10-20ml /kg of 4%  gelofusine. Group 2: 1-2 units of fresh frozen plasma.	Chest tube drainage, transfusion, drug dosages. Ventilation time, intensive care stay, hospital stay.	Using prophylactic fresh frozen plasma in the prime is not beneficial over using gelofusine. Recovery time, post-op bleeding, transfusion requirements and drug usage is comparable between groups.
Golab et al. (2009), Interact Cardiovasc Thorac Surg.	73 patients weighing under 10kg.	Colloid osmotic pressure, blood loss, urine production, transfusion requirements, haematological data and ventilation length.	Low pre-operative colloid osmotic pressure will result in low post-operative colloid osmotic pressure (<15mmHg). Suggests a colloid osmotic pressure strategy based on individual patient.   Using human albumin during bypass to increase the colloid osmotic pressure is not effective. Fluids given during induction are an important cause of low colloid osmotic pressure.	Patients have been all grouped together, it might be useful to separate between 3-5kg, 5-7kg and 7-10kg as dilution will have a different effect.
Haneda et al. (1985), Tohoku J. exp. Med.	55 infants undergoing a surgical repair of transposition of the great arteries   Group 1. Ringers based prime (1975-1980). Group 2. Prime comprised of whole blood and plasma (1981-1983).	Colloid osmotic pressure, fluid balance and type, hospital stay,	Lower fluid balance, CICU stay and mortality is related to patients who had a prime consisting of whole blood and plasma.   Colloid prime reduces fluid accumulation. Colloid osmotic pressure is linked to survival.	Perfusion technology would have developed between the first patients in the study-1975 and the final patients in 1983.
Hindman et al. (1990), Anestesiology.	14 rabbits   Group 1: 7 rabbits had a Hydroxyethyl starch prime, maintained normal colloid osmotic pressure Group 2: 7 rabbits had prime leading to low colloid osmotic pressure. Group 3: 9 rabbits were used as a control, and received no cardiopulmonary bypass. Just had measurements taken.	Temperate, central venous pressure, mean arterial pressure, blood chemistry, haemodynamics	Group 2 required extra fluid and bicarbonate through cardiopulmonary bypass. Tissue water was identical in group 1 and 3.   Brain fluid measurements remain consistant during cardiopulmonary bypass.	Study on small amount of rabbits.
Blunt et al. (1998), Anaesthesia.	145 patients in adult intensive care unit; 66 non survivors, 79 survivors.	Serum albumin, colloid osmotic pressure,	No correlation found between colloid osmotic pressure and survival.   Lower serum albumin levels are correlated to lower survival. Serum albumin usually makes up 40% of the colloid osmotic pressure, in the study patients albumin contributed 17% of the colloid osmotic pressure.
Hoeft et al. (1991), Br j Anaesth.	20 Coronary artery bypass grafting   Group 1: Ringers cardiopulmonary bypass prime. Group 2: Albumin and a nearly physiological colloid osmotic pressure.	Colloid osmotic pressure, trans capillary gradient and pulmonary wedge pressure.	Group 1 resulted in drop of 50% colloid osmotic pressure pre bypass to on pump, and increase of extra vascular lung water by 60%. Group 2 resulted in a drop of 30% of colloid osmotic pressure and only slightly increased extravascular lung water. No difference in haemodynamic and respiratory data post operatively.	Small patient numbers.
Liu et al. (2011), Eur J Cardiothoracic Surg.			Suggests a difference between adult and paediatric colloid osmotic pressure-maybe because of cyanotic and acyanotic patients.   Suggests using modified ultrafiltration to increase colloid osmotic pressure post-operatively.
Kamra et al. (2013), Perfusion	20 CABG patients   Group 1: 2.5 g of human albumin in the prime Group 2: protein free.	Blood samples for platelet function, platelet count and haemoglobin.	Using albumin results in a lower amount of fluid in the chest drains.	Small sample size. It would be useful if patients were analysed based on co-morbidities. Some unusual data.
Tigchelaar et al. (1998), Perfusion	33 patients undergoing cardiopulmonary bypass,   Group 1 (10 patients): 2.5% HES starch Group 2 (11 patients): 3% gelatin Group 3 (12 patients): 4% Human albumin	Total blood loss, fluid balance, origin of grafts, CVP after protamine,	2.5% Hydroxyethyl is a suitable alternative for human albumin.   Less blood needed in human albumin group (not statistically significant). Gelofusine group had higher colloid osmotic pressure.
Golab et al. (2014), Perfusion	55 patients under 10kg, who were filtered to target the haematocrit.	Post-operative urine production, blood loss, length of stay in intensive care and in hospital. NIRS, haematocrit, colloid osmotic pressure, blood products used, haemoglobin, haematocrit, platelet count and lactate levels.	Filtration did not affect duration of ventilation, length of need for chest drains, length of ICU and hospital stay.   Filtration resulted in higher colloid osmotic pressure.
Shin’oka et al. (1998), The Society of Thoracic surgeons.	26 piglets   Group 1: Control group; blood and crystalloid prime and haematocrit of 20%. Group 2: Blood and Hydroxyethyl prime, haematocrit of 20%. Group 3: Blood and pentafraction prime, haematocrit of 20% Group 4: Blood and crystalloid prime with 10 minutes of modified ultrafiltration Group 5: Whole blood prime and a haematocrit of 30%.	Body weight gain, neurological deficit.	Higher haematocrit and colloid osmotic pressure improve cerebral recovery after deep hypothermic circulatory arrest.   Higher haematocrit does not reduce total body oedema. It is more effective to have a higher haematocrit and colloid osmotic pressure in the prime than to modified ultrafiltration after. Group 5 had the best neurological outcomes	One of the only studies that measures neurological deficit. Small sample size; 5/6 in each study group. Animal study model.
Farstad et al. (2005), J Thorac Cardiovasc Surg.	24 piglets   Group 1: 10 individuals control group Group 2: 7 individuals, albumin group Group 3: 7 individuals HES group	Fluids added, plasma volume, colloid osmotic pressure in interstitial fluid and plasma. Hct levels and tissue water content.	Albumin is more effective than HES in limiting the amount of cold induced fluid shift during cardiopulmonary bypass. Both are better than crystalloid prime.	Animal model, small number in each group.
TÖrnudd et al. (2014), Clinics	10 patients	Haemoglobin, serum albumin, sodium concentrations.		Small sample size
Bland (1972) N Engl J of Med. Boston.	The cord blood of 2,200 newborn patients.	Total protein content, birth weight, ventilation requirements.	Low levels of total protein content of cord blood can indicate a risk of idiopathic respiratory-distress syndrome.
Gallagher (1985), Crit Care Med.	18 patients   Group 1: 9 patients aged 49 +/- 2 years. Group 2: 9 patients aged 65 +/- 2 years.	Extravascular lung water, colloid osmotic pressure, cardiac index, Pulmonary shunt function, PaO2.	Before during and after cardiopulmonary bypass Extravascular lung water was correlated with age (lower in group 1). Extravascular lung thermal volume was increased but returned to normal by the following morning, colloid osmotic pressure and cardiac index dropped during cardiopulmonary bypass but returned to normal the following day.	Small sample sizes. It is not explained whether the colloid osmotic pressure and extravascular lung water drop the same percentage in both ages.
Simonardottir et al. (2001), Perfusion.	10 consecutive patients undergoing cardiopulmonary bypass.	Plasma colloid osmotic pressure before during and after cardiopulmonary bypass. Compartment pressure for 48 hours (beginning at start of surgery), calf circumference. Arterial pressure, CVP, temperature. Serum albumin, total protein, urine output, blood results, post-operative bleeding, body weight.	Compartment pressure increases during and after cardiopulmonary bypass.   Colloid osmotic pressure is not linked to compartment pressure. Colloid osmotic pressure drops during cardiopulmonary bypass and returns to normal after 6 hours.	Small patient size, but three types of surgical intervention. Two patients did not get a full dataset.
Miao et al. (2014a), Perfusion	60 paediatrics aged 3-18 months.   Group 1:30 patients with HES prime Group 2: 30 patients with human albumin	Perioperative haemodynamics, plasma colloid osmotic pressure, renal function, blood loss, allogenic blood volumes and plasma volume substitution.	Colloid osmotic pressure was higher than normal physiological values in the Hydroxyethyl group after cardiopulmonary bypass. Colloid osmotic pressure was not higher than normal in the albumin group.   After 6 hours colloid osmotic pressure had returned to normal. Hydroxyethyl is an alternative for human albumin.	All prime volumes were 450ml, but patients’ aged ranged from 90 days to 451 so the amount of haemodilution seen would have varied (although there was no significant difference in the weight between the two groups). Good methodology.
Gupta et al. (2002), Arch Dis Child.			Low serum albumin is a marker of disease severity not merely colloid osmotic pressure levels. No studies have investigated albumin transfusions to improve an illness in children
Verheij et al. (2006), Br J Anaesth	67 patients   Group 1: Saline Group 2: Gelatin Group 3: HES Group 4: Albumin	Haemodynamics, ventilation variables, chest radiograph, pulmonary leak index, extravascular lung water, plasma colloid osmotic pressure, lung injury score	Colloid osmotic pressure increased in the groups receiving colloids and reduced in patients receiving crystalloids.   Saline needed greater infusion volumes. Patients receiving colloids has an increased LIS due to a decrease in respiratory compliance.
Schupbach et al. (1978) Vox Sang.	75 rabbits   Groups not obvious.	Colloid osmotic pressure, tissue oedema, urine production, protein content.		Not enough information is given into the method used, making the procedure not replicable.
Eising et al. (2001) Eur J Cardiothorac Surg.	20 adults undergoing coronary artery bypass grafting with cardiopulmonary bypass   Group 1: Colloid prime Group 2: Crystalloid prime	Colloid osmotic pressure, lung water, cardiac index, fluid added, hemodynamics, DO2.	Hydroxyethyl starch resulted in better colloid osmotic pressure than colloid, decreased lung water, cardiac index and fluid balance. No difference was found in haemodynamics.
Jin et al. (2014) J Chin Med Assoc	40 adults undergoing coronary artery bypass grafting with cardiopulmonary bypass   Group 1: Colloid based prime Group 2: Crystalloid based prime	Fluid balance, colloid osmotic pressure, extravascular lung water.	A crystalloid prime resulted in a more extreme colloid osmotic pressure drop than the colloid based prime.   The crystalloid prime caused a 21% increase in extravascular lung water, colloid prime remained the same	Some text very similar to the Eising, et al. (2001) paper.

Risks of changes in colloid osmotic pressure

The composition of the blood will affect the colloid osmotic pressure and therefore when a patient is put on the heart-lung bypass machine their circulating blood will be mixed with the priming fluids and thus will be changed. Golab, et al. (2011; 2009) kept the colloid osmotic pressure over 15 mmHg and also investigated a second group with a higher colloid osmotic pressure- above 18mmHg. The group with a colloid osmotic pressure above 18mmHg needed less artificial ventilation and had lower lactate levels.

If the colloid osmotic pressure drops, then the oedema that is seen can result in a weight gain. Loeffelbein, et al. (2008) found that adding album into the cardiopulmonary bypass prime resulted in a lower amount of weight gain, due to a lower amount of capillary leakage. Tassani, et al. (2007) disagrees and suggests that low cardiopulmonary bypass seen due to cardiopulmonary bypass is not linked to capillary leakage, but hypothermia and reduced lymphatic drainage are also causes for oedema. Loeffelbein. et al, (2008) studied 20 patients and Tassani, et al. (2007) studied 11. Both papers are comparable so more research should be carried out. Fluid shifts can be minimised through higher colloid osmotic pressure levels was one of the conclusions from Darling, et al. (2000), which studied 10 patients in each of two sample groups.

Premature infants had their colloid osmotic pressure monitored by Bhat, et al. (1981).  Connections were found between patients who were premature and survived and patients who did not survive. The patients who did not survive saw their colloid osmotic pressure drop below 10mmHg. Samples were taken from patients within an hour of death; however there were only 5 patients in the healthy preterm group, this makes the data less precise. These patients had not undergone cardiopulmonary bypass so it is likely that the patients’ underlying disease was the cause of the colloid osmotic pressure and as the patients’ condition deteriorated the colloid osmotic pressure fell alongside it. The patients’ group size was not consistent in this paper.

Paediatric patients have a higher permeability of the capillary than adults (Rosenthal, et al. 1975), this is due to the effect of age on transvascular fluid movement and the circuit and volume of priming fluids being larger compared to the body size and circulating volume of a neonatal patient.

The baseline measurements of the colloid osmotic pressure for healthy infants was measured by Sussmane, et al. (2001), they were reported as increasing with age from 1-9 months and was comparable with recorded numbers for the adult baseline (25.1 ± 2.4mmHg for paediatrics and 25mmHg in adults). No differences were found here between male and females. This study measured 37 healthy patients (Bland, 1972).

Guthe, et at. (2015) studied 99 healthy infants from 2 to 10 years old and discovered that colloid osmotic pressure significantly increases with age, from 24.6mmHg ± 3.2 at 2-3 years and 28.mmHg ± 4.2 at 8-10 years. The average seen over all age ranges (25.6mmHg ± 3.3) is comparable with figures given for adult baseline studies. The paper also reports that the patients are ‘healthy’, however all patients have been hospitalised for tonsillectomy, adenotomy or tympanic paracentesis. Patients were also fasted for a minimum of 8 hours in preparation for the general anaesthetic. These conditions could cause the patients to have artificially high or low colloid osmotic pressure baselines. Patients had their colloid osmotic pressure measured after anaesthetic intubation, and Golab, et al. (2009) recorded this having the biggest impact on a postoperative colloid osmotic pressure. This data was reported in patients undergoing cardiopulmonary bypass and these patients were not.

A larger study measuring the umbilical cord blood in 2,000 newborns over an 8 month period was undertaken by Bland (1972).  This study found that idiopathic respiratory distress syndrome is seen in 0.5% of the neonatal population with normal protein content. However, patients with a protein content less than 4.6g/100ml was linked to a 30% incidence in neonatal patients born at a normal gestation and 50% of the patients who were born prematurely or those with a low birth rate. In a letter to the editor Liu, et al. (2011) disagreed with the findings that colloid osmotic pressure was comparable with adult and paediatric patients. Instead suggesting that previous data (Golab, et al. 2010) should be separated into a group of patients who have cyanotic disease and a group of acyanotic patients to understand any differences here.

Method of birth

Sola, et al. (1981) studied the colloid osmotic pressure of new born infants and found a discrepancy between those born naturally and those born via caesarean section. The numbers are statistically significant and showed those born via caesarean section had lower colloid osmotic pressure. Premature infants had even lower colloid osmotic pressure (although this was not separated into natural birth and compared these with caesarean births). When seen in the context of the Zimmerman, et al. (1997) paper with a link suggested between colloid osmotic pressure and respiratory illness, and a connection drawn between colloid osmotic pressure at birth and low blood pressure (possibly due to decreased contractility of the heart and viscosity of the blood) this allows us to understand how method of birth can link to respiratory illness.

The method of birth was also linked to protein content.  Patients who had been born via caesarean section had a lower protein content when compared to those born naturally.  This is in line with the differing levels of colloid osmotic pressure (Bland, 1972). Looking at these papers it would suggest that naturally born infants will have the best respiratory health and higher blood pressure, patients born via caesarean will have lower respiratory health and blood pressure and premature infants will have a decreased level of respiratory health. The gestation length for caesarean and natural births are not listed. This would be interesting as the specific hospital might have protocols that ensure caesarean births might need to be carried out on a specific gestation length. If this is earlier than the average of the natural birth then the lungs might be underdeveloped and so effect the colloid osmotic pressure.

Crystalloid and colloid. Haneda, et al., (1985) studied the differences for a patients’ fluid balance and mortality for patients undergoing cardiopulmonary bypass with a priming fluid based on Ringers lactate or with a colloid based prime (consisting of whole blood and plasma). The patients who had a colloid prime had a lower fluid balance post-surgery, their stay within intensive care was halved (from 20.6 days to 10.9 days) and the mortality rate dropped from 42% for patients having a crystalloid prime to 23% for patients having a colloid prime. This data is now over 30 years old and was collected between 1975 and 1983 and the rates of survival are alarming when compared to a modern paediatric cardiac centre.  The data is however significant and if all other parts of the surgical procedure was carried out without changes then the numbers are still important to include in the review. Nothing is mentioned about the risks of transfusion reactions in this paper, but if the fluids that have been given have effectively halved the mortality numbers then the small risk associated with transfusion of blood problems is probably ignored. Modern data reported by SHOT in June 2017 show a risk of major morbidity as 1 in 21,000 transfusions, and would have been higher at the time of study.

Albumin based prime

The fluids that are added to the heart-lung machine for priming can have an effect on the colloid osmotic pressure during cardiopulmonary bypass. Riegger, et al. (2002) measured colloid osmotic pressure for 86 paediatric patients weighing less than 14kg. Patients who were given a prime that consisted of 5% albumin had a net negative fluid balance after bypass whereas those that had a crystalloid prime ended bypass with a net positive fluid balance. The albumin prime was linked to a higher colloid osmotic pressure and serum albumin levels.  The risk associated with the albumin priming method found in this study is the increased use of transfusion products.  Given the relatively short term effects of this- the colloid osmotic pressure and haematocrit had returned to normal after 24 hours -using albumin in the prime might not be beneficial for the patient.

In patients not undergoing cardiopulmonary bypass the results of giving albumin to improve the colloid osmotic pressure has been investigated by Gupta, et al. (2002). They consisted the effects of albumin on adult health and whether increases in colloid osmotic pressure lead to an increase in survival.  Patients in this study had not been placed on a heart-lung bypass machine, so the low colloid osmotic pressure was due to their underlying health condition and not haemodilution. The results concluded that in patients with a critical illness, only small amounts of albumin contributes to the colloid osmotic pressure and that artificially increasing the colloid osmotic pressure would not improve a state of disease if the low colloid osmotic pressure had been caused by the illness.

In 2010 Golab, et al. further investigated the use of albumin as a part of the cardiopulmonary bypass prime and recoded peri- and post-operative data for 73 consecutive patients and found that adding 5% albumin to the prime would allow the perfusionist to run the cardiopulmonary bypass with a higher colloid osmotic pressure. When the colloid osmotic pressure was kept above 18.0mmHg patients had a lower duration of ventilation post-operatively and a lower post-operative lactate levels. These patients were all below 10kg, but it would have been useful to have data presented in regards to the circuit size, priming volume and haemodilution effect.

Fresh frozen plasma prime

Miao, et al. (2014b) investigated the theory that the administration of fresh frozen plasma might be beneficial to a patient.  They looked at adding prophylactic fresh frozen plasma into the cardiopulmonary bypass circuit as a priming fluid and compared this to the use of gelofusine. Miao, et al. (2014b) found the prime constituents to be comparable when measuring recovery time, post-op bleeding, transfusion requirements, drug usage and therefore there was no advantage to using the more expensive product as a priming fluid, especially given the risks associated with blood product transfusions. This study followed 80 patients (split into 40 with fresh frozen plasma and 40 without) and is one of the larger studies reported.

Hydroxyethyl prime

Miao, et al. (2014a) also looked into using hydroxyethyl starch prime (hydroxyethyl) as a priming fluid and compared it with an albumin prime. In patients that had been treated with a hydroxyethyl prime the colloid osmotic pressure was higher than physiological values post-operatively and patients who received prime including albumin had colloid osmotic pressure levels that were physiologically comparable.

Since 2013 there have been contraindications attached to the use of hydroxyethyl in cardiopulmonary bypass circuits for cardiac surgery due to prolonged bleeding and for those with pre-existing renal insufficiency. The patients monitored by Miao, et al. (2014a) had slightly more postoperative chest drain bleeding; however, this was not significant and the amount of blood products needed was significantly more in the patients receiving an albumin prime.  They therefore suggest that there is no association between the use of hydroxyethyl and renal function. Other studies including by Akkucuk, et al. (2013) and Hanart, et al. (2009) found there to be no negative links between the use of hydroxyethel on the renal function in a paediatric setting.

Filtration through bypass machine

Another way that colloid osmotic pressure can be increased during cardiopulmonary bypass is through filtration. After cardiopulmonary bypass modified ultrafiltration has more benefits than just increasing the colloid osmotic pressure.  Ziyaeifard, et al. (2014) suggested performing modified ultrafiltration will increase haemodynamic, pulmonary, coagulation and organ function.  It also leads to a decrease in blood transfusions, reduction of body water and blood loss after surgery. Elliot (1993) also agreed that modified ultrafiltration is beneficial and increases haematocrit, removes accumulation of water (and so increases the colloid osmotic pressure), reduces blood loss and the need for blood transfusion. Further benefits include an increase in cardiac index, decrease in heart rate, decrease in pulmonary vascular resistance along with an increase in contractility and a decrease in myocardial wall volume. This study investigated the effects of 400 patients and agreed that modified ultrafiltration improved the patient’s post-operative condition. Darling, et al. (2000) linked modified ultrafiltration to an increase in total protein content and showed that increased colloid osmotic pressure can prevent fluid shifts but warn that a hyperoncotic fluid might damage kidney function.

In 2014, Golab et al. disagreed with the positive factors that have been attributed to modified ultrafiltration post operatively and reported that filtration did not affect length of ventilation, length of chest drain insertions, duration of time spent in intensive care and total hospital stay.  However it was found that filtration while on cardiopulmonary bypass did result in a higher colloid osmotic pressure and resulted in a lower transfusion requirement, although if these are the only benefits it is still worth carrying out filtration. Fifty five patients weighing less than 10kg were recruited for this study.

If there are no underlying co-morbidities and a neonatal patient is coming for cardiopulmonary bypass, then I would suggest an investigation into the patients’ gestational age at birth and birth method be carried out If the patient was delivered via caesarean the colloid osmotic pressure should be measured more closely and the aim should be to have a patient leave the operating theatre with a colloid osmotic pressure that is normal for a healthy neonatal patient.

Few studies have separated the patients out in groups depending on the type of disease. Haneda, et al. (1985) investigated the colloid osmotic pressure changes for 45 patients undergoing cardiopulmonary bypass for transposition of the great arteries, a cyanotic cardiac diagnosis. These patients who had a colloid based prime left hospital earlier.

One paper records the ethnicity of the patients and finds that one patient who has a significantly lower colloid osmotic pressure has a different ethic groups, although this just one patient, it might be interesting for further study (Sussmane, et al., 2001).

Modifying the circuit. Darling, et al. (2002) suggests that modifying circuits and prime volumes should be carried out periodically to improve performance. After the team at Duke University carried this out the amount of fluid that was given to a patient through the duration of bypass dropped from 363.5ml to 245.1ml. Here only a small sample of patients were measured; 10 before the modifications and 10 after, but resulted in benefits to the patient.

Adult colloid osmotic pressure

Tonnesen, et al. (1977) identified a link between colloid osmotic pressure and survival in 84 adults that were admitted into an intensive care unit. The colloid osmotic pressure was measured through a patients stay and analysis of this data showed a relationship between the patients’ colloid osmotic pressure and survival. A colloid osmotic pressure of 15.0mmHg was identified as the marker for 50% survival. These were not post cardiac patients; however the link between the colloid osmotic pressure and survival might be useful. In this study patients’ colloid osmotic pressure was not grouped depending on diagnosis, which may have been useful in identifying patients who might benefit from close colloid osmotic pressure monitoring. In a study of 20 adults undergoing coronary artery bypass graft surgery by Eising, et al. (2001), a hyper oncotic bypass prime prevented water accumulation but did not have an effect on the pulmonary function. The colloid prime here had a colloid osmotic pressure of 48mmHg.  It is suggested that this could benefit patients who a congestive heart failure condition.

Several papers have linked colloid osmotic pressure to survival. A study of 145 patients situated within an intensive care unit (ITU) for prolonged critical illness (who had not undergone surgery with cardiopulmonary bypass) found no relationship between colloid osmotic pressure and survival. The most common primary diagnosis in this patient group was blunt trauma, cardiopulmonary arrest, acute respiratory failure and gunshot wounds. The study by Blunt, et al. (1998) found that levels of serum albumin were linked to survival-patients with lower mean serum albumin through their stay in the ITU had a higher mortality. In these critically ill patients, the albumin made up 17% of the total colloid osmotic pressure, not the 40% that albumin constitutes towards colloid osmotic pressure in healthy patients.

In a further study of patients in a shock unit by Morrissette, et al. (1975), 99 consecutively admitted patients were studied. In this ward the colloid osmotic pressure was used as an indicator as multisystem failure. Again the conditions that patients were admitted for were wide ranging, the most common were: acute myocardial infarction, drug overdose and hypovolemia, these contributed 49 of the 99 patients. Patients who did survive had a significantly higher colloid osmotic pressure than those who did not. The 50% survival rate here is recorded as a colloid osmotic pressure of 14.1 mmHg. Twenty-one patients had a colloid osmotic pressure that dropped below 10.5mmHg and none of these patients survived. It is suggested that the advancement of cardiopulmonary failure caused the decline of colloid osmotic pressure. If patients colloid osmotic pressure drops below 15-16mmHg, increased interstitial fluids are observed and the lymph system is not able to compensate for the extra fluid.

Compartment pressure

Simonardottir, et al. (2001) looked for a link between colloid osmotic pressure and compartment pressure and how this changes through cardiopulmonary bypass. The study of 10 patients found that the colloid osmotic pressure dropped during cardiopulmonary bypass and the compartment pressure increased during and after cardiopulmonary bypass, reaching the maximum recorded just after cessation of cardiopulmonary bypass, however there was no correlation between the two pressures.  Colloid osmotic pressure had returned to normal after 6 hours but the compartment pressure was recorded for 48 hours post bypass and was still elevated.

Lung water. Cardiac patients and non-cardiac patients who have oedema both have increased levels of extravascular lung water. Gallagher, et al. (1985) investigated the effects that age has on extravascular thermal volume in the pulmonary system and found that in the pre, peri and post-operative periods the extravascular thermal volume was lower in patients who fell into the older age group (65 ± 1.2 years, against 49 ± 2 years). The amount of extravascular thermal volume increased while on cardiopulmonary bypass. Interestingly the cardiac index dropped in the younger group of patients, but remained stable in the older group and colloid osmotic pressure dropped in both patients, these values had returned to normal by the following morning. Correlation was found between the colloid osmotic pressure, cardiac index and extravascular thermal volume, however no correlation was seen between PaO2 and pulmonary shunt fraction. Lung water was also measured by Lumb (1987) who reported that extravascular lung water is associated with a drop in serum oncotic pressure. Through surgery the extravascular lung water was elevated, but this was transient as the numbers were not significantly above baseline when recorded in the intensive care unit. Ten patients were given cardiopulmonary bypass primes enhanced with albumin and 10 had 500ml hydroxyethyl added to the prime. The small patient numbers studied here makes it difficult to put much weight on the numbers. The second group (with hydroxyethyl added to the prime) had an additional total priming volume of 350ml when compared to the albumin prime group; which might have effected fluid dynamics.

Oedema and colloid osmotic pressure

Sanchez, et al. (1982) studied 16 patients and identified no link between colloid osmotic pressure and lung oedema.  This patient selection consisted of 12 patients having coronary artery bypass grafts, two having mitral valve replacements and two having aortic valve replacements. The duration of the cardiopulmonary bypass varied from 50 and 130 minutes. COP returned to normal after 6 hours. This group reported a drop in COP can be tolerated without pulmonary oedema. This study has wide ranging variables just 16 patients to draw conclusions from including a variety of surgery types that might not be comparable (baseline colloid osmotic pressure in patients with valve insufficiently might be not within normal range). The lengths of cardiopulmonary bypass vary widely as does the body surface area of the patients (1.55m2 – 2.10m2) and the age range of patients (a range of 20 years over 16 patients), which was identified by Gallagher, et al. (1985) to be a wide enough age range to see differences in baseline measurements.

A large study monitoring the haemodynamics for 100 off-pump coronary artery bypass graft and 97 on bypass coronary artery bypass grafting found that off-pump patients had better gas exchange and earlier extubation. Surprisingly the off pump surgical procedures required more fluids through the operation. It is thought this is due to maintaining the pressures during the posterolateral grafting. No difference in survival rates, pneumonia, pleural effusion or pleural oedema was found. This data was published in 2005 by Stanton, et al.