Parenteral fluid therapy
The rational administration of parenteral fluids is one of the most significant advances in the care of acutely ill patients in the twentieth century. However, despite advances
in the monitoring of cardiovascular variables, the questions of what? when? and how much? remain areas of enormous controversy.
In the 1930s Blalock suggested that it was the loss of blood rather than the ‘release of evil humors’ that led to death after major trauma and recommended treatment by the administration of intravenous fluids. It was not until the 1940s and World War II that blood and plasma were widely used for the treatment of blood loss. Ironically, the same period highlighted the problems of inducing anaesthesia in vasoconstricted hypovolaemic casualties. When thiopentone was used as a sole intravenous anaesthetic agent at Pearl Harbour on 9 December 1941 there were numerous deaths as a result of vasodilatation and cardiovascular collapse. In the 1940s and 1950s descriptions of postoperative retention of salt and water as part of the metabolic response to surgery led to a widespread reluctance amongst surgeons and anaesthetists to administer crystalloids to their patients. The 1960s brought a swing back the other way when Shires and others demonstrated an increased survival in experimental animals that were bled and then reinfused blood plus additional crystalloid. The ensuing enthusiasm for the infusion of crystalloids led to the publication of Moore and Shires’ now famous article calling for ‘moderation’.
Colloids or crystalloids?
The colloid—crystalloid argument rages to this day: if one was clearly better than the other there would not be a controversy. The common link in the majority of articles on this subject is the final conclusion that it is dose not choice of fluid that is the real issue — and ‘the proper dose of any drug is enough’ (Dr J.H. Drysedale). Certainly if the aim is to restore the circulating blood volume this will be achieved with a smaller volume and thus more rapidly by using a colloid.
Hypovolaemia can occur as a consequence of a wide variety of pathological processes. The nature of the fluid lost should dictate the choice of replacement fluid. Clinical history, whether first or second hand, in combination with appropriate laboratory investigations, should be the most useful guides to a rational fluid regimen. However, hypovolaemia from whatever cause is an acute medical emergency. Any degree of hypovolaemia jeopardises oxygen transport and increases the risk of tissue hypoxia and the development of organ failure. The greater the degree and duration of hypovolaemia the greater the risk. Therefore, the initial treatment is to restore the circulating volume as quickly and effectively as possible.
Hypovolaemia may be divided into three categories: covert compensated hypovolaemia, overt compensated hypovolaemia and decompensated hypovolaemia.
Covert compensated hypovolaemia
This is the commonest yet least often diagnosed form of hypovolaemia. It refers to the presence of a reduced circulating blood volume without very obvious associated physical signs. Price found that healthy volunteers could have 10—15 per cent of their blood volume removed with no significant change in heart rate, blood pressure, cardiac output or blood flow to the splanchnic bed (gut, etc.). However, splanchnic blood volume was reduced by 40 per cent. The subjects in his study had essentially autotransfused and were maintaining the systemic circulating volume at the expense of the splanchnic circulating volume. This same process happens when we donate a unit of blood with no obvious adverse effects. Over the course of the next few hours we feel thirsty and therefore drink more, we also ingest salt and at the same time reduce urine output of salt and water. We make new proteins and blood cells, and very soon everything has returned to normal with no sequelae. In sick patients, however, many of the natural compensating mechanisms malfunction and this, coupled with the fact that fluid intake is being determined by a second party, namely the physician, makes hypovolaemia common.
Covert compensated hypovolaemia is extremely difficult to diagnose. In the conscious patient central nervous system (CNS) symptoms are the best guide. In the experiments performed above, all of the subjects developed CNS symptoms ranging from drowsiness and nausea to hiccoughs. Any thirsty patient should be assumed to be hypovolaemic until proven otherwise. Urinalysis showing an increased urinary osmolality and decreased sodium concentration is the most useful laboratory investigation.
Although covert compensated hypovolaemia is common and probably contributes significantly to morbidity, the majority of patients withstands the insult. If the hypovolaemia persists consequent end organ hypoperfusion may be present for many days before it manifests itself as organ dysfunction. By this time the patient is usually in a state of overt compensated hypovolaemia.
Overt compensated hypovolaemia
Here there is hypovolaemia to an extent that the reflex mechanisms required to maintain perfusion to vital organs are obvious on clinical examination but the blood pressure is maintained. As before, clinical history is essential. On examination the patient will demonstrate the manifestations of an increased sympathetic drive with tachycardia, a wide arterial pulse pressure, and typically increased systolic blood pressure and cool clammy skin, particularly at the hands and feet.
There may be other evidence of an inadequate cardiac output such as drowsiness, confusion and an increased respiratory rate. If the diagnosis is uncertain additional dynamic bedside tests can be performed such as gentle head-down bed tilting, leg raising or the administration of a bolus of intravenous fluid. If the diagnosis of hypovolaemia is correct then the increase in venous return may result in a reduction in heart rate, narrowing of pulse pressure, reduction in respiratory rate and overall improvement in well-being. If the diagnosis remains uncertain, or coincidental medical problems such as heart or lung disease make performing or interpreting such tests difficult, then more complex investigations may be required.
With the exception of electrolyte and blood gas analysis the majority of laboratory investigations is of little use in the acute phase. Arterial blood gas analysis can be performed rapidly; hypovolaemic patients are commonly hypoxaemic and may have a metabolic acidosis as a consequence of an inadequate cardiac output. Urinalysis, as described above, may support the diagnosis of hypovolaemia but no single test is diagnostic. Rapid determinations of total blood volume are not yet available.
Except for extreme cases the clinical interpretation of CVP by examination of the jugular venous waveform is unreliable and has no place in the management of hypovolaemic patients. If there is any doubt about the diagnosis, particularly in patients with cardiorespiratory disease, the patient needs a CVP catheter. The insertion and, indeed, interpretation of the information available from central venous catheters carries a significant morbidity and mortality so they should only be inserted and managed by experienced clinicians. For a more detailed account of central venous catheterisation readers are referred to Rosen et al. As a general rule the right internal jugular approach is favoured. The subclavian vein may be particularly difficult to locate in the hypovolaemic patient and the risk of arterial cannulation, haemorrhage and pneumothorax is then greatly increased. If during insertion steep head down tilt produces no adverse effects and it is difficult to visualise or palpate neck veins, the judicious administration of at least 500 ml of colloid by a peripheral route before proceeding is sensible and safe. In the unlikely event that fluid administration produces a deleterious effect the infusion can be stopped easily, the head-down tilt corrected and the patient sat up.
In cases of ventricular dysfunction and/or severe pulmonary disease there will be a misleading discrepancy between right and left atrial filling pressures. If the information obtained from the CVP catheter is confusing, it may be necessary to insert a pulmonary artery flotation (Swan-Ganz) catheter. The pulmonary artery occlusion pressure (PAOP) provides an index of left ventricular filling pressure and may help to clarify the situation. It is important to realise that right atrial pressure and PAOP are influenced not only by the circulating volume but also by the degree to which the circulation is constricted, the compliance of the right and left heart, as well as pain, agitation, etc., causing increases in sympathetic tone. Low values are sensitive indicators of hypovolaemia, but high values do not necessarily mean the patient is well filled.
using fluid challenges give much more information and should always be tried in patients with evidence of an inadequate circulation. The administration of 200—5 00 ml of colloid over 5—10 minutes and comparison of the CVP or stroke volume (not the cardiac output) before the challenge and 5—10 minutes after the infusion has finished is the most useful guide. A sustained rise in CVP or PAOP of 3 mmHg and failure of the stroke volume to increase suggest the circulation is well filled.
This is what many people refer to as shock. The degree of hypovolaemia is such that reflex redistribution of blood flow is insufficient to compensate and vital organs are no longer adequately perfused. The mean arterial blood pressure falls and may be difficult to record as peripheral pulses are often impalpable. The blood supply to the heart and lungs is compromised, which further reduces cardiac output, causes ventilation/perfusion (V/Q) mismatching and compounds the problem. Tachycardia changes to bradycardia as myocardial oxygenation becomes critical and the conscious level is severely obtunded. If untreated this clinical state rapidly progresses to total circulatory arrest. No special equipment or investigations are needed to make the diagnosis of decompensated hypovolaemia and to start aggressive volume replacement therapy. Misdiagnosis and inappropriate over transfusion is an overrated problem. Delay in the treatment of hypovolaemic shock greatly reduces the chances of successful resuscitation. Most causes of hypovolaemic shock carry a far better prognosis than any condition that presents in a similar fashion but would be made worse by a fluid challenge.
The consequences of hypovolaemia
Decompensated hypovolaemia will result in endorgan damage and death if it is not treated rapidly and completely. Probably a far more common and insidious source of morbidity and mortality is the compensated hypovolaemias. As described above, a small reduction in circulating blood volume rapidly results in a far more significant reduction in splanchnic blood volume and in particular the supply to the innermost layer of the gut lumen, the mucosa. It is becoming increasingly clear that hypoperfusion of the gut mucosa is of fundamental importance in the pathogenesis of multiple organ dysfunction (see below).
Therefore, hypovolaemia is a potential killer in any disease process. The manifestations of persistent covert compensated hypovolaemia may not be seen for many days. Once a patient has overt hypovolaemia the chances of successful treatment are already significantly reduced with the exception of simple acute haemorrhage. Most patients are referred to intensive care units once they have progressed to decompensated shock with established organ failure. By that stage it is probably too late to make a significant difference to outcome. The early recognition and treatment of hypovolaemia is essential in any disease process.
Treatment of hypovolaemia
Very few patients benefit from fluid restriction; if there is evidence of hypovolaemia it should be treated. Ionotropes should be used only when the circulating volume has been corrected.
Occult hypovolaemia is very difficult to diagnose. Therefore, in conscious patients who can drink the most rational approach is to be generous with fluids. Access is important, as is strength and volition. The patient with a full water jug and a raging thirst is commonplace. The aim should be an asymptomatic patient (i.e. no thirst) with good urine volumes (in excess of 0.5 mI/kg/hour) and normal urinalysis. The overriding principle is that fluid overload is easy to treat, whereas fully established organ failure is incurable.
Overt hypovolaemia should be considered a medical emergency and treatment is required urgently. The intravascular space must be resuscitated in minutes to hours not hours to days, as is currently common practice. Restoration of total body water and electrolytes will he slower. Treatment should be started following a presumptive diagnosis of hypovolaemia; by all means send laboratory investigations but do not wait for the results before starting treatment.
High-flow oxygen therapy should be given to all hypovolaemic patients until arterial blood gas analysis confirms normoxia. A pulse oximeter is useful if available. Venous access should be secured with short, large-bore cannulae, allowing large volumes to be infused rapidly. Ideally a 14G cannula in an arm vein should be used. These allow flow rates twice those of a 16G cannula. CVP catheters are of very limited use in the early phase of resuscitation, and are difficult and hence more dangerous to place. They should only be used if the diagnosis of hypovolaemia is in doubt or if no other access is available.
The initial choice of fluid in overt hypovolaemia should be a colloid for the reasons stated above. The need for blood (see below) should not delay initial resuscitation. Cardiac arrest due to a low haemoglobin concentration is very unusual; cardiac arrest due to hypovolaemia is relatively common. Resuscitation should he a continuous process with the doctor at the bedside re-evaluating the patient. Failure of a fluid challenge to secure improvement requires the use of more invasive monitoring. Each fluid challenge must be seen to produce a definite improvement. Precharted fluid regimens and remote management cost lives.
Just as patients compensate for volume loss in the early stages of hypovolaemia, so an apparently resuscitated patient may still have a significant volume deficit. The aim for immediate resuscitation should he normal measures of pulse, blood pressure and CVP, urine output > 0.5 mI/kg/hour with normal urinary osmolality and sodium concentration. Any metabolic acidosis should be seen to be correcting. Thereafter, one must try to maintain normovolaemia. This is a continuous process. Critically ill patients may have capillary leak and will therefore have a continuing colloid requirement. Gelatins, being small molecules, are poorly retained and can be replaced by hydroxyethyl starch, plasma or blood at this stage. In sepsis this requirement may be very large (see below).
The importance of blood in immediate resuscitation, the threshold at which one should transfuse urgently (i.e. consider using group compatible, uncross-matched blood or even Group 0 blood) and even the target haemoglobin level are controversial. Resuscitation should not be delayed whilst waiting for blood to be grouped; if acute anaemia is secondary to the bleeding resuscitation should be with Group 0 blood or group-compatible blood as it becomes available. Otherwise colloid should be used initially and cross-matched blood and relevant blood products should be used when they are ready. Packed cells are not colloid and have little plasma expanding effect; transfusions with large amounts of packed cells will require supplementation with colloid. The age of the blood is important (old blood is acidic, with decreased oxygen-carrying capacity and poor red cell deformability) —use the youngest possible blood and whole blood if it is available.
Hypovolaemia and the surgical patient
Hypovolaemia is extremely common among patients undergoing surgery. It remains standard practice in the UK to deny patients food or drink for a minimum of 6 hours prior to elective surgery in an attempt to reduce the risk of pulmonary acid aspiration syndrome. It is not uncommon for this to extend to 10 or even 20 hours due to unforeseen delays. Preoperative fluid restriction is currently a matter of debate. Indeed, there is evidence suggesting that the administration of oral fluids to patients until 2 hours prior to elective surgery has produced a more favourable effect on gastric contents than total starvation. Yet no attempt is routinely made to maintain normal hydration in preparation for surgery, despite numerous previous studies demonstrating the benefits of fluid administration for even the most minor surgical procedures. Recently a study on fit young patients having elective laparoscopic sterilisation under general anaesthesia demonstrated a reduced morbidity by the administration of crystalloid during the operation. It is commonly taught that as part of the stress response to surgery patients have increased levels of ADH and aldosterone postoperatively and thus retain salt and water. As a result of this the overzealous administration of intravenous fluids is feared. Whilst it is probably true that ADH levels do rise in all postoperative patients, the presence of hypovolaemia per se may be responsible for much of the increase. A reduction in urine output in the first 24 hours after surgery may be acceptable, but a fall to oliguric levels (<0.5 ml/kg/hour) is not. Using the gastric tonometer, reduced splanchnic perfusion as a consequence of hypovolaemia has been demonstrated to be common during major surgery and associated with the development of postoperative organ failure. It is difficult to find any objective evidence to support the hypothesis that peripheral oedema or the accumulation of extravascular lung water, as opposed to
pulmonary oedema due to left ventricular failure, has any significant adverse effects. On the contrary, the evidence supporting the prophylactic administration of intravenous fluids to patients having major surgery is impressive.
Hypovolaemia and cardiogenic shock
Conventional management of the patient with acute pulmonary oedema is still all too often based on diuretics. Unlike congestive cardiac failure, the usual problem in acute left ventricular failure is not an excess of total body salt and water but an acute redistribution of a normal quantity. This leads to an effectively hypovolaemic patient, particularly if they have been treated enthusiastically with diuretics. The most appropriate treatment is to reduce pre- and after-load with infused vasodilators such as GTN to encourage redistribution of fluid in a more normal fashion and to consider using ionotropes and even judicious colloid challenges guided by a pulmonary artery catheter if vasodilators do not produce a rapid improvement. This is particularly important in right ventricular infarction, where the effect of hypovolaemia and poor right ventricular function is to underfill the left ventricle which in itself is usually underperforming.
The treatment of patients with congestive cardiac failure is more difficult. The basic principle should remain the maintenance of an adequate circulation with the use of vasodilators to improve overall haemodynamics supplemented with diuretics as necessary. The fact that so many patients tolerate diuretic-based regimens so well reflects the inherent robustness of the average human being!
No comments yet.