589

POST-OPERATIVE FLUID COMPLICATIONSBy B. J. HOUGHTON, M.B., B.S.

Lately Clinical Research Fellow, Surgical Unit, St. Mary's Hospital

Changes in the water and electrolyte content ofthe body tissues after surgical operations occur asthe result of two factors. On the one hand,there are the various changes consequent uponaltered intake and excessive loss of these materials,as, for example, in disorders affecting the gastro-intestinal tract and with haemorrhage, and on theother hand, there are several less specific butequally important changes which occur as thephysiological response to trauma. The latter,which were first observed in detail by Cuth-bertson (1930) and have been studied minutely inrecent years (Moore and Ball, 1952; Wilkinsonet al., I950, and others), are reviewed elsewhere inthis issue and will only be mentioned here inrelation to the genesis and treatment of individualfluid and electrolyte disturbances. It is oftenargued that patients survived major surgical opera-tions and illnesses long before these newer ideaswere conceived, but, of course, the same is truefor antibiotics and modern anaesthesia, the valueof which no one would deny, and prevention of acomplication still remai'ns better than its cure.Mortality figures can be improved by such means,as they were amongst cases of infantile diarrhoeawhen potassium supplements were first tried(Darrow, I946). It is probably no exaggerationto say that many patients who in the past suc-cumbed to the after-effects of an operation did sonot from ' toxaemia ' or' shock' but from a thenunrecognized biochemical complication. This,then, is the justification for yet another review ofour present knowledge in this field, which is theconcern of clinician, pathologist and physiologistalike.

General ConsiderationsClaud Bernard emphasized the importance to

living organisms of maintenance of the constancyof the ' milieu interieur,' and during the courseof evolution an elaborate system of homeostaticmechanisms has developed by which this isachieved. Consequently, we find in man aremarkable constancy in the water and electro-lyte content, acid-base balance and volume of thebody fluids, and it is clearly necessary to have

some familiarity with the factors involved in orderto understand the nature of the disorders seenafter he has been submitted to a major surgicaloperation. Although the existence of ' osmo-receptors ' has been clearly indicated followingthe pioneer work of Verney (I947), there is, asyet, little known about the mechanism of controlof the overall fluid volume, although a ' volumereceptor ' was postulated in I948 by Peters andBorst. In the normal course of events it appearsthat the body maintains the pH and osmolarity(osmotic solute content) of the fluids bathing thecells with more vigour than the total volume, andthis is, perhaps, not surprising in view of thechanges in the latter that occur with intestinalabsorption after feeding. However, there areclinical situations, for example, severe sodiumdepletion and haemorrhage, in which maintenanceof constancy of volume eventually takes prece-dence over osmolarity and acid-base balance con-trol, and similar observations have been madeunder experimental conditions (McCance, I936).The disturbances of fluid balance that occur in

surgical conditions can be grouped in severalways, their essential nature having been elucidatedand described by a large number of workers(Gamble, Hartmann, Darrow, Coller and Mad-dock, Cuthbertson, Marriott, McCance, Elkintonand many others). In the first place we mustconsider the disorders leading to alteration oftotal fluid and electrolyte content, and, therefore,changes of osmolarity. Later it will be necessaryto review the changes in acid-base balance and thespecial role of potassium, and to consider thecomplications due to disorders of cardio-respira-tory, renal and suprarenal function in altering theresponse of the organism. The final and majorobject of this discussion must be to outlinemethods of avoiding these hazards in routinesurgical practice.

Disturbances of the Total Body Fluid andElectrolyte Content

These may consist of a deficiency or excess ofwater or electrolytes, or both, and a mixture ofthese conditions is commonly encountered.

by copyright. on M

arch 31, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.386.589 on 1 D

ecember 1957. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

The Depletion SyndromesThe clinical state generally described as ' de-

hydration' may consist either of loss of water byitself (the true meaning of the term) or loss ofwater and variable amounts of sodium salts. Inview of the marked clinical differences betweenthese states it is probably best to avoid the termdehydration altogether and to use Marriott'sclassification, as follows here.

(i) Water Depletion. Loss of water occurs intwo ways: namely, by extrarenal routes as vapourin the expired air, insensible skin loss, sweat, andwater in the faeces, and by renal excretion asurine. The insensible skin loss and respiratorywater loss was estimated by Dubois in 1927, whofound it to be an average of 0.5 g. of water perkg. body weight per hour in adults, and this hasbeen confirmed by others. Thus for a 70-kg. manthe total loss is approximately 840 ml. per 24hours, and this is increased in febrile conditionsand elevation of the metabolic rate. Faecal lossof water clearly varies considerably, depending onmany factors, but allowing an average of I50 ml.per 24 hours, and not less than 5oo ml. of urineover the same period, the minimum obligatoryloss of water in health is generally accepted asapproximately I,500 ml. per 24 hours. Hence ifthe available water is less than this volume, orthere is excessive loss from one or more of theseroutes, water depletion is bound to occur.

Intake of water may be deficient due to comaor inadequate parenteral administration where theoral route is contraindicated, or it may be rela-tively insufficient compared with the intake ofexcretable solutes such as electrolytes and, in thecase of the diabetic patient, glucose.

Excessive loss may be from the skin in fever ora raised atmospheric temperature, hyperventila-tion, and polyuria as seen, for example, in renalinsufficiency, glycosuria and diabetes insipidus.The latter, of course, has a particular surgicalimportance in relation to hypophysectomy, asnowadays practised for some forms of advancedmalignant disease.The cardinal feature of water depletion is

thirst, and it is this symptom which generallyguides the clinician in assessing his patient's waterrequirements. Clinical signs are relatively slightat first, but as the condition becomes more severeand the body is depleted of 2 litres or more ofwater, dryness of the tongue and mucous mem-branes is apparent, together with the productionof only small volumes of urine with a high specificgravity. A high potassium concentration may befound in the urine, and this is a reflection of thetransfer of water and potassium from the reservoirof cell fluid, the ' toxic' potassium being hastilyexcreted. In severe cases restlessness, a per-

sonality change, impaired mental function leadingto stupor, and a marked rise in the plasma sodiumconcentration (and other evidence of haemo-concentration) may occur. If treatment is notcommenced immediately, hyperpyrexia and deathensue.

Administration of water in adequate quantitiesproduces a complete regression of symptoms andsigns, and the restoration of a normal urine output.

(ii) Sodium Depletion (Salt Depletion). Theaverage normal diet for man contains 8o to 150mEq. of sodium per 24 hours. Apart from asmall amount (3 to 5 mEq.) which is lost in theskin secretions and also in the faeces (3 to 5 mEq.),an equivalent amount is lost in the urine. Inhealth the body can maintain homeostasis on afraction of this intake, and that this is possible isdue to an efficient sodium conservation mechanismin the kidney, which is apparently mainly underthe control of the hormones of the adrenal cortex.In renal disease such as chronic pyelonephritisand chronic glomerulonephritis, and in adrenalcortical insufficiency (Addison's disease) excessiveand uncontrolled sodium losses in the urine mayoccur, and therefore it is important to rememberthis effect when such a patient has to have anoperation, for in contrast to the normal personwho excretes very little sodium in the urine in thefirst four or five days, as part of the physiologicalresponse to trauma, these patients may continueto lose large amounts.As apart from these special conditions a deficient

sodium intake is compensated by a greatly reducedoutput of sodium in the urine, it follows that themajority of cases of clinical sodium depletionoccur as the result of excessive extrarenal lossesand mainly, therefore, by vomiting, diarrhoea, orpooling of fluid in the gastro-intestinal tract orserous cavities. If the patient's fluid losses arerepaired by water alone, or, at most, by sodium-containing solutions of low concentration, then astate of sodium depletion (salt depletion) mayoccur whilst the total water content of the body isnormal. As might be expected, the plasmasodium, chloride and bicarbonate concentrationsare found to be low (the sodium may be as lowas izo mEq. per litre) and, in fact, the volume ofthe extracellular compartment is also reduced.A relatively high haematocrit is found as evidenceof haemoconcentration.

Reduction in the extracellular fluid volumecauses lassitude and weakness, signs of loss ofskin elasticity and tissue turgor, reduced bloodpressure and tachycardia, with a tendency tofainting when sitting up. Oliguria and a rise inblood urea follow, and if the condition is nottreated immediately, death results from oligaemia

December 1957590by copyright.

on March 31, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.33.386.589 on 1 Decem

ber 1957. Dow

nloaded from

HOUGHTON: Post-operative Fluid Complications

and uraemia ('pre-renal' or 'extra-renal'uraemia).

(iii) Mixed Water and Sodium Depletion. Infact these pure depletion syndromes described arerelatively uncommon, and the great majority ofcases which are clinically ' dehydrated' are infact suffering from combined water and salt deple-tion in varying proportions. The clinical featuresare mainly those of sodium depletion although thepatient may complain of thirst, and the bloodanalysis findings vary with the nature of thepredominant deficits. In all cases a raisedhaematocrit will be found, unless, of course, thepatient was grossly anaemic before his illness orlost excessive blood during his operation.The treatment of sodium depletion is adequate

replacement with this ion in the form of sodiumchloride. In most cases this will be given inisotonic solution (normal saline) because of theaccompanying deficiency of water, but in cases ofpure sodium depletion with a normal body watercontent, administration of a hypertonic solutionwill be desirable. The quantities and principlesof calculation will be considered later.

The Syndromes of Water and ElectrolyteOverloadingMuch has been said and written in the past by

clinicians about the dangers of excessive intra-venous infusions, but all too commonly the result-ing condition is described loosely as ' water-logging' or ' overhydration' without statingwhich components are in excess. However, theeffects of excess water or sodium, or both, givejust as definite clinical disorders as do the corre-sponding depletions, and should always be con-sidered from this standpoint.

(i) Clinical Effects of Excess of Water. It isdifficult to produce symptoms in a normal healthyperson by excessive administration of water,whether by oral or parenteral routes, unless thenormal diuretic response is artificially inhibitedby injection of posterior pituitary extract. How-ever, in the early post-operative period it is all tooeasy, owing to the operation of an antidiureticmechanism which is thought to be mediated viathe supraoptico-hypophyseal mechanism. Apartfrom the obvious risk of too large an intravenousinfusion of dextrose-in-water solutions, the dangerof rectally administered water is insufficientlyrealized, and this is due to a widely-held buterroneous belief that the body only absorbs therequired amount of water from the bowel.Unfortunately, the behaviour of the bowel in thisrespect is extremely variable in different subjectsand completely unpredictable, and some of themost severe cases of water intoxication have beenseen in these circumstances (de Takats, 193 I).

Excess water is distributed throughout all thebody fluid compartments, diluting them to anequivalent extent in terms of osmotic strength(Wynn, I955). In the normal way, therefore,approximately two-thirds of the added wateraccumulates inside the cells. The extent ofdilution of the body fluids is reflected in a loweredplasma sodium, chloride and bicarbonate, andthe symptoms generally appear when the sodiumis below I2o mEq. per litre. Initial symptoms arelassitude, mental depression and confusion, nauseaand vomiting. Later, restlessness, rapid respira-tions, spontaneous muscular twitching and drowsi-ness appear and, if untreated, the patient slips intocoma and may have generalized convulsions.The reflexes may be exaggerated.

Treatment depends on whether the patient iscapable of having a spontaneous diuresis, butassuming this is not possible, these being thecircumstances in which it usually occurs, then itis necessary to infuse hypertonic sodium chlorideintravenously at a slow rate.

(ii) Clinical Effects of Excess of Water andSodium. This is the commonest form of fluidoverloading and results when isotonic saline anddextrose-saline solutions are given in excess of thepatient's maximum output. It is therefore likelyto occur chiefly during the first five post-operativedays or in cases of cardiac or renal failure. Theexcess of sodium salts is almost entirely confinedto the extracellular fluid compartment, which isthereby expanded, causing haemodilution andoedema with the appropriate risk of pulmonaryoedema. Unless there is a relative excess ofwater compared with the extra sodium, thepatient's plasma sodium concentration will benormal. Treatment is, of course, to stop furthersodium administration, restrict water intake tothat necessary to cover the obligatory losses, and,if necessary, to adopt the measures for the treat-ment of congestive cardiac failure.

(iii) Clinical Effects of Excess of Sodium SaltsAlone. This situation is relatively uncommonexcept where an unconscious patient has had hisentire daily fluid requirements supplied in theform of isotonic saline and is unable to expresshis thirst. Insensible water vapour loss con-tinues to deprive him of about 8oo to I,ooo ml. ofwater each 24 hours, and although oliguriadevelops, with a high electrolyte output, theinevitable result is that the body fluids becomehypertonic. The clinical condition is essentiallythe same. as that of water depletion, except thatwhen the patient is given the extra water that heneeds, slight oedema may appear temporarily untilexcretion of the excess sodium (and water) iseffected.

December 1957by copyright.

on March 31, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.33.386.589 on 1 Decem

ber 1957. Dow

nloaded from

592 POSTGRADUATE MEDICAL JOURNAL December 1957

THE HYPERNATRAEMIC SYNDROMES(Oliguria, dry mucous membranes)

w (High plasma sodium) r

0z /ater deeletion t lectrolyte 0o sdiu excess Zz >rMixed NORMAL SODIUM Saline '

O ddepletions s CONCENTRATION *overloadings O

Sodium depletion ; Symptomless Water excess <uu ghyponatraemia I-- -

I (Oliguria, uraemia,) (Water diuresis later) ui ( hypotension ) 4' (Normal blood urea) I

(Low plasma sodium)

THE HYPONATRAEMIC (HYPOTONICITY)SYNDROMES

FIG. i.-The 'tetrad' of osmotic syndromes.

Chronic AdjustmentsThe various disorders of osmolarity which have

been described can be represented as a tetrad(Fig. i), showing the differential diagnosis andsimilarities of the low and high osmolarity states,depletions and overloading syndromes. There isone other condition which should be mentionedin connection with the low osmolarity states, andthat is one which was observed by Coller andMaddock in 1940. It is now variously describedas asymptomatic hyponatraemia, symptomlesshypotonicity or chronic low salt syndrome, andit is found in many chronic wasting illnesses, suchas tuberculosis, carcinomatosis, post-operativesepsis, and fistulas. It is characterized by a lowplasma sodium, which cannot be raised by ad-ministering additional sodium in the diet, and,surprisingly, the absence of the symptoms of acutesalt depletion. It is probably related to chronicpotassium and protein depletion and disappears ifand when the patient recovers from his illness.Recently, however, it has been suggested that itrepresents a chronic excess of water and that itcan be improved by the administration of corti-sone, which facilitates a spontaneous diuresis.Thus it might be related to chronic suprarenalhypoactivity.

Disturbances of Potassium BalanceThe clinical importance of potassium has

mainly been recognized since Darrow (1946)described the beneficial effects of its administra-tion in cases of infantile diarrhoea. It is, ofcourse, the predominant intracellular cation andis only present in the blood and extracellular

fluid in low and carefully limited concentrations,where it appears to be functionally related tonerve and muscle excitability. The exact con-centration is only of limited significance in relationto the total body content, for raised plasmapotassium levels may be found in the presence ofmarked potassium depletion, and low levels innormal persons, but it appears to be related to thegeneral cell metabolism according to whether thecells are retaining or shedding it. Generally,however, in surgical patients constant low plasmaconcentrations are a feature of overall potassiumdeficiency, and elevated levels are found in statesof renal failure.

Potassium DeficiencyAs excretion of potassium continues in all cir-

cumstances except anuria, depletion of the bodycontent can occur just as a result of deficientintake, in contrast to sodium. Recent work, how-ever, has demonstrated that in conditions ofpersistent low intake, the potassium excretionmay be reduced considerably below normal.

Urinary loss of potassium is increased in waterdepletion, alkalosis or following rapid intravenousinfusions of isotonic saline or dextrose solutions,and large quantities may be lost in gastro-intestinalsecretions. (The accompanying table shows theaverage composition of various gastro-intestinalfluids.) Evidence is accumulating (Darrow et al.,1948; Darrow and Pratt, I950; Cooke et al.,1952; and others) that the lost intracellular po-tassium is replaced in part by sodium and hydrogenions, and that this results in an extracellularalkalosis. Another form of loss, in direct propor-

by copyright. on M

arch 31, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.386.589 on 1 D

ecember 1957. D

ownloaded from

HOUGHTON: Post-operative Fuild Complications

tion to nitrogen loss, occurs as the result ofstarvation and tissue catabolism following injury,and in these circumstances alkalosis is uncommon.The most useful evidence in diagnosing po-

tassium depletion is a long history of loss ofsecretions. The clinical features are variable butessentially consist of weakness, lassitude, apathy,slurred slow speech, drowsiness and personalitychanges. In addition, post-operative ileus mayfail to recover, urinary incontinence may occurand the reflexes may be weak. The bloodpressure is often low, and E.C.G. changes con-sisting of reduction in voltage, diminution orinversion of the P waves, S-T depression andinversion of T waves, prolongation of the Q-Tcintervals and the appearance of U waves aredescribed, but are not necessarily present.Treatment consists of replacement with po-

tassium salts, usually the chloride, which shouldpreferably be given by mouth, but may have tobe administered intravenously when oral feedingis impracticable.

Potassium ExcessIt is not possible to elevate the plasma potassium

in healthy persons to dangerous levels by excessiveadministration of potassium salts, and this isbecause of an efficient excretory mechanism in thekidneys. It follows that cases of potassium in-toxication post-operatively occur as the result ofrenal failure or gross functional oliguria. Thereare few clinical signs except for a slow, irregularheart beat, and this is commonly anticipated byE.C.G. changes consisting of peaked T waves,prolonged PR interval and widened QRS com-plexes, when the potassium concentration exceeds7 mEq. per litre.

Acid-base Balance DisordersIn the normal person regulation of acid-base

equilibrium is maintained by the kidneys and, toa lesser extent, by the respiratory system, anddisease of either may lead to profound biochemicalderangement. The acidosis consequent uponuncontrolled diabetes mellitus and upon renalfailure is well known, as is the alkalosis found inpatients vomiting from pyloric stenosis, but theacidosis and alkalosis found in respiratory dis-orders are less familiar to the general surgeonthan to his physician colleagues. Essentially, theacid-base disturbances can be grouped into thoseof metabolic origin and. those due to respiratorydisorders, and each will be mentioned separately.

Metabolic AcidosisThe classical example of this condition is, of

course, uncontrolled diabetes mellitus, in whichacid ketone bodies accumulate in the blood and

extracellular fluid as a result of deranged carborhydrate metabolism. The hydrogen ions fromthese acids neutralize some of the plasma bicAr-bonate producing a fall in its concentratiQn(reduced ' alkali reserve' or carbon dioxide com-bining power) and a fall in the blood pH.; Withthe low bicarbonate the plasma electrolyte pattermalso reflects the acidosis in that the total of.Cl+HCO3, which normally is about 8 to I1 mEqlper litre less than the sodium level, is reduced;indicating the presence of the ketone anions,Renal failure gives a similar set of changes, theunusual anions in this case being excess of phos-phate and sulphate, although there is, of courmo,an accompanying rise in the blood urea. Also, asmentioned previously, sodium depletion can bcthe cause of this extrarenal uraemia and metabolicacidosis. It is obviously of great importance towatch for this state of affairs in surgical patientsas a slight degree of metabolic acidosis is verycommonly found after any major operation, withthe attendant low calorie intake and ketosis in th.immediately succeeding days.A metabolic hyperchloraemic acidosis due to

the accumulation of chloride ions may occur innephrotic patients, in cardiac cases after ammo-nium chloride administration and following trans-plantation of the ureters into the lower bowel.

Clinical features are those of the causative con-dition, but when acidosis is advanced the patientbecomes disproportionately ill, has a low bloodpressure and may have deep rapid respirations.The latter are due to the reduced plasma bicar-.bonate (alkali reserve) which is unable to bufferthe accumulating carbonic acid in the tissue fluidsand blood, and as the result the respiratory centre*is stimulated. Remembering that at the normal

BHCO3==20pH of blood, 7.4, the proportion H2C03 w ?

it is apparent that when the BHCO3 is lowered,as in metabolic acidosis, a corresponding reductionof carbon dioxide tensi'on in the alveolar air byhyperventilation, and therefore the reduction ofblood carbon dioxide tension and H2CO3 concen-tration, is necessary to facilitate the correction ofthe pH towards normal.

Metabolic AlkalosisThis may result from subtraction of acid, as in

the vomiting due to pyloric obstruction, and fromexcessive administration of alkali as, for examplr,in ulcer patients treated continuously with alkalinepreparations, and it is therefore important to tbhesurgeon chiefly as a pre-operative condition wliieh,should be corrected before submitting the patientto a surgical procedure. When due to alkaliingestion the body attempts a4justment by

December 1947 $93by copyright.

on March 31, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.33.386.589 on 1 Decem

ber 1957. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

secreting an alkaline urine containing sodium andbicarbonate ions, but when the alkalosis is due tovomiting gastric juice the paradox of secretion ofan acid urine is often found. This is due toactiVe sodium retention consequent upon loss ofextracellular fluid, continued small losses ofpotassium in the gastric secretions and obligatorypotassium loss in the urine, frequently resultingin an intracellular acidosis (Cooke et al., 1952).This biochemical complication is frequently foundin patients vomiting post-operatively owing to atemporary pyloric or stomal block.The alkalosis is characterized by irritability and

mental depression, cyclical periods of shallowrespiration, almost amounting to apnoea, andspontaneous tetany and myotatic irritability. Theplasma bicarbonate is proportionately high (up to66 mEq. per litre), the chloride correspondinglylow, and the plasma sodium may be elevated asthe result of excessive sodium intake or excesswater loss in vomiting. It is now thought thatpotassium depletion plays a major part in thecausation of the alkalosis and that treatmentshould be, primarily, its correction with potassiumchloride.

Respiratory AcidosisFor the surgeon this is a relatively new hazard

which has only come to the forefront since theintroduction of relaxant drugs in anaesthesia.These are particularly dangerous to patients withchronic respiratory insufficiency, e.g. chronicbronchitis and emphysema, and more than everbefore the patient is at the mercy of his attendants.If he is returned to the ward before the effect ofthe last injection of a relaxant agent has worn off,or has been countered by prostigmine, he may'have a sufficiently reduced respiratory excursionto,lead to a serious accumulation of carbon dioxideand a consequent respiratory acidosis. The per-sistence of a pink colour, due to vaso-dilatation,is a deceptive feature in a patient intoxicated withcarbon dioxide and the diagnosis may be missed,for when the acidosis is severe the patient becomesdeeply comatose, stops breathing and may developsigns suggestive of a cerebro-vascular accident,including papilloedema and a raised C.S.F.pressure.Treatment consists of artificial ventilation with

a respirator until the excess of carbon dioxide hasbeen blown off and the patient regains con-sciousness.

Chronic respiratory acidosis as in bronchiticsubjects is characterized by an accumulation ofbicarbonate in the plasma to buffer the excessretained carbonic acid and, thus, restore the pro-

BHCO3pdrtion H C and pH towards normal. It is essen-*H2C03

tially a medical condition, and it would be inappro-priate to discuss it here except to say that suchpatients may be precipitated into the acute phasevery easily by infections, any post-operative chestcomplication, or relaxant drugs.

Respiratory AlkalosisThis is the opposite to respiratory acidosis and

occurs in anaesthetized patients who are over-ventilated by the anaesthetist, and in conditionswhere the respiratory centre is stimulated directly.The most familiar example of the latter seen bythe surgeon is in hepatic coma, where hyper-ventilation is common and is due to primarystimulation ofthe respiratory centre by an unknownmechanism. In these circumstances excessivecarbon dioxide is blown off with a resulting fall inthe tension of the gas in the alveolar air and blood.A compensatory mechanism of metabolic acidosisand excretion of an alkaline urine subsequentlyresult in a return of the pH towards normal anda fall in the plasma bicarbonate to levels below thenormal range.The conditions described above had a fourfold

relationship to one another, as have the syndromesof disorders of osmolarity, and may be portrayedin a similar manner (Fig. z).

ShockThe physiological and pathological responses

to acute haemorrhage and shock are well known,and so we can confine our attention here to theprofound retention of sodium and water withresulting oliguria that occurs probably as theresult of the operation of haemodynamic andhormonal factors.The prolonged hypotension which may follow

shock and oligaemia may, of course, lead to renalischaemia and tubular necrosis, a serious corn-plication which in spite of careful conservativemanagement may be followed by a fatal issue.The principles of management in these circum-stances are now firmly established along the linessuggested by Borst (1948), Bull, Joekes and Lowe(I949), and others, and essentially consist ofrestriction of water intake to I,OOO ml. plus theurine volume, limitation of electrolyte intake tothat necessary to cover measured losses, and theprovision of a high glucose intake to supplysufficient calories and reduce tissue catabolism.This regime can be given by intravenous orintragastric routes.

The Influence of Co-existing MedicalConditionsChronic Cardiac DisordersThe effect of cardiac insufficiency on the post-

operative salt and water balance is important to

December, 1957S94by copyright.

on March 31, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.33.386.589 on 1 Decem

ber 1957. Dow

nloaded from

December I957 HOUGHTON: Post-operative Fluid Complications 595

THE SYNDROMES OF ACIDOSIS

Metabolic acidosis Respiratory acidosis ui

tn Z °@tloo d .05>>, j <I L<uZ

0 C lo O

- ° NORMAL k"4 i-OO X- BICARBONATELU0 CONCENTRATIONuu high Z )Iqj

pHCOa4Metabolic alkalosis I Respiratory alkalosis

THE SYNDROMES OF ALKALOSISFIG. 2.-The 'tetrad' of acid-base balance syndromes.

remember when a patient with this conditionneeds to have an operation, and this is a frequentproblem nowadays following the introduction ofvalvulotomy operations. A characteristic featureof the disorder is the tendency to retain sodiumand water, and this, of course, is the basis forusing a low salt diet in the medical control ofcardiac oedema. It is clearly vital not to loadsuch a patient with more sodium and water thanhe can excrete in the post-operative period, andit is equally important to treat any decompensationof cardiac function that may occur. In spite ofthe presence of oedema it is vital for these patientsto be given the normal amount of water to covertheir obligatory losses, or water depletion may beadded to their troubles.

Chronic Respiratory DiseaseThis has already been discussed under acid-

base disorders.

Chronic Renal DiseaseThis, of course, may take many forms and is in

any case often a contraindication to any opera-tion other than an urgent lifesaving procedure.The importance of recognizing urinary sodiumloss in chronic renal disease has been mentioned,and the need to supply appropriately increasedamounts of sodium salts in the post-operativeperiod should be assessed on the basis of the24-hour urinary sodium output.

In the nephrotic syndrome and some cases ofchronic glomerular nephritis there is a markedsodium retention, and these cases must bemanaged with the same careful consideration asin the cardiac cases. Again the 24-hour urinary

sodium output is a guide to the amounts that canbe replaced, although any extra-renal losses mustbe fully covered.

Diabetes MellitusThe fluid complications due to uncontrolled

diabetes in the post-operative period are wellknown, and have been mentioned earlier. Essen-tially they consist of water, sodium and potassiumdepletion from polyuria with a superadded meta-bolic acidosis.

Suprarenal DisordersThe importance of recognizing Addison's

disease and of the careful post-operative manage-ment with respect to adequate sodium replacementhas also been discussed earlier, but should beemphasized again. Similar principles shouldgovern the management of patients subjected tobilateral adrenalectomy, and in all these casesadequate substitution therapy with cortisone anddesoxycorticosterone acetate is, of course, essentialif an Addisonian crisis is to be avoided.

Diseases Requiring Cortisone TherapyThese are discussed elsewhere in this issue.

Liver DiseaseOliguria is a well-recognized feature in acute

hepatic failure, but there are also importantchanges in chronic cirrhotic disease. Althoughthe mechanisms are far from clearly understood,the cardinal features are formation of ascites andoedema, consequent upon a raised portal venouspressure and low plasma protein levels, and atendency to retain sodium and water. The main-

by copyright. on M

arch 31, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.386.589 on 1 D

ecember 1957. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

tenance of a low urinary sodium output in spiteof a normal intake may be due to increased pro-duction or decreased inactivation of sodium-retaining hormones (Bongiovanni and Eisen-menger, 1951; Goldman and Bassett, 1952), andincreased amounts of an anti-diuretic substance(Hall et al., 1949; Sims, I950) may account forthe persistence of an oliguria in spite of variationsin the intake. The low plasma protein levelsundoubtedly assist the formation of peripheraloedema, but it may be that hormonal factors areinvolved in the initial retention of salt and water.So far as the surgeon is concerned this means

that after operations on cirrhotic patients thesodium and water intake must be very carefullysupervised, as one would in the case of cardiacdisease.

Special Problems Associated with SurgeryHypothermia

Artificial cooling of patients being subjected tomajor cardiovascular surgery and certain otheroperations is a relatively recent development andhas brought some special biochemical problems ofits own. Wynn (1954) has produced evidence thatthe metabolism of glucose is considerably reducedin hypothermia, and that the intravenous infusionof glucose or glucose-saline mixtures in thesecircumstances can lead to very high blood-sugarlevels, even in excess of i,ooo mg. per ioo ml.Because the glucose is not taken up by cells underhypothermic conditions, the added glucose re-mains in the extracellular fluid and draws waterout of the cells by osmotic forces, thereby render-ing the cells relatively water-depleted. In orderto avoid this complication it has been recom-mended that 2.5 per cent. glucose solutions, tohalve the glucose load, or preferably plain salinesolutiop3 should be used.

Another complication is that citrate ions, fromthe anticoagulant in donor blood bottles, are lesseasily metabolized and may accumulate in thepatient's blood. They have recently been blamedin cases of profound hypotension and cardiacarrest where decreased ionized calcium levels havebeen demonstrated (Bunker et al., I955). Cal-cium salts have proved disappointing in the treat-ment and prevention of this condition.

Free potassium ions may be present in storeddonor blood in relatively high concentrations (upto io mEq. per litre) and thus rapid and largebulk infusions in the hypothermic subject, whoseglucose metabolism is depressed, may be dan-gerous. Freshly-drawn blood is safer.

UreterocolostomyThis operation is now well established for cases

where total cystectomy has to be performed for

malignant disease and it has been shown to leadto an interesting biochemical complication, namely,that of hyperchloraemic acidosis. Many theorieshave been advanced to explain this condition, butrecent work by Parsons et al. (I952) has shownthat chloride is reabsorbed from the urine accu-mulating in the colon more quickly than is sodium.Patients affected by this condition may havenausea and vomiting, weakness and lassitude.Poor appetite leads to inadequate food intake andsignificant potassium depletion may occur. Reliefis usually affected easily by continuous drainageof the lower bowel with an indwelling tube, andthe biochemical changes regress. However, occa-sionally it may be necessary to administer potas-sium, and this is best given as potassium citratemixture, 30 gr. three times a day (Wilkinson, 1954).

Principles of Post-operative FluidManagementThe first principle in preventing post-operative

fluid complications is to have the patient in acorrectly balanced state when he goes to operation,and however urgent the surgical procedure itssuccess can always be made the more certain byspending a short while pre-operatively correctingblood loss or gross fluid depletion. Post-opera-tively, the objectives are threefold: to correct lossof circulating blood volume with blood, plasma,dextran, etc., as the situation demands; to coverobligatory fluid and electrolyte losses; and toreplace losses incurred by vomiting and suchlike.It is with the latter two aspects that we are con-cerned here.

Fluid Requirements in Relation to the Stages of theMetabolic Response

(i) First 48 Hours. As mentioned earlier, in thefirst 24 to 48 hours after any major operation orgeneral anaesthetic there is oliguria apparentlydue to primary water retention mediated by thepituitary antidiuretic mechanism. The smallvolume of urine secreted is characterized by ahigh specific gravity and high electrolyte content,but as the volume passed in the first 24 hours isseldom more than 6oo mL., it is uncommon forthe total electrolyte loss to be more than about30 mEq. of sodium and 40 mEq. of potassium,and it is usually less. The insensible fluid lossis often high in the first 24 hours, covering theperiod in the operating theatre, and may amountto 900 to I,200 ml. The sodium loss from theskin is about 5 mEq. Hence the overall obliga-tory fluid loss in the first 24 hours is about I,500to I,8oo mi., and the electrolyte loss relativelysmall. Of course, vomiting,- diarrhoea, the drain-age of fistulas, ascites and the like will add to the

596 December 1957by copyright.

on March 31, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.33.386.589 on 1 Decem

ber 1957. Dow

nloaded from

HOUGHTON: Post-operative Fluid Complications

fluid and electrolyte loss, but this must be assessedand replaced appropriately to the individual case.The obligatory losses in each 24 hours can be

covered by I,500 to i,8oo ml. of water andapproximately 70 mEq. of sodium, given assodium chloride, to replace sodium plus potassium.Thus, if the intravenous route is to be used,500 ml. of normal saline may be given, and therest of the fluid as dextrose in water solutions. Insituations where intravenous therapy is not neces-sary, because the patient will be able to commenceoral fluid feeding within 48 hours or so and elec-trolyte losses are small, as, for example, afteruncomplicated partial gastrectomy, many surgeonsfavour the rectal route and 3 pints (1,500 ml.) ofwater per 24 hours is a common and satisfactoryprescription. Whenever possible, the patientshould be allowed to take his fluids by mouth.

Clearly during this phase of primary waterretention the risk of excessive administration isgreatest, and water intoxication is a real dangerwhen fluid intake is not carefully supervised.

(ii) Third to Sixth Post-operative Days. Afterthe first 24 to 48 hours the oliguria passes off, butthe characteristic finding at this stage is that thesodium content of the urine is low (averageIo mEq. per litre or less) and that the principalcation excreted is potassium. The amount of thelatter which is lost seems to bear a rough directrelationship to the severity of the operation andmetabolic response, and in some cases very largeamounts (up to 150 mEq. per day) may be ex-creted in the urine. It is not surprising, therefore,that many patients are quite markedly depleted ofpotassium when they reach the sixth day of intra-venous therapy unless some replacement of thation has been prescribed, and it has been sug-gested, for example, with some good evidence,that delay in recovery from post-operative ileus isoften due to potassium deficiency (Streeten andWard-McQuaid, 1952; and others). Sodiumshould be adequately replaced if vomiting, diar-rhoea and fistulas are producing sizeable losses,for otherwise there is a risk of salt depletion and,if a sudden increase in water intake occurs inthese circumstances, superadded water intoxi-cation.

Fluid prescriptions should aim to supply i,8ooml. of water for obligatory losses (in additionsome I00 to 200 ml. of water is formed in 24 hoursfrom oxidation processes) and the appropriatesodium replacement for the urinary electrolytelosses. The commonly described ' sodium para-dox' (Wilson et al.) of a falling plasma sodiumwhilst the patient is kept in sodium and waterbalance may be observed if insufficient sodium isgiven to corer also the potassium losses, sodiumand potassium being osmotically equivalent (Wynn

and Houghton, in the press). As the total urinary(Na+K) losses usually amount to 6o to I00 mEq.per 24 hours, 750 ml. of normal saline per 24 hoursgenerally proves sufficient, with the additionalwater supplied as dextrose solutions. Potassiumcan be added to infusion solutions to the extentof 25 to 50 mEq. per 24 hours when large lossesare occurring, but this is safe only when thepatient is passing more than 8oo mL. of urine in24 hours. The accompanying table lists theaverage composition of secretions from the gastro-intestinal tract and replacement of losses shouldbe arranged appropriately, using normal or half-normal sodium chloride solutions with potassiumchloride supplements. These figures are intendedonly as a guide, for the exact composition variesin individuals and direct analysis of these fluidsshould be performed whenever possible.

Correction of ComplicationsThe first essential is, of course, the establish-

ment ofa correct clinical and biochemical diagnosis,for the aim in treatment must be the correctionof the condition which has led to the clinical andbiochemical state observed in the patient, and notjust an attempt to ' put the figures right.' Withthis in view it is clearly vital that accurate fluidbalance records should be kept for any patientsubjected to a major surgical procedure, and alsothat whenever possible the patient should beweighed before operation. A detailed clinicalexamination and reference to the fluid balancecharts will generally lead to the correct diagnosis,although in some cases this is not possible untilthe results of blood and urine analyses are avail-able. In the absence of recent tissue wasting thepresence in the adult of signs of loss of skinelasticity represents the deficiency of not lessthan 2 litres of extracellular fluid. Estimationsof the plasma Na, K, Cl, HCO3 and urea shouldbe done whenever possible, and in the case of adepleted patient will give information as to therelative deficiency of water and sodium; e.g. ifthe plasma sodium is normal or low and the urearaised, sodium depletion is present, and a fallingbicarbonate concentration represents a developingmetabolic acidosis. On the other hand, a highbicarbonate and low chloride in this patient wouldsuggest a metabolic alkalosis from loss of acidgastric secretions or from potassium depletion.

Let us consider first a patient who has oli-guria, signs of loss of skin elasticity, dry mucousmembranes and hypotension, with a plasmasodium of I26 mEq. per litre, potassium 5.5 mEq.per litre, chloride go mEq. per litre and bicar-bonate 20 mEq. per litre. The clinical findingsindicate a loss of extracellular fluid, and the lowplasma sodium shows that the electrolyte loss is

December 1957 597by copyright.

on March 31, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.33.386.589 on 1 Decem

ber 1957. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

relatively greater than that of water. The loweredbicarbonate and relatively greater reduction inanalysed anions compared with cations suggestsa metabolic acidosis. There are two aspects tothe fluid replacement in this case: first, the bulkreplacement of the missing extracellular fluid withan isotonic sodium solution; and secondly, theprovision of additional sodium salts in appropriateamounts to elevate the concentration in the existingbody fluids. However, as during the period ofintravenous correction therapy the patient willneed water to cover the obligatory losses, thisadditional sodium can usually be supplied in theform of an isotonic solution if the infusion isspread over more than 48 hours. In the normalcourse of events 2 litres of normal saline may begiven fairly rapidly to restore extracellular fluidvolume, and the clinical signs of depletion shouldregress. A third, or even a fourth, litre may begiven if clinically indicated whilst therapy is inprogress. In order to correct the existing lowsodium concentrations further calculations areneeded, as it is necessary to ascertain the amountof sodium that must be added to raise the concen-trations in all the body fluid by the desiredamount. If calculated on the basis of the extra-cellular fluid volume the infusion fails to producethe expected rise in plasma sodium because in-creased electrolyte concentrations in the extra-cellular compartment result in the shift of waterfrom the cells and equilibration as if the sodiumwere distributed in the entire body water. Letus suppose that our patient, a male, weighed70 kg. before operation and his total body fluidvolume was therefore about 42 litres (6o per cent.of body weight), and that it is desired to raise hisoverall plasma sodium concentration to 140 mEq.per litre. Sufficient sodium must be infused toraise the existing 40 litres of body fluid by(I4o-i26=4) mEq. per litre, and, therefore(40X14=56o) mEq. of sodium are necessary.This amount could be given as a hypertonicsolution over a period of, say, 24 hours; or, better,could be infused slowly as an isotonic solutionover a period of some 48 hours. It is advisableto repeat the blood analyses 24 hours after com-mencement of intravenous therapy, for in this wayerrors due to estimation of, say, the initial totalbody water will not incur clinical risks. It is alsoof vital importance to watch the state of thecirculation, especially in elderly patients in whomthe dangers of pulmonary oedema are great. Itis generally wise to lengthen the time period overwhich replacement is carried out in such patients.A further advantage of a slower infusion is that itallows an opportunity to observe the recovery ofrenal function and if a serious renal shutdownsuch as in tubular necrosis has occurred adjust-

ments to fluid prescriptions can be made beforethe patient is filled with electrolytes which he willbe unable to excrete. Conservative treatmentcan thus be commenced early.

Similar principles govern the calculations* forthe correction of all disorders, but one furtherexample, namely, acute water intoxication, maybe mentioned. Let us consider a female patientwho has had an elective partial gastrectomy andwho, in the first 48 hours after operation, hasreceived a sufficiently large excess of water (byintravenous or rectal routes) to induce drowsinessand other signs of water intoxication. Let us,furthermore, suppose that she is found to have aplasma sodium of i i8 mEq. per litre and that herpre-operative body weight was 6o kg. As thefemale generally has a slightly greater proportionof body fat compared with a male of the sameweight we will assume that the initial total bodywater was approximately 50 per cent. of the bodyweight, i.e. in our patient the total body waterwas approximately 30 litres. If the patient'splasma sodium was normally I40 mEq. per litre,it follows that [(I40-I i8) x 30] mEq. or 66o mEq.of sodium have been transferred into the addi-tional water, the volume of which was approxi-mately (66o +-I8-5.6) litres. The amount ofsodium chloride necessary to restore the initialsodium concentration of I40 mEq. per litre to theexpanded total body water (35.6 litres) is thus[(I40-1I8)X35.6]=783 mEq., and it must obvi-ously be in hypertonic solution, say, approximately3 to 5 per cent. w/v. The infusion should begiven into a large vein over a period of not lessthan six hours, and the clinical improvement isusually quite dramatic. As a general rule it iswise to give only about two-thirds of the calcu-lated amounts of sodium in the first 24 hours andto wait for a simultaneous diuresis.

In situations where the plasma is hypertonicwith respect to sodium the simplest method is tocalculate the total excess of sodium salts presentand thus the volume of water necessary to dilutethem down to normal plasma concentrations.

Summary and ConclusionsThe post-operative period carries certain special

hazards in relation to fluid and electrolyte balance,and disorders of body fluid volume, osmolarityand acid-base balance may occur.The fluid and electrolyte requirements of a

patient post-operatively are dictated by the meta-bolic response, and the general principles havebeen considered. It is important, however, toremember the influence of co-existing cardio-

*The hasic formulae have been simplified in thisaccount.

598 December 1957by copyright.

on March 31, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.33.386.589 on 1 Decem

ber 1957. Dow

nloaded from

66o POSTGRADUATE MEDICAL JOURNAL December 1957

respiratory, renal and other diseases in modifyingthe response of the patient, and newer surgicaltechniques involving the use of relaxant drugsand hypothermia bring their own peculiarproblems.

Awareness of possible fluid complications is thesurest way of preventing them, and care taken inpre-operative assessment and treatment is wellrepaid in the smooth post-operative course.When fluid and electrolyte losses have occurred,accurate quantitative assessment of the bio-chemical derangement must be the prelude toplans for corrective therapy.

In this way the surgeon building his prescrip-tions on sound clinical, chemical and mathe-

matical foundations can face the metabolic hazardsof surgery with confidence, anct avoid fluid com-plications in his patients.

AVERAGE COMPOSITION OF SECRETIONS FROMTHE GASTRO-INTESTINAL TRACT

Composition(nimEq./litre)

NatureNa K Cl

Gastric 50 10 150Biliary I45 5 100Pancreatic .. 'I45 5 75Upper intestinal 1.40 10 100Ileostomy (established) 50. 5 25Diarrhoea fluid 1.45 15-20 75

BIBLIOGRAPHYBONGIOVANNI, A. M., and EISENMENGER, W. J. (i9si),

J7. cin. Endocr. II, 152.BORST, J. G. G. 41948), Lancet, i, 824.BORST, J. G. G. (r948), Acta. med. scand., I30, Suppl. 207.BULL, G. M., JOEKES, A. M., and LOWE, K. G. (I949), Lancet,

ii, 229.BUNKER, J. P., STETSON, J. B., COE, R. C., GRILLO, H. C.,

and MURPHY, A. J. (xgSS), J. Amer. med. Ass., 157, I6 I36I.COOKE, R. E., SEGAR, W. E. CHEEK, D. B., COLVILLE,

F. E., and DARROW, D. C. (i952), J. cUn. Invest., 31, 798.CUTHBERTSON, D. P. (1930), Bichem. Y., 4, I244.CUTHBERTSON, D. P. (1936), Brit. Y7. Surg., 23, 505.DARROW, D. C. (1946), J. Pediat., 28, Si5.DARROW, D. C., SCHWARTZ, R., IANUCCI, J. F., and

COLVILLE, F. (I948), 7. clin. Invest., 27, 198.DARROW, D. C., and PRATT, E. L. (Ig9o), J. Amer. med. Ass.,

143, 36% and 432.GOLDMAN, R., and BASSETT, S. H. (I952), J7. clin. Invest.,

29, 973.

HALL,C. A., FRAME, B., and DRILL, V. A. (I949), Endocrinology44, 76.

McCANCE, R. A. (1936), Proc. roy. Soc. Med., 119, 245.MOORE, F. D., and BALL, M. R. (1952), 'The Metabolic

Response to Surgery,' Springfield Illinois, C. C. Thomas.PARSONS, F. M., PYRAH, L. N., POWELL, F. J. N., REED,

G. W., and SPIERS, F. W. (i952), Brit. 7. Urol., 24, 317.PETERS, J. P. (I948), New Engl. . Med., 239, 353.SIMS J. L. (I950) -7. Lab clin. Med., 36, 990.STREETEN, D. UI. P., and WARD-McQUAID, J. N. (I952),

Brit. med.a., i, 587.DE TAKATS, G. (i 3 i), Amer. Y. Surg., II, 39.VERNEY, E. B. (1947), Proc. roy. Soc. B, 135, 25.WILKINSON, A. W., BILLING, B. H., NAGY, G., and

STEWART, C. P. (Is95o), Lancet, 1, 533; Ibid., ii, 135.WILKINSON, A. W. (I9S4), Postgrad. med.!Y., 30, 405.WILSON, G., BROOKS, L., and MOORE, F. (1953),J. clin.

Invest., 32, 612.WYNN, V. (I9s4), Lancet,,ll, 575.WYNN, V. (I955), Cisn. Sci., 14, 669.

THYROID DISEASE(Postgraduate Medical Journal, July 1957)

Price: 3s. 9d., post free

DIAGNOSTIC PROCEDURES IN THYROID RECENT WORK ON THYROID HORMONESDISEASE J. H. Wilkinson, B.Sc., Ph.D., F.R.I.C.Russell Fraser, M.D., F.R.C.P., D.P.M.

SOME UNUSUAL MANIFESTATIONS OFTHE PLACE OF RADIOACTIVE IODINE IN THYROID DISEASETHETREATMNTHOFOTHYRIDSDISEASTHETREAI NTOFTHYROIDDISEASE W. R. Trotter, D.M., M.R.C.P.

E. E. Pochin, M.D., F.R.C.P.

ANTITHYROID DRUGS CARCINOMA OF THE THYROIDJames Crooks, M.B., M.R.C.P. (Lond. and John E. Piercy, F.R.C.S., F.R.C.S.E.

Ed.), F.R.F.P.S.G.LYMPHOID GOITRES

SUBACUTE THYROIDITIS T. Levitt, M.A., F.R.C.S.Eng., F.R.C.S.Ed.,Selwyn Taylor, M.Ch., F.R.C.S. F.R.C.S.I.

Published by

THE FELLOWSHIP OF POSTGRADUATE MEDICINE

60, Portland Place, London, W.1

by copyright. on M

arch 31, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.386.589 on 1 D

ecember 1957. D

ownloaded from