6i

THE TREATMENT OF URAEMIAWITH SPECIAL REFERENCETO ACUTE RENAL FAILURE

By A. M. JOEKES, M.A., B.M.Department of Medicine, Post Graduate Medical School of London

The diagnosis of uraemia is often consideredsynonymous with a hopeless prognosis, or, at best,a condition from which a patient may recover buton which treatment has no specific effect. Thispaper attempts to set out a rational system oftreatment for uraemia. The clinical picture,aetiology of the renal failure, diagnosis and prog-nosis are briefly considered before treatment isdiscussed.

Fishberg1 has defined uraemia as' the symptom-complex resulting from renal insufficiency andaccompanying the retention of urinary con-stituents in the organism.' It should not beforgotten that water is the most important urinaryconstituent, and many of the symptoms of oliguricuraemia result from its retention.

Clinical PictureTo await a clinical condition which allows of a

confident bedside diagnosis of uraemia is to courtdisaster. The classical textbook picture wouldleave little room for doubt. A pale, drowsy orcomatose patient with slightly yellowish tinge tohis skin, showing generalized oedema, hiccough-ing, the breath smelling uriniferous, the tonguedry and brown, involuntary twitchings of thelimbs, positive Chvostek and Trousseau signs,possibly a bleeding tendency or generalized con-vulsions will complete an unmistakable clinicalpicture. In practice, the clinical condition of auraemic patient is determined largely by the fore-going treatment, probably most of all on themaintenance or otherwise of the fluid balance. Ifideally treated, a patient with a blood urea of300 mgm. per cent. or more may present with aclean moist tongue, normal mental faculties, nooedema and, in short, with no abnormal physicalsigns.

So favourable a clinical picture is not likely tooccur in the absence of very careful treatment;there will almost certainly be an acidosis, with analkali reserve below 40 vols. per cent. (resulting inincreased depth of respiration) and evidence ofcardiac failure as judged by a raised venouspressure, although there may be little or no

oedema unless fluid intake has been forced.Nevertheless, such a picture does not call to mindthe uraemic syndrome as generally described and,unless the reduced urinary output is taken intoaccount, the diagnosis will be overlooked. Indeed,the clinical picture in uraemia should be used moreas a guide in management, rather than as a meansof diagnostic labelling; recognition must largelydepend upon the clinical history.Diagnosis

It is impossible to discuss the.diagnosis ofuraemia without stressing continually the im-portance of watching the urine output in any con-dition that may be complicated by renal failure.Not only must the volume passed in the 24 hoursbe taken into account but, especially in thepresence of existing renal disease, the urea contentbe borne in mind.Some classification of the cause of uraemia is

useful in the approach to a case with renal failure,and a table is given based on Fishberg's1 threemain groups. No attempt is made to give a com-prehensive list but it may be useful as including thecauses of renal failure most likely to be met within practice.Prerenal.

Postrenal.I. Stone.2. Prostatic obstruction.3. Obstruction of ureters by carcinoma of the cervix.4. Obstruction of the ureters by uterine prolapse.

Renal (Chronic).i. Glomerular nephritis.2. Pyelonephritis.3. Nephrosclerosis (benign and malignant).4. Polycystic kidneys.5. Amyloid kidneys6. Renal tuberculosis.7. Renal calcinosis.

Renal (Acute).i. Glomerular nephritis.

*2. Chemical poisons. Mercury. Carbon tetrachloride,*3. Sulphonamides.

(a) Blockage(b) Sensitivity.

by copyright. on A

ugust 26, 2019 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.25.280.61 on 1 F

ebruary 1949. Dow

nloaded from

POST GRADUATE MEDICAL JOURNAL

*4. Pigment nephrosis.(a) Mismatched transfusion.(b) Crush syndrome.(c) Haemolytic infections.(d) Severe burns with haemolysis.(e) Ischaemic muscle necrosis.(f) Blackwater fever.

*5, Shock kidney.6. Cortical necrosis.* Lower nephron nephroses (LuckeS3.)

PrerenalThe group of prerenal causes of uraemia is not

further subdivided owing to its complexity.Renal failure due to causes outside the kidney aretypically seen after severe haernetemesis, in thecrisis of Addison's disease, and after continuedvomiting or diarrhoea. The main factors resultingin the renal failure would seem to be a fall in bloodpressure, loss of water and loss of electrolytes.The importance of these various factors is, how-ever, far from clear at present, and there may beothers not yet appreciated. If the replacement ofblood loss and the correction of dehydration orelectrolyte disturbance in the presence of uraemiadoes not improve renal function, it must beassumed that there is a true renal cause for theuraemia, which may have been precipitated by aperiod of hypotension (shock kidney).

Acidosis is the result rather than the cause ofrenal failure. Whether alkalosis in itself, apartfrom the accompanying dehydration and reducedplasma chlorides, reduces renal function is notcertainly established.The mechanism of the anuria of the ' hepato-

renal' syndrome is not known.

PostrenalConsideration of the postrenal group should

prevent the oversight of obstruction of the uro-genital tract, as by stones, by prostatic or urethralstricture obstruction in the male and, more easilyforgotten, ureteral blockage either by carcinoma ofthe cervix or due to prolapse of the uterus in thefemale. Any such obstruction may be accom-panied by urinary infection and, indeed, the latterin the presence of oliguria should lead one tosuspect an obstruction. It should be rememberedthat albumen and white cells in the urine are therule in oliguria and do not in themselves denoteinfection. If urinary infection is present withoutevidence of obstruction and cannot be laid at thedoor of repeated instrumentation, previouslyexisting pyelonephritis, tuberculosis or polycystickidneys should le considered.The diagnosis of postrenal or obstructive

uraemia must not be missed as surgical relief ofthe block or pyelostomy should be performed atthe earliest opportunity.

RenalRenal failure due to a primary disturbance

within the kidney itself may conveniently begrouped into two main categories, old-standingrenal disease in which some added insult has re-duced effective renal function below the necessaryminimum, and acute renal disease or catastropheoccurring in previously healthy kidneys. It maybe exceedingly difficult to differentiate betweenthese two groups clinically; and it is usually theclinical history rather than the examination whichallows one to make a tentative diagnosis. Evidenceof previously existing hypertension should leadone to suspect old renal damage and sometimesthe urinary sediment may be of assistance.The differential diagnosis of the causes of acute

renal failure must, again, depend largely onobtaining an accurate history, although a fewclinical points may be helpful. Jaundice or araised serum bilirubin may be found in the firstday or so of the pigment nephroses. A raisedleucocyte count is found in cortical necrosis, butthis occurs normally during the first two or threedays post-partum, and, of course, if there is sepsis.A raised blood pressure may be suggestive ofglomerular nephritis, but can also occur withlower nephron nephroses.

It is extremely common to have a true renaluraemia complicated by vomiting, resulting in theloss of water and salt, the patient presenting de-hydrated and with a low blood pressure, so thatwhen first seen both a prerenal and a renal causefor the condition exist. In such a case it may beimpossible to diagnose the presence of an intra-renal cause until it is found that correcting theextrarenal cause does not restore kidney function.Such an instance is typically seen in a case withchronic dyspepsia, taking alkalis over manymonths, who presents dehydrated, alkalotic andwith a raised blood urea and lowered plasmachloride; the intravenous administration ofnormal saline will in many cases restore a renalfunction adequate to correct the alkalosis and re-duce the urea to normal. In a proportion of suchcases the administration of intravenous saline doesnot allow the kidney to restore to normal the bloodchemistry, and it can then be assumed that there ispresent severe chronic renal damage.

Presented with a case of renal failure which haspersisted for several days and in which con-siderable nitrogen retention already exists, theprimary problem should be treatment which maybe urgent, and unless successful, detailed diagnosisis of only academic interest. Once, however, allaspects of treatment that will be discussed havebeen attended to, then every effort should be madeto make as precise a diagnosis as possible. This isof importance in relation to prognosis and it is

62 February 1949by copyright.

on August 26, 2019 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.25.280.61 on 1 February 1949. D

ownloaded from

JOEKES: The Treatment of Uraemia

only by so doing that one may hope to unravelsome of the obscure oligurias. Unfortunately themodern techniques for investigating renaldynamics tell one little more than the degree ofrenal damage and nothing about the aetiology, norare they helpful in prognosis, except in so far asthey allow one to follow the recovery of functionin greater detail than a simple urea clearance test.In the presence of anuria no renal function testscan, of course, be applied; indeed there is norenal function. It is therefore not known whetherthere is some glomerular filtration with quantita-tive return of this filtrate into the blood stream, orwhether there is complete absence of filtrateformation.

Pregnancy AnuriasThe anurias related to pregnancy constitute the

largest group met in practice. O'Sullivan andSpitzer2 have reviewed the literature of renalfailure complicating abortion. With septic abor-tion, the kidney lesion is usually a pigmentnephrosis due to infection with haemolytic or-ganisms. In the case of non-septic abortion, thelesion may be either a pigment nephrosis or abilateral cortical necrosis. Post-partum complica-tions are related most commonly to haemorrhageor toxaemia. As after any severe blood loss, post-partum haemorrhage may be followed by oliguriasecondary to the fall in blood pressure and bloodvolume. Concealed accidental (ante-partum)haemorrhage is usually preceded by the signs oftoxaemia even if only for a few days, and if com-plicated by renal failure the lesion in fatal cases isfound to be bilateral cortical necrosis. Deliveryafter toxaemia without placental separation has alsobeen recorded as complicated by cortical necrosis.

PrognosisSevere renal failure occurring in the presence of

chronic renal damage can be an extremely difficultprognostic problem. An acute episode of glomeru-lar nephritis complicating chronic nephritis,whether severe oliguria occurs or not, may settle,leaving the patient with a further reduction infunctioning kidney tissue. The ultimate prog-nosis must depend largely on the residual renalfunction and the severity of the cardiovascularstrain. With other forms of chronic renal damagean episode of acute failure is usually brought aboutby some prerenal factor or a superimposed urinarytract infection. If the immediate cause of theaggravation can be treated, the prognosis is thatof the underlying renal condition.

If anuria occurs in malignant hypertension, thismust be considered as a terminal event.

If with chronic renal damage severe nitrogen re-tention occurs the prognosis depends largely on

the ability of the kidneys to excrete an adequatevolume of urine, iLe. 2 litres or more in the 24hours. If, despite sufficient water and salt ad-ministration, the urine volume cannot be increasedenough to compensate for the poor concentratingpower of the kidneys the prognosis is hopeless.Of the acute causes of uraemia only in the case

of bilateral cortical necrosis is there no convincingevidence that the kidneys will recover if the patientsurvives sufficiently long, something of the orderof three weeks in the most severe cases. Duff andMore3 reviewed the reported cases of corticalnecrosis up to I941 ; of 77, 48 had complicatedpregnancy. Such a review is necessarily based onpost-mortem evidence, and therefore fatal cases;Gibberd4 and others5 6 7 have claimed that thediagnosis of bilateral cortical necrosis can be con-fidently mnade on occasions, and he described twocases so diagnosed with post-partum anuria whoboth recovered. Acute glomerular nephritis onlyfalls into this group when severe oliguria oranuria persists for more than two or three days.Such an oliguric episode is not necessarily anindication of a severe renal damage. If the patientcan be tided over the oliguric episode the prog-nosis is probably the same as for an attack in whichsevere oliguria does not occur, in other words,an 85 per cent. chance of complete recovery is tobe expected

It should therefore be assumed that in anycase of anuria or oliguria if the patient can bekept alive for three weeks, kidney function willhave recovered sufficiently to support life. Themajority of lower nephron nephroses will havestarted a diuresis in 14 days. This will not betrue for, at least, most cases of cortical necrosis.As the latter diagnosis cannot be proved in life,treatment should be as active as for any acuterenal failure; it is moreover possible that somedegree of recovery might take place if life be pro-longed beyond i2 days, within which time thegreat majority of the recorded cases have died.Any case of lower nephron nephrosis, that is anycondition falling into subdivisions 2,'3, 4 and 5 ofthe causes of acute renal failure, which has sur-vived the acute episode will show a slow improve-ment of renal function reaching normal or slightlybelow normal values in anything up to ninemonths.

TreatmentAs the first step in treatment the possibility of

postrenal obstruction must be considered; Ifsuch an obstruction exists and cannot be relieved,nephrostomy should be performed immediately.If sulphonamides have been administered up tothe onset of oliguria and might be a cause forobstruction, ureteric catheters should be passed

Febrgary I949 61by copyright.

on August 26, 2019 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.25.280.61 on 1 February 1949. D

ownloaded from

POST GRADUATE MEDICAL JOURNAL

and the renal pelves -washed out. Obviouslysulphonamides must not be administered in thepresence of oliguria.Of the renal causes of uraemia there are two

main groups of patients: one in which anuria oroliguria persists for some days despite treatment sothat insignificant amounts of breakdown productsare excreted; the other in which the concentratingpower of the kidneys has become grossly impairedbut in which a large urine volume can be main-tained with correct management. There is noessential difference in the treatment of the twogroups. In the presence of anuria the retention ofpotassium plays an important part, whereas theability to excrete potassium is usually maintainedwhile the kidney can excrete water.

Emergency TreatmentApart from the replacement of blood loss after

haemorrhage and the correction of dehydrationand electrolyte depletion, there are only two groupsof kidney catastrophe in which immediate treat-ment may be of decisive importance.

In the pigment nephroses, the renal lesion isthought to be dependent on the precipitation ofhaemoglobin or myohaemoglobin in the tubules.This precipitation probably varies with the con-centration of the pigment in the affected tubules,with the change in pH of the filtrate to acid, andpossibly with the concentration of sodium chloride.Clinically, as judged by the onset of oliguria, therenal damage is complete within 24 and possiblywithinl 2 hours. It would seem rational thereforeto attempt to dilute the urine maximally in thefirst I2 hours after the transfusion or injury and toadminister alkalis sufficient to render the urinealkaline. Bywaters8 advised for the crush syn-drome cases that, as an immediate first aidmeasure, one pint of fluid with one teaspoonful ofbicarbonate be given by mouth every hour for thefirst I2 hours, or, if the patient is in hospital,intravenous isotonic (I.87 per cent.) sodiumlactate in the same quantities. It should be added,however, that excess of total fluid intake over out-put should not at any time be allowed to be morethan I,500 cc. (2 pints), unless an existing de-hydration has to be corrected. Very good experi-mental evidence is offered for the use of isotonic(4.3 per cent.) sodium sulphate in the anuria pro-duced by haemoglobinaemia secondary to severeburns9. If this work is accepted as valid, sodiumsulphate should be used in place of sodium lactate.If in the first 24 hours a successful diuresis isobtained, the fluid intake should be such as tomaintain a daily urine. volume of about 3 litres;this will require 4 litres (7 pints) intake. To keepthe urine alkaline about 2 gm. (30 gr.) of bi-

carbonate are required in the day. It must bestressed that to- attempt to force a diuresis bygiving large quantities of fluid after the first 12hours is exceedingly dangerous unless the urineoutput keeps pace with the fluid intake. Anypatient with lower nephron nephrosis, particularlythose due to chemical poisons, may have a diuresisin the first 24 hours and then become anuric.The other group of renal catastrophe which

may benefit by emergency treatment is acutemercury poisoning. If adequate treatment withBAL is instituted within three hours of theadministration of the mercury, the prognosis isvery much improved'0. Of the total course ofBAL, up to 2 gm., 6oo to 750 mgm. should begiven in the first I2 hours after the event.

General TreatmentTreatment of a uraemic patient can be con-

sidered under several main headings:-i. Diet.2. Water balance.3. Correction and maintenance of electrolyte

pattern.4. Prevention or treatment of intercurrent in-

fection.5. Blood, peritoneal and intestinal dialysis.6. Correction of anaemia.

DietThe symptom complex of uraemia is probably

dependent on the retention in the body of manysubstances normally excreted in the urine,aggravated by disturbances in the electrolytepattern and the acid-base balance. Endogenous orexogenous protein breakdown is almost entirelyresponsible for the accumulation of the toxic sub-stances. A starving man will katabolize about50 gnm. of endogenous protein a day, whereas, ifgiven no protein but an adequate calorie intake inthe form of carbohydrate and fats, he will kata-bolize less than 2 gm. of nitrogen or about 12 gm.protein. In other words the normal daily ureaexcretion of about 20 g. can be reduced by ahigh-calorie, low-protein diet to 5 g. per day.Expressed as the theoretical daily rise in bloodurea level in an anuric patient, this would beequivalent to a io mgm. per cent. as opposed to40 mgm. per cent. rise.Kempner" and Borst'2 13 have stressed the

importance of giving adequate-calorie, low-proteindiets to uraemic patients. Below is given a list ofthe protein content of some of the main foods(extracted from Borst)

Februory 1049by copyright.

on August 26, 2019 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.25.280.61 on 1 February 1949. D

ownloaded from

JOEKES: The Treatment of Uraemia

Alcohol..SugarButterCornflourTapiocaSago JFruitCream (40 per cent.'PotatoRice (raw)Bread (old)MilkEgg (2- IOC gMi)GreensMeat (lean)

Gm.proteinper Ioogm.00

0.44

Caloriesper xoo

grn.7004008oo

Gm.protein

per 2,00ocals.

00

T . I

0. 5 350 2.75

. . 0.25-1.0fat) 2.4

I.44.. 6.8.. 9.1

3.25. . 11.9.. .0-5.0. 22.0

30-l004008o

360250545170IO'0o12$

1212

363873125200200

350

It will be seen that butter and sugar must formthe main basis of any diet, with cornflour, tapioca,sago and fruit as the vehicles for making somethingpalatable. Rice, although having a fairly highprotein content, is extremely useful owing to itshigh calorie value and can be used as the mainstayof one hot meal in the day. Salt must not be givenas a rule while the patient is anuric, which is afurther difficulty in making the diet sufficientlypalatable. In practice, in this country, it has beenfound impossible to get patients to take a calorieintake of 2,000 per day, which is aimed at, theaverage intake with a co-operative patient workingout at I,ooo calories.

In the presence of anuria the intake of potassiummust be kept to a minimum. Milk contains ahigh potassium content; not more than 500 gm.of fruit should be given daily. Synthetic fruitjuices may contain more than IO mgm. per cent.of potassium and must not be given14.

Recently it has been found that an adequatecalorie intake can be assured by feeding bystomach tube a mixture of fat and sugar in wateras a continuous nasal drip15 16. This method alsoallows the fluid and salt intake to be very accuratelycontrolled.

Water BalanceIn reviewing treatment of 33 cases of urinary

suppression Lattimer17 found that of those re-ceiving less than 2,000 cc. total fluids in the 24hours none died, whereas of those in whom 3,500cc. or more was given 75 per cent. died. He sumsup his view by stating: ' the body is not analogousto a tank into which water can be forced until itfinally bursts out through the kidneys.'In all probability most uraemic deaths are due

to cardiac failure, in which overloading withwater with consequent pulmonary oedema is themost important precipitating factor36 37. Electro-lyte disturbances, in particular retention ofpotassium, may seriously affect the heart action.

If no water is lost from the body through thekidneys, the only losses are by way of the lungs and'skin, and in the faeces and vomit. The expectedwater loss by these routes, in the absence ofvomiting or diarrhoea,.totals about one litre a day.and this'volume of water will be required by ananuric patient to maintain fluid balance. Kugel"8suggested 700 cc. of normal saline as the basicdaily requirement and stresses the possible im-portance of endogenous water produced bymetabolism and cell breakdown. The administra-tion of salt to anuric patients, 700 cc. normal salinegiving 6.3 gm., cannot, however, be acceptedunless dehydration is present. The route ofadministration is not important but the totalamount of fluid administered by mouth, intra-venously or by rectum, must be taken into account.Whole blood given intravenously should not beincluded as fluid intake. From the onset ofanuria a patient should be given a total fluid intakein 24 hours of not less than 700 cc. and not morethan 1,500 cc. If vomiting or diarrhoea arepresent the amount of fluid thus lost should begiven in addition to the above, and the loss ofsodium chloride will have to be replaced as well.In the case of vomit, half the volume of fluid lostshould be given as normal saline.When once a diuresis starts, or in cases already

passing urine, the fluid intake should total I,500 Cc.plus the volume of the previous 24 hour urineoutput. It is essential that accurate fluid intake-output charts are kept whenever a case of renalfailure is treated, and these must be continueduntil the blood urea is reduced to normal.Although a diuresis has started'after a period ofanuria this does not mean that the patient is outof danger, and it is still easily possible to overloadwith water or, if insufficient fluid is administered,to prevent the kidneys from excreting sufficientamounts of katabolic products. Following aperiod of oliguria it is the rule for the kidneys tobe unable to concentrate for some weeks. Duringthe first week or so of diuresis the urine urea con-centration will be about the same as in the bloo'd,i.e. between 2oo and 500 mgm. per cent., and itshould be assumed that a litre of urine will containless than 5 gm. of urea. It is therefore not sur-prising that the blood urea may continue to risefor some days after the start of a diuresis. Thisgives no cause for alarm, but illustrates the needfor continuing extremely careful managementuntil the patient's blood urea is normal.

Before discharging a patient who has had anepisode of renal failure, the concentrating ability ofthe kidneys should be assessed. From the ureaconcentration in a 24-hour specimen of urine therequired minimum daily urine volume on a givenprotein intake can be calculated:

Febnzrary x949by copyright.

on August 26, 2019 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.25.280.61 on 1 February 1949. D

ownloaded from

POST GRADUATE MEDICAL JOURNAL

e.g. urine urea concentration 8oo mgm. per cent.daily protein intake . . 40 gm.

If it is assumed that i gm. urea is equivalent to3 gm. of protein, the minimum required 24 hour

urine vol. .84X X 100 CC.

11,670 cc.

These figures are of the order expected ineither a recovering acute lesion or in a chroniclesion with enough function to avoid increasingnitrogen retention.

In such cases it is safer not to allow the volumeof urine passed in the 24 hours to fall below aminimum of two litres. A Winchester bottle(2 litres) can be used for a patient to collect a24-hour urine once or twice a week, and unlessthis is filled in 24 hours, the quantity must beconsidered insufficient, and the patient must takemore fluids. Where the renal failure was theresult of one of the ' catastrophes' in previouslynormal kidneys, it can be confidently expectedthat renal function will continue to improveuntil, between four and nine months after theevent, almost normal function will have returned.The necessity of maintaining a large daily urineoutput may then only persist for some two monthsafter the start of diuresis.

Correction and Maintenance of ElectrolytePatternPatients with gross renal failure, whatever the

aetiology, usually present with a disturbance ofthe electrolyte pattern. In the group of renalcatastrophes with oliguria it is common to find areduction in the plasma chloride level, notnecessarily explicable by loss in vomit, and thebicarbonate may also be low.

Despite a very low plasma chloride, patients maynot present with clinical dehydration. With chronicrenal insufficiency the bicarbonate is almost in-variably low and the chloride level is more nearlyrelated to the loss of chloride in vomit. In thelatter group the kidney is probably able to excretewater and it is then possible to administer saltwith water as a vehicle to correct a low plasmachloride level. But in the presence of anuria, theadministration of salt is extremely difficult as onehas at most i,500 cc. per day in which to give it;if more than half strength saline is given by mouthvomiting is likely. A low chloride level may wellplay an important part in maintaining a dis-turbance of renal dynamics, and a gross reductionin the presence of oliguria may be an indication forsome form of dialysis with which it is possible toright an electrolyte imbalance without water-logging the patient.

In those cases in which a large urine volume hasto be maintained because of poor concentratingability, it is most important to consider the loss ofsodium chloride in the urine. Obviously leastosmotic work is demanded of the kidneys if theconcentration of sodium chloride is the same inthe urine as in the plasma, i.e. of the order of6oo mgm. per ioo cc. If. the urine volume is2,500 cc. in the day, this would represent a lossof 15 gm. of salt. In the presence of poor renalfunction the patient will be put on a low-proteindiet to reduce the work in excreting urea, and thediet is unlikely to contain much salt. Added saltmay therefore have to be given to cases witheither chronic renal damage or, in the recoverystage of an acute failure; the amount is mostsafely established by estimating the chloride con-tent of a 24-hour specimen of urine. The extrasalt administered can be either as tablets or incapsules, but should be given with meals to re-duce the tendency to vomiting.The reduction of the plasma bicarbonate may

in itself not be of serious importance, but rather areflection of the retention of acid radicles. Dialysiscan correct a low bicarbonate value to normal,presumably removing retained acids at the sametime. The presence of an acidosis, as judged bythe bicarbonate level, is not sufficient indicationfor active intervention by dialysis. Whether anattempt to correct the acidosis by the administra-tion of alkalis is of value, as is particularly stressedby Kirk19 for acidosis of any origin, primarilyrenal or not, is still uncertain. There can, how-ever, be no doubt that the maintenance of thefluid balance in oliguria is of far greater im-portance than the correction of an acidosis. Manypatients survive after being severely acidoticduring a period of anuria, few survive when oncegross oedema has developed. It has yet to beshown that the administration of alkalis to thepatient with chronic renal damage and acidosis isof greater benefit than the administration of salt.

Potassium retention can be of great significancein oliguria, and may be the main immediate causefor death in cardiac failure20. It is difficult toexplain the very sudden death that may occur inan anuric patient with only a moderately raisedblood urea value, e.g. 200 mgm. per cent., and aspotassium intoxication, as seen in the electro-cardiogram, is of grave prognostic significance, afurther rise in serum potassium due to cell break-down could conceivably cause cardiac arrest. Themain electrocardiographic change is in the Twaves, best seen in Lead II and V3; the T wavebecomes sharp and rises abruptly from the S-Tsegment, which is usually isoelectric and occasion-ally depressed, although it may be terminally raisedand slurred into the T wave. Late changes

66 February 1949by copyright.

on August 26, 2019 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.25.280.61 on 1 February 1949. D

ownloaded from

JOEKES: The Treatment of Uraemia

described are widening of the QRS c6mplex,and disappearance of the P waves21.

Potassium retention is particularly likely to be ofimportance in cases of anuria related to pregnancywith an involuting uterus, or in patients withextensive sepsis. The avoidance of food anddrinks containing much potassium has beenmentioned in the section on diet.

Apart from the use of insulin combined withcarbohydrate administration to attempt to depositpotassium in the tissues with glycogen22, dialysisis the only means of reducing a dangerous level ofserum potassium.A reduction of free calcium ions, or the

accumulation of a substance counteracting theireffect such as guanidine, may give rise to tetany andprolong cardiac systole as shown by a lengthenedQ-T interval in the electrocardiogram. Calciumgluconate will abort tetanic attacks, although thesemay recur within a relatively short time.An increase in plasma phosphate roughly

parallels the urea retention, and, except for re-ducing the ionized calcium, is apparently notharmful. The rise parallels the nitrogen retention.

It is considered that no attempt should be madeto right an electrolyte imbalance at the expense ofupsetting the water balance. Potassium retentionis the only absolute indication for active inter-vention with dialysis.

Prevention or Treatment of IntercurrentInfectionAny uraemic patient is extremely liable to

infection, which would increase the katabolicprocesses in the body. It is therefore of thegreatest importance to guard against the pos-sibility of infection. Humphrey and Avery-Jones'4 advocated the routine use of preventivepenicillin in acute renal failure, and found thatan adequate blood level was maintained in thepresence of anuria with a single dose of 45,000units repeated every fifth day.The treatment of an established infection must

be vigorous. The drug of choice is penicillin andrelatively enormous doses may be. safely given ifan exceptionally high blood level is indicated for apartially resistant organism.

Sulphonamides must be avoided unless there isa good urine flow, preferably above two litres inthe day.

Blood, Peritoneal and Intestinal DialysisSweating and purging were the traditional

methods of treating uraemia, but not only areinsignificant amounts of urea removed by theseroutes, but the-danger of causing serious waterand electrolyte balance disturbances is very great.The belief that any toxic effects of retained

protein breakdown products can be lessened by adilution effect, that is by increasing above normalthe water content of the body, ignores the extremedanger of waterlogging an anuric patient.

In the last few years increasing attention hasbeen given to the possibility of removing retainedproducts by means of dialysing the patient's bloodagainst a physiological saline solution. Thesemi-permeable membrane could either be theperitoneum or the gut, or the blood could be ledthrough a Cellophane tube.

Artificial KidneyKolff23 24 was the first to develop extra-

corporeal blood dialysis into a practical method.Blood flows outside the body through a cello-phane membrane immersed in a bath containinga solution of sodium chloride, sodium bicarbonate,potassium chloride and glucose, the latter insufficient concentration to compensate for theosmotic pressure of the plasma proteins. Thepatient is heparinized, blood flows from a cannulain a radial artery into a hundred foot length cello-phane tube, which is wound in a spiral on awooden lath drum. This drum is half immersed ina bath of ioo litres, and revolves so that the bloodtravels in a thin film along the cellophane tubingby the principle of- an Archimedes screw, and isthen returned into a vein by means of a pump.

Alwall25 and Gordon Murray and Delorme26have independently developed blood dialysers forthe treatment of uraemia similar in principle butdifferent in detail from Kolff's.Using the Kolff type of artificial kidney it is

found that in a single passage through the Cello-phane the urea content of a sample of blood canbe reduced from over 400 mgm. per cent. to40 mgm. per cent. or less. Similarly, the bi-carbonate and chloride content can be brought towithin normal limits whatever their values in theblood passed into the ' kidney.' Creatinine, uricacid and phenols are also removed by dialysis.Owing to a certain amount of haemolysis whichalways occurs with the dialysis, it is difficult toobtain figures on the removal of the potassium ionfrom th plasma, but there can be no doubt thatthis is achieved.

Both Kolff and Alwall claim that generalizedoedema can be reduced by dialysis, but where itis not possible to, weigh these patients before andafter without a weighing table, no convincingfigures have been obtained in humans. Workingwith animals Alwall was able to reduce the bodyweight by i per cent. per hour27.

Peritoneal DialysisThe use of the peritoneum as a semi-permeable

membrane for the removal of toxins has long been

February 1949 67by copyright.

on August 26, 2019 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.25.280.61 on 1 February 1949. D

ownloaded from

68POST GRADUATE MEDICAL JOURNAL

considered. In 934 Balasz' and Rosenak usedperitoneal lavage *in the treatment of a case ofanuria resulting from mercury poisoning28. Frank,Seligman and Fine29 performed animal experi-ments with 'peritoneal lavage in induced uraemia,and went on to treat a patient with uraemia. Sincethen a large number of cases have been treated byperitoneal irrigation. In short, the method con-sists of running through the peritoneum, eithercontinuously or intermittently, a fluid verysimilar to that described as used in the bath waterfor the artificial kidney. Blockage of the outflowtube, in the case of intermittent perfusion the sameas the inflow tube, is one of the main technicaldifficulties with the danger of retention of excessfluid.

Kop"0, using the, technique for continuousperitoneal dialysis described by Kolff24, hasdialysed 2I patients on 35 occasions.' In only twowas there any resulting infection of the peri-toneum, and in neither was this severe enough tobe the cause of death.

Apart from estimating blood urea and electro-lyte levels before and after dialysis, it is moredifficult to assess the effect of peritoneal than ofblood dialysis, in which a sample of blood can betaken before and after passing through the cello-phane tubing. With a flow speed of about 2 litresper hour through the peritoneum an original bloodurea level of 300 mgm. per cent. might be broughtdown in Io hours perfusion to 200 mgm. per cent.Increased serum potassium levels can be reduced:in one case the level before dialysis was 56.6 mgm.per cent. and was brought down to 34.9 mgm. percent.80. Other electrolyte disturbances are alsorighted.

Intestinal DialysisThe intestine can be used as a semi-permeable

membrane in several ways. A multilumen tubemay be passed into the small intestine, a perfusingfluid is run in through one lumen high up in thegut, and is sucked out low down through anaperture just above the end of the tube'1. Perfus-ing fluid mav be run into the colon and collectedthrough an appendicostomy opening24. Forchronic renal insufficiency it has been suggested*that a permanent double enterostomy with a loopof ileum could be used for intermittent irriga-tion82 24 With all these methods of intestinaldialysis, the difficulty arises of preventing excesswater and salt absorption from the gut, apart fromthe relatively poor rate of urea removal. Itcannot be considered that intestinal dialysis has yetan established place in the treatment of uraemia.

Replacement TransfusionIt is obvious that replacement of a uraemic

patient's blood by fresh donor blood will removeprotein breakdown products. The latter are,however, distributed throughout the extra-cellularwater if not throughout the entire body water, andan exchange transfusion to be effective will re-quire a very large volume of donor blood, of theorder of 6o or more pints. This is impracticableonly on the score of the difficulty of obtainingsufficient blood; this is the greater in view of thehigh proportion of anurias related to pregnancyand mismatched transfusions in which Rh negativeblood may be imperative.

Indications for DialysisIt may be accepted that extra-renal removal of

protein breakdown products and the correction ofelectrolyte imbalances by some form of dialysishas a place in the management of acute renalfailure. To define the indications for its use isextremely difficult. While admitting that as ageneral index of the severity of uraemia the bloodurea level is very useful, it must be emphasizedthat some patients may have the blood urea risingto over 500 mgm. per cent. and recover, whileothers die with a blood level of only 200 mgm.per cent. If it could be argued that there were norisks or disadvantages attached to the use of anyform of dialysis, the logical step would be toprevent the blood urea rising even over 200 mgm.per cent., and preventing any electrolyte imbalanceoccurring. Any uraemic patient is, however, veryliable to infection, and interference must increasethis risk. Borst has suggested, without any directevidence, that both peritoneal and blood dialysisresult in increased body protein breakdown18.The oliguric patient in whom the management

has been poor, will present with gross oedema,very acidotic, with low blood chlorides, raisedvenous pressure, pulmonary oedema and a highblood urea. Such patients can no longer withsafety be treated expectantly, and dialysis shouldbe performed in the hope that the clinical con-dition can be righted sufficiently for the -moreconservative methods then to be employed.As previously stated, the presence of marked

potassium intoxication as judged by the electro-cardiogram is probably an absolute indication fordialysis, whatever the blood urea level.The relative merits of peritoneal dialysis and the

artificial kidney are not yet clear. Theoretically itis attractive to dialyse the blood direct by runningit through a semi-permeable membrane, but thisrequires full heparinization of the patient, which initself must carry a risk, however small. Possiblesudden changes in blood volume during dialysiswith the Kolff artificial kidney have been con-sidered an additional risk, particularly where mostof the patients have some degree of heart failure33.

68 February ir949

by copyright. on A

ugust 26, 2019 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.25.280.61 on 1 F

ebruary 1949. Dow

nloaded from

JOEKES: The Treotment of Uraemia

Peritoneal dialysis is a somewhat simpler pro-cedure, but the efficiency as judged by removal ofurea per hour is less, and the danger of infectingthe peritoneum is considerable, although Kop'sresults are very encouraging. Meteorism andabdominal pain are frequent complications ofuraemia, and it is not always possible to performperitoneal dialysis.Whichever method of dialysis is chosen, it is an

operation that can be undertaken only with themost careful technical detail and biochemicalcontrol. The possible dangers in either methodare so great that dialysisOmay get into disrepute ifperformed without rigid controls.

Correction of AnaemiaChronic renal disease with nitrogen retention is

usually accompanied by moderately severeanaemia. The anaemia is normocytic and ortho-chromic, and is thought to be due to a depressanteffect of some retained product on the bonemarrow. As a part of routine treatment patientswith chronic renal disease and anaemia should begiven a transfusion of packed cells, with the usualcare against overloading the heart, once every twoto three months.The acute renal failures are commonly associated

with a rapidly developing anaemia down to ahaemoglobin level of 5 gm. or even less. In manyinstances this severe anaemia is unexplained.Where haemolysis may be the immediate cause ofrenal failure, the transfusion of blood must onlybe done with every known precaution. Pre-ferably packed cells should be used to lessen thelikelihood of imposing further strain on theheart, if it is decided to transfuse during theperiod of oliguria. Whether the correction of theanaemia should be attempted while oliguria is stillpresent is undecided. With the passing of thenitrogen retention the anaemia tends to right itselfspontaneously.

Renal Decapsulation and Splanchnic BlockLarge numbers of cases of oliguric uraemia have

been treated by renal decapsulation 34 35 or someform of nerve block, most commonly a splanchnicblock2. These procedures are presumably basedon the belief that the primary reason for the con-tinued renal failure is a reduction in the renalblood flow, and that if this can be restored thekidney function will recover.The results of these operations are extremely

difficult to assess as they are usually only one ofmany forms of treatment used in the same patient.There are undoubtedly cases in whom a diuresisstarts from the time of decapsulation or splanchnicblock; frequently these have been used only inthe last resort at a time when a lower nephron

nephrosis, i.e. either a toxic or pigmert nephrosis,might be expected to be recovering some renalfunction. If any such case can be kept alive for14 days from the onset of anuria a diuresis willprobably have set in.There are, however, many qases in which either

splanchnic block or decapsulati9n have failed toproduce a diuresis, and such negatiyS results makeit certain that no very specific effect can be claimedfor these manoeuvres. It is not- considered thatthere is any place for either in the treatment ofacute renal failure.

DiureticsThere is no evidence that any diuretic has the

least value in instigating a diuresis in a patientwith anuria due to a renal cause. Their use in thefirst 24 hours of a pigment nephrosis is based onthe maintenance of a diuresis rather than a specificeffect on a damaged kidney.

ConclusionFrom the point of view of treatment, renal

failure is divided into two main groups. In theone group correct management will ensure asufficient daily volume of urine to enable thekidneys to excrete an adequate quantity of proteinbreakdown products despite a poor concentratingpower. The amount of protein in the diet iscalculated from the quantity of urea that can beexcreted in the 24 hours; i gm.' of urea may betaken as equivalent to 3 gm. of protein. To reducethe osmotic work by the kidneys or to replace saltloss by kidneys unable to conserve it, salt may haveto be administered in excess of what is taken inthe diet. This group consists of chronic renaldamage and the recovery stage of the acute renalcatastrophes. In the terminal stage of chronicrenal failure the urine concentration will not onlybe poor, but it may be impossible to increase theurine volume above a litre or so, and unless a highcalorie and minimal protein diqt is given, in-creasing nitrogen retention is inevitable.The other group forms the main subject of this

paper, and comprises acute renal failure withanuria or oliguria, whether this -has,arisen on anexisting chronic renal damage or in previouslyhealthy kidneys. Any oliguria due to an intra-renal cause lasting more than a few days, unless aterminal event in an old lesion, is probably relatedto a renal ' catastrophe,' and in the great majorityof cases is a recoverable lesion.

In any case of oliguria obstructiop to the urinarytract must be excluded-a plain X-ray of theabdomen can be very helpful-and if suchobstruction exists, it must either be relieved or anephrostomy be performed immediately.Lowered blood pressure, loss of blood and

February I949 69by copyright.

on August 26, 2019 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.25.280.61 on 1 February 1949. D

ownloaded from

70 POST GRADUATE MEDICAL JOURNAL February 1949

depletion of body water or salt are prerenal factors'which must be righted if possible. Renal failureentirely conditioned by prerenal factors will notpersist, at any rate as far as the excretion of wateris concerned, if these factors are righted. Fullrenal function, and the ability to concentrate theurine, may not be recovered for some weeks.Acute renal failure due to a primary disturbance inthe kidney may frequently be complicated by pre-renal factors, but the failure will persist despitetheir correction as long as the renal cause remains.A chronic renal disease may easily be tipped intofailure by a prerenal factor, and if the latter isrighted renal function may return to its previouslevel.The treatment of oliguric renal failure is

directed towards keeping a patient alive longenough for renal function to recover sufficientlyto support life. Most patients dying in uraemiadie of heart failure as the immediate cause. Thetwo main contributory factors to the onset ofheart failure are overloading of the circulation byincreasing above normal the water content of thebody, as a result of ' forcing fluids,' and the re-tention of potassium with abnormally high serumlevels.The main features of the management are:i. Strict maintenance of fluid balance. Only

sufficient fluid is given to replace the water lost tothe body. One litre may be taken as the amountlost in the 24 hours by way of the lungs, skin andin the faeces.

2. To reduce the accumulation of proteinbreakdown products a high calorie, non- or low-protein diet is given.

3. The minimum potassium should be given indrinks and food. In particular to be avoided areany synthetic fruit drinks, which may containlarge amounts of potassium.

4. Electrolyte imbalances should only be cor-rected if this can be done without upsetting thefluid balance. No potassium salts must be usedtherapeutically.

5. Infection should be prevented if possible andis an indication for routine penicillin administra-tion. If infection is present this must be vigorouslycombated.The indications for intervention in a case of

anuria with either the artificial kidney or peritonealdialysis are extremely difficult to define. Whetheran ideally managed case of anuria should needdialysis is very doubtful, as it should be possible tokeep any such patient alive for I4 days, by whichtime the large majority of recoverable lesions willhave started a diuresis. Gross oedema and highserum potassium levels will only be found in

cases which have not been carefully managed, andthese are probably the main clinical indications forthe use of any form of dialysis at the present time.

In conclusion it is stressed that the majority ofcases of acute renal failure with oliguria have arecoverable lesion. Some renal function re-covery should be expected within I4 days from theonset of the failure, and correct managementshould keep a patient alive for at least this lengthof time.

REFERENCES

I. FISHBERG, A. M. (1939), 'Hypertension and Nephritis,'Philadelphia.

2. O'SULLIVAN, J. V., and SPITZER, W. (1946), J. Obstet.Gynaec., 53, I58.

3. DUFF, G. L., and MORE, R. H. (I94I), Amer. J3. Med. Sci.,201, 428.

4. GIBBERD, G. F. (1936), J. Obstet. Gynaec., 43, 6o.5. CROOK, A. (1926-7), Proc. Roy. Soc. Med., 20, 1249.6. SCRIVER, W. de M., and OERTEL, H. (1930), J. Path.

Bact., 33, 1071.7. GROEN, J., and LINDEBOOM, G. A. (I940), Nederl. Tydschr.

Geneisk, 84, 688.8. BYWATERS, E. G. L. (1945), Brit. Med. Bull., 694.9. OLSON, W. H., and NECHELLS, H. (I947), Surg. Gynec.

Obstet., 84, 283.io. LONGCOPE, W. T., and LUETSCHER, J. A. (1946), J. Clin.

Invest., 25, 557.iI. KEMPNER, W. (1946), Bull. New York Acad. Med., 22, 358.12. BORST, J. G. G. (I947), Nederl. Tydschr. Geneisk., 91, 2718.13. BORST, J. G. G. (1948), Lancet, i, 824.14. HUMPHREY, J. H., and AVERY-JONES, F. (I947), Clin. Sci.,

6, 173.15. KOLFF, W. J., Personal communication.I6. BULL, G. M., JOEKES, A. M., and LOWE, K., Unpublished

work.17. LATTIMER, J. K. (I945),J7. Urol., 54, 312.I8. KUGEL, V. H. (1947), Amer. J7. Med., 3, I88.19. KIRK, E. (1946), Suppl. I83, Acta Med. Scand.20. HOFF, SMITH and WINKLER (194), .7. Clin. Invest., 20,

607.21. FINCH, C. A., SAWYER, C. G., and FLYNN, J. M. (1946),

Amer. Y. Med., i, 337.22. BYWATERS, E. G. L. (I944), Y. Amer. Med. Ass., 124, 1103.23. KOLFF, W. J., and BERK, H. T. J. (1944), Acta. Med.

Scand., 117, 121.24. KOLFF, W. J. (1947), 'New Ways of Treating Uraemia,'

Churchill, Londor.25. ALWALL, N. (I947),, Acta. Med. Scand., 128, 317.26. MURRAY, G., DELORME, E., and THOMAS (I947),

Arch. Surg., 5S, 505.27. ALWALL, N., and NORVIT (1947), Acta. Med. Scand.,

Suppl. I96, 250.28. BALASZ and ROSENAK (I934), Wien. Klin. Wchnschr., 47,

85I.29. FINE, J., FRANK, H. A., and SELIGMAN, A. M.-(1946),

Y. Amer. Med. Ass., 130, 703.30. KOP, P. S. M. (1948), Thesis, 'Peritoneal Dialyse,' Kampen.31. ODEL, H. M., and FERRIS, D. 0. (1948), Proc. Mayo Clin.,

23, 201.32. SELIGMAN, A. M., FRANK, H. A., and FINE, J. (1946),

Y. Clin. Invest., 25, 211.33. BYWATERS, E. G. L., and JOEKES, A. M. (1948), Proc.

Roy. Soc. Med., 41, 420.34. ABESHOUSE, B. S. (I94S),J. Urol., 53, 27.35. PETERS, J. T. (I945), Ann. Intern. Med., 23, 221.36. MUIRHEAD, E. E., et al. (1948), Blood, Special issue No. 2,

IOI.1

37. COLLER, F. A., CAMPBELL, K. N., and IOB, V. (1948),Ann. Surg., 128, 379.

38. LUCKE, B. (1946), Mil. Surg., 9g, 371.

by copyright. on A

ugust 26, 2019 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.25.280.61 on 1 F

ebruary 1949. Dow

nloaded from