Battle Casualties in Korea: Studies of the Surgical Research Team, Volume IV
The Management of Acute Renal Insufficiency*
Major W. H. Meroney, MC, USA
First Lieutenant R. F. Herndon, MC, USAR
General Considerations
Renal insufficiency has been described in association with numerous pathologic conditions which do not involve the kidneys primarily. Insults to non-renal systems of the body may provoke compensatory responses which protect the organism as a whole but damage the kidneys secondarily. The prime example of this phenomenon is shock, in which blood is shunted away from organs whose functions are not immediately necessary for survival. By the time critical functions are restored, organs with low priority for blood, such as the kidneys, may have undergone ischemic changes of severe degree. Insults to non-renal systems may release to the plasma pigment or other intracellular materials which are concentrated in the kidneys and produce obstruction or cellular degeneration. There are numerous other ways in which kidneys may be damaged, but these two are considered to be of greatest importance in war or traumatic disaster.
Once renal damage has occurred, it may assume greater importance than the original condition as a factor in prognosis unless specialized treatment is instituted. Comprehensive studies by the Board for the Study of the Severely Wounded in World War II1 revealed that the degree of oliguria was correlated with the mortality rate. In a group of 186 severely wounded men with an over-all mortality rate of 35 per cent, those with a urinary output of 100 to 600 cc. per 24 hours had a mortality rate of 47 per cent, and those with a urinary output of less than 100 cc. per 24 hours had a mortality rate of 91 per cent. The researches conducted by these and many other investigators provided a greater understanding of renal failure which allowed development of newer methods of treatment which lower the mortality rate.
In the Korean War renal failure as a complication in the severely wounded was comparable with that seen following shock in World War II. Pigment nephropathies were not recognized as significant
*Previously published in modified form in The Journal of the American Medical Association 155: 877, 1954.
52
causes of renal failure in this study, a fact which is a tribute to the excellence of the blood replacement program and which probably also stems from the infrequent occurrence of crushing injuries. Members of the Surgical Research Team observed that oliguria occurred most frequently in the individuals who were most severely wounded and whose resuscitation was most difficult. Of the first 10 such patients observed, 9 died of potassium intoxication.
In April 1952, an artificial kidney was put into operation by the Surgical Research team at the Renal Insufficiency Center at Wonju, about 75 miles behind the main line of resistance. This unit was attached to the 11th Evacuation Hospital and was operated as a joint effort of the Army Medical Service Graduate School in Washington, the 406th Medical General Laboratory in the Far East Command, and the 8th Army Medical Service in Korea. During the 16 months of its operation this unit received approximately 160 patients who had, or initially were suspected of having, renal insufficiency. The greatest number of patients cared for at any one time was 11, and the average number was about 4. However, the intensive care and study given these patients required that three to five internists, one surgeon, four to six nurses, six to nine corpsmen, and two to four laboratory technicians work full time on this effort. In addition, all of the personnel of the hospital, particularly the anesthesiologist, pathologist, laboratory technicians, supply officer and utilities officer, contributed significantly to this effort. The operation of the unit centered around the artificial kidney and the laboratory, but all of the personnel and facilities utilized in the care of any seriously ill patients were even more concerned for these patients.
The immediate net result of this investment was a reduction in over-all mortality rate from about 90 per cent to about 55 per cent. When the biochemical abnormalities resulting from renal failure were controlled, the degree of renal failure no longer paralleled the mortality rate. A less tangible but probably more important result of the control of renal failure was the dissociation of the effects of uremia from the effects of other clinical disorders. Upon the completion of 6 hours of treatment with the artificial kidney, a patient is momentarily relieved of uremic symptomatology, and the remaining symptoms can be traced to their proper source.
If the source of such symptoms can be located and corrected, the symptoms resulting from uremia alone can then be assessed as they reappear in ensuing days of oliguria. The importance of such maneuvers lies in the difficulty of assessing symptoms and assigning priority to therapeutic procedures in a man with multiple abnormalities in addition to oliguria. If one is not to be confounded by the complexities of such a patient, the clinical patterns to be expected from each
53
disorder alone must be established. It was not uncommon in Korea for a moribund patient to be rushed as an emergency to the Renal Center when, in fact, none of his symptoms were the result of uremia. The surgical care which may have represented the patient`s only chance for survival was not given by those best qualified and in the best situation to give it. Other patients whose threats to life were primarily of renal origin could have been treated more satisfactorily if the information gained at the Renal Center had then been available in the forward area. It is the purpose of this report to outline the concepts and methods employed at the Renal Center which appear to have the greatest practical application, with specific reference to 48 renal and 11 non-renal patients observed by the authors. The system presented is subject to revision, as analyses now pending are completed, as further information becomes available from other workers, and as any divergent opinions are resolved. A more detailed account of individual cases and raw data is recorded elsewhere.2
Recognition and Transfer
The diagnosis of renal insufficiency should be suspected if the urine output falls below 500 cc. per 24 hours, about 20 cc. per hour. The patient and his records should then be re-examined with the following possibilities in mind:
Reflex Oliguria. This term is used to designate the transient oliguria which occurs following surgery or other trauma. The condition may not be reflex in origin, but it appears to be a normal response to this type of stress and lasts only a few hours.
Hypotension. The blood pressure in the renal artery is an essential component of renal filtration pressure. When hypotension is observed in the arm, renal filtration can be expected to be diminished. The resulting oliguria causes an increase in plasma NPN which can be misleading, because a patient with persistent hypotension may have central nervous system signs which resemble those of uremia at a later stage. It is important to distinguish the conditions because the treatment is quite different. The most practical approach to the differential diagnosis is to defer a diagnosis of renal failure until normal blood pressure is restored. The condition which produced the hypotension should receive first attention, because it likely will kill or be cured before renal failure, if it be present, requires any treatment.
Dehydration. A severe degree of dehydration is required to produce oliguria of 20 cc. per hour, and either history or physical examination should confirm this unusual diagnosis. If any question remains, the
54
high specific gravity of urine excreted by normal kidneys during dehydration should settle the point, because the urine of acute post-traumatic renal failure characteristically has a low specific gravity. The practice of administering a fluid load as a test for dehydration is dangerous and should not be necessary.
Obstruction. In the post-traumatic state urinary tract obstruction is suspected when there is flank pain with genital radiation, when the scanty urine contains crystals of drugs or heme casts, or when there is total anuria. Even the most severe cases of renal insufficiency usually are painless and the patients excrete 30 to 40 cc. of urine per day. If there is no urine whatever, first attention should be given the urethral catheter, for obstruction by mucus is not uncommon.. Daily irrigation of the catheter with a bland fluid which is measured when instilled and recovered should prevent this complication. If obstruction is still suspected, cystoscopy and catheterization of the ureters are indicated.
Acute Tubular Nephrosis. If a patient with the usual state of hydration excretes less than 20 cc. of urine per hour after 5 to 10 hours of normal blood pressure, and there is no evidence of obstruction, the diagnosis of acute renal failure is justified. Pathologically, this lesion was first described as epithelial necrosis in the lower segment of the tubule, but subsequent studies indicate that the necrosis occurs throughout the length of the tubule.3 Functionally, there is serious impairment of the capacity of the kidneys to excrete or conserve selectively the substance in excess or in scarce supply. During the acute stage the problem is excretory. All of the plasma substances which normally appear in the urine accumulate in the plasma, where some of them are toxic. The renal lesion is potentially reversible, and if the patient can be maintained for the few days or few weeks required for regeneration of the tubular epithelium, spontaneous diuresis will occur. During the diuretic phase, the problem is lack of conservation. Scarce and essential substances are washed out in the copious urine flow, and serious chemical deficits can occur.
Once the diagnosis of renal failure has been made, it is essential that the symptomatology be re-appraised to establish which components are uremic in origin. When one is faced with a deteriorating clinical status in a patient with serious wounds and oliguria, there is a strong tendency to blame uremia for all of the symptoms and to neglect other disorders. With two exceptions, the symptoms of uremia do not appear, in our experience, until the fifth to eighth day of anuria when the NPN is 250 to 300 mg. per 100 cc. The first exception is overhydration, which is not really a symptom of the pa-
55
tient`s disease but of the doctor`s error. The second exception is severe potassium intoxication, which only occurs as a sequel of incomplete care of other features of the patient`s condition. If neither overhydration nor potassium intoxication is present, anuria must persist for many days before it provokes symptoms.
It follows that other pathologic processes should be sought in a patient who is symptomatic during the first few days of anuria. Symptoms which resemble those of uremia are observed in patients with wound infection, generalized sepsis, or shock, but the offending agents cannot be dialyzed from the patient`s plasma. There is no gain but potentially total loss in transferring such a patient to a renal center. If the basic condition cannot be corrected, there is even less chance of success after the delay and trauma of evacuation. If such a patient is to survive, resuscitation and débridement must be completed before transfer, and he must be able to maintain a blood pressure of more than 100 mm. Hg throughout the journey without stimulants or infusions. Once these conditions are fulfilled, the oliguric patient can be transferred for definitive care of uremia with good prospects for success.
The methods for management of acute renal failure and allied conditions which were developed from the Korean experience have general application and are reviewed below. The problems in management are considered in order requiring first action by the physician.
Fluid Balance
First and continuing consideration should be given to the prevention of overhydration. The evils of water intoxication are well known, yet this preventable complication is common. The total fluid intake per 24 hours should be in the range of 500 to 600 cc. plus the measured output, increasing to about 700 to 800 cc. plus the measured output if the weather is hot or the patient is feverish. The measured output usually is the total of urine and gastric suction, although diarrhea, not present in our cases, might increase the output significantly. This average intake allowed a daily weight loss of 1/2 to 1 lb. (0.23 to 0.5 kg.) without producing clinical evidence of edema or dehydration.
The best route for administration of the fluids is oral, if the patient is able to tolerate them. In our patients, however, fluids administered orally, even by duodenal tube, provoked vomiting. This compounded the problems of fluid and electrolyte balance and also added the risk of aspiration pneumonia; thus the intravenous route was used almost exclusively. Restriction of orally administered fluids can be quite difficult. Patients so restricted often suffer a cruel thirst, yet their thirst is not an accurate gauge of their needs. If allowed to drink freely,
56
they will literally drown themselves. When denied fluids, they develop great craftiness in prevailing on compassionate neighbors and attendants for small sips of any fluid or for pieces of ice, a bountiful and easily overlooked source of water. When unobserved, they will quaff heartily from flower vases, emesis basins, or urinals with great stealth and cunning. It is the physician`s responsibility to prescribe the proper amount of water and to insure that no violation of his order occurs. Careful oral hygiene and unlimited quantities of chewing gum, together with an explanation of the reasons that compel the apparent cruelty, are usually adequate to control thirst. The amount of water prescribed must contain all the solids to be administered during the 24 hours, these will be discussed later.
Potassium Intoxication
The only chemical abnormality that is likely to cause death in the first week of uremia is potassium intoxication. Normally, potassium exists in high concentration inside cells but does not exceed a level of 5.5 mEq. per liter in the plasma. Normal daily catabolism of cells provides the plasma with a small quantity of potassium that is readily excreted. This amount of potassium can be handled by an anuric patient for many days, even several weeks, without accumulation of significant quantities in the plasma.4 However, the basic condition that originally produced the renal insufficiency is often characterized by excessive loss of potassium from cells. This was especially evident in these patients, many of whom had suffered extensive tissue damage from trauma or infection and in whom the plasma potassium rose to high levels as early as the second day of oliguria. Devitalized tissue, whether permanently destroyed or temporarily embarrassed by trauma, infection, chemical or physical agents, or hypoxia, gives up potassium to the plasma.5 In the plasma excess potassium is exceedingly toxic, and the first and most important evidence of toxicity is cardiac.
The toxic effect of potassium on the heart is recorded on the electrocardiogram long before any other signs or symptoms appear. The degree of electrocardiographic abnormality produced by a given excess of plasma potassium will vary widely, however, depending on the activity of other factors. Under experimental conditions, such as the infusion of potassium into the plasma of normal dogs, it has been shown that the electrocardiographic effects of progressive increments of plasma potassium are mostly consistent6; in clinical situations, however, the electrocardiogram has not previously been shown to be a
57
reliable gauge of the plasma potassium level. The function of the heart with respect to potassium is the result not only of the absolute level of plasma potassium but also of factors that influence the effects of potassium. The electrocardiogram is, in fact, an accurate gauge of the plasma level of potassium in man when there is no abnormality except hyperpotassemia.7 The discrepancies stem from the rarity with which pure hyperpotassemia occurs. In the oliguric patient, substances that normally would be excreted in the urine are retained within the body, and some of these substances affect the behavior of potassium.
Calcium. The retention of inorganic phosphate, though not itself harmful, is associated with a fall in the level of plasma calcium during oliguria. Figure 1 demonstrates the consistency with which a given excess of plasma phosphate is associated with a predictable deficit of plasma calcium. A deficit of plasma calcium is of cardinal importance during oliguria because calcium is a specific antagonist of potassium and hyperpotassemia and hypocalcemia occur at the same time.
As the plasma potassium level rises, the degree of toxicity recorded by the electrocardiogram is consistent with it only if the plasma cal-
FIGURE1. Graph of random pairs of phosphate and calcium determinations in oliguric patients.
58
cium level is maintained; otherwise, the electrocardiographic abnormality and the threat to the patient`s life are greatly increased. Since in oliguria the plasma calcium level regularly is depressed by a rise in the plasma phosphate level at the same time that the plasma potassium level rises, the electrocardiogram of the untreated patient bears little relationship to the plasma level of potassium. Replacement of the calcium deficit produces a striking improvement in the electrocardiogram, which then is reverted to that degree of abnormality characteristic of the plasma potassium level(Fig. 2).
The improvement following a single intravenous injection of calcium, however, is very transient. The plasma level of calcium falls quickly from the high level immediately after injection to the pre-injection level, which is governed by the phosphate concentration. Although a rise in the plasma phosphate level reduces the plasma calcium level, the converse is not true; phosphate, unlike calcium, cannot be driven into the body repositories by such a maneuver. Intravenous
59
dextrose infusion will cause a slight reduction in the plasma phosphate level, but not enough to allow a rise in the plasma calcium level sufficient to antagonize potassium. Short of hemodialysis, the only effective way to maintain a normal level of calcium in plasma that is high in phosphate is by continuous intravenous infusion of calcium.
The effect of calcium on potassium toxicity is purely one of antagonism. The measurable level of potassium, as well as phosphate, is unaffected by raising the plasma calcium level. Once the calcium deficit is replaced, the electrocardiogram reflects the plasma level of potassium rather accurately, although still other factors less apparent in these patients have been shown to be influential.8 Digitalis, which was rarely indicated in our patients, antagonizes potassium9 in a manner similar to calcium. Digitalis and calcium appear to be additive in this respect, and great care should be used if both agents are administered to the same patient. There is particular danger if the excess potassium is suddenly removed by hemodialysis,10 allowing digitalis, enhanced by a high calcium concentration, to exert its toxicity without the opposing action of potassium.
Small excesses of potassium produce no electrocardiographic abnormality except elevation and peaking of the T waves, best seen in the precordial leads (Fig. 3). As the plasma potassium level increases further, the T wave abnormalities progress in the precordial leads and become obvious in the limb leads, but they have no quantitative significance in severe hyperpotassemia. Progressive changes in the QRS complex, best seen in the limb leads, are much more ominous. In the limb leads, the angle between the S wave and the ST segment widens and encroaches on the horizontal component of the ST segment until it is obliterated. P waves disappear; the T waves diminish in height and become rounded at the top; and finally, the R-S angle increases, and the smooth biphasic curves resemble a sine wave.
When mild intoxication, indicated by T wave abnormality only, suddenly progresses to the severe intoxication indicated by QRS widening, it usually is the result of a fall in the plasma calcium level. Infusion of calcium should cause instant reversion of the tracing to its former degree of abnormality. If calcium does not produce an immediate effect, this indicates that the plasma potassium level has increased. Sudden rises in the plasma potassium level occur during oliguria when tissue cells are subjected to stress. Hemolysis infection or trauma may allow the release of all the potassium contained in the affected cells. Smaller amounts of potassium may temporarily leave hypoxic cells, only to be recovered by the cells when normalcy is restored. Hypoxia from hypotension, convulsions, certain types of
60
anesthesia, pulmonary edema and simple breath-holding has been observed to cause such a reversible shift of potassium.11 In our experience, a rapid increase in the electrocardiographic evidence of potassium intoxication that was not responsive to calcium invariably was associated with a rapid increase of the plasma potassium level associated with one of the above conditions.
Progressive potassium intoxication caused no symptoms in these patients until it had reached the stage at which the electrocardiogram showed severe deterioration similar to that of the fourth tracing in Figure 3. Patients who had symptoms with a less abnormal electrocardiogram were found to have other causes for their symptoms.
61
Many of the disorders associated with hyperpotassemia, such as those noted above, produce severe symptoms. Correction of the basic condition relieved the symptoms in our patients, although in many instances the plasma potassium was still greater than 8 mEq. per liter and the electrocardiogram was appropriately abnormal. At the time of the tracings shown for case 1 in Figure 2, the patient had neither signs nor symptoms. If one attempts to follow the course of the intoxication in patients with this condition by changes in tendon reflexes, respiratory symptoms, or any other known means except electrocardiography or blood chemistry, the patient will probably be dead before corrective action can be taken.
Sodium. Another ion that should be used in the practical management of potassium intoxication is sodium. Sodium and potassium are inversely related in the plasma of the oliguric patient; raising the plasma sodium concentration causes a fall in plasma potassium concentration and modifies the electrocardiographic effects of potassium intoxication (Fig. 4). Potassium concentration is depressed with a rise in sodium concentration independently of the other ions involved, indicating that the agent responsible for the change is sodium and not the anion. An alkaline salt of sodium is preferred, however, because the retention of organic acids regularly produces acidosis in these patients. The changes in the concentration of potassium and the other substances are not attributable to dilution. Although tissue analyses were not performed, it is presumed that raising the plasma sodium concentration forces potassium back into cells.
The improvement in the electrocardiogram that followed a rise in the plasma sodium level and a fall in the plasma potassium level was sometimes greater than would be expected from the lower level of potassium. This suggests that sodium, in addition to depressing the plasma concentration of potassium, may also have some antagonistic effect similar to that of calcium. Figure 5 shows the beneficial effect of raising the plasma calcium level followed by still further improvement from administration of sodium chloride. The last tracing in this series is almost normal, although the potassium concentration still is 7.2 mEq. per liter. Obviously, all of the factors influencing the response of the electrocardiogram to a given potassium concentration have not been considered. Infusions of hypertonic solutions of sodium salts, do, however, provide an effective and practical means for modifying potassium intoxication. Again it is emphasized that all of the patients in this series were young men without known previous cardiovascular disease, a fact that undoubtedly influenced their responses and that should be considered if these methods are applied to dissimilar patients.
62
Dextrose. The plasma potassium level may also be reduced by the infusion of a hypertonic dextrose solution.12 Its metabolism, which can be hastened with exogenous insulin, removes potassium and phosphate from the plasma. While calcium antagonizes the effects of potassium without changing the quantity present in the plasma and sodium forces potassium into cells by some physicochemical means, glucose carries potassium into cells by a more active process. As glycogen is formed, potassium, as well as phosphate, is incorporated into the carbohydrate complex; this is an effective, though slow, method of controlling potassium intoxication. Results similar to those of a dextrose-insulin combination could be expected from the use of fructose, which does not require insulin for its early metabolism, but this compound was not used.
63
If dextrose is to be used effectively by the intravenous route, it should be given continuously. Intermittent intravenous injection of dextrose causes a sharp spike in blood sugar level followed by hypoglycemia. The hypoglycemic period has the double disadvantage of failure to remove potassium during that period, and provocation of glycolysis with further release of potassium to the plasma. Also, if nutrition is limited to intermittent intravenous feedings of dextrose, the period between infusions is one of relative starvation, which is characterized by cell destruction and release of potassium.
On the basis of the above observations, the following standard solution was devised and was found to be effective in controlling potassium intoxication for many days.
|
Calcium gluconate 10% |
100 cc. |
|
Sodium bicarbonate 7.5% |
50 cc. |
|
Dextrose 25% in H2O (Containing 25 units of regular insulin) |
400 cc. |
|
Isotonic sodium chloride solution or 1/6 M sodium lactate |
Volume of output |
64
This solution should be given intravenously, preferably through a catheter, in a large vein at a constant rate of about 25 cc. an hour. Water-soluble vitamins should be added to this basic solution. Our patients received daily 2 gm. of ascorbic acid, 50 mg. of thiamine chloride, and 20 mg. of vitamin K, but this vitamin prescription was chosen in the interests of a concurrent study of wound healing and is not necessarily optimal.
If the laboratory can provide frequent determinations of plasma sodium levels, the sodium content of the fluid should be varied to maintain a plasma level of about 150 mEq. per liter. If not, the amount suggested can be given empirically without fear. The only complication arising from administration of more than the desired concentration is aggravation of thirst. This is uncomfortable for the patient, and if he is allowed excess water pulmonary edema may be provoked. None of the dire effects attributed to sodium administration13 were observed in these patients, and it would appear that it is not the sodium alone but the excess water that may accompany it that is dangerous. In older patients or in those who have cardiovascular disease, additional hazards may be associated with a high plasma sodium level. It is unlikely, however, that the dosage recommended would have adverse effects. A likelier source of harm is overhydration; the necessity for water restriction must be kept constantly in mind.
If frequent determinations of the plasma potassium level show that it is not elevated, it would not be necessary to use calcium in the infusion, as its purpose is to antagonize an elevated plasma potassium level. The only other reason for giving calcium would be to prevent tetany if the plasma calcium level were depressed. However, in patients with this condition the plasma calcium level is not depressed unless the potassium level is elevated because hypocalcemia is caused by an elevation of the plasma phosphate level, and phosphate rises at the same time as potassium; therefore, calcium should be given only if the plasma potassium level is elevated. If frequent determinations of potassium cannot be done, the electrocardiographic changes noted are sufficiently specific to guide logical therapy.
The constituents of the basic solution should be tailored for the individual patient when possible. If an item is omitted, its volume should be replaced by an isotonic solution or distilled water.
Ion-Exchange Resins. Ion-exchange resins are effective agents for the withdrawal of potassium from the plasma into the intestine under certain circumstances.14 In this series of patients, however, those whose potassium levels were highest were unable to take anything by mouth. Rectal instillation of resins was attempted repeatedly, but
65
such intractable concretions were formed that the amount of water necessary to remove them caused water absorption and overhydration. An attempt was made to contain resins in a silk tube that could be inserted and extracted from the rectum mechanically, but the procedure was very painful for the patient; effective exchange occurred only at the surface of the resin bolus, even when a silk tube of only 5 mm. diameter was used; and only 7 mEq. of potassium was extracted by 30 gm. of resin left in the rectum for 24hours. It was intended to utilize a colostomy for this method but a suitable patient was not found. The impression was gained that resins would not be more effective than the intravenous therapy described.
Dialysis. Artificial-kidney dialysis is the most effective treatment known for removing potassium, but dialysis for hyperpotassemia alone is, in our experience, rarely indicated. The use of this technic is discussed below under the heading, Clinical Uremia. We have had no experience with other methods of dialysis.
Less Urgent Problems in Management
Nutrition. Principally fat and carbohydrate are recommended during the oliguric phase of acute uremia, because of the potential toxicity of protein.15 Maximal caloric intake is desirable to prevent the patient from burning his own tissues to supply his caloric needs, but the exact quantity of fat or carbohydrate necessary is unknown. Various high-calorie mixtures, such as olive oil and glucose, frozen butter balls, and commercial fat emulsion preparations, were given by mouth and by gastric and duodenal tube. Unfortunately, the patients who needed it most were nauseated by any of the forms of oral feeding. Intravenous administration of fat was not done in these patients because a satisfactory preparation was not available.16 Caloric intake, in the main, was limited to intravenous administration of dextrose, which was given as outlined above. This is far from ideal but will sustain the patient. Current studies with newer antiemetic drugs offer hope for successful oral or tube feeding.
Clinical Uremia. After a few days of oliguria (in this series5 to 8 days), the syndrome of clinical uremia develops. The patient gradually becomes lethargic and shows evidence of mental torpor, yet his extremities are tremulous and hyperreflexic. The nausea that may have been present for days now become active vomiting and retching. Intractable hiccups are usual. These clinical manifestations first appeared in our patients when the plasma nonprotein nitrogen level was about 250 mg. per 100 cc. The consistency with which this figure was
66
associated with the clinical findings was remarkable, and it came to be useful in allowing one to predict from a nonprotein nitrogen value of about 200 mg. per 100 cc. that that patient would show the symptoms of uremia on the following day.
At this juncture artificial-kidney dialysis is strongly indicated. Six hours of dialysis will produce dramatic relief of symptoms as well as restoration of electrolyte balance. It does nothing for renal function, of course, but it clears the plasma of the substances that normally would have been excreted in the urine. The patient is then ready to start anew on his course of uremia; if diuresis does not occur within several days, dialysis should be repeated each time the symptoms recur until renal function is recovered. Dialysis was performed one or more times in 27 of our 46 patients; 14 underwent dialysis once, 6 twice, and 7 three times.
It is important to recognize the syndrome of clinical uremia and to know when to expect it, because similar signs and symptoms can be produced by hypotension or sepsis. In our patients with renal insufficiency the nonprotein nitrogen rose gradually to the critical level of 250 mg. per 100 cc. over a period of about a week. This level or symptoms of uremia did not occur before the fifth day of oliguria, and the average was later. When symptoms appeared dialysis usually was performed, eliminating the symptoms. If dialysis was not performed the symptoms progressed gradually, and or 3 days elapsed before such threatening clinical manifestations as coma, convulsions, and pericarditis appeared. In patients with prerenal azotemia secondary to shock or sepsis, symptoms appeared at any time, usually within 1 or 2 days, and were totally unrelated to the nonprotein nitrogen level. The onset often was abrupt and the progression rapid. Dialysis was quite successful in restoring chemical balance but did not improve the clinical manifestations. The rate at which plasma potassium and phosphate levels rose after dialysis also was quite different from that seen in patients with renal insufficiency. Within a day, or even a few hours, of a normal post-dialysis level, great increments of potassium and phosphate were again found in the plasma. The condition that had produced the symptoms was still present, and it was associated with devitalization of tissue, which gave up its intracellular materials to the plasma. Dialysis, therefore, was virtually useless, and time was lost in this procedure that should have been spent in treating the basic condition of the patient.
The measures outlined for the treatment of potassium intoxication were effective, in patients without necrotic tissue, for the period prior to the appearance of clinical uremia. In some instances dialysis was avoided because diuresis occurred before clinical uremia appeared, and, in others, the use of these measures made fewer dialyses neces-
67
sary. In the presence of clinical uremia no attempt was made to control potassium further by medical means, because the dialysis that was otherwise indicated was superbly effective in removing potassium.
When an artificial kidney is not available, medical management of potassium intoxication may save the patient`s life for the moment but should be used with a view to providing time for transport to a renal center. Where an artificial kidney is available, this method will allow dialyses to be performed on a reasonable schedule for clinical uremia only and should obviate emergency dialyses for potassium intoxication.
Anemia and Hemorrhage. Of the various causes of anemia in post-traumatic renal insufficiency, the only one studied in this series of patients was severe purpura similar to that seen in chronic uremia. Bleeding into the skin, nasopharynx, and intestine appeared after 12 to 15 days and was related only to the duration, not the severity, of the uremia. Unlike the other manifestations of uremia, the hemorrhagic tendency was entirely unaffected by artificial kidney dialysis or by diuresis; to the contrary, the severest bleeding occurred several days after the onset of diuresis. The only positive temporal relationship was the cessation of bleeding with the resumption of normal diet; however, the absence of hemorrhagic diathesis during pure starvation discourages speculation on this point.
A search for the cause of the bleeding revealed no intrinsic clotting defect, only capillary fragility. The condition was unrelated to abnormality of plasma electrolytes and was unaffected by vitamins C or K or fresh blood transfusions. The replacement of the blood lost was considered desirable, yet transfusions were feared because a minor reaction might have exaggerated significance in patients so seriously ill. Nevertheless, small transfusions, 200 to 300 cc. daily, were given regularly without apparent harmful effect.
Bleeding was originally considered a contraindication to artificial kidney dialysis, because of the necessity for heparinization before and during the procedure. On several occasions, however, dialysis was performed as a lifesaving procedure on patients who were bleeding. During the procedure large quantities of blood were available for immediate use if bleeding should be aggravated, and after the neutralization of heparin with protamine at the end of the procedure, packed erythrocytes were given. No alarming bleeding occurred, and so it became the practice to perform dialysis whenever indicated, regardless of hemorrhagic tendencies. The risk of dialysis in the presence of hemorrhage was considered to be less than the risk of the complications of uremia if dialysis were withheld, and subsequent experience supported this view.
68
Infection and Antibiotics. The specific infections encountered in battle wounds in these patients will not be reviewed here. In general, their successful management depended mainly on thorough and frequent dressings. An amount of devitalized tissue and infection that may be acceptable in a patient with normal renal function may be lethal in the oliguric patient. Small amounts of tissue, particularly muscle, contain sufficient potassium to kill if it is released into plasma that is not being cleared by the kidney. Removal of such tissue must be prompt and thorough, and the approach must be much more radical than is customary with good surgeons. If the point at which devitalized tissue merges with normal tissue is not apparent to the surgeon, he should débride farther. Excessive débridement may cost the patient precious tissue, but inadequate débridement may cost his life.
Remnants of devitalized tissue usually can be suspected if the plasma potassium and phosphate levels are inordinately high with respect to nonprotein nitrogen. The measures outlined above will control the potassium level; the phosphate level, however, is little affected by such measures, and it is a useful index of the presence of necrosis. A plasma phosphate level reaching 10 mg. per 100 cc. before the non-protein nitrogen level reached 150 mg. per 100 cc. indicated severe tissue destruction in these patients, and on several occasions deep necrosis underlying a wound with a clean surface was first suspected from this relationship.
Antibiotics cannot be relied upon to control infection in devitalized tissue. Also, the proper methods for administration of antibiotics have not been clarified. The principal route for excretion of antibiotics is through the urine, and during oliguria tremendous concentrations accumulate in the blood if usual doses are given. The importance of this fact was not investigated in these patients, but recent reports of serious intoxications from certain antibiotics require that great caution be observed.17
The systemic infection oftenest seen in patients with this condition is pneumonia. This complication is particularly likely to occur if the syndrome of clinical uremia is present and is not interrupted by hemodialysis. The hazard of pneumonia was combated in these patients by deep breathing exercises, voluntary coughing, frequent turning, and good general nursing care. Patients whose cooperation was doubtful because of associated wounds or illnesses were placed on Stryker frames so that turning and drainage could be assured.
The diagnosis of oliguria carries with it a poor prognosis, although the renal lesion itself is usually reversible. The onset of infection may well be enough additional insult to cause a death that would not other-
69
wise have occurred. The seriousness and the potential reversibility of the illness demand that no effort be spared in preventing or treating the complication of infection.
Salt Wasting During Diuresis. The onset of diuresis was arbitrarily defined in these patients as the day on which urine volume exceeded 1 liter. Once this volume was reached, the output increased rapidly, often 100 percent or more on successive days, until a peak volume of 3 to 6 liters was attained. As the urine volume increased, the urinary concentration of nonprotein nitrogen, sodium, potassium, and phosphate increased similarly, and the total output of salts reached high values within 1 or 2 days. At this stage the kidneys excrete salts wantonly, without regard for the body`s needs. Fortunately, appetite and food tolerance have returned at this stage, and salt depletion is partially offset by the food intake. Daily supplements of sodium and potassium, however, must be added to prevent serious depletion. Sodium chloride, 4 to 6 gm., and potassium chloride, 1 to 2 gm., added to a regular diet daily for about a week were found to be adequate. Fluids were administered freely without noticeable aberrations of fluid balance.
Summary
In a study of acute renal insufficiency in war casualties, it was found that total fluid intake per 24 hours should vary between 500 and 800 cc. plus the output. Potassium intoxication, the only chemical abnormality that threatens the patient`s life during the first week of oliguria, produces no clinical signs or symptoms in patients with this condition until death is imminent. The electrocardiogram shows early and continuing evidence of the intoxication, and if other plasma chemicals are at normal levels the electrocardiogram is a rather accurate gauge of the plasma potassium concentration. Plasma calcium is depressed, however, when potassium is elevated, because of the associated retention of inorganic phosphate. As calcium is a specific antagonist of potassium, its administration causes striking modification of the electrocardiographic effects of potassium intoxication. Administration of sodium or of dextrose with insulin lowers the plasma potassium level. Continuous infusion of a standard mixture of calcium, sodium, and dextrose will protect the patient from potassium intoxication for many days, allowing transportation of the patient to a center where hemodialysis can be performed if necessary.
References
1. The Board for the Study of the Severely Wounded: The Physiologic Effects of Wounds. U. S. Govt. Printing Office, Washington, D. C., 1952.
70
2. Meroney, W. H.: Activities Report of the Renal Insufficiency Center, Feb. to Aug., 1953. Army Medical Service Graduate School, Walter Reed Army Medical Center, Washington,
D. C.
3. Moon, V. H.: Acute Tubular Nephrosis, a Complication of Shock. Ann. Int. Med. 39: 51, 1953.
Oliver, Jean: Correlations of Structure and Function and Mechanisms of Recovery in Acute Tubular Necrosis. Am. J. Med. 15:535, 1953.
Lucké, B.: Lower Nephron Nephrosis. Mil. Surg.99: 371, 1946 (renal lesions of the crush syndrome, of burns, transfusions, and other conditions affecting lower segment of nephrons).
4. Welt, L. G., and Peters, J. P.: Acute Renal Failure: Lower Nephron Nephrosis. Yale J. Biol. & Med. 24: 220, 1951.
5. (a) Fenn, W. O.: The Role of Potassium in Physiological Processes. Physiol. Rev. 20: 377, 1940. (b) Meroney, W. H.: Unpublished data.
6. Winkler, A. W., Hoff, H. E., and Smith, P. K.: Electrocardiographic Changes and Concentration of Potassium in Serum Following Intravenous Injection of Potassium Chloride. Am. J. Physiol. 124: 478, 1938.
7. Herndon, R. F., Meroney, W. H., and Pearson, C. M.: Unpublished data.
8. (a) Darrow, D. C.: Medical Progress; Body-Fluid Physiology: The Role of Potassium in Clinical Disturbances of Body Water and Electrolytes (part 1), New England J. Med. 242: 978, 1950; (part 2), ibid. 242:1014, 1950.
(b) Abrams, W. B., Lewis, D. W., and Bellet, S.: The Effects of Acidosis and Alkalosis on the Plasma Potassium Concentration and the Electrocardiogram of Normal and Potassium Depleted Dogs. Am. J.M. Sc. 222: 506, 1951.
(c) Keating, R. E., and others: The Movement of Potassium During Experimental Acidosis and Alkalosis in the Nephrectomized Dog. Surg., Gynec. & Obst.96: 323, 1953. Fenn (5a).
9. Hopper, J. Jr., O`Connell, B. P., and Fluss, H. R.: Serum Potassium Patterns in Anuria and Oliguria. Ann. Int. Med. 38:935, 1953.
10. Lown, B., and others: Effects of Alterations of Body Potassium on Digitalis Toxicity, read before the American Society for Clinical Investigation, Atlantic City, N. J., May 5, 1952, abstracted, J. Clin. Invest. 31: 648, 1952. Meroney (5b).
11. Stewart, J. D., Potter, W. H., Hubbard, R. S., and Andersen, M. N.: Potassium Movement in Acute Liver Damage. Ann. Surg. 138:593, 1953. Fenn (5a). Meroney (5b).
12. Fenn, W. O.: Deposition of Potassium and Phosphate with Glycogen in Rat Livers. J. Biol. Chem. 128: 297, 1939. Fenn (5a). Darrow (8a).
13. Swann, R. C., and Merrill, J. P.: The Clinical Course of Acute Renal Failure. Medicine 32: 215, 1953.
14. Elkinton, J. R., and others: Treatment of Potassium Retention in Anuria with Cation Exchange Resin. Am. J. M. Sc. 220:547, 1950. Danowski, T. S., and others: The Use of Cation Exchange Resins in Clinical Situations. Ann. Int. Med. 35: 529, 1951.
15. Kolff, W. J.: Forced High-Calorie, Low-Protein Diet and the Treatment of Uremia. Am. J. Med. 12: 667, 1952. Welt and Peters (4).
16. Creditor, M. C.: Some Observations of Effects of Intravenous Fat Emulsions on Erythrocyte Fragility. Proc. Soc. Exper. Biol. & Med. 82:83, 1953.
17. Bateman, J. C., and others: Fatal Complications of Intensive Antibiotic Therapy in Patients with Neoplastic Disease.
A. M. A. Arch. Int. Med. 90: 763 (Dec.), 1952.
