Fluid therapy for ketoacidosis and renal failure (Proceedings)

When presented with animals with evidence of renal dysfunction (as evidenced by elevated serum creatinine concentration), there are a few considerations. One is whether the abnormal lab test represents functional impairment only; meaning that at this moment the animal has an abnormally high serum creatinine concentration in spite of an adequate mass of functional renal tissue. This is typically because the kidneys are not receiving enough blood flow to maintain normal plasma concentrations of compounds excreted by the kidney (a process that produces the phenomenon of 'pre-renal azotemia'), or has lower urinary obstruction, or has uroabdomen. Another consideration is whether or not the kidneys are sustaining acute, active damage by some disease process such as toxicity, infection, thrombosis, outflow obstruction, or others that will require specific therapy intended to remove the cause. Another is whether or not there is impaired renal blood flow subsequent to circulatory problems (or lower urinary obstruction) that is sufficiently severe to cause collateral damage to the kidney by ischemia. Yet another is whether systemic illness that may be causing renal hypoperfusion is damaging the kidney by other mechanisms – for example, severe sepsis, wherein the kidneys do suffer from underperfusion but the organ injury and tubular epithelial apoptosis is being driven by many other factors (mostly associated with inflammation) as well. And finally, what was the baseline level of health of the kidney – if the patient with acute renal injury began the illness with a functional renal mass that was 20% of what it was 5 years earlier, that animal has less potential for recovery of adequate function compared to what it would have had at the younger age. This is especially pertinent in animals that present sick (uremic) with chronic renal disease (CRD), as they often have had chronic, insidious, compensated loss and all it took to make them symptomatic was a tiny acute further reduction in function.

Some specific points

1. Fluid therapy is essential to prevent or limit ischemic injury in vulnerable kidneys. In the face of hypovolemia or other causes of poor circulation, healthy young kidneys have enormous biological reserve to maintain function, and if pushed to extremis with ischemia can sustain a large amount of tubular epithelial loss and still do their job. But, if the kidneys have limited biological reserve – for example overwhelming acute nephron loss or senescent kidneys in geriatrics or chronically diseased kidneys - they are more likely to fail both *functionally* (as defined by reduced glomerular filtration rate [GFR]) and are perhaps also more likely to develop temporary or permanent *nephron loss* (due to epithelial apoptosis or necrosis, depending on the situation) when renal blood flow is compromised.

2. The only therapeutic effect of fluid therapy is to restore and maintain renal perfusion. Fluids do nothing else. When indicated to correct hypovolemia, fluid therapy can be life saving as no other treatment is as effective at preventing or limiting acute ischemic injury. However, unless hypovolemia is present fluids will not prevent kidney injury due to most nephrotoxins, and, once renal blood flow is restored, extra fluid will not provide any 'extra' renal benefit. So prompt recognition of impaired renal blood flow (assumed to be present in an animal with clinical signs of hypovolemia) and RAPID restoration of extracellular fluid and renal blood flow are huge, but once extracellular fluid volume and plasma volume are restored fluid therapy goals should immediately shift toward simply maintaining homeostasis.

3. Drug therapy for acute renal failure is a mixed bag. Although drug therapy of any specific cause of acute renal injury is essential (e.g., antibiotics for pyelonephritis or leptospirosis, 4MP to prevent conversion of ethylene glycol to oxalate), there are no great drug therapies to limit renal injury from poor perfusion, sepsis, or other shock syndromes. The only serious exception to this is drug therapy for circulatory support: in the world of critical care, restoring renal blood flow often requires drug therapy on top of fluid therapy to restore renal perfusion (for example, dobutamine and norepinephrine to animals with septic shock). Diuretics do not prevent acute renal injury and do nothing to help renal function. For all the interest in drugs like calcium channel blockers, dopamine receptor agonists and the like there is currently no good evidence to suggest that these have even a tiny fraction of the benefit of fluids and circulation support. Dopamine does not preserve renal tissue (and is a very weak diuretic).

4. Creatinine and BUN are not 'uremic toxins'. These substances are used as convenient chemical markers of GFR, but they are not toxins that contribute to uremic syndrome. At biologically relevant concentrations creatinine appears to be almost completely nontoxic, and urea nearly so. For example, Balestri et al(1) performed unilateral nephrectomy on 6 dogs and injected them with 3-4 grams of urea SQ every 6 hours for weeks. In spite of BUN concentrations that averaged 600 mg/dl two hours after each injection the dogs were essentially asymptomatic. Similarly, when urea is added to the dialysate for people on hemodialysis to maintain their urea concentrations > 200 mg/dl, these people are clinically indistinguishable from people whose urea concentrations are maintained lower on conventional dialysate. There are a couple hundred other known molecules that ARE felt to contribute to uremic syndrome (for a review, see Vanholder et al.(2)), and for many if not most of them their renal clearance is similar to creatinine – almost completely dependent on glomerular filtration rate (GFR). If products such as Azodyl™ are eventually shown to improve quality of life for uremic patients, this will likely be based on factors other than a reduction in BUN per se.

5. Once you've corrected any hypoperfusion, fluid therapy in excess of what is required to maintain homeostasis has only slight effects on renal function or the clinical condition of the patient, and no effect on recovery of renal tissue. Fluid administration designed to mildly expand the extracellular fluid compartment and force a water diuresis will increase GFR (and reduce serum creatinine concentration) by a few percent. This is similar to the postprandial 'bump up' in GFR of around 10-15% that occurs following a meal as renal blood flow increases and superficial cortical nephrons are recruited from a quiescent state to help eliminate excess solute. Although this mechanism is blunted in animals with renal failure, it still occurs to a limited extent. Once fluids are discontinued (or the animal returns to a fasted state), the GFR returns back down to baseline. As an example, if your patient has a stable serum creatinine of 2.0, volume expansion with a salt-containing fluid might increase GFR by 10%, yielding a new stable creatinine of 1.8 as long as fluids are continued. Once fluids are cut off and the need to excrete excess solute and water abates the creatinine goes back to 2.0. A similar albeit less substantial reduction in creatinine (and other molecules whose clearance proportional to GFR) concentration will be seen in animals with CRD and a higher baseline creatinine concentration. For example, if a dog with a stable serum creatinine concentration of 5.0 mg/dl has an 8% increase in GFR when volume-expanded with saline, its serum creatinine concentration will fall to about 4.6 mg/dl until the fluids are turned off. Once the fluids are turned off, GFR and serum creatinine return to baseline. But in contrast to creatinine, urea concentration tends to fall much more with water dieresis for two reasons: 1) urea is partially eliminated in the collecting tubule system, not just via glomerular filtration, so at any given level of kidney function the capacity for elimination of urea is greater than creatinine, and 2) it is normally recycled by the kidney (even sick ones) under the influence of ADH. When hypothalamic release of ADH is suppressed by administration of excess water, urea clearance in the collecting tubule system increases with water clearance. This is one reason urea tends to rise disproportionately to creatinine in pre-renal azotemia (when water and urea are intensely reclaimed from the collecting tubule), and is also the reason you've seen BUN continue to fall in some renal failure patients on fluid therapy while the creatinine holds steady. But as suggested above this may well have nothing to do with whether the animal feels better or not. So a pertinent question regarding continued volume expansion and forced diuresis is whether improvement in GFR of a very few percent will create enough improvement in how the animal feels that it will promote a transition to self-sufficiency. That is, will this slight improvement change a patient with anorexia/vomiting/hypodipsia into one that is willing and able to eat and drink enough to support itself? Although it might for the rare case, in my experience this is unusual.

6. Once ECF is restored, continued administration of a high-sodium fluid requires a renal response to excrete the massive load of sodium. The minimum daily dietary requirement (as established by the AAFCO) for sodium for adult dogs and cats is 0.06% and 0.08%, respectively, of food on a DM basis. This translates to about 0.4 mEq of dietary sodium per kg per day for cats and slightly less (0.3) for dogs. Around 3 times as much is needed for growth/reproduction. Commercial dog and cat foods supply 4-15 times the minimum (dry foods are at the lower end of this range) and reduced sodium type diets provide around 2-3 times the minimum. Consider a 10 kg dog with chronic renal disease and a previously stable creatinine of 3.0 mg/dl that now presents with a week's history of some vomiting, dehydration, and azotemia and the serum creatinine concentration has risen to 8 mg/dl. The typical approach is to rapidly correct the dehydration with a high-sodium replacement fluid (e.g., LRS) to restore plasma volume and RBF to eliminate the pre-renal component, then give the dog the same fluid afterwards at a higher-than-maintenance rate to provide its maintenance needs for water and to compensate for some polyuria from renal failure (say, for example, enough to increase the daily amount by ~50% above maintenance). The water maintenance requirement for dog this size may be as low as 19 ml/hour (~450 ml/day) if sick and inactive to as high as 32 ml/hour (770 ml/day) for an active dog jumping around in the cage. If we estimate a daily maintenance need of 600 ml and increase this by 50% to 900 ml per day, the patient will receive ~ 12 mEq/kg/day of sodium, or ~40x maintenance! For a 10 kg dog to eat this much sodium it would have to ingest 92 pounds of an AAFCO minimum sodium diet every day, or about 18 pounds of a dry food diet that contained 5 times the AAFCO-recommended minimum. Clearly this overdose usually works out in noncritical animals without left sided CHF or else we'd see a lot more immediate evidence of harm. Also, it seems that in spite of the CRF dog's blunted response to excreting an IV sodium load(3), dogs with surgically-induced renal failure can adapt to high sodium intakes within a few days(4). Abruptly stopping salt administration (or tapering over just a few days) may require a reversal of sodium excretion that some animals cannot achieve: Animals given enormous sodium loads with 'diuresis' efforts are usually given only a few days (at best) to abruptly reverse course and 'taper' their renal sodium excretion in response to withdrawal of fluids. Healthy dogs can respond with rapid adaptation - for example, when abruptly switched to a sodium-free diet they can drop their total daily urinary sodium excretion to < 0.2 mEq/kg/day within 2 days(5). However, it is unclear whether animals with renal disease can do the same, especially when overdosed with sodium. Humans with moderate - to - severe chronic renal disease can reduce their renal sodium excretion when weaned to a sodium restriction diet protocol (mean total of 5 mEq sodium/day), but in one small study(6) the weaning process took up to 14 WEEKS and if their sodium intake was 'tapered' too fast some were hospitalized with isotonic dehydration and pre-renal azotemia. So I wonder: how many feline patients that 'need' to be chronically treated with subcutaneous fluids administered at home are actually sodium 'addicts' of our own creation? These animals may well need more daily sodium than a healthy cat: Buranakarl et al(7) demonstrated that cats with induced renal failure do not do well - at least over a week – when switched (presumably from a high-sodium commercial maintenance diet) to a dietary sodium intake of 2.2 mEq/kg/day (however, the authors did not state how much sodium the cats were ingesting before switched to the test diet, so it is possible that part of their troubles were also related to an abrupt reduction in sodium intake from the pre-test diet). Many veterinary patients with renal disease are pushed to 1.5 – 3 x maintenance water and 10-20 mEq/kg/day (or more) of sodium for days, and then forced to adapt to rapid reductions in sodium excretion. Those that can't, come back to us in 3-7 days with pre-renal azotemia and our assessment that this animal 'just can't make it on its own'. If animals with spontaneous CRD are like humans – and this is unknown – then forcing them to transition from massive sodium excess to moderately restricted dietary intake within days is a recipe for development of hypovolemia and re-presentation within a few days for another uremic crisis.

So what to do? Here's my approach to support of the sick animal with renal failure, done in addition to any specific treatments aimed at the underlying disease, and controlling hypertension.

1. Rapidly (within minutes to hours, depending on the urgency and my level of confidence that they need the fluid and I won't hurt them if I overestimated their need) correct any hypovolemia. This is your only opportunity to modify the course of acute renal injury. This is accomplished with a high-sodium replacement fluid like LRS or saline and the goal is to restore the animal back to a NORMAL extracellular volume and renal perfusion. If fluids alone are not the answer (for example, hypotension and poor renal perfusion in animals with septic shock) then drug therapy is added as indicated by patient condition.

2. Once volume status is restored, immediately switch from resuscitation to maintenance. This means using a low sodium fluid like Normasol-M™, provided at a rate designed to maintain only a slightly positive water balance. For a quietly resting animal at room temp this will be around 24 ml/kg/day (for a 50 kg animal) to around 70 ml/kg/day (for a 3 kg animal) This figure is multiplied by no more than 1.5 – 2x to provide extra water for the polyuria of renal failure. The only patients that ever need more than this are the post-obstruction diuresis animals that are more polyuric for a few hours.

3. In animals with concurrent GI losses (due to continued vomiting or diarrhea) those losses are estimated and replaced within hours of the loss with a replacement, or at least half-strength replacement, fluid (GI losses tend to be high in sodium).

4. Remember that the potential for recovery of function lost to ischemia is relatively good, but in my experience routinely requires 7-10 days to achieve. If you think the injury is reversible, the animal is hanging on with hospitalization support, and the owner can afford it, give it that time. Remember that most initial reduction in creatinine (first 3 days) is likely due to correction of any pre-renal component, not improvement of intrinsic renal function.

5. Once the animal is ready to take over sustaining itself with oral intake (or tube feeding), fluid therapy is 'tapered' in proportion to the increase in oral intake of both water and salt (this means you have to read the ingredients label!) while monitoring the patient carefully to confirm that it is successfully making the transition. Pay attention that the oral salt intake is comparable to what was being administered with fluids. If you put the animal on a renal diet, you may need add a *calculated* amount of sodium chloride and taper it down to nothing over days to weeks.

6. Some animals are indeed so fragile that minor disruptions in environment or health will cause their oral intake to fall off and their water and/or electrolyte losses to exceed what they take in, setting them up for recurrent episodes of 'pre-renal' azotemia, decompensation, further reductions in oral intake, and clinical sickness. These are the ones that truly benefit from chronic SQ fluids. I do not recommend the chronically implanted SQ catheters, as I cannot imagine a world where these do not build up a bacterial biofilm that gives the animal with a chronic inflammatory stimulus. I would try to avoid LRS or saline, and stick with a lower-sodium fluid. Options include half-strength saline or half-strength LRS with 2.5% dextrose, Normasol-M™, or 5% dextrose in water supplemented with potassium chloride (10-25 mEq/L) and sodium chloride. 20-40 mEq/l). Patient tolerance for these will vary! Normasol-M uses acetate as an alkalinizing agent, and this substance can sting. Nonetheless this product is often well tolerated, and contains a lot less acetate than does Normasol-R™.

7. If you are going to 'routinely' administer SQ sodium, pay attention to the finding that chronic excess intake has been associated with acceleration of renal failure. Kirk et al(8) demonstrated progressive worsening of azotemia in cats with CRD when fed a diet that provided about 25 mEq of sodium per day. If your patient is getting 10 mEq/day of sodium from diet and you have the owner administer 100 ml of SQ saline each day you are duplicating this model.

8. Should you instruct owners to stop giving SQ fluids to animals already on chronic therapy? Absolutely not! If the owners are happy and the pet is doing well, don't mess with it – UNLESS the daily salt intake is high, in which case I would try to cut back on sodium over a period of a few weeks. But keep the principle of sodium intake in mind for future patients – particularly if condemning them to dependency on SQ fluids might prompt euthanasia instead.

Diabetic Ketoacidosis (DKA)

The terminal stage of diabetes mellitus, often occurring weeks or months after the onset of clinical signs. Because it represents decompensation in a previously stable diabetic, diagnosis of DKA should prompt a search for underlying causes such as infection, pancreatitis, or environmental stressors. One of the most common mistakes made in therapy of DKA is to assume that insulin is administered to treat hyperglycemia. In fact, insulin is administered to treat ketoacidosis, and glucose is monitored to a) avoid iatrogenic hypoglycemia, and b) help verify a biological effect of administered insulin.

Common complications of DKA include hyperglycemia, academia, and hyponatremia. Depletion of potassium, phosphorus, and magnesium are routine. Although the patient may initially present with hyperkalemia, hyperphosphatemia, or hypermagnesemia, those elevations may quickly reverse to dangerously low serum concentrations once renal perfusion is restored and insulin therapy is begun.

Baseline laboratory studies to obtain include

• Chemistry panel: Serum glucose, sodium, chloride, bicarbonate, BUN, creatinine, magnesium, calcium, and phosphate

• Urinalysis: to screen for urinary tract infections

• CBC

• Blood gas analysis: to confirm or rule out significant acidemia

• Appropriate samples for culture, based on suspicion of infection

• Radiographs/ultrasound

• After initial stabilization; look for evidence of pneumonia, urinary calculi, pyelonephritis, pancreatitis, peritonitis. Thoracic radiographs are also useful to screen for metastatic neoplasia

Monitoring

Should aim to detect responses to correction of hypovolemia, ketoacidosis, and the underlying cause. Continuous ECG monitoring is used as surveillance of the heart rate and to screen for signs of electrolyte imbalance. Hyperkalemia produces tenting of T waves, reduced amplitude of P waves, and widening of the QRS complex. Hypokalemia predisposes to arrhythmias and prolongation of the Q-T interval (to > 250 msec). Hypophosphatemia may manifest as arrhythmias. Arterial blood pressure is used to direct fluid therapy and treatment of septic shock if present. Although noninvasive pressure monitoring is adequate if the animal is hemodynamically stable, invasive monitoring by direct arterial catheterization may be required. Urine production measurement is used to assess renal perfusion and measurement of urine ketones and glucose monitors the response to insulin. Blood glucose concentration should be assessed frequently (often at least hourly in the first few hours) to confirm a biological response to insulin and to monitor against development of hypoglycemia. A handheld glucometer allows relatively inexpensive frequent testing. Serum and urine ketones and arterial or central venous blood pH are measured to monitor the response to insulin. A urine dipstick can be used on serum to assess ketosis, and serum electrolyte concentrations assess the need for supplementation.

Patient support methods include

1. Central venous catheter to allow painless collection of multiple blood samples and administration of maintenance fluids, electrolytes, and glucose. It also facilitates measurement of central venous pressure if needed.

2. Peripheral vein catheter: Large gauge if resuscitation is needed, e.g., 20-18 gauge if < 10 kg, 10-16 gauge in larger dogs. A smaller gauge is acceptable if it is to be used primarily for the administration of insulin.

3. Assistance with thermoregulation using blankets +/- a heat source.

4. Nursing care to assist with hygiene and skin care.

5. Oxygen therapy if hypoxemic

6. Analgesia may be necessary in the presence of painful infections or systemic inflammation.

7. Antibiotic therapy is based on identifying bacterial infection, its location, and Gram stain results; it is refined by culture/sensitivity results.

8. Antiemetic drug therapy +/- a gastric sump (nasogastric tube for continuous or intermittent evacuation of the stomach).

Priorities for treatment

Restoration of perfusion: patients with physical signs or laboratory evidence (azotemia) of hypovolemia require rapid resuscitation with a high-sodium (replacement type) fluid such as lactated Ringer's, 0.9% saline, or an equivalent. 5-10 ml/kg of warmed fluids is administered rapidly (< 5 minutes) with assessment of physical signs before and after. This is repeated until signs of hypovolemia are resolved. Although the patient may have lactic acidosis, lactated Ringer's solution is often effective as it restores tissue perfusion and eliminates the source (tissue hypoxia) of lactic acid. Patients with septic shock may require advanced support of circulation (e.g., vasopressors, cardiac drugs).

Maintenance fluids: Once initial resuscitation is accomplished, administer fluids at a rate sufficient to provide maintenance needs and to make up for any ongoing GI losses or polyuria. In the absence of GI loss, a commercial maintenance fluid containing 30-40 mEq/l of sodium is ideal. This fluid may contain 5% dextrose (see below). If there is ongoing GI loss or polyuria, equal volumes of a higher sodium fluid (e.g., 0.45% NaCl/2.5% dextrose or even LRS) may be used to provide maintenance needs and replace ongoing losses. This fluid should be administered at a constant rate to facilitate addition of glucose and electrolytes.

Insulin therapy is delayed until hypovolemia is corrected (this should happen in < 30 minutes!) and therapy for hypokalemia if present has already begun. Regular insulin is ideal for therapy. Our practice is to dilute to 1 IU/ml in 0.9% NaCl or D5W, and administer with syringe pump at a starting dose of 0.1- 0.2 IU/kg/hour. The relatively high concentration (equivalent to 1000 units/liter) minimizes dosage error due to adsorption of insulin onto plastic fluid disposables. As an alternative, regular insulin may be administered IM at an initial dosage of 0.1- 0.2 IU/kg IM every hour. One should avoid diluting insulin solutions in IV fluid bags due to the significant loss of insulin to adherence to the plastic bag and administration set.

If baseline blood glucose is > 600 mg/dl, avoid overly rapid reductions of serum glucose (keep under 100 mg/dl/hour). If the blood pH or serum bicarbonate does not increase significantly within 4-6 hours, we increase the insulin dose in 0.1 IU/kg/hour increments. If pH measurement is unavailable, monitor the serum and/or urine ketone concentration. An initial increase within the first 24 hours is not uncommon and may represent resolving acidosis and a shift toward acetone production. This is not necessarily a reason to increase insulin administration unless accompanied by signs of deterioration.

The insulin infusion rate may be reduced as the acidosis (not the hyperglycemia) resolves.

Once serum glucose falls below 250 mg/dl, supplement maintenance fluid with dextrose to a 5% concentration. This may be accomplished with a commercial maintenance fluid with 5% dextrose already added, e.g., Normasol-M™ (Abbott). Alternatively, one can create a 5% dextrose solution by removing 100 ml of fluid from a 1 liter bag and adding 100 ml 50% dextrose. Some animals will require 7.5 – 20% dextrose solutions in fluid administered at a maintenance rate. Concentrations above 7.5% should be administered through a central venous catheter to avoid phlebitis. The blood glucose concentration should be monitored frequently. If hypoglycemia develops, administration of 0.2 ml/kg lean body weight of 50% dextrose will transiently raise the blood glucose concentration by 50 mg/dl. Once the patient is improving clinically and the acidosis is resolving, insulin therapy may be reduced and shifted to a maintenance regimen. We typically calculate the total daily insulin dosage being administered as an infusion at that time, and give this amount as a slow-release form of insulin divided BID.

Potassium therapy is routinely required in most animals with DKA, as they typically have significant and sometimes severe potassium depletion. Even if they present with hyperkalemia from poor renal perfusion, marked hypokalemia may follow initiation of insulin therapy. If they are initially hyperkalemic, serum [K] is monitored every 1-2 hours after insulin is begun. Once the serum [K] falls to < 4.0 mEq/l we begin KCl supplementation at a concentration in the maintenance fluid designed to administer 0.2 mEq/kg/hour. If hypokalemia is more severe or if clinical signs are present, we administer 0.25 – 0.5 (or higher) mEq/kg/hour. For example, in a 10 kg dog with a serum [K] = 2.8 mEq/l we may choose to supplement potassium at 0.35 mEq/kg/hour, or 3.5 mEq/hr total. At a maintenance fluid rate of ~ 15 ml/hour, a 1 liter bag will last 1000/15 or 66 hours. 66 hour's worth of therapy with an hourly potassium administration rate of 3.5 mEq/hour = 230 mEq/l (!!!) Because this is such a high potassium concentration several precautions are used:

• Concentrations > 60 mEq/l Must be given in a central vein to avoid phlebitis.

• Accidental increase in fluid rate will be fatal! Avoid this by using a fluid pump and an in-line burette which holds only 100 cc of fluid mix at a time. In this case, 230 mEq/l = 23 mEq/dl. To make 100 ml of fluid, inject 23 mEq KCl (11.5 ml) into the burette and add 88.5 ml of IV fluid (assuming there is no potassium in the fluid already).

• Frequent checks!

Magnesium depletion is caused by the same factors that cause potassium depletion. Hypermagnesemia may be found initially until normal renal perfusion is restored. Magnesium depletion may make hypokalemia more difficult to correct. This condition is treated with either magnesium sulfate or magnesium chloride, with the latter being preferred. If one uses magnesium sulfate the loading dose is 30 mg/kg IV over 10-30 minutes, followed by 30 – 60 mg/kg/day. For magnesium chloride, divide these figures by a factor of 3.

Phosphorus: Hypophosphatemia is common following initiation of insulin, as insulin stimulates the phosphorylation of glucose to form glycogen and increases ATP stores. The generally recommended dose of 0.03 mmol/kg/hour is too low for most DKA patients. We use KPO4 to supplement both potassium and phosphorus in hypophosphatemic patients. In patients with mild to moderate hypophosphatemia we use a 50:50 mix of KCl and KPO4 administered at a rate to supply K needs (0.2 – 0.5 mEq/kg/hour of potassium). In those with more severe hypophosphatemia (< 2.5 mg/dl), we use straight KPO4 to supply potassium needs. This provides MUCH more phosphorus than 0.03 mmol/kg/hour, but the author is not aware of any diabetic patient that became hyperphosphatemic. The rate or administration and /or the ratio of KCl:KPO4 is adjusted based on monitoring both electrolytes.

Bicarbonate: DON'T! Proper administration of insulin will result in ketone metabolism to bicarbonate and restoration of serum bicarbonate. Administration of bicarbonate may produce complications including intracellular acidosis and an 'overshoot' alkalemia when ketosis is resolved.

References

Balestri PL, Rindi P, Biagini M. Chronic urea intoxication in dogs. Experientia 1971 Jul;27(7):811-2.

Vanholder R, De SR, Glorieux G, Argiles A, Baurmeister U, Brunet P, et al. Review on uremic toxins: classification, concentration, and interindividual variability. Kidney Int 2003 May;63(5):1934-43.

Bourgoignie JJ, Kaplan M, Gavellas G, Jaffe D. Sodium homeostasis in dogs with chronic renal insufficiency. Kidney Int 1982;21:820-6.

Coleman TG, Guyton AC. Hypertension caused by salt loading in the dog. 3. Onset transients of cardiac output and other circulatory variables. Circ Res 1969 Aug;25(2):153-60.

Spangler WL. Pathophysiologic response of the juxtaglomerular apparatus to dietary sodium restriction in the dog. Am J Vet Res 1979;40(6):809-19.

Danovitch GM, Bourgoignie J, Bricker NS. Reversibility of the "salt-losing" tendency of chronic renal failure. N Engl J Med 1977 Jan 6;296(1):14-9.

Buranakarl C, Mathur S, Brown SA. Effects of dietary sodium chloride intake on renal function and blood pressure in cats with normal and reduced renal function. Am J Vet Res 2004 May;65(5):620-7.

Kirk CA, Jewell DE, Lowry SR. Effects of sodium chloride on selected parameters in cats. Vet Ther 2006;7(4):333-40.