Metabolic acidosis is common in ill and injured animals although the actual incidence of this disorder has not been described in human or veterinary patients. Metabolic acidosis has been found to have diagnostic, therapeutic, and prognostic implications in human medicine. Base excess on admission correlates with outcome in several studies of human intensive care patients, and base excess and pH are significantly lower in nonsurvivors compared with survivors of major vascular trauma.[1-4] There is some evidence that metabolic acidosis may have similar clinical relevance in veterinary patients. Base excess was correlated with survival in dogs with diabetic ketoacidosis, and bicarbonate concentration has been inversely correlated with mortality in cats.[5, 6]
Metabolic acidosis occurs when the accumulation of nonvolatile acids or the loss of bicarbonate exceeds the buffering capability of the body. On acid base analysis, metabolic acidosis is characterized by a decrease in serum bicarbonate concentration or a decrease in base excess. Base excess is calculated by blood gas machines and often is reported as standardized base excess (SBE). Metabolic acidosis can originate in 1 of 3 scenarios. First, it can be a primary acid base disorder that is associated with a corresponding decrease in pH. Second, it may be compensatory for primary respiratory alkalosis, in which case the pH would be higher than normal. Third, it can occur in combination with a respiratory acid base abnormality, resulting in a mixed disorder. In these mixed disorders, the pH may be far lower than expected when occurring with a respiratory acidosis, or it may remain in the reference range when occurring with a respiratory alkalosis. As a consequence, the pH alone cannot be used as a screening tool to detect the presence of metabolic acidosis.
Acid base analysis is becoming readily available in the clinical setting. Blood gas machines generally require very small quantities of blood and provide rapid, valuable results, making acid base analysis an ideal diagnostic and monitoring tool in emergent and critically ill animals. To utilize this information effectively, an understanding of the importance of acid base abnormalities in dogs and cats is required. The aim of this study was to describe the incidence, magnitude, and etiology of metabolic acidosis in a large group of dogs and cats treated at a veterinary teaching facility.
- Top of page
- Materials and Methods
The results of this study show that metabolic acidosis is a common abnormality in dogs and cats in which blood gases are evaluated. Primary metabolic acidosis was the most common disorder identified in dogs, and a mixed acid base disorder of metabolic acidosis in conjunction with a respiratory acidosis was the most common disorder evident in cats in this study. Hyperchloremic metabolic acidosis was more common than high AG metabolic acidosis. Metabolic acidosis as a compensatory response to primary respiratory alkalosis was uncommon in dogs and rare in cats. Twenty-five percent of dogs and 34% of cats could not be classified as having either a hyperchloremic metabolic acidosis or a high AG metabolic acidosis. These results suggest that hyperchloremia is common in small animal patients and the underlying causes and clinical relevance of this abnormality should be investigated. The routine categorization of metabolic acidosis based on the presence of high AG or hyperchloremia may be inadequate in many cases, indicating that more advanced acid base analytical tools have a role in patients with complex disease.
Arterial and venous blood samples were analyzed together in this study, using venous results for the acid base diagnosis. Because arterial acid base parameters, in particular PCO2, are different from venous parameters, there is the potential that respiratory alkalosis may have been falsely diagnosed in some arterial samples. Interestingly, when the venous samples were analyzed separately, there was an increase in the proportion of cats and dogs classified as having a combined metabolic acidosis and a respiratory alkalosis. Almost all of the arterial samples were collected under anesthesia, and the PCO2 is likely to have been influenced by respiratory depressant drugs. Accordingly, when the venous samples were analyzed separately from the combined arterial and venous samples, the major difference was a reduction in the number of cases diagnosed with metabolic acidosis and concurrent respiratory acidosis. This is not an unexpected finding, but must be taken into account if trying to extrapolate our findings to another patient population.
Metabolic acidosis was evident in 49% of the animals in which a blood gas was performed during the 13-month period of this study. These patients included those presenting to the emergency room, intensive care unit patients, anesthetized patients, and specialty service patients. A wide variety of primary disease processes was present (Table 2). The most common complaints in combined arterial and venous samples of dogs were neoplasia and surgery. On review of the records, the most common reason for surgery in this population was neoplasia, and blood gas analysis was performed intraoperatively. Not surprisingly, in the canine group of venous samples, surgery was an uncommon problem and neoplasia was far less frequently reported, although it remained the most common problem of the canine group. There may be several causes of metabolic acidosis in patients with neoplasia including Type B lactic acidosis and secondary organ function impairment. In contrast to dogs, renal disease was the most frequent disease process in cats in this study. Renal disease is well recognized to cause metabolic acidosis by several potential mechanisms. It cannot be determined if the acid base abnormalities identified in this study were primarily due to the underlying diseases present, the treatment provided, or both. Unfortunately, it is not possible to assess the incidence of metabolic acidosis in specific disease states in the current study, but these results do indicate that metabolic acidosis is a common acid base abnormality in those patients for which clinicians elect to perform blood gas analysis.
The ideal marker of metabolic acidosis remains controversial.[12, 13] Serum bicarbonate concentration is the traditional indicator of metabolic acid base balance. It is most commonly calculated from the measured pH and PCO2 by a blood gas machine, although some chemistry laboratories measure it directly. The bicarbonate concentration has been criticized in the literature because it is not independent of changes in the PCO2.[14, 15] As PCO2 increases, the bicarbonate concentration increases and it is possible that an increased bicarbonate concentration could be mistaken for metabolic alkalosis, when in fact it had occurred subsequent to respiratory acidosis. The calculated value of SBE offers a measure of metabolic acid base balance that is standardized to a PCO2 of 40 mmHg and was used as the diagnostic marker of metabolic acidosis in the present study. The strong ion approach also offers a method of metabolic acid base analysis that is considered independent of changes in PCO2, but it requires more measured variables than those provided by the blood gas machine and many complex equations, making it difficult to apply in the clinical setting.
The SBE is calculated from the pH, PCO2, and hemoglobin concentration representing the quantity of acid or base that would need to be added to a liter of blood to normalize the pH with a PCO2 of 40 mmHg. A simple metabolic acidosis exists when there is a decrease in bicarbonate and a more negative SBE in conjunction with a lower than normal pH. The PCO2 also is expected to decrease as a compensatory response. When mixed disorders exist (ie, concurrent derangements in both the metabolic and respiratory acid base systems), the change in PCO2 is not within the range expected. If a respiratory alkalosis coexists with a metabolic acidosis, the pH may not be decreased as expected. For this reason, we used the change in SBE to identify patients with metabolic acidosis in this study, regardless of the concurrent pH, an approach that has been utilized in previous studies.[3, 16] As a result, the patients identified included those with simple metabolic acidosis, those with mixed disorders, and those with metabolic acidosis that was compensatory to primary respiratory alkalosis. Complex metabolic acid base disorders also occur in which concurrent acidotic and alkalotic processes can coexist to result in a relatively normal SBE.[10, 17] A quantitative approach to acid base analysis is required to identify such disorders involving measurement of serum albumin and phosphorous concentration at the same time as acid base analysis is performed. In the current study, many animals did not have these variables measured, and when they did, it was impossible to know the time frame in which the samples were collected in reference to the acid base analysis. As a result, patients with complex metabolic acid base disorders may have been missed using the inclusion criteria of this study.
In this study, compensation in dogs was determined using previously published criteria. Compensation in cats is not fully understood, and the studies available to date suggest that cats do not develop respiratory compensation to a metabolic acidosis.[18, 19] For this reason, compensatory responses were not evaluated for cats in the present study. Mixed acid base disorders in cats in this study were classified as all cases in which there was a change in both PCO2 and SBE.
The magnitude of the acid base abnormalities identified in this study were moderate overall but varied with underlying acid base disorder (Tables 3 and 4). The degree of change in pH will vary depending on compensation and the presence of complex disorders, as shown by the results of this study. For example, the mixed disorder of metabolic acidosis with concurrent respiratory acidosis caused the most extreme abnormalities in pH, whereas primary metabolic acidosis was typified by mild decreases in pH and animals with a combined metabolic acidosis with respiratory alkalosis had normal pH. As a result, pH alone cannot be utilized to determine the severity or nature of an acid base disorder. This also may explain why pH has been reported to correlate poorly with outcome compared with measures of metabolic acid base balance such as bicarbonate or SBE.[20, 21]
Primary respiratory alkalosis will result in compensatory metabolic acidosis in dogs, and the severity of this response will vary with the chronicity and severity of the respiratory alkalosis. For the purposes of this study, the degree of metabolic compensation considered appropriate for a given respiratory alkalosis was anything that fell in the range of that reported for acute or chronic disease. Because there was no way to accurately determine the duration of the primary disease process in this study, it is possible that some cases were incorrectly classified as a simple respiratory alkalosis. Primary respiratory alkalosis was only evident in 6% of the dogs in this study, making it an unusual cause of metabolic acidosis (Table 3). It should be noted that the enrollment criterion for this study was an SBE outside of the reference range. As a result, dogs with a respiratory alkalosis and no metabolic compensation (ie, no change in SBE) would not have been identified. Metabolic acidosis with concurrent respiratory alkalosis (ie, a greater decrease in PCO2 than that appropriate for compensation) was present in 22% of dogs. This suggests that respiratory alkalosis is not uncommon in dogs but tends to occur in conjunction with other disease processes.
The limited information available on metabolic compensation in cats would suggest that it is similar in magnitude as dogs. The single cat in this study with respiratory alkalosis had a bicarbonate concentration consistent with compensation for a chronic process. With 1 case of respiratory alkalosis of 134 cats included in this study, primary respiratory alkalosis seems to be an unusual abnormality in cats. Thirty-six cats (27%) with metabolic acidosis had a PCO2 below the reference range, classified as a concurrent metabolic acidosis and respiratory alkalosis for the purposes of this study. In a manner similar to dogs, it would appear that respiratory alkalosis is relatively common in cats in conjunction with complex disease processes.
Primary (or simple) metabolic acidosis was a common acid base disturbance (36%) in dogs, but it was not as common as mixed acid base disorders which were evident in 58% of dogs included in this study (Table 3). These results were similar in cats although direct comparison of the acid base results in cats is challenging given our lack of understanding of compensation in cats. Because a large number of acid base evaluations in our hospital are performed in animals presenting to the emergency room and intensive care unit, this finding is not surprising, because these patients tend to have complex disease processes. A high incidence of complex acid base disorders has been demonstrated in human emergency room and intensive care unit patients.[2, 22, 23] These results suggest that recognition of mixed acid base disorders is important and that there may be a role for semiquantitative analysis of acid base balance in critically ill or injured patients.[10, 17]
Calculation of the AG can aid in determination of the underlying cause of a metabolic acidosis. Metabolic acidosis associated with a high AG is due to the gain of acids such as acids associated with uremia, lactate, ketones, and the metabolites of ethylene glycol. An increased AG was identified in 28% and 34% of dogs and cats, respectively, whereas an increased lactate concentration was evident in almost half of the cases in this study (Tables 5 and 6). The normal AG is largely due to the presence of albumin, an anion the quantity of which is not accounted for directly; hence, it is one of the unmeasured anions included in the AG. The presence of an increased lactate concentration with a normal AG is likely to be due to the presence of hypoalbuminemia, lowering AG, and reducing its sensitivity of detecting unmeasured anions in the system. When a patient has a low albumin concentration, other unmeasured anions such as lactate may accumulate and yet the total calculated AG still can fall within the reference range.[24, 25] As a result, it is important to recognize the limitations of AG to avoid missing underlying causes of metabolic acidosis. Unfortunately, in this study, we did not have measurement of serum albumin concentration and could not evaluate its influence on AG.
Hyperlactatemia was the second most common condition associated with metabolic acidosis in dogs and the most common cause identified in cats in this study. Hyperlactatemia is most commonly associated with inadequate oxygen delivery to the tissues. There is a growing body of evidence in the human and veterinary literature that serum lactate concentration also can have important prognostic value, in particular lactate clearance over time.[27, 28] This study was not designed to evaluate the impact of acid base status on outcome, but the frequent occurrence of hyperlactatemia in this patient population suggests that lactate could be a useful diagnostic and prognostic tool. Additional studies on the relevance of serum lactate concentration in small animal patients are required at this time.
Metabolic acidosis commonly is divided into 1 of 2 categories: those associated with a gain in acid and an increased AG versus those associated with a loss of bicarbonate, characterized by hyperchloremia and a normal AG. Hyperchloremia was a common abnormality associated with metabolic acidosis in dogs and cats in this study. It also has been reported to occur frequently in hospitalized human patients.[29, 30] Chloride-rich fluids are a considered an important cause of hyperchloremia. This may occur in the form of saline administration or IV fluids supplemented with substances such as potassium chloride. Renal disease could be another potential cause of hyperchloremic metabolic acidosis. Metabolic acidosis associated with differential shifts of sodium and chloride from the intracellular space have been described in an experimental model of endotoxemia although the mechanisms responsible for these changes could not be identified. There is some controversy as to the clinical relevance of the presence of hyperchloremia. Several studies in humans have reported an association between hyperchloremia and increased mortality, whereas other studies have failed to demonstrate this association.[3, 31, 32] There have been no studies evaluating hyperchloremic metabolic acidosis in veterinary patients.
An interesting finding in this study was the high incidence of cases with both normal AG and normal chloride concentration, when both measured chloride and corrected chloride were evaluated (Table 5). There are several possible explanations for this seemingly contradictory finding. The presence of hypoalbuminemia is likely to be 1 contributing factor, as discussed previously. Animals with free water excess can be misdiagnosed as having a normal chloride concentration unless the chloride concentration is corrected appropriately.[10, 33] As Table 5 shows, the number of dogs in this study with normal chloride concentration and normal AG is far less when the corrected chloride concentration was used. This indicates that there were a substantial number of dogs with free water excess in this patient group. In contrast, the number of cats with a normal chloride concentration and normal AG did not change appreciably when the corrected chloride concentration was used, indicating free water excess was not a common finding in cats.
This study has several limitations. As previously mentioned, we used normal venous values to diagnose acid base abnormalities for both arterial and venous blood gas samples. Analysis of the venous samples separately showed that the overall acid base interpretation of the results of this study is minimally impacted by this approach. We generated a reference range of values for our blood gas machine for both dogs and cats from which we generated the diagnostic criteria used in this study. This reference range was based on a small group of animals and would be strengthened if a larger population of animals had been included. Given the retrospective nature of this study, it was not possible to determine exactly when blood gas samples were drawn with respect to treatment and other interventions. Furthermore, the reasons a clinician elected to perform acid base analysis were unknown, which somewhat limits the interpretation of the information presented. Another issue is that we only evaluated the first blood gas measured per visit, and metabolic acidoses that developed during hospitalization may not have been identified in this study.
Metabolic acidosis was found commonly in this group of patients at a University Teaching Hospital in which blood gas analysis was performed, and it was associated with a wide variety of disease processes. Mixed acid base disorders were frequent in this population, and routine categorization of metabolic acidosis based on the presence of high AG or hyperchloremia was misleading in many cases. Given the frequency of metabolic acidosis identified in this study, further investigations evaluating the diagnostic and prognostic relevance of acid base disorders in small animal patients are warranted.