Comparison of different methods of acid-base analysis in canine parvoviral enteritis

Abstract

Acid-base disorders are common in critical care patients, and thus understanding these derangements is important in critical care medicine1-3. Disturbances should be expected in animals presenting with gastro-intestinal, renal, respiratory and neurologic diseases, as well as in shock4. Routine serum biochemistry results may be suggestive of an acid-base disturbance and abnormalities should prompt clinicians to investigate the patients’ acid-base status further, through blood gas analysis5. Identification of acid-base disorders is valuable in the clinical appraisal of a patient and informing treatment regimens6. Since canine parvoviral enteritis (CPE) is a disease associated with extensive fluid, protein and electrolyte losses through vomiting, diarrhoea, sepsis and malabsorption, the associated losses contribute to acid-base imbalances, most importantly in the metabolic compartment7-11. Several methods have been used to assess acid-base status in dogs. These include the traditional Henderson-Hasselbalch approach (HH), and the quantitative approaches, namely the Stewart strong ion model and various adaptations of this approach. The hypothesis of this study was that the strong ion approach would identify acid-base changes that would not be appreciated by the HH model. Accordingly, blood was collected from 41 puppies with confirmed CPE. Blood was collected for venous blood gas and serum biochemistry, and all samples were collected at admission prior to any therapeutic interventions. Each patient’s acid-base status was assessed according to the HH model, the base excess algorithm (BEA) and a simplified strong ion approach (SSA). Most data were not normally distributed as determined by Shapiro Wilk and as such all comparisons made use of non-parametric methods (Mann Whitney U for the comparisons between medians). The control group of dogs were compared to the CPE group for all measured and calculated variables and a p- value <0.05 was regarded as significant. The HH model detected acid-base abnormalities in 41% of dogs, the SSA demonstrated derangements in 46% of dogs and the BEA displayed abnormalities in 89% of dogs. Acid-base disorders may be assessed utilising different methods with variations within each of the methods themselves. There are many ions, proteins and buffers that all affect the acid-base balance. The acid-base approaches discussed in this study namely, the HH, SSA and BEA, place emphasis on different components of a complex system in an attempt to understand and dissect the underlying pathogenesis. The contribution of one component may also dampen the effect of a component having an opposing effect on acid-base balance (for example the opposing effect of hypoalbuminaemia on the effect of phosphorus). The study also highlights the complexity of methods that use multiple variables to generate a final result (particularly the SSA). The opposing or antagonising effects of these individual variables may mask and falsely diagnose acid-base disturbances. The current study was chiefly descriptive. Despite this, the methods were compared and the level of discordance between the models was evaluated. A significant discordance was demonstrated between the cases diagnosed as normal by each method. In the assessment of the discordance between the methods, the HH demonstrated discordance of 42% with the SSA and 92% to the BEA. Similarly, the SSA demonstrated discordance of 36% between the HH and 86% between the BEA. Lastly, the BEA demonstrated a discordance of 50% between the HH method and 25% between the SSA. Based on these levels of discordance it becomes clear that the different methods of acid-base analysis cannot be used interchangeably. In conclusion, the acid-base changes in CPE are complex and not easily understood using the HH model. The pathophysiologic mechanisms are more easily explained by the SSA and BEA approaches. However, it is unclear if using these methods of diagnosis change the way that CPE should be managed in clinical practise.