Objective: To develop a diagnostic analysis software for determining the type of acid-base balance disorder.
Methods: Mathematical models were built based on Henderson-Hasselbalch equations and compensation formulas, to determine the important parameters of acid-base balance disorder, and to develope acid-base balance disorder analysis process. The software was compiled using the Visual Basic.NET programming language, and the installation package was generated after debugging. Acid-base balance disorder cases were searched by PubMed, Wanfang and CNKI databases from 1980 to 2015, and the blood gas parameters [pH, arterial partial pressure of carbon dioxide (PaCO2), HCO3- and anion gap (AG)] and the types of acid-base imbalance (literature results) were recorded. All cases were reanalyzed by software and the type of acid-base balance disorder was determined (software diagnostic type). Kappa-test and McNemar-test were performed for the two diagnostic results.
Results: The "four parameters-four steps" analysis method was used as the analysis process to judge the types of acid-base balance disorder. "Four parameters" included pH, PaCO2, HCO3- and AG. "Four steps" were outlined by following aspects: (1) according to the pH, combined with PaCO2 and HCO3-, the primary types of acid-base balance disorder was determined; (2) according to the compensation situation, double mixed acid-base balance disorder (DABD) was determined; (3) according to AG value, three mixed acid-base disorders (TABD) were determined; (4) the ratio of ΔAG ↑ /ΔHCO3- ↓ was also calculated to determine whether there was normal AG metabolic acidosis or metabolic alkalosis. The software had the characteristics of simple interface, convenient operation, rapid judgment, and comprehensive analysis. It could judge all acid-base balance disorder types excepted "AG normal metabolic acidosis combined metabolic alkalosis". The software was used to reanalyze 112 cases of acid-base balance disorder reported in the literature, with a consistent rate of 87.50% and better consistency of the diagnostic results (Kappa test: κ = 0.84, P < 0.01; McNemar test: χ2 = 0.87, P = 0.65).
Conclusions: The software can be used as an important tool to judge the type of acid-base balance disorder, and provide clinicians with diagnostic reference, which have practical value and application prospect.