Alterations and diagnostic performance of capillary ketonemia in pediatric acute appendicitis: a pilot study

This is a subanalysis of the BIDIAP study [13, 14], an observational prospective study aimed to assess the diagnostic performance of several serum biomarkers in PAA. The objective of this subanalysis was to characterize the alterations in CK of patients with PAA and calculate its diagnostic performance. Two groups of pediatric patients were included in this study: (1) patients with acute abdominal pain who were admitted at the emergency department in whom there was a diagnostic suspicion of acute appendicitis and in which it was finally excluded—also defined as non-surgical abdominal pain (NSAP), and (2) patients with histologically confirmed diagnosis of acute appendicitis. For further analysis, patients in group 2 were stratified based on histopathological classification of the appendix in uncomplicated PAA (congestive, phlegmonous or suppurative appendicitis) and complicated PAA (gangrenous or perforated appendicitis). All patients in group 1 were contacted 2 weeks after their inclusion in the study to confirm that they had not been diagnosed with PAA in that period.

Sociodemographic and clinical variables were collected at baseline through a paper-based questionnaire completed by parents or legal guardians. Information on analytical, surgical, radiological and histological variables was extracted from participants’ clinical records.

Patients were recruited when the personnel conducting the investigation were available at the center. The recruitment period extended from February to December 2021. Inclusion and exclusion criteria are shown in Supplementary file 1.

In all patients, the blood sample for the determination of CK was taken at the time of inclusion in the study, upon arrival at the pediatric emergency department and before initiating intravenous hydration. Blood sample was extracted by finger prick and placed on a blood beta-ketone test strip (Abbott ©, Abbott Diabetes Care Ltd). The analysis was performed by electrochemical study in a FreeStyle Optimum Neo device.

For descriptive purposes, we used mean and standard deviation or median and interquartile ranges for quantitative variables and proportions for categorical ones. Kolmogorov–Smirnov test was used to assess the normality of quantitative variables. Sociodemographic, clinical and analytical variables were compared between groups using the Mann–Whitney U test. To calculate the discriminative capacity of the biomarkers, we calculated the area under the receiver operating characteristic curves (ROC). Statistical significance was settled in a p-value < 0.05. Statistical analysis was performed with STATA 17.0 (StataCorp LCC).

We calculated the Youden Index for each value with the following formula:

The biomarker value with the highest Youden Index was considered the optimal cut-off point.

Pearson’s or Spearman’s correlation tests were used to assess the correlation between CK and both clinical and analytical variables.

This study, part of the BIDIAP cohort [13‐15], included 151 patients, divided into two groups: (1) patients with non-surgical abdominal pain in whom the diagnosis of PAA was excluded (n = 53) and (2) patients with a confirmed diagnosis of PAA (n = 98). In 80 patients (group 1 = 23 patients, group 2 = 57 patients) a capillary ketone blood determination was obtained at the time of diagnosis. In the remaining 71 patients, no such determination was obtained, either because the parents and/or patients expressly refused the capillary finger prick or because of errors in the reading of the test strip.

Given the high drop off rate of the cohort (47%) and to assess the possibility of selection bias we performed a detailed comparative analysis between the two subpopulations of the cohort: patients with capillary ketonemia determination present (n = 80) and patients without such determination (n = 71). This analysis, which included all the sociodemographic and clinical variables presented in this paper, found no statistically significant differences in any of the analyses performed.

Patients in group 2 (n = 57) were further subdivided into complicated PAA (n = 18) and uncomplicated PAA (n = 39). Sociodemographic characteristics of patients by group are shown in Table 1. Significant differences between groups were observed for height (p = 0.05) and weight (p < 0.01).

Clinical characteristics of patients by group are shown in Table 2. No significant differences were observed for the time of pain evolution, the prevalence of hyporexia or the presence of fever at home. However, compared with patients with NASP, those with PAA had significantly more emetic episodes (p < 0.01). In relation to analytical determinations (Table 2), we found that patients in group 2 had significantly higher levels of total leukocytes (p < 0.01), total neutrophils (p < 0.01) C- reactive protein (CRP) (p < 0.01) and procalcitonin (PCT) (p < 0.01).

Table 3 shows the comparison of clinical variables between complicated and uncomplicated PAA groups. No significant differences were observed for the time of pain evolution, the prevalence of hyporexia or the number of emetic episodes. However, patients with complicated PAA were more likely to have fever at home than those with uncomplicated PAA (p = 0.01). The comparison of analytical determinations between the complicated and uncomplicated PAA groups (Table 3) showed significantly higher levels of total leukocytes (p < 0.01), total neutrophils (p = 0.01), CRP (p = 0.02) and PCT (p < 0.01) in patients with complicated PAA.

None of the patients included in this study presented hypoglycemia. No significant differences in serum sodium or creatinine levels were found in either of the two univariate comparative analyses performed.

CK values were 0.3 [0.1–0.9] mmol/L in group 1 and 0.7 [0.4–1.4] mmol/L in group 2 (p = 0.01). Regarding the type of PAA, CK values were 0.6 [0.4–0.9] mmol/L in uncomplicated PAA and 1.2 [0.8–1.4] mmol/L in complicated PAA (p = 0.02). Figure 1 shows the graphical representation of the determinations of CK in groups 1 and 2 (2 outliers from group 1 and 2 outliers from group 2 were removed for a better visualization of data distribution).

Regarding the capacity of CK to discriminate between patients from groups 1 and 2, we found an AUC of 0.68 (95% CI 0.53–0.82) (p = 0.24) (Fig. 2). The cut-off point with the best Youden Index resulted in 0.4 mmol/L, with a sensitivity of 80.70% and a specificity of 52.17%.

Table 4 summarizes the alternative proposed cut-off points with their corresponding values of sensitivity and specificity.

Regarding the capacity of CK to discriminate between complicated and uncomplicated PAA, we found an AUC of 0.69 (95% CI 0.54–0.85) (p = 0.04) (Fig. 3). The cut-off point with the best Youden Index resulted in 1.1 mmol/L, with a sensitivity of 61.1% and a specificity of 76.9%.

Table 5 summarizes the alternative proposed cut-off points with their corresponding values of sensitivity and specificity.

The results of the correlation analysis between CK and both clinical and analytical variables is summarized in Table 6. We found weak correlations between CK and total neutrophil count (r = 0.22), number of emetic episodes (r = 0.25), presence of nausea (rho = 0.25), hours of pain evolution (r = 0.24) and serum creatinine (r =  – 0.28). Moderate correlations were found between CK and presence of fever at home (rho = 0.44), axillary temperature at the ER (r = 0.33), INR (r = 0.37), serum level of sodium (r =  – 0.46), days of admission (rho = 0.32) and CRP level (r = 0.54). No significant correlations were found between CK and days of hospitalization or post-surgical complications.