In the Asian population, diverticulosis affects approximately 25.1% of the population, accounting for 87.9% of the colonic diverticulosis cases involving the right side, which is significantly higher than in Western countries.[18] ARCD is reported to occur at a relatively young age and the overall prevalence of diverticulitis is 75% in the Asian population.[19, 20] This study population was limited to Asians and 81.1% of colonic diverticulitis cases were ARCD.
Our results showed that more surgical treatment was provided for AA, whereas conservative management was provided for ARCD, consistent with a previous study.[7] Therefore, distinguishing the two diseases will be important for determining a therapeutic plan and avoiding unnecessary surgery for ARCD patients. A classical tool for distinguishing appendicitis from other abdominal diseases is the Alvarado scoring system.[17] However, even appendicitis can show equivocal Alvarado scores and some studies revealed that ARCD patients had a higher or broader range of Alvarado scores.[21–23] Therefore, it may not be enough to use this scoring system as a tool to distinguishable between these diseases. Although the Alvarado scores showed a difference between the two groups in our single variable analysis, with a 6-point median value in AA and a 5-point median value in ARCD, it would be difficult to assign clinical meaning because scores of 5 to 6 have an equivocal probability for appendicitis.[24] Currently, few studies have reported clinical differences in the symptoms and signs (such as longer symptom duration associated with ARCD and nausea or vomiting, anorexia, migration pain, and RLQ pain associated with AA) between AA and ARCD.[12, 13, 21] Much of the differences in the symptoms are hypothesized to be due to the different pathophysiologies and elapsing course of the two diseases. Although both diseases have similar final symptoms due to localized peritonitis, appendicitis has a sequential reaction with prodromal symptoms due to the blockage and dilatation of the appendix first, then the increased intraluminal pressure results in wall necrosis.[25] Each subjective indicator might have a risk of bias by clinicians or patients. Otherwise, there are few studies that reported some objective factors (such as neutrophilia and high CRP) associated with AA than ARCD.[12, 13] Like most previous studies, we consistently found that a previous history of diverticulitis was a predictor of ARCD and anorexia and neutrophilia were predictors of AA.
An insufficient number of studies has reported the usefulness of leukocytosis for differentiating the two diseases and thus, it remains controversial.[21, 26] Shin et al.[12] reported that an elevated proportion of lymphocytes and a near-normal proportion of segmented neutrophils was seen in ARCD, even though no hypothesis was suggested to explain this phenomenon. Our results supported the relationship between neutrophilia and AA and higher fractions of lymphocytes were also related to ARCD. Sasaki et al. [13] categorized high serum CRP as > 3.0 mg/dL and reported that high serum CRP was associated with ARCD. However, our result did not support that finding. There was no statistical difference in our results between the AA and ARCD groups, even when serum CRP values over 3.0 mg/dL were used as criteria, as in the other study (39.0% vs 34.7%, p = 0.544). Serum CRP is known to peak after 48 hours due to its response to inflammation.[27] Our study and the study by Sasaki differed in the time interval from onset-to-visit in the ARCD patients (24 h vs. 48 h) and this could affect the blood sample collection time after infection. Therefore, it may not be appropriate to compare the two studies. Moreover, it has been revealed that serum CRP levels increase over time in appendicitis, questioning whether high CRP levels are more relevant to AA than ARCD if the confounding time factors related to the CRP increase are not removed.[28]
Our results revealed that ketonuria was one of the objective predictors of AA. We presented a boxplot for the distribution of urine ketone values between the two groups in Fig. 1. A few studies reported that ketonuria was seen in 12–45% of the AA cases, but there have been no comparative studies between AA and ARCD.[29, 30] Ketones are known as the end-product of fatty acid metabolism and ketonuria indicates that the body is excessively using fat over carbohydrates as the major source of energy.[31] Ketonuria could arise from dietary conditions, such as fasting, nausea and vomiting, and anorexia, and metabolic conditions, including type-1 diabetes, fever, and pregnancy.[32] In our study, there were no patients with type-1 diabetes or pregnant women and there was no difference in body temperature between the two groups. Therefore, we could suggest that the probable cause of ketonuria in appendicitis is related to dietary conditions such as anorexia because these conditions were found more in the AA group than in the ARCD group. When we made a simple rule for predicting ARCD that included objective clinical factors, such as no ketonuria and no neutrophilia and the presence of a history of diverticulitis, the calculated positive predictive value (PPV) was 87.5%, the negative predictive value (NPV) was 82.6%, and the specificity was 99.5%. Thus, this rule might be a useful tool for distinguishing ARCD from AA.
This study had some limitations. First, this study was conducted at a single center in an Asian population in the Republic of Korea. Therefore, the results may not represent all races and nations. However, the presentation of diverticulitis in the Asian population is unusual because it most commonly involves the right side of the colon and the exact pathological mechanism of diverticular disease is unclear, although several theories related to genetics, diet, motility, and microbiome that may be affected by individual races and cultures have been presented.[33] Therefore, this study had some clinical implications because not many studies have been reported in an Asian population. Second, because this study was a retrospective study rather than a confirmative study, data or cases might be missing. Specifically, for subjective symptoms and signs, it is difficult to accurately describe the intensity or presence without a proper prospective protocol. Third, ketonuria is roughly correlated with serum ketone concentrations, but the absence of ketonuria does not mean the absence of blood ketone bodies. However, we could not obtain the serum ketone concentrations to distinguish between AA and ARCD as that analysis was not included in the routine ER blood tests because it is more expensive than urine analysis and has not been proven useful in the diagnosis of these diseases. Further studies quantitative analyzing these factors should be conducted.