Delta neutrophil index as a prognostic marker in emergent abdominal surgery

Background Delta neutrophil index (DNI) is the fraction of circulating immature granulocytes provided by a routine, complete blood cell analyzer. It is known to be a useful prognostic marker of sepsis. The aim of this study was to evaluate the role of DNI in the diagnosis and prognosis of patients who had undergone emergent surgery for an acute abdomen. Methods A total of 694 patients who had visited the emergency room for acute abdominal pain and undergone emergent abdominal surgery from May 2015 to September 2016 were retrospectively reviewed. Clinical characteristics, laboratory findings on the day of hospital visit, hospital stay, postoperative complications, and 30‐day mortality were investigated. Results In the analysis of patients who had undergone an operation for acute peritonitis, the DNI was a good predictor for predicting 30‐day mortality rate (area under the curve [AUC]: 0.826). It was not inferior to other laboratory values, including activated partial thromboplastin time (AUC: 0.729), C‐reactive protein (AUC: 0.727), albumin (AUC: 0.834), prothrombin time (AUC: 0.816), and creatinine (AUC: 0.837) known to be associated with sepsis. Patients with high DNI displayed higher incidence of bacteremia and sepsis, longer hospital stay, higher postoperative complication rate, and higher 30‐day mortality rate than patients with low DNI. Among patients diagnosed with acute appendicitis, the DNI was a useful marker for differentiating appendiceal perforation. Conclusion The DNI was a practical and useful marker for predicting the prognosis of patients who needed emergent abdominal surgery.

intervention to prevent progression into poor outcome, although some cases require a few hours or days after admission to decide surgical management.
Delta neutrophil index (DNI) is the fraction of circulating immature granulocytes. It has been reported to be a useful prognostic marker of infection or inflammation. [3][4][5] DNI can be assessed as the difference between leukocyte subfraction determined by cytochemical myeloperoxidase reaction and leukocyte subfraction determined with nuclear lobularity assay by a reflected light beam using an automated blood cell analyzer. It is included in routine, complete blood count (CBC) tests. Along with inflammatory serologic markers, including white blood cell (WBC) counts, C-reactive protein (CRP), and procalcitonin, DNI serves as a diagnostic tool that can predict mortality in patients with sepsis, disseminated intravascular coagulation, and bacteremia. [6][7][8] In gastrointestinal diseases, increased DNI values are independently associated with mortality in patients with acute upper gastrointestinal bleeding. 9 DNI could differentiate perforated appendicitis from non-perforated appendicitis. 10 However, the clinical utility of DNI in patients undergoing emergent abdominal surgical procedures for acute abdomen has not been reported yet.
Thus, the objective of this study was to evaluate the role of DNI in diagnosing and predicting the prognosis of patients who underwent emergent surgery.

| Patients
Medical records of 896 patients who visited the emergency department for acute abdominal pain and underwent emergent abdominal surgery at the Department of Surgery, Hallym University Sacred Heart Hospital, Anyang, Korea, between May 2015 and September 2016 were retrospectively reviewed. Among these patients, 142 patients who were younger than 18 years old, 26 patients who had undergone surgery for trauma, 12 patients who had hernia surgeries, 10 patients who had perianal surgeries, six patients who had liver or kidney transplantation, three patients who had removed foreign bodies on the abdomen, and three patients who had been treated with cellulitis of the abdominal wall were excluded (Figure 1)

| Data collection and DNI measurement
Blood tests of all patients were evaluated in the first blood samples collected at the emergency department. CBC counts, chemistry, prothrombin time, activated partial thromboplastin time (aPTT), and CRP were calculated. DNI was examined with a blood cell analyzer (ADIVA 2120i, Siemens Inc, Forchheim, Germany).  However, postoperative ICU stay of the mortality group was significantly longer than that of the non-mortality group (8.2 days vs 1.0 days, P = 0.018).

| Comparison of clinical characteristics and laboratory findings in patients with acute peritonitis (Group I) between 30-day mortality and nonmortality groups
In laboratory findings, WBC and platelet counts were not significantly different between the two groups. Prothrombin time, aPTT, CRP, and creatinine of the mortality group were significantly higher than those of the non-mortality group while albumin was significantly lower in the mortality group. The DNI value was significantly higher in the mortality group than that in the non-mortality group (4.20% vs 1.46%, P = 0.036). Bold defined that P-value of variables was < 0.05.

| Receiver operating characteristic analysis for predicting 30-day mortality in Group I
We evaluated the relationship between 30-day mortality and labo-  The best cutoff level of DNI for the prediction of 30-day mortality in acute peritonitis was 0.9 or greater with a sensitivity of 100.0% and a specificity of 67.2%. ROC curves using variables are plotted in Figure 2.

| Comparison of clinical characteristics and laboratory findings between patients with high and low DNI
We divided subjects into two groups according to whether their DNI level was 0.9% or greater or <0.9%.

| Comparison of clinical characteristics and laboratory findings between patients with acute perforated appendicitis and those with nonperforated appendicitis
Clinical characteristics and laboratory findings were compared between 78 patients diagnosed with acute perforated appendicitis and 305 patients with acute non-perforated appendicitis (Table 4).
Patients with acute perforated appendicitis were older and maledominant. They had more frequent diabetes than patients with acute non-perforated appendicitis. The postoperative hospital stay of the perforated group was significantly longer than that of the non-per- In laboratory findings, WBC and platelet counts were not significantly different between the perforated and non-perforated groups.  DNI value of the perforated group was higher than that of the nonperforated group (1.36% vs 0.72%, P = 0.008).

| D ISCUSS I ON
The present study demonstrated that DNI value was correlated with severe infection and poor prognosis in patients with acute abdomen.
Patients with high levels of DNI (≥0.9%) displayed higher incidence of bacteremia and sepsis, longer hospital and ICU stay, and higher rate of postoperative complications than patients with low DNI levels (<0.9%). In addition, 30-day mortality was higher in patients with high DNI values. In acute peritonitis, DNI could predict 30-day mortality. It was not inferior to other laboratory markers associated with infection. In acute appendicitis, DNI value was a useful marker for appendiceal perforation. Our results indicated that DNI could be a useful tool for predicting severity and prognosis in acute abdomen.
According to recent studies, DNI is a predictive marker of histological chorioamnionitis in patients with preterm premature rupture of membranes. 13 A higher DNI is a prognostic marker of out-of-hospital cardiac arrest 14 and an independent factor of mortality in septic acute kidney injury patients with continuous renal replacement therapy. 15 Septic condition of patients who visited the emergency room is an important factor for predicting their prognosis and mortality. Therefore, there have been efforts to find proper biomarkers associated with sepsis. The utility of DNI value in patients with sepsis and bacteremia has been reported in several studies. [6][7][8]16,17 In a previous study, the DNI value was used as an early marker of disease severity in critically ill patients with sepsis. 18 However, in another study, the use of DNI for predicting bacteremia or sepsis was limited to immunocompromised cases. 19  The first limitation of the present study was that it was retrospective in nature with subjects from in a single center. Such limitation might have resulted in selection bias. Second, many infectious diseases of the abdomen were included in the group of acute peritonitis. Because the severity and activity were diverse for these diseases, assessments of acute peritonitis might not have been consistent. Third, DNI was only measured upon arrival at the emergency room. Serial changes in DNI values according to aggravation or improvement of the infection were not examined in this study. Fourth, other inflammatory serology markers such as ESR or procalcitonin were not evaluated in the present study. These markers were not routinely checked in our hospital.
In conclusion, DNI is a valuable prognostic marker in patients who visited the emergency room complaining of acute abdominal pain. Patients with DNI level of 0.9% or greater who needed emergent abdominal surgery or required surgical intervention for acute peritonitis should be monitored closely with appropriate treatment strategies. DNI could be helpful for selecting high-risk patients and deciding therapeutic modalities such as emergent operation or intensive care unit treatment.