The aim of this study is to investigate the value of C-reactive protein (CRP) and of other laboratory parameters obtained during the initial evaluation of the patient in the prediction of the clinical severity of acute diverticulitis. The records of patients treated for acute diverticulitis at the Oulu University Hospital from December 2006 to December 2008 were retrospectively reviewed. Mild disease was defined when conservative treatment was successful. Severe acute diverticulitis was considered when percutaneous drainage of an abscess and/or surgery was necessary. From the 182 patients considered for analysis, 158 (87%) had mild disease, whereas 24 (13%) had severe. CRP (P = 0.034) and the Hinchey classification (P = 0.006) were shown to be independent risk factors for severe acute diverticulitis in the logistic regression analysis. The receiver operating characteristic curve showed that a CRP cutoffvalue of 170 mg/L significantly discriminated severe from mild diverticulitis (87.5% sensitivity, 91.1% specificity, area under the curve 0.942, P < 0.00001). CRP is a useful tool in the prediction of the clinical severity of acute diverticulitis. A mild episode is very likely in patients with CRP less than 170 mg/L. Those with higher CRP values have a greater probability to undergo surgical treatment or at least a percutaneous intervention.
ACUTE DIVERTICULITIS OF the colon is a common condition in Western societies and it is caused by the development of inflammation resulting from perforation of a diverticulum. Epidemiologic data show that the rate of hospitalization for acute diverticulitis has increased since the mid-1990s and accordingly the disease constitutes a substantial economic burden for health systems.1, 2
Conservative treatment is successful in most of the cases of acute diverticulitis.2 However, some patients will present with an attack that will eventually necessitate an emergent or semiurgent intervention at the same hospitalization such as percutaneous drainage of an abdominal abscess and/or surgery. Early identification during the evaluation in the emergency department of those patients who will subsequently develop severe disease is essential for decision-making and constitutes a challenge for the physician.
Computed tomography (CT) is an important tool for the confirmation of the diagnosis of clinically suspected acute diverticulitis.3 Moreover, imaging-severity criteria including the Hinchey classification and its successive modifications4, 5 have been shown to assist in the prediction of the clinical course of the disease.6 However, in a notable group of patients, there is not a precise correlation between the clinical severity and the CT grading of acute diverticulitis.7, 8
In addition, research concerning the association of inflammation markers or other laboratory factors with the clinical grading of acute diverticulitis is yet incomplete. A study has reported that initial C-reactive protein (CRP) value is significantly increased in patients with a clinically severe episode.9 Nevertheless, other studies have mainly focused on the correlation of CRP values with the CT grading of acute diverticulitis.10, 11
The purpose of this research is to investigate the value of CRP and of other laboratory parameters obtained during the evaluation of the patient at the emergency department in the prediction of the clinical severity of acute diverticulitis.
Patients and Methods
This retrospective cross-sectional study was carried out at the Oulu University Hospital, Oulu, Finland. All patients with a discharge diagnosis of acute diverticulitis during the period starting from December 1, 2006, to December 31, 2008, were considered for review. The patients were identified using a computergenerated search from a register that contains prospective data concerning diagnoses coded according to the International Classification of Diseases, 10th Revision.
Electronic records were reviewed to retrieve the characteristics and laboratory values at the initial evaluation of the patients, the diagnostic methods, the number of previous admissions resulting from acute diverticulitis, the abdominal or systemic complications, the specific operative or interventional procedures, and mortality. The serologic markers obtained at the presentation of the patient in the emergency department were recorded and comprised the CRP, white blood cells (WBCs), creatinine, and albumin. The modified Hinchey classification was used to stratify the imaging severity of the disease in those patients with an initial CT scan (Table 1).5 Distinction between Hinchey III and IV was based on surgical reports of patients who underwent laparotomy as a result of generalized peritonitis.
Patients who presented with a CT showing acute diverticulitis or with leftlower abdominal pain, fever, leukocytosis, CRP values over 10 mg/L, history of known diverticulosis, and existence of diverticula in the follow-up colonoscopy were included in the study. Individuals who had a concomitant disease and/or were receiving drugs that potentially affect inflammation markers' serum levels were not included in the analysis. Therefore, a past or present history of any cancer or autoimmune disease, detection of neoplastic disease of the large bowel in the follow-up colonoscopy, liver dysfunction, and glucocorticoid therapy were among the exclusion criteria. Finally, patients with chronic symptoms of diverticulitis were not considered for this research.
The therapeutic protocol for acute diverticulitis in our institution followed the practice parameters suggested by the guidelines of the American Society of Colon and Rectal Surgeons.12 Therefore, the indication for emergent surgery consisted of signs of obvious generalized peritonitis, whereas disease unresponsive to initial conservative treatment was subsequently addressed with percutaneous intervention and/or surgery. Accordingly, patients who did not show clinical improvement within 48 to 72 hours from admission underwent a primary, if not obtained earlier, or a sequential CT scan to document imaging stage and to guide therapeutic decisions. In this series, all patients received antibiotic therapy. Follow-up colonoscopy or CT colonography confirming the presence of diverticulosis was performed at least in those who did not undergo laparotomy. Late complications, elective operations, and clinical recurrence of the disease were retrieved as well.
In this study, acute diverticulitis was classified as mild when conservative treatment was successful and the patient did not necessitate rehospitalization as a result of recurrence or complication of the disease within 30 days of discharge. A severe attack was designated when percutaneous CT-or ultrasound-guided drainage and/or surgical intervention were necessary.
The statistical software package SPSS 17.0 was used for the analysis. Risk factor analysis was performed using exclusively the baseline laboratory and CT imaging data obtained at the presentation of the patients at the emergency unit. The association between categorical variables was assessed with the Fisher's exact test and the x2 test. Means were compared with the analysis of variance. Receiver operating characteristic (ROC) curve was performed to detect the best prognostic CRP cutoffvalue based on the maximum specificity and sensitivity sum. Standardized cutoffvalues were applied for age (75 years), WBC (15.0 109/L), creatinine (150 mg/L), and albumin (35 g/L). A P value < 0.05 was considered significant. Statistically significant variables from the univariate analyses were used in the logistic regression method to depict independent risk factors for severe disease and for secondary end points.
One hundred eighty-two patients with acute diverticulitis were evaluated. Demographic, clinical, and laboratory data are depicted in Table 2.
Severity of Acute Diverticulitis
One hundred fifty-eight patients were treated conservatively, whereas 24 patients underwent surgical or percutaneous intervention. The laparotomies included 16 Hartmann resections, three sigmoid resections with primary anastomosis ± prophylactic ileostomy, and one laparotomy for peritoneal lavage and drainage. In three patients, percutaneous drainage was performed.
The ROC curve analysis showed that a CRP cutoffvalue of 170 mg/L significantly discriminated severe from mild diverticulitis with 87.5 per cent sensitivity and 91.1 per cent specificity (area under the curve 0.942; standard error 0.027; 95% confidence interval, 0.888 to 0.996; P value < 0.00001) (Fig. 1). The scatterplot of CRP values in patients with clinically mild and severe disease is depicted in Figure 2.
Univariate analysis showed that CRP greater than 170 mg/L (P 4 0.000), Hinchey classification greater than I (P < 0.0001), age older than 75 years (P < 0.0001), guarding (P 4 0.005), WBC greater than 15.0 109/L (P < 0.0001), serum albumin less than 35 g/L (P40.001), and serum creatinine greater than 150 mg/L (P < 0.0001) were related to severe acute diverticulitis.
After inclusion of the previous variables in the logistic regression model, CRP (P 4 0.034) and the Hinchey classification (P 4 0.006) were shown to be independent risk factors for severe acute diverticulitis.
Duration of Hospital Stay
In univariate analysis, CRP greater than 170 mg/L (P < 0.0001), WBC greater than 15.0 109/L (P < 0.0001), age older than 75 years (P < 0.0001), serum albumin less than 35 g/L (P < 0.0001), serum creatinine greater than 150 mg/L (P 4 0.009), and the Hinchey classification (P < 0.0001) were related to increased length of in-hospital stay.
Including the previous parameters in the logistic regression model, CRP greater than 170 mg/L (P 4 0.048) and the Hinchey classification (P 4 0.035) were shown to be independent predictors for increased in-hospital stay.
Univariate analysis showed that CRP greater than 170 mg/L (P < 0.0001), WBC greater than 15.0 109/L (P 4 0.012), age older than 75 years (P 4 0.003), serum albumin less than 35 g/L (P 4 0.029), serum creatinine greater than 150 mg/L (P < 0.0001), and the Hinchey classification (P < 0.0001) were related to increased extraintestinal complication rate.
However, analyzing the previous risk factors with the exclusion of the Hinchey classification in the logistic regression model, serum creatinine greater than 150 mg/L (P40.033) was shown to be an independent predictor for increased extraintestinal complication rate.
In univariate analysis, death was affected by CRP greater than 170 mg/L (P 4 0.04), age older than 75 years (P40.014), serum albumin less than 35 g/L (P4 0.003), serum creatinine greater than 150 mg/L (P < 0.0001), and the Hinchey classification (P40.003). The logistic regression analysis did not find any independent predictor for death.
Discussion and Conclusions
This study shows that CRP levels obtained during the initial evaluation of a patient presenting at the emergency department with acute diverticulitis may assist in the prediction of the clinical severity of the episode. Patients with a CRP greater than 170 mg/L have a greater probability to experience severe disease and may necessitate surgical treatment or at least a CTor ultrasound-guided percutaneous intervention to drain an abdominal abscess. Conversely, patients with lower levels of CRP are very likely to undergo a mild attack of acute diverticulitis and therefore antibiotic therapy or supportive care will suffice.
Accordingly, two previous reports showed that the CRP levels of patients with acute diverticulitis who required percutaneous or surgical intervention were significantly higher compared with those of patients with a mild attack.9, 10 However, these reports did not provide a suggestion concerning the best predictive value of CRP that could discern between the two aforementioned groups of patients.
A similar difference like in our study between the CRP values of patients with severe acute diverticulitis and those with mild disease was therefore observed in a prospective series of 42 patients conducted by John et al. (281 vs 58 mg/L, respectively).9 The authors state that the CRP samples were obtained within 4 hours from admission. The corresponding CRP values provided by the retrospective analysis of 247 patients performed by Käser et al. were 172 versus 101 mg/L, respectively.10 A shortcoming of this study is that the sample of the CRP considered for analysis was the latest obtained during an interval of 48 hours before the CT scan for each patient. This may have created some degree of bias as a result of a dishomogeneous pool of CRP values because there is evidence that CRP levels in patients with acute diverticulitis show significant variation within this time range.13
It is worth mentioning that most studies in the literature focus on the association of the CRP level with the CT severity in patients with acute diverticulitis. In a recent retrospective research of 426 patients, it was shown that those presenting with a Hinchey class over Ia had significantly higher CRP and a cutoffvalue of 175 mg/L was suggested for CT-complicated disease.11 Unfortunately, further information concerning the clinical outcome of the patients was not provided. In the report of Käser et al., patients with CRP less than 50 mg/L were likely to have a confined pericolic inflammation, whereas those with CRP greater than 200 mg/L belonged to a Hinchey class greater than Ia.10 Lastly, in a study of 50 patients with acute diverticulitis CRP, WBC, fibrinogen levels as well as the clinical presentation differed between various CT stages.14
In this study, it is confirmed that the Hinchey classification plays an important role in the prediction of the clinical severity of the disease. CT scan is an established modality to appreciate the risk of percutaneous drainage and/or surgery in patients with acute diverticulitis. Nevertheless, various reports including this study suggest that a substantial number of patients with CTsevere disease will not necessitate any kind of intervention during the hospital stay for the initial acute episode.7, 8, 15 In our series, half of the patients with Hinchey II disease were treated with antibiotics alone.
This report has some considerations that should be taken into account. An initial CT scan for the diagnosis of acute diverticulitis was not performed in a substantial number of cases. The rest of the patients had clinical features16 and a history strongly suggestive of acute diverticulitis. However, in case of a doubtful differential diagnosis including acute appendicitis, urologic, and gynecologic disease, an abdominal ultrasound and/or a subsequent CT was obtained to confirm acute diverticulitis. Furthermore, as a result of the retrospective nature of this study, we cannot disregard the possibility that some patients with mild acute diverticulitis may have been erroneously classified and treated as having severe disease. In such case, the actual CRP cutoffvalue that discriminates a mild from a severe attack may be higher than the one proposed here. However, the latter consideration may not decisively influence the negative predictive value for severe disease of CRP values below 170 mg/dL or indeed it would increase it.
In conclusion, our study shows that the CRP level at the initial evaluation of the patient may be helpful in the prediction of the clinical severity of acute diverticulitis. CRP is inexpensive and simple to use in clinical practice. Therefore, an early CRP sample should be ordered as a routine. Patients with a low CRP can be treated in primary care, whereas those with CRP greater than 170 mg/L should necessarily undergo a CT scan and be referred for hospitalization in a tertiary healthcare unit. Further investigation to better define the role of CRP in relation to the clinical outcome and potentially to improved management of the disease as well as prospective validation of these results is needed.
1. Etzioni DA, Cannom RR, Ault GT, et al. Diverticulitis in California from 1995 to 2006: increased rates of treatment for younger patients. Am Surg 2009;75:981-5.
2. MäkeläJ, Kiviniemi H, Laitinen S. Spectrum of disease and outcome among patients with acute diverticulitis. Dig Surg 2010; 27:190-6.
3. Cho KC, Morehouse HT, Alterman DD, et al. Sigmoid diverticulitis: diagnostic role of CT-comparison with barium enema studies. Radiology 1990;176:111-5.
4. Hinchey EJ, Schaal PG, Pischards GK. Treatment of perforated diverticular disease of the colon. Adv Surg 1978;12: 85-109.
5. Klarenbeek B, de Korte N, van der Peet DL, et al. Review of current classifications for diverticular disease and a translation into clinical practice. Int J Colorectal Dis 2012;27:207-14.
6. Lohrmann C, Ghanem N, Pache G, et al. CT in acute perforated sigmoid diverticulitis. Eur J Radiol 2005;56:78-83.
7. Ambrosetti P. Value of CT for acute left-colonic diverticulitis: the surgeon's view. Dig Dis 2012;30:51-5.
8. Brandt D, Gervaz P, Durmishi Y, et al. Percutaneous CT scanguided drainage vs. antibiotic therapy alone for Hinchey II diverticulitis: a case-control study. Dis Colon Rectum 2006;49:1533-8.
9. John SKP, Teo NB, Forster AL. A prospective study of acute admissions in a surgical unit due to diverticular disease. Dig Dis 2007;24:186-90.
10. Käser SA, Fankhauser G, Glauser PM, et al. Diagnostic value of inflammation markers in predicting perforation in acute sigmoid diverticulitis. World J Surg 2010;34:2712-22.
11. van de Wall BMJ, Draaisma WA, van der Kaaij RT, et al. The value of inflammation markers and body temperature in acute diverticulitis. Colorectal Dis 2013;15:621-6.
12. Rafferty J, Shellito P, Hyman NH, et al. Standards Committee of the American Society of Colon and Rectal Surgeons. Practice parameters for sigmoid diverticulitis. Dis Colon Rectum 2006;49:939-44.
13. Elsing C, Ernst S, StremmelW. Value of lipopolysaccharide binding protein, interleukin-6 and C-reactive protein as biomarkers of severity in acute diverticulitis: a prospective study. Clin Lab 2012;58:145-51.
14. Tursi A, Brandimarte G, Giorgetti G, et al. The clinical picture of uncomplicated versus complicated diverticulitis of the colon. Dig Dis Sci 2008;53:2474-9.
15. Kumar RR, Kim JT, Haukoos JS, et al. Factors affecting the successful management of intra-abdominal abscesses with antibiotics and the need for percutaneous drainage. Dis Colon Rectum 2006;49:183-9.
16. Laméris W, van Randen A, van Gulik TM, et al. A clinical decision rule to establish the diagnosis of acute diverticulitis at the emergency department. Dis Colon Rectum 2010;53:896-904.
ARISTOTELIS KECHAGIAS, M.D.,* TERO RAUTIO, M.D., PH.D.,* GEORGIOS KECHAGIAS, M.D.,[dagger] JYRKI MÄKELÄ, M.D., PH.D.*
From the *Department of Surgery, Division of Gastroenterology, Oulu University Hospital, Oulu, Finland; and the [dagger]Second Department of Gastroenterology, Evaggelismos Hospital, Athens, Greece
Address correspondence and reprint requests to Aristotelis Kechagias, M.D., Kountourioti 20, Neo Iraklio Attikis, 141-22, Greece. E-mail: email@example.com.