| Background:Acute pancreatitis (AP) is one of the most common diseases presented with acute abdomen, and the disease incidence has kept increasing during the last 30 years. Most of the patients are self-cured, but there are still about 20 to 30 percent of all the patients who suffer from a severe course of disease. Therefore, the total case fatality of AP still reaches to 5～10%, and that of the severe acute pancreatitis (SAP) is still higher(20～50%). The clinical laboratory has played an important role in diagnosing AP. The serum amylase(SAmy), urine amylase(UAmy) and lipase(Lip) are widely used in clinic, but the diagnostic sensitivity and specificity are not so satisfying. The prognosis of SAP is highly connected with the early diagnosis, but the score system widely used now in clinic(Ranson's score and APACHEⅡ) are of shortages of complication and inaccuracy. In the recent years, some researches have tried to determine the possibility of early diagnosis of AP and SAP by chemical markers, during which the trypsinogen-2(TPS-2) and C-reactive protein(CRP) are the most popular. However, the results seem not to agree with each other. No systematic review is available now about TPS-2 and CRP in diagnosing AP and SAP. Objective:Carry out the systematic review of TPS-2 and CRP in early diagnosing AP and SAP according to the evidence of all the corresponding clinical trials around the world. Provide the best evidence of formulating the clinical guidelines or performing the rule-in administration for Medical Administrative Command. Provide the best evidence of clinical decision for doctors and pathologists in diagnosing or auxiliary diagnosing AP and SAP according to TPS-2 and CRP.Methods:Retrieve the databases of MEDLINE, EMBASE, Chinese biomedical literature database(CBMDISC),and CNKI from 1970 to 2006, and ask for the full text literature from the authors. 57 articles about TPS-2 in diagnosing AP and 14 articles about CRP in diagnosing SAP were included according to the including criteria. The quality of the articles was evaluated by the tool of QUADAS and only the qualified articles were ruled in according to the evaluation results. Heterogeneity analysis and meta analysis were conducted by the software of Metadisc. SROC curves of TPS-2, SAmy, LIP and UAmy in diagnosing AP were built respectively, and the posterior probabilities were calculated. SROC curves for CRP in diagnosing SAP in patients of different admission time and with different cutoff values were built respectively. The results were compared with those of Ranson's score and APACHEⅡtest which were widely used in clinic nowadays.Results:Slight heterogeneity (P=0.0987, I~2=20.3%) was found in the results of 55 researches of TPS-2 in diagnosing AP, the summary sensitivity was 92.2% (95%CI: 91.1～93.2%), the summary specificity was 94.1% (95%CI: 93.4～94.7%), and the area under the SROC curve (AUC)was 0.9784, SE=0.0023.As for the 44 researches of TPS-2 and SAmy in diagnosing AP, moderate heterogeneity (P=0.0417, I~2=28.7%) was found in the former and no significant heterogeneity (P=0.1133, I~2=21%) was found in the latter. The summary sensitivity was 93% (95%CI: 91～94%) in the former and 81% (95%CI: 79～83%) in the latter. The summary specificity was 94% (95%CI: 93.0～95.0%) in the former and 86.0% (95%CI: 85.0～87.0%) in the latter. The SROC AUC was 0.9788 (SE 0.0058) in the former and 0.8965 (SE 0.0058) in the latter.As for the 39 researches of TPS-2 and UAmy in diagnosing AP, no significant heterogeneity (P=0.05641, I~2=0)was found in the former and moderate heterogeneity (P=0.0086, I~2=38.5%) was found in the latter. The summary sensitivity was 93% (95%CI: 92～94%) in the former and 78.0% ( 95%CI: 76.0～80.0%) in the latter. The summary specificity was 94% (95%CI: 94～95%) in the former and 82.0% (95%CI: 81.0～83.0%) in the latter. The SROC AUC was 0.9788(SE=0.0024) in the former and 0.8680(SE=0.0084) in the latter. The heterogeneity became insignificant (P=0.6152, I~2=0.0%) after exclude 5 highly heterogeneity researches, the SROC AUC was 0.8702(SE=0.0076), which had no significant difference with that before the exclusion.As for the 14 researches of TPS-2 and LIP in diagnosing AP, High heterogeneity (P=0.0009, 0.0013; I~2=62.8%, 61.6%) was found in both of the markers. The summary sensitivity was 88% (95%CI 85～90%) in the former and 84.0% (95%CI 81.0～86.0%) in the latter. The summary specificity was 93% (95%CI 92～95%) in the former and 88.0% (95%CI 86.0～90.0%) in the latter. The SROC AUC was 0.9704(SE=0.0090) in the former and 0.9134(SE=0.0133) in the latter. The heterogeneity became insignificant (P＞0.10) after exclude 3 highly heterogeneity researches, the SROC AUC was 0.8932 and SE=0.0113, which had no significant difference with that before the exclusion.The differences were all found to be significant when comparing the AUCs of SAmy,LIP,UAmy with that of TPS-2. The posterior probability of TPS-2 in diagnosing AP was 55% when the result turned to be positive, and was 0.69% when the result turned to be negative; while the posterior probabilities of SAmy and LIP in diagnosing AP were both 30% when the result was positive, and were both 1.96% when the result was negative; the posterior probability of UAmy was 28% when the result was positive, and was 2.38% when the result was negative.The summary sensitivity and the summary specificity were 66% and 75% for CRP in diagnosing SAP in patients one day after admission, which were both highly heterogeneity. The SROC AUC was 0.7894. The corresponding results changed to 83%, 68% and 0.8682 in patients two days after admission.When the cutoff value was defined as 100-130mg/L, the summary sensitivity and the summary specificity changed to 63% and 73% for CRP in diagnosing SAP in patients one day after admission, which were still highly heterogeneity. The SROC AUC changed to 0.7805. The corresponding results were 78%, 78% and 0.8496 in patients two days after admission.The diagnostic value of CRP in early prognosis of SAP in AP patients seemed to be better when comparing with that of APACHEⅡtest(the SROC AUC was 0.8007 and 0.6644 respectively).The result seemed to be of the same value when comparing with that of the Ranson's score(the SROC AUC was 0.7746 and 0.7749 respectively).Conclusion:Obvious defect was found in the researches about diagnostic test, especially those of the domestic researches. High diagnostic value was found in diagnosing AP with urine TPS-2, which was detected by the immunochromatographic method. TPS-2 detection was recommended to be the screen assay in diagnosing AP. Probability of about 99.31% could be achieved to exclude the diagnosis of AP when the assay result was negative. The diagnostic value were equal for the SAmy and LIP assay in diagnosing AP, which were widely used in clinic nowadays. SAmy assay was recommended according to the technical and economical factors. SAP can be prognosed efficiently in the AP patients with early admission(1～2days) by serum CRP assay. The application value could be great according to the convenient, cheap and easily-enforce of the CRP assay. |
PDF Full Text Request