| Objective Now the diagnosis of invasive aspergillosis mainly depended on the principles described by EORTC/IFICG and blood disease or cancer diagnostic criteria for invasive fungal infection and treatment principles in our country (revision). In the clinical diagnostic criteria, the principles mentioned the detection of antigen in the serum, but it didn't point out the specific method or the positive cut-off point. In abroad, the galactomannan (GM) determination by ELISA method was widely used. But in our country, the technology was not yet fully mature. In particular, the selection of positive cut-off point was in conclusive. In this study, our objective was to discuss the selection of cut-off point, compare with traditional diagnostic methods, refine and give supplement to the traditional criteria, observe the coincidence between the results and the clinical signs, estimate the clinical availability.Methods Blood samples were obtained from 47 patients in the first hospital of Jilin university from 2009.7~2010.2. All the patients were consistent with one of the factors following:①Neutropenia (neutrophil count less than 1×109/l);②Febrile patients were persistent fever after 4 days broad-spectrum antibiotic therapy. All the blood samples were obtained from 47 patients, male: 28, female: 19. Aged 14 to 74 years old, mean age was 44 years old. The proportion of various diseases: 1case of postoperative chemotherapy of lung cancer, 1case of postoperative chemotherapy of bladder cancer, 1case of postoperative chemotherapy of prostate cancer, 13 cases of acute lymphoblastic leukemia, 21 cases of acute myeloid leukemia, 2 cases of chronic leukemia, 2 cases of multiple myeloma, 2 cases of myelodysplastic syndrome, 2cases of lymphoma, 2 cases of cytopenia of unknown origin. Eliminating all the factors leading to false positive results, we collected the serum samples and tested I value with the kit by sandwich enzyme-linked immunosorbent assay (platelia aspergillus), sensitivity, specificity, predictive values and agreement rate were calculated according to the traditional criteria. At last, we estimated the availability of this test.Results1. Sensitivity, specificity and other values with different positive cut-off point : We selected 1.5, 1.0, 0.8, 0.5,0.3as positive cut-off point. With 1.5 as cut-off point, the sensitivity was 25%, specificity was 100%, false negative rate was 75%, false positive rate was 0, negative predictive value was 82.35%, positive predictive value was 100%, agreement rate was 83.33%. With 1.0, 0.8 as positive cut-off point, the sensitivity, specificity, false negative rate, false positive rate, negative predictive value, positive predictive value and agreement rate were the same with the results when selected 1.5 as cut-off point. When with 0.5 as cut-off point, the sensitivity was 50%, specificity was 100%, false negative rate was 50%, false positive rate was 0, negative predictive value was 87.50%, positive predictive value was 100%, agreement rate was 88.89%. When with 0.3 as cut-off point, the sensitivity was 75%, specificity was 57.14%, false positive rate was 42.86%, negative predictive value was 88.89%, positive predictive value was 33.33%, agreement rate was 61.11%.2. The difference of I value between clinical diagnostic and suspicious group. According to the blood disease or cancer diagnostic criteria for invasive fungal infection and treatment principles in our country (revision), by traditional diagnostic methods, confirmed case was 0, clinical diagnostic cases were 4, suspicious cases were 29, exclueded cases were 14. Combined with the GM test results when set 1.5, 1.0, 0.8, 0.5 as positive cut-off point, clinical diagnostic cases increased to 5, 6, 8, 13 respectively. The clinical diagnosis rate increased from 8.51% to 10.64%, 12.77%, 17.20%, 27.66% respectively. By traditional diagnostic methods, the mean value( x±S)of I value in the clinical diagnostic, suspicious and excluded group were 0.89±0.93 , 0.60±0.98 , 0.27±0.09; Combined with the GM test results when selected 1.5 as cut-off point, the mean value( x±S)of I value in the clinical diagnostic, suspicious and excluded group were 1.82±2.23, 0.42±0.30,0.27±0.09; Combined with the GM test results when selected 1.0 as cut-off point, the mean value( x±S)of I value in the clinical diagnostic, suspicious and excluded group were 1.74±2.00, 0.39±0.25 ,0.27±0.09;Combined with the GM test results when selected 0.8 as cut-off point, the mean value( x±S)of I value in the clinical diagnostic, suspicious and excluded group were 1.55±1.73, 0.34±0.18, 0.27±0.09; Combined with the GM test results when selected 0.5 as cut-off point, the mean value( x±S)of I value in the clinical diagnostic, suspicious and excluded group were 1.20±1.40, 0.26±0.10, 0.27±0.09. When used the traditional methods or traditional method combined with GM test results in which selected 1.5, 1.0, 0.8, 0.5 as positive cut-off point, we made t test about I value between the clinical diagnostic and suspicious group. There was no significant difference between these two groups(t test, P>0.05)by traditional methods. There was significant difference between these two groups (t test, P<0.05)when combined with the GM test results in which we selected 1.5, 1.0, 0.8, 0.5 as positive cut-off point.Conclusion 1.In this study, with 1.5 as cut-off point, the sensitivity was 25%, specificity was 100%, false negative rate was 75%, false positive rate was 0, negative predictive value was 82.35%, positive predictive value was 100%, agreement rate was 83.33%. With 1.0, 0.8 as positive cut-off point, the sensitivity, specificity, false negative rate, false positive rate, negative predictive value, positive predictive value and agreement rate were the same with the result when selected 1.5 as cut-off point. When with 0.5 as cut-off point, the sensitivity was 50%, specificity was 100%, false negative rate was 50%, false positive rate was 0, negative predictive value was 87.50%, positive predictive value was 100%, agreement rate was 88.89%. When with 0.3 as cut-off point, the sensitivity was 75%, specificity was 57.14%, false positive rate was 42.86%, negative predictive value was 88.89%, positive predictive value was 33.33%, agreement rate was 61.11%. Can be seen as a positive cut-off of 0.5, the sensitivity and the diagnostic rate increased, while the specificity could be maintained to 100%, the sensitivity was high. 2. There was no significant difference between he clinical diagnostic and suspicious group(t test,P>0.05)by traditional methods. There was significant difference between the clinical diagnostic and suspicious groups (t test,P<0.05)when combined with the GM test results in which selected 1.5, 1.0, 0.8, 0.5 as positive cut-off point. The results suggested that: the combination of traditional criteria and I value analysis of GM test had an important clinical significance for the judgments of clinical diagnostic group and suspicious group.
|
PDF Full Text Request