ABSTRACT
This article reviews the emerging artificial intelligence (AI) technology and its application in the field of aneuploidy prenatal screening from the aspects of the research methods of AI, the status of prenatal screening, and the role of AI in the integration of the screening markers, improvement of the screening performance, and optimization of screening strategy, etc. There is no doubt that AI has great potential in improving the ability of disease prediction through, integrating various screening data, discovering additional value of the data, and reducing social medical expenses. However, AI technology should be viewed and used in a scientific, rational, and comprehensive way to achieve the ideal effect in the field of prenatal screening.
ABSTRACT
Objective:To investigate the use of the reference intervals for blood cell counting and the reference of industry standard in China.Methods:Information from all laboratories was collected using online questionnaire in 18 reference intervals survey in blood cell counting in 2019. The information includes the source of the reference intervals, the verification of the reference intervals, and the upper and lower limits of the reference intervals, the method used, the instrument, the reagent and the calibrator. Microsoft Excel 2007 software was used to analyze the results of all laboratories. The median and 95% confidence interval were calculated. The distribution of the reference intervals for blood cell counting and their conformance to industry standards were analyzed.Results:2, 869 labs reported the data. The main sources were industry standards and National Guide to Clinical Laboratory Procedures. The proportion was 33.30%-35.02% and 28.55%-30.90% respectively. 49.44%-55.13% of laboratories validated the reference interval when citing industry standards. The reference interval grouping of most laboratories (89.37%-91.69%) cited in RBC, Hgb and Hct were consistent with the industry standards. We compared the upper and lower limits of the reference intervals with that given by the industry standards, when the lower limit of the reference intervals of mean corpuscular hemoglobin concentration, absolute neutrophils count, absolute basophils count, absolute monocyte count, and lymphocyte percentage were compared. The upper limit of reference intervals of neutrophils percentage as well as upper and lower limits of reference intervals of mean corpuscular volume, mean corpuscular hemoglobin, absolute eosinophil count, basophils percentage, and monocyte percentage were also compared. The median and mode were equal and consistent with industry standards. For other labs, the upper and lower limits of the reference intervals were not consistent with the reference intervals given by the industry standards.Conclusion:The use of reference intervals for blood cell counting was not the same, and the implementation of industry standards was not optimistic. A considerable number of laboratories had not verified the reference intervals, so it was necessary to promote the industry standards for reference intervals.
ABSTRACT
Objective To eva1uate the external quality assessment results of prenatal screening for maternal serum inhibin A in the second trimester in 2018 and to improve the accuracy of prenatal screening.Methods National Center for Clinical Laboratories provided three batches of quality control urine sample (Lot:201811-201813) to 94 prenatal screening laboratories nationwide in March 2018.Laboratories participated in the assessment voluntarily and reported the results,methods,equipment,reagents and other related information as required.Clinet EQA and Microsoft Excel 2010 were used for statistical analysis of the laboratory test results and for descriptive evaluation of the accuracy rate.Results A total of 55 laboratories submitted their testing results giving a return rate of 58.5％ (55/94),of which 52 (94.5％) were consistent with the expected results,while none of the results submitted by the other three laboratories was accurate.At the mean time,the bias of all three batches in each laboratory fell into the same side (two laboratories showed negative bias and one positive bias).Conclusions The results of the external quality assessment of prenatal screening for maternal serum inhibin A are generally satisfactory except for a few laboratories.It is necessary to incorporate prenatal screening for maternal serum inhibin A in the second trimester into the formal external quality assessment plan and regularly monitor the level of its detection quality.
ABSTRACT
Objective To investigate and analyze the source and upper and lower limits of the reference interval of children's complete blood count in 110 maternal and child health hospitals and chil-dren's hospitals nationwide. Methods Laboratories submitted the data of reference intervals via external quality assessment (EQA) software which was based on the web. To collect the results of reference intervals questionnaires on complete blood cell counting of children in laboratory departments of 110 maternal and child care service centers and children's hospitals in China in 2017. Questionnaires include information on the source of reference intervals for 18 items of complete blood count, whether to verify, upper and lower lim-its, grouping, methods used, instruments, reagents, and calibrators. Data was analyzed using Microsoft Excel 2007 and SPSS 22.0. The median, P25, P75were obtained, and rank sum test were used to determine wheth-er there were statistical differences between groups. Results The results of 110 laboratories were obtained after rejecting invalid data. The reference intervals were mainly derived from operating procedures and the laboratories themselves, of which 50.5％-53.6％ of the laboratories were validated. The white blood cell counting reference intervals gradually decreased from birth to adolescence, and the value was close to that of adults. The reference intervals of red blood cell counting and hemoglobin were close to that of adults except in the neonatal period. The value of the reference intervals of hematocrit slightly decreased with age. The dif-ference between RBC, Hgb, HCT in reference intervals between groups was statistically significant when grouped by gender(P<0.05). The reference intervals of white blood cell counting were less grouped by gender (5 laboratories), and the difference among groups was not statistically significant. There were no statistically significant differences in the reference intervals of elements between the two measurement systems that Sys-mex XN series and Sysmex XS-800i/XS-1000i/XS 500i/XS 900i series. Conclusions The establishment of reference intervals for children's complete blood cell counting was urgently needed. The reference inter-vals of the complete blood cell counting item had statistical significance in both age and gender. It was sug-gested that the industry standard of children's complete blood cell counting reference intervals should be es-tablished based on age and gender.
ABSTRACT
Objective To investigate the status of blood specimen acceptability for clinical chemistry tests in routine medical laboratories of China. Methods The questionnaires were assigned to the laboratories which participated in the routine chemistry exter-nal quality assessment (EQA) programs proposed by National Health Commission for Clinical Laboratory. The questionnaires included general information of participants and information about unacceptable blood specimens. Participants were required to record all the in-formation concerning unacceptable blood specimen received from 1stto 31stJuly, 2017. The data from each laboratory were reported and collected via special online system.Results A total of 866 valid questionnaires were collected.Of 15 981 752 specimens received dur-ing the data collection period unqualified 122 00 specimens were rejected with overall rejection rate of 0.076%. The main reasons for unacceptable specimens were hemolysis (33.98%), insufficient specimen quantity (10.78%) and chylemia/lipemia (10.62%). The rejected specimens were related to the original laboratories, types of container and specimen, transportation manner and operating staff of blood collection. Conclusion Certain problems existed in the receiving and management system for unqualified blood specimen in our country and remaining to be perfected. The clinical laboratories should pay more attention for pre-examination stage, including routinely monitoring unacceptable specimens, analyzing related data at the most possible granular levels, identifying the main problem and taking effective measures.
ABSTRACT
Objective To investigate the implementation of professional standards for reference intervals in routine biochemistry laboratories.Methods The relevant data of reference intervals from different laboratories including upper and lower limits,sources of reference intervals,verification and grouping rules,etc was submitted by using the software of interval quality assessment (EQA) via WEB application.The background monitor saved all the data as Microsoft Excel 2007 documents.The implementation of reference intervals of professional standards for decreed 18 routine biochemical tests was analyzed.Results The proportion of laboratories in which the reference intervals were derived from professional standards increased largely in 2016 compared with those in 2014.The data of 2016 showed that the reference intervals derived from professional standards were verified in 58.9％ to 67.5％ of laboratories before they were used in clinical diagnosis.The grouping rules for reference intervals in most laboratories (59.1％ to 83.3％) were in accordance with professional standards,except for individual items,including urea (43.4％),creatinine (40.1％) and alkaline phosphatase (35.8％).There were significant differences for the upper and lower limits of the most items between the laboratories using professional standards and those without using professional standards (P ＜ 0.05),except for some items,including lower limit value of kalium,upper limit of phosphorus,upper limit of amylase,upper limit of aspartate aminotransferase and lower limit of ferrum.There was no significant difference in most items between the the upper and lower limits of the reference intervals in the laboratories using professional standards and the reference intervals defined by professional standards,except for individual items,including upper limit of total protein,lower and upper limit of creatine kinase,upper limit of urea and upper limit of creatinine (P ＜ 0.05).Conclusion The implementation of reference intervals from professional standards in most routine biochemistry laboratories could not be satisfactory.Only a small part of laboratories used the professional standards,and the reference intervals of professional standards,were not verified before use in quite a part of laboratories.
ABSTRACT
Objective To evaluate the results of 2017 external quality assessment for newborn hemoglobinopathyand improve the quality of disease screening .Methods Each of 26 participating laboratories testing newborn hemoglobinopathy across the country received 5 batches of quality control blood spots ( Lot 201711-201715 ) in octorber 2017.Laboratories voluntarily participated in the survey and reported the results, methods, equipments and reagents information .Clinet EQA, and Microsoft Excel 2010 were used to perform statistical analysis on the laboratory test results .The rates of accuracy ( number of correct results/total number of submitted results ) were used for evaluating the performance of laboratories . Results 24 laboratories submitted the testing results with a return rate of 80.8%(21/26).The rates of accuracy for each lot were 100%(21/21), 90.5%(19/21), 90.5%(19/21), 57.1%(12/21) and 100%(21/21 ) respectively.Conclusions The results of this external quality assessment for newborn hemoglobinopathy is generally satisfactory , except for HbBarts′and HbA2.The screening laboratories should improve their quality control system , take timely measures to correct mistakes during the analytic period and improve the accuracy of screening tests for newborn hemoglobinopathy.
ABSTRACT
The aims of this article is to provide the quality control requirements of preanalytical variation for the determination of trace element aluminum,arsenic,chromium,cadmium and mercury in samples of human origin,reduce the influence of preanalytical variation on the test results.Refer to the Clinical and Laboratory Standards Institute (CLSI) documents,Control of Preanalytical Variation in Trace Element Determinations and other references and guidelines,the methods of quality control of aluminum,arsenic,chromium,cadmium and mercury determination have been made,including:the factors needed to be considered in collection,preservation,transportation and other preanalytical factors,the abilities and considerations of laboratory staff,etc.Hope this article provide some useful suggestions and help to the laboratories of determination of aluminum,arsenic,chromium,cadmium and mercury in samples of human origin.
ABSTRACT
Objective To analyze the reference intervals of serum total bilirubin (TBIL) and serum direct bilirubin (DBIL) by all Chinese clinical laboratories and make a comparison with the upcoming part 4 of Industry Standard WS/T 404.Methods Relevant information about reference intervals of all clinical laboratories participating in TBIL and DBIL testing items of 2014 national external quality assessment scheme of clinical routine chemistry was collected by a web-based external quality assessment software,including source,grouping,verification,upper and lower limits of reference intervals and instruments,reagents,methods and calibrators used.Microsoft Excel 2010 and SPSS 19.0 were applied to statistical analysis.The comparison between reference intervals used and the upcoming Industry Standard was conducted by the simple mean t test.Laboratories derived from the highest percentage of source distribution whose reagent and instrument were matching with each other were grouped according to the test systems used and the differences of reference intervals between three mainly used test systems and the upcoming Industry Standard were also compared with the single sample mean t test.Results The number of laboratories participated in the investigation about source distribution of reference intervals for TBIL and DBIL was 749 and 709 respectively.For these two items,the highest sources were both instructions of reagent(TBIL 58.08％,DBIL 58.67 ％),next were both National Clinical Laboratory Operation Rules (3rd Edition) (TBIL 29.64 ％,DBIL 28.91 ％),the percentages of other sources were all less than 10％.Besides,there were respectively 379 (50.60％) and 354 (49.93％) laboratories verifying their reference intervals for TBIL and DBIL.The difference of the comparison between reference intervals used and the upcoming Industry Standard had statistical significance (P＜0.05).The using reference intervals were narrower than the upcoming Industry Standard for both items.In all laboratories with reference intervals from instructions of reagent,themating laboratories respectively accounted for 41.88％ and 41.48％％ for item TBIL and DBIL.As the single sample mean t test showed,the comparison of the reference intervals between three mainly used test systems in these mating laboratories and the upcoming Industry Standard had statistical significance (P＜0.05),except for the lower limit of DBIL with Beckman test system (P value was 0.068) and the upper limit of DBIL with HITACHI test system (P value was 0.087).Conclusion Currently,the using situation of reference intervals about TBIL and DBIL by all Chinese laboratories was not scientific and rational enough and has significant difference with the upcoming Industry Standard.Should publish the part 4 of Industry Standard WS/T 404 as soon as possible,which would help clinical laboratories establishing suitable reference intervals and promote the standardization of its usage.
ABSTRACT
Objective To investigate the dates of 477 Chinese prenatal screening centers for previous half year analyse prena-tal screening status and provide recommendations for quality control.Methods All China prenatal screening centers were sent the data via the National Quality Assessment Scheme.This covered the software used,the risk cut-offs,monthly sample throughput,monthly median MoM of AFP,hCG,β-HCG,freeβ-HCG and uE3,monthly screening positive rate for trisomy 21,trisomy 18 and Open Neural Tube Defect.Results Screening protocol were versatile,with 73.48% (133/181)used two-marker model,24.31% (44/181)used three-marker model and 2.21% used four-marker model.About the software used, 350 laboratories never updated the screening parameters,89 laboratories had updated their median or parameter by manufac-turers,24 laboratories had updated the parameters by themselves.Cut-offs differed between laboratories.59.91% (275/459) used 1/270 as their cutoffs for trisomy 21.66.22% (296/447)used 1/350 as their cutoff for trisomy 18 and 96.52% (361/374)used cutoffs between 2.0~2.5 MOM for ONTD.Results of monthly median MOM:the percentage of laboratories that all six monthly median MOM within the target of 0.90~1.10 was 46.69% (155/332)for AFP,20.0% (4/20)for hCG, 29.17% (28/96)forβ-HCG,15.66% (31/198)for freeβ-HCG and 4.82% (11/228)for uE3.The percentage of laborato-ries that all six monthly median MOM within the target of 0.95~1.05 was 14.16% (47/332)for AFP,0% (0/20)for hCG,4.17% (4/96)forβ-HCG,12.63% (25/198)for freeβ-HCG and 4.82% (11/228)for uE3.About screening positive rate,there were difference of trisomy 2 1 positive rate in the same laboratory within the six month.Conclusion There were variations types of screening protocol,different kinds of soft platform,randomness of choice or update of medians or other important parameters and great difference in awareness of quality control.It would be taken effective and practical quality control measures to help laboratories improve quality control.
ABSTRACT
Eight newborn screening quality indicators were developed though expert discussing organized by National Center for Clinical Laboratories according to NBS quality management in USA .These indicators can not only be used to evaluate the performance throughout critical aspects of pre -examination, examination and post-examination processes for routine tests , but also be applied to monitor a series of processes from specimen collecting to results interpretation in newborn screening .Meanwhile, NCCL can use these quality indicators to carry out External Quality Assessment programs in order that the total service quality of newborn screening can be improved.
ABSTRACT
Objective Discussthe reference intervals and their sources of lymphocyte subsets by flow cytometry tests by all Chinese clinical laboratories at 2014 with a comparison of difference between different test systems.Methods Questionnaires about reference intervals were distributed to 212 clinical laboratories participating in October 2014 national external quality assessment scheme of lymphocyte subsets by flow cytometry tests of National Center for Clinical Laboratory.Relevant information about reference intervals of 5 relative counting items CD3 + (％),CD3 + CD4 + (％),CD3 + CD8 + (％),CD3-CD16 + CD56 + (％) and CD3-CD19 + (％) was collected by a internet-based external quality assessment software systems.Microsoft Excel 2010 and SPSS 19.0 were used to sort out raw data,eliminate unreasonable or incorrect data,and only analyze source distribution,arithmetic average,median,minimum value,maximum value and percentiles (P2.5 and P97.5) of upper and lower limits about healthy adults' reference intervals.Laboratories derived from the highest percentage of source distribution whose reagent and instrument were matching with each other were grouped according to the test systems used and the differences of reference intervals between different systems were compared with Mann-Whitney test.Results For five items collected,in the clinical laboratories who accepted statistical analysis,respectively 39.32％ (81/ 206),37.44％ (79/211),37.68％ (78/207),37.11％ (72/194),36.98％ (71/192) of them had verified the reference intervals they used.The top three sources of reference intervals were instructions of reagent and instrument manufactures (49.70％-52.54％),calculations by laboratories themselves (18.64％-20.00％),and calculations by other laboratories (14.12％-15.43％).The arithmetic averages and medians of upper and lower limits about every item were relatively close.But the minimum and maximum values of upper and lower limits of certain items had moderate differences.As Mann-Whitney test later showed,all items with the P values of upper and lower limits between different test systems,except for the lower limit of CD3 + CD4 + (％) (P value is 0.052) and the upper limit of CD3-CD19 + (％) (P value is 0.117),were lower than 0.05 which meant an existing statistical difference.Conclusions The sources of lymphocyte subsets by flow cytometry tests in all Chinese clinical laboratories were different and the instructions of reagent and instrument had the highest percentage,but the reference intervals among differenttest systems had statistically significance.
ABSTRACT
Objective To investigate and analyze the current situation of reference intervals and decision limits of reticulocyte count in China.Methods The information related to upper/lower limits,decision limits,data resources,method principles, instruments and reagents of reticulocyte count in 2014 was collected via the external quality assessment software system based on website.The participants were divided according to their methods and instruments.The SPSS 1 9.0 was used for conducting the data analysis and statistics of results for each group.Results There were 276 laboratories submitting the val-id results.The major resources for reference intervals came from national guide to clinical laboratory procedures (53.99%), instructions of instrument manufacturers (16.30%),textbook (9.78%),instructions of reagents manufacturers (9.06%) and others.The major resources for decision limits were national guide to clinical laboratory procedures (56.06%),instruc-tions of instrument manufacturers (16.67%),textbook (7.58),instructions of reagents manufacturers (7.58%)and others. The medians and averages of upper/lower limits and decision limits among subgroups were close,whereas the P2.5 and P97.5 were significant discrepancies.Only 45.65% and 48.48% of participant laboratories had verified reference intervals and deci-sion limits before clinic use.The lower limits and decision limits among subgroups had nostatistical difference (P>0.05), where as the upper limits of different instrument groups had significant differences (P<0.001).Conclusion The current situation of reference intervals and decision limits of reticulocyte count is unsatisfactory.It is certain to establish a uniformed reference interval and decision limits for reticulocyte count,which provides a reference for the standardization of reticulocyte count testing.
ABSTRACT
Objective To describe the reference intervals and their sources of tumor markers in clinical laboratories all over China.And make a comparison of difference between different testing systems.Methods The questionnaires about reference intervals of tumor markers testing were distributed to participants.The information was collected by external quality assessment software system based on website, which including upper and lower limits, sources, validation information and testing systems.The analytes were AFP, CEA, total-PSA, CA125, CA153, CA199, ferritin and free-PSA. The participants were classified according to the testing systems they used.The mean, median, maximum and minimum of each group were calculated using Microsoft Excel 2007.The difference of reference intervals between different testing systems were compared by Kruskal-Wallis test.Results The main source of reference intervals was instructions of testing system manufactures ( 83.1% for mean of 8 ratios ) .The next in sequence were instructions of reagent manufactures(8.4% for mean), National Guide to Clinical Laboratory Procedures (4.6%for mean), determined by their own laboratory(2.0% for mean) and the rest (1.9% for mean). There were 48.0% ( 1 906/3 967 ) of analytes whose reference intervals had been validated.Difference of reference intervals which was with P value of Kruskal-Wallis test with <0.05 was found between different testing system groups except the upper limit of free prostate specific antigen ( PSA ) .Conclusions Most clinical laboratories establish the reference intervals of tumor markers on the basis of instructions of testing system manufactures.The reference intervals among different testing systems have statistically significance.
ABSTRACT
Objective To carry out the current status of internal quality control by statistical analysis of the Internal Quality Control (IQC)data of homocysteine in 2014 March.Methods Web-based External Quality Assessment (EQA)system was used to collect IQC data of homocysteine from 292 EQA participant laboratories nationwide.The data include thecoefficient of variation (CV)of IQC data under control in March 2014 and long-term cumulative data.Acceptable rates of CVs of two-lot internal quality controls in homocysteine were calculated according to 5 criteria,that were 1/3TEa,1/4TEa and the speci-fications based on biological variation including the minimal,appropriate and optimal allowable imprecision.The instrument the participant laboratories used were sorted into 6 groups and the passing rate of each group were calculated by the 5 crite-ria.Results 292 laboratories reported the data of level 1 IQC for homocysteine,106 of which reported the data of level 2 IQC.The passing rate was different according to different criteria.The passing rate had few difference by the criteria of 1/3TEa and the minimal allowable imprecision based on biological variation,which were from 63.36% to 76.42%.It was also true for 1/4TEa and the appropriate allowable imprecision based on biological variation,which were from 34.25% to 57.55. The passing rate was only 10.62%~16.98% by the criteria of appropriate allowable imprecision based on biological varia-tion.Statistical results showed that instruments the participant mainly used were HITACHI (77/292),Olympus (61/292), Roche (19/292),Beckman (14/292),Abbott (10/292)and Simens (10/292).The passing rates of all the instrument group had few difference except Simens group.Conclusion Most of the laboratories could meet the criteria of 1/3TEa and the min-imal allowable imprecision based on biological variation.But less than half of the laboratories could meet the criteria of 1/4TEa and the appropriate and optimal allowable imprecision based on biological variation.The precision performance of ho-mocysteine among laboratories needs further improvement.
ABSTRACT
Objective To calculate and describe the decision limits of 2014 external quality assessment(EQA) programs in D-Dimer of 225 labs.Methods To collectedthe information of decision limits which come from the 2014 External Quality Assessment ( EQA) programs in D-Dimer.All data from maternal and child care service centers , children′s hospitals and corporations were excluded .Then all the abnormal values and errors were eliminated either .Data of 225 labs statistics were processed by Microsoft Excel 2007, which include the mean, median, maximum value, minimum value, P2.5 and P97.5.Weanalyzed the sources of decision limits and then get the data filtered .The reagent manufacturer 's instructions data are reservedas a reference source and data are analyzed according to differentreagents .Results According to the surveys of decision limits sources ,most of thereference sources use the reagent manufacturer′s instructions. According to the surveys,the mean and median of each test project is close .The P2.5 and P97.5 distribution ranges are: D-Dimer(DDU) 0.3 -1.33 μg/ml,D-Dimer(FEU) 0.23 -1.50 μg/ml,with significant differences.According to the surveys , most P2.5 and P97.5 reagent distributions are significantly different . Conclusion The decision limitsof the mean and median of each test item are approximate ,and the P2.5 and P97.5 distributions of each test item are poorly consistent .
ABSTRACT
Objective To investigate the current application status of reference intervals in routine chemistry and compare with the current health industry standards draft for approval.Methods By using web-based external quality assessment (EQA)software system,collected the submitted data from the laboratories which attended national reference intervals inves-tigation,used professional statistical software to analyze the data.Results 385 laboratories had submitted the investigation results.The vast majority of the analytes of attended laboratories had significant statistical differences in comparison with the health industry standards draft for approval in no grouping and not filled groups[t = - 55.435 ~ 17.070,P < 0.05, 86.1% (31/36)].Only a few of laboratories’reference interval had been grouped (approximately 20%).There were some differences in grouping rules between laboratories and standards.Even though some of the laboratories adopt the grouping rules as same as the standards,there were still big differences in reference intervals between the laboratories and the stand-ards[t=-39.365~13.155,P <0.01,62.5% (10/16)].Conclusion The reference intervals of routine chemistry analytes suggested by the health industry standards draft for approval had quite big differences from the reference intervals used in current clinical laboratories daily work.It is important to propel the authority health industry standards to use in the daily work of clinical laboratories.The evaluation and validation of using them should be done first.
ABSTRACT
Objective To evaluate hemoglobin A2 and F (HbA2 and HbF)assays in part of Thalassemia screening laborato-ries in China.Methods Two lots of controls were sent to 50 Thalassemia screening laboratories.Test results of HbA2 and HbF were collected,analyzed and evaluated by different method groups.Results 49 of all laboratories reported the results and the rate of return was 98%.The proportion of laboratories reporting acceptable results for HbA2 and HbF testing was 42.9~92.3% and 27.3% ~ 84.6% respectively.HbA2 test results from 3 laboratories for Lot 201311 and 4 for 201312 were unsatisfactory.And HbF test results from 5 laboratories for Lot 201311 and 3 for 201312 were unsatisfactory.Conclu-sion Conclude that quality in the measurement of HbA2 and HbF should be improved.
ABSTRACT
Objective To compare the current application status of reference intervals in dry chemistry to the reference inter-vals of anlytes of wet chemistry in the same hospital.Methods By using web-based external quality assessment (EQA) software system,collecting the submitted data from the laboratories which attended national reference intervals investigation of reference intervals of analytes in dry chemistry and wet chemistry,using professional statistical software for analyzing the data,including paired-samples t test and Yates’Chi-square test.Results 231 laboratories submitted their investigation re-sults including the reference intervals of 23 analytes of dry and wet chemistry.69.6% (32/46)of reference intervals of ana-lytes in dry chemistry have significant statistical differences compared with the wet chemistry in paired-samples t tests. 80.4% (37/46)of medians of reference intervals of analytes in dry chemistry had significant statistical differences in loca-tions and distributions compared with medians of reference intervals of wet chemistry in the same institute.Conclusion The reference intervals of analytes of dry chemistry have differences compared with the wet chemistry in the same hospital.The medians of reference interval’s lower and higher limits of dry and wet chemistry have difference in locations and distribu-tions.Proposed to create our own reference intervals of dry chemistry in China.
ABSTRACT
<p><b>OBJECTIVE</b>To analyze the sources of reference intervals and distribution of medical decision levels in NT-proBNP, BNP and homocysteine (Hcy) based on the feedback of laboratories which participated in the 2014 External Quality Assessment (EQA) program in China.</p><p><b>METHODS</b>Information of reference intervals and medical decision levels was collected from the 2014 EQA programs in NT-proBNP, BNP and Hcy.All the abnormal values and errors were eliminated.Data were analyzed by SPSS 13.0 and Microsoft Excel 2007 in terms of referring to sources of reference intervals and distribution of medical decision levels.Values were further grouped based on instrument system for NT-proBNP and BNP, reagent manufacturers for Hcy.</p><p><b>RESULTS</b>There were 330, 119 and 339 laboratories submitting the data on the reference intervals of NT-proBNP, BNP and Hcy. Only 133, 54 and 102 laboratories provided information about medical decision levels.Most reference sources were derived from reagent manufactures (NT-proBNP 76.97%(254/330), BNP 89.08%(106/119), Hcy 83.19%(282/339)). Approximately, 36.53%(122/334), 45.38%(54/119) and 47.95%(164/342) of the laboratories validated the reference intervals of NT-proBNP, BNP and Hcy.According to the surveys, the mean and median values were not close to each other (NT-proBNP 507.99 ng/L and 300.00 ng/L, BNP 342.72 ng/L and 100.00 ng/L, Hcy 17.34 µmol/L and 15.00 µmol/L). Variation of P2.5 and P97.5 was as follows: (NT-proBNP 125.00-1 800.00 ng/L, BNP 87.50-4562.50 ng/L, Hcy 4.15-50.00 µmol/L). The ranges were NT-proBNP 60.00-7 000.00 ng/L, and BNP 80.00-7 000.00 ng/L, and Hcy 1.5-52.00 µmol/L.P values of rank sum test for NT-proBNP, BNP and Hcy were 0.449, 0.208 and 0.754.</p><p><b>CONCLUSIONS</b>Medical decision levels of NT-proBNP, Hcy and BNP showed large variation among laboratories and the sources of reference interval were not unified.Thus, urgent efforts should be made to establish appropriate reference intervals for Chinese people and promote harmonization and standardization of reference intervals and medical decision levels on BNP, NT-proBNP and Hcy.</p>
