ORIGINAL ARTICLE Evaluation of Analytical Performance of the D-100 He mo globin Testing System for Hemog

Size: px

Start display at page:

Download "ORIGINAL ARTICLE Evaluation of Analytical Performance of the D-100 He mo globin Testing System for Hemog"

서겨 부
4 years ago
Views:

1 ORIGINAL ARTICLE Evaluation of Analytical Performance of the D-100 He mo globin Testing System for Hemoglobin A1c Assay Kyunghoon Lee 1,2, Sung Min Kim 3, Sun Hee Jun 3, Sang Hoon Song 1,2, Kyoung Un Park 1,3, and Junghan Song 1,3 1 Department of Laboratory Medicine, Seoul National University College of Medicine; 2 Department of Laboratory Medicine, Seoul National University Hospital, Seoul; 3 Department of Laboratory Medicine, Seoul National University Bundang Hospital, Seongnam, Korea Corresponding author: Junghan Song Department of Laboratory Medicine, Seoul National University Bundang Hospital, 82 Gumi-ro 173 beon-gil, Bundanggu, Seongnam 13620, Korea Tel: Fax: songjhcp@snu.ac.kr Background: The hemoglobin A1c (HbA1c) level is widely used to diagnose and monitor glycaemic control in people with diabetes mellitus, and various methods are used for its determination. The D-100 hemoglobin testing system (Bio-Rad Laboratories, USA) is a fully automated, high-throughput glycohaemoglobin analyzer based on an ion-exchange highperformance liquid chromatographic method. Here, we evaluated the analytical performance of a newly developed HbA1c analyzer. Methods: Precision, linearity, and comparison to the Variant II Turbo analyzer (Bio-Rad Laboratories, USA) were evaluated according to the Clinical Laboratory Standards Institute guidelines. Carryover, bias from the value assigned by the HbA1c Network Laboratory of Korea Centers for Disease Control and Prevention, and the vulnerability to interference by hemoglobin variants frequently found in Korea were also assessed. Statistical analyses were performed using Excel 2010 (Microsoft Co., USA) and MedCalc ver (MedCalc Software bvba, Belgium). Results: The coefficients of variation for repeatability and within-device precision were less than 1.08% in National Glycohaemoglobin Standardization Program (NGSP) unit and less than 1.68% in international system of unit at all three levels. The calibration curve was linear, with R 2 =0.996 in the range of.6% to 15.% in NGSP unit. The results highly correlated with those produced by Variant II Turbo (r=0.998). The 95% confidence interval for differences from the assigned values was 3.3% to 2.9%. No significant interferences of haemoglobin variants were observed except for Hemoglobin Yamagata. Conclusions: The D-100 hemoglobin testing system showed excellent precision, linearity, and good correlation with the Variant II Turbo analyzer and agreement with the assigned values. Therefore, its analytical performance is satisfactory for diabetes diagnosis and treatment monitoring. (J Lab Med Qual Assur 2016;38:95-101) Key Words: Glycated hemoglobin, Hemoglobin A1c, Analytical performance, Diabetes mellitus, High pressure liquid chromatography pissn: eissn: Received January 15, 2016, Revision received April 29, 2016, Accepted May 11, 2016 서론 전세계적으로당뇨병의유병률은계속적으로증가하고있다. 200년도의보고에의하면 2030년에는 3억5천만명까지증가할것이라고추정하였으나 [1], 이미 201년도에 3억8천7 백만명에도달하였다. 우리나라의경우질병관리본부에서 30 세이상성인을대상으로 2007 년부터 2010 년도까지시행한 국민건강영양조사자료에의하면당뇨병유병률은 10.1%, 당뇨병전단계인공복혈당장애유병률은 19.9% 이었고, 2050 년도우리나라당뇨병환자수는약 600만명으로늘어날것이라고추정하고있다 [2]. 혈중당화혈색소 (hemoglobin A1c, HbA1c) 는최근약 3개월간의평균혈당치를반영하기때문에환자의혈당상태를파악하여당뇨병의진단, 치료효과판정및모니터링에중요한역할을하는검사이다 [3]. 따라서당 Copyright 2016 Korean Association of External Quality Assessment Service 95

2 뇨병환자의정확한진단과치료지침결정을위해서는정확한검사결과가필수적이기때문에 HbA1c 검사결과의표준화 (standardization) 를위해전세계적으로많은노력이진행돼오고있다 [,5]. 우리나라에서는대한임상정도관리협회가 2007 년 HbA1c 의신빙도조사를시작하였고, 2009 년부터는표준방법으로측정한참값을기준으로평가하는정확도에기반하는신빙도조사를 (accuracy-based proficiency testing) 시행하고있다 [6]. 2012년이전에는참값을구하기위해외국표준검사기관에검사를의뢰하였는데, 2012년 10월에우리나라질병관리본부의국가표준의학검사실이 International Federation of Clinical Chemistry HbA1c Network Laboratory로승인받은이후에는국내에서도참값측정이가능해지는등 HbA1c 검사결과의표준화를위한사업이활발히진행되고있다. HbA1c 측정법을검사원리에따라서이온교환고성능액체크로마토그래피법 (ion-exchange high-performance liquid chromatography), 면역측정법 (immunoassay), 그리고친화크로마토그래피법 (affinity chromatography) 으로크게 3 가지로구분할수있다. 최근에개발된 D-100 hemoglobin testing system (D-100 system; Bio-Rad Laboratories Inc., Hercules, CA, USA) 는이온교환고성능액체크로마토크래피법를이용한 HbA1c 자동분석장비로이에대한분석능력및유용성에대한평가는국내에아직보고되지않았다. 그리하여 D-100 system에대한정밀도, 직선성, 검체간교차오염률, 상관성, 질병관리본부국가표준진단의학검사실에서측정한참값과일치도, 헤모글로빈변이형 (hemoglobin variant) 이미치는영향을평가함으로써임상검사실에서의유용성을확인하고자하였다. 재료및방법 본연구는 201년 11월 1일분당서울대학교병원의임상연구심의위원회 (institutional review board) 심의를얻어진행되었다 ( 승인번호 : B ). 1. 검체 2015년 3월부터 월까지분당서울대학교병원에서 HbA1c 검사결과보고후남은잔여검체총 197개, 질병관리본부국가표준진단의학검사실에서측정한참값이있는 HbA1c 검체 8개, 그리고분당서울대학교병원을포함한 개의병원에서 HbA1c 분석장비에서헤모글로빈변이형이의심되었던검체중 DNA 서열분석을통하여확인한 1검체 (hemoglobin Chad 2개, hemoglobin Fort de France 1개, hemoglobin Yamagata 1개, Hemoglobin G-Coushatta 10개 ) 를이용하여평가하였다. 2. 장비및시약 HbA1c는 D-100 system 과 Variant II Turbo (Bio-Rad Laboratories) 장비를사용하여 EDTA 전혈검체로측정하였다. 검사는각제조회사의지침에따라시행하였으며, 모두제조사전용키트의동일한로트시약을사용하였다. 3. 정밀도정밀도는 Clinical and Laboratory Standard Institute (CLSI) EP05-A3 지침에따라낮은농도, 중간농도, 높은농도의환자검체를각각하루에 2회씩반복하여 20일간측정하여평가하였다 [7]. Repeatability, 검사일간정밀도와검사기기내 (within-device) 정밀도의변이계수를구하였다.. 직선성직선성은 CLSI EP06-A 에따라 Variant II Turbo 장비에서,.6% 의낮은 HbA1c 농도를보이는검체와 15.% 의높은농도를보이는검체를각각 0:, 1:3, 2:2, 3:1, :0으로혼합하여조제한 5개농도의검체를각각 회씩반복하여측정하였다 [8]. 선형회귀분석을이용하여회귀방정식과결정계수 (R 2 ) 를각각구하였다. 5. 검체간교차오염률높은 HbA1c 농도검체와낮은농도검체를 회연속측정하여검체간교차오염률을확인하였다. 식 %carryover=[l1- (L3+L)/2] 100/[(H2+H3)/2-(L3+L)/2] 를이용하여계산하였으며, 1.0% 를기준치로선정하였다. 6. 상관성상관성평가는 CLSI EP09-A3 지침에따라수행하였다 [9]. HbA1c 농도가.8%-12.3% 범위안의환자검체 190개를선정하여 Variant II Turbo와 D-100 system 장비로각각측정하였고, Passing-Bablok 회귀분석으로 Pearson 상관계수 (r) 와상관방정식을구하였으며, Bland-Altman agreement 분석을실시하였다. 7. Hemoglobin A1c 참값과일치도평가질병관리본부국가표준진단의학검사실에서측정한 HbA1c 참값을알고있는 8검체를이용하여참값과의일치도를평 96 J Lab Med Qual Assur 2016;38:

3 가하였다. 참값과 D-100 system에서측정한값의일치도평가를위하여 Passing-Bablok 회귀분석과 Bland-Altman agreement 분석을실시하였다. 8. 헤모글로빈변이형검체의평가 DNA 염기서열분석을통하여확인된총 1검체를이용하여질병관리본부국가표준진단의학검사실에서측정한값과 D-100 system에서나오는결과값및 flag를확인하였다. 9. 통계분석통계분석은 Microsoft Excel 2010 (Microsoft Co., Redmond, WA, USA) 과 MedCalc ver (MedCalc Software bvba, Ostend, Belgium) 프로그램을사용하였다. 결과 Repeatability의변이계수와검사기기내정밀도의변이계수는 international system of unit (mmol/mol) 을적용하였을때 1.05%-1.68% 이었으며, National Glycohaemoglobin Standardization Program (NGSP) unit (%) 을적용하였을때 0.80%-1.08% 이었다 (Table 1). 직선성검증을위해 5가지농도의검체를 회반복측정한결과,.6%-15.% HbA1c 농도범위에서결정계수 (R 2 ) 가 0.996이었으며, 회귀방정식은 y=0.989x이었다 (Fig. 1). 높은농도검체 (18.9%) 와낮은농도검체 (3.8%) 를각각 회연속측정하여계산한검체간교차오염률은 0.65% 였다. Variant II Turbo와 D-100 system에서동시에 2번반복측정한 HbA1c 농도는.8%-12.3% 범위에서우수한상관성을보였다 (r=0.998). Passing-Bablok 회귀분석을시행하여얻은 y=0.95x+0.187의상관식을얻었다 (Fig. 2). 질병관리본부국가표준진단의학검사실에서측정한참값과 D-100 system에서측정한결과간의 Bland-Altman plot에서결 과값 % bias의평균은 -0.2% 였고, 95% 신뢰구간은 -3.3%- 2.9% 였다 (Fig. 3). 1개의헤모글로빈변이형검체를측정하였을때 hemoglobin G-Coushatta 10검체의 bias 평균과표준편차는 -0.28%±0.2% (NGSP unit) 이었다. Hemoglobin Chad 2검체와 hemoglobin Fort de France 1검체의 bias 는각각 0.10%, -0.08%, 0.16% 였다. 그러나 hemoglobin Yamagata 1검체의 bias는 30.08% (NGSP unit) 였다. 고찰 본연구에서는 D-100 system의정밀도, 직선성, 검체간교차오염률을통하여분석능을평가하였다. Repeatability와검사기기내정밀도의결과를보면미국임상화학회에서당뇨병에관련한임상검사지침에서제시하고있는검사실내정밀도 2% 미만보다도낮은수치를보이고있어서권고안을만족하 Measured concentration (HbA1c%) y=0.989x 2 R = Estimated concentration (HbA1c%) Fig. 1. Linearity of HbA1c values using the D-100 hemoglobin testing system. Abbreviation: HbA1c, hemoglobin A1c. Table 1. Precision of the D-100 hemoglobin testing system International system of unit (mean mmol/mol) National Glycohaemoglobin Standardization Program unit (mean %) Low (30.55) Medium (5.9) High (76.16) Low (.9) Medium (6.) High (9.1) Repeatability 1.68 ( ) 1.20 ( ) 1.05 ( ) 0.95 ( ) 0.99 ( ) 0.80 ( ) Between-day 0.00 ( ) 0.26 ( ) 0.9 ( ) 0.00 ( ) 0.5 ( ) 0.38 ( ) Within device 1.65 ( ) 1.5 ( ) 1.16 ( ) 0.96 ( ) 1.08 ( ) 0.88 ( ) Values are presented as CV% (95% confidence interval). J Lab Med Qual Assur 2016;38:

4 A B HbA1c% (D-100 system) Difference (D-100 system-variant II Turbo) SD 0.08 Mean SD HbA1c% (Variant II Turbo) Mean of HbA1c (Variant II Turbo and D-100 system) Fig. 2. Comparison of HbA1c levels obtained by the D-100 haemoglobin testing system (test) and the Variant II Turbo (comparative) by the Clinical Laboratory Standards Institute EP9-A3. (A) Scatter plots with ordinary linear fit with the Passing- Bablok regression analysis. (B) Bland-Altman plots. Abbreviation: HbA1c, hemoglobin A1c. % Bias SD 2.9 Mean SD Assigned HbA1c level (National Glycohaemoglobin Standardization Program unit) Fig. 3. Comparison of HbA1c levels obtained by the D-100 hemoglobin testing system (test) and assigned HbA1c levels by the Korea Centers for Disease Control and Prevention reference laboratory using Bland-Altman plots. %Bias was calculated using the following formula: [(obtained HbA1c level assigned HbA1c level)/assigned HbA1c level] 100. Abbreviation: HbA1c, hemoglobin A1c. 는것을확인할수있었다 [10]. 그리고기존에보고된동일기 기와다른기기보다작거나비슷한정밀도 (0.60%-1.6%) 를 가졌다 [11-1]. 기존에보고된논문에서는 Bio-Rad 사의정도 관리물질을사용하여보다넓은범위 (3.5%-19.0%) 의직선성 을평가하였다 [11]. 본연구에서는환자검체를사용하여상대 적으로좁은범위에서의직선성을평가하였지만임상검사실 에서검사를수행하기에충분한넓은범위로생각된다. 검체간교차오염률은기준치 1% 보다작은값이어서검체간교차오염은미미한것으로판단되었다. 동일한방법으로측정하는 Variant II Turbo와비교하였을때상관계수 (r) 값이 0.998로우수한상관성을보였다. 다만 Fig. 2의 (B) Bland-Altman plot에서볼수있듯이 HbA1c 값이 9.0% 에서점점증가할수록차이가점점증가하는경향을보였으며, 이러한경향은이전의보고에서도동일하게나타나고있었다 [11]. 그러나의학적의사결정농도인 6.5% 와 5.7% 에서의추정 bias와이에대한 95% 신뢰구간은각각 (-0.13 to -0.10), (-0.09 to -0.06) 으로 CLSI EP31-A-IR [15] 에서생물학적변이를고려한 HbA1c 검사법간의허용가능한차이 1.8% 를기준으로계산한각농도의허용가능한차이 0.11 및 0.10과비교하였을때, 두검사법은동등한성능을보였다. 질병관리본부국가표준진단의학검사실에서측정한참값과 D-100 system에서측정한결과의일치도를평가하였을때 9.0% 이상의값에서는값이증가할수록그차이가점점증가하는경향은상관성평가와동일하게나타나고있었다. HbA1c 11.0% 이상한검체에서는 -3.3% 를벗어났지만 95% 신뢰구간범위는다른연구에서제시하고있는 ±6.7% 보다낮았다 [16]. Braga 등 [17] 은 optimal, desirable, minimum 총허용오차 (total allowable error) 로각각 1.97%, 3.9%, 5.92% 를제시하고있는데, desirable quality level보다는작은값이어서임상검사실에서사용가능한것을판단하였다. 우리나라에서흔히발견되는헤모글로빈변이로는 hemo- 98 J Lab Med Qual Assur 2016;38:

5 globin G-Coushatta, hemoglobin Yamagata, hemoglobin Chad, hemoglobin Fort de France 등이알려져있고 [18,19], 이들변이가 HbA1c 결과에미치는영향에대해평가하였다. Hemoglobin Yamagata 검체의경우 bias값은 30.08% (NGSP unit) 로매우큰값이었으며, 동일한방법으로사용하고있는 Variant II Turbo에서도 bias 값이 30.9% (NGSP unit) 로비슷한경향을보였다. 그러나이경우결과보고서에 should not report HbA1c. Suspect variant, high HbA1c 라는 flag note를각각확인하여헤모글로빈변이형에대한영향을인지할수가있었다. 그외다른헤모글로빈변이형에대한 bias는상대적으로낮은값을가지고있었다. Hemoglobin G-Coushatta 10검체의 bias 평균과표준편차의값은 -0.28% ±0.2% (NGSP unit) 이었으며, 기존의보고된이온교환고성능액체크로마토그래피법으로측정하였을때나타난 bias의값이음의값을가지고있었지만값은적었다 [19]. 결과보고서에서 hemoglobin G-Coushatta의경우 suspect variant 라는 flag note를, hemoglobin Chad의경우에는 should suspect variant peak(s) >10% 라는 flag note를각각확인할수있었기때문에헤모글로빈변이형에대한영향을인지할수가있었다. 하지만 hemoglobin Fort de France 의경우아무런 flag note를확인할수없었지만참값과의 bias가 0.16% 로낮아문제가되지않았다. 전세계적으로더흔한 hemoglobin S, hemoglobin C, hemoglobin D, hemoglobin E에대해서는검체를확보할수가없어서결과에미치는영향을직접확인할수가없었지만기존의연구에서 HbA1c 값이위에언급한헤모글로빈변이체들에의한영향을받지않는다는것을확인하였으며, carbamylated hemoglobin나 hemoglobin F에의한영향은거의받지않는다고보고하였다 [11]. 결론적으로 D-100 system은높은정밀도와직선성및낮은검체간교차오염률및다른기기와의우수한상관성과표준검사실의결과값과의일치도를나타내었으므로, 임상적으로당뇨병의진단, 치료효과의추적관찰및합병증예방을위하여 HbA1c의결과를적용할수있는장비라고판단된다. 또한기존장비와의우수한상관성을보이고있으므로기존의장비를충분히대체할만한성능을가지고있다는것을확인할수있었다. 감사의글 본논문은분당서울대학교병원위탁연구비 ( ) 지원에의해이루어진것이다. REFERENCES 1. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for Diabetes Care 200;27: Korean Diabetes Association, Korea Centers for Disease Control and Prevention. Diabetes fact sheet in Korea sheet2012.pdf (Accessed December 22, 2015). 3. American Diabetes Association. Standards of medical care in diabetes Diabetes Care 201;37 Suppl 1: S Trevino G. Consensus statement on the Worldwide Standardization of the Hemoglobin A1C Measurement: the American Diabetes Association, European Association for the Study of Diabetes, International Federation of Clinical Chemistry and Laboratory Medicine, and the International Diabetes Federation: response to the Consensus Committee. Diabetes Care 2007;30:e Weykamp C. HbA1c: a review of analytical and clinical aspects. Ann Lab Med 2013;33: Chung S, Jun SH, Song WH, Song J. Six years experience of accuracy-based proficiency testing for HbA1c in Korea. J Lab Med Qual Assur 2015;37: Clinical and Laboratory Standards Institute. Evaluation of precision of quantitative measurement procedures: approved guideline. 3rd ed. EP5-A3. Wayne (PA): Clinical and Laboratory Standards Institute, Clinical and Laboratory Standards Institute. Evaluation of the linearity of quantitative measurement procedures: a statistical approach: approved guideline. EP6-A. Wayne (PA): Clinical and Laboratory Standards Institute, Clinical and Laboratory Standards Institute. Measurement procedure comparison and bias estimation using patient samples: approved guideline. 3rd ed. EP9-A3. Wayne (PA): Clinical and Laboratory Standards Institute, Sacks DB, Arnold M, Bakris GL, Bruns DE, Horvath AR, Kirkman MS, et al. Guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Diabetes Care 2011;3:e Jaisson S, Leroy N, Guillard E, Desmons A, Gillery P. Analytical performances of the D-100 TM hemoglobin J Lab Med Qual Assur 2016;38:

6 testing system (Bio-Rad) for HbA1c assay. Clin Chem Lab Med 2015;53: Park HD, Kim HJ, Kim MS, Lee SY, Kim JW. Evaluation of hemoglobin A1c on the Cobas Integra 800 Immunoassay and Tosoh HLC-723 G8 HPLC Analyzer. J Lab Med Qual Assur 2009;31: Kim T, Kim S, Chang HE, Song SH, Park KU, Song J, et al. Performance evaluation of HbA1c test on the Toshiba 200FR NEO using AutoLab HbA1c reagent. J Lab Med Qual Assur 2010;32: Choi Q, Han M, Chang HE, Song SH, Park KU, Song J. Performance evaluation of the ADAMS A1c HA-8180 Analyzer for HbA1c. J Lab Med Qual Assur 2012;3: Clinical and Laboratory Standards Institute. Verification of comparability of patient results within one health care system: approved guideline (interim revision). EP31- A-IR. Wayne (PA): Clinical and Laboratory Standards Institute, Glad Working Group A1c Delegates WG, Mosca A, Branca MT, Carta M, Genna ML, Giorda CB, et al. Recommendations for the implementation of international standardization of glycated hemoglobin in Italy. Clin Chem Lab Med 2010;8: Braga F, Dolci A, Montagnana M, Pagani F, Paleari R, Guidi GC, et al. Revaluation of biological variation of glycated hemoglobin (HbA(1c)) using an accurately designed protocol and an assay traceable to the IFCC reference system. Clin Chim Acta 2011;12: Lee ST, Kim MS, Choi DY, Kim SK, Ki CS. Incidence of variant hemoglobin (Hb) and increased fetal Hb concentrations and their effect on Hb A1c measurement in a Korean population. Clin Chem 2006;52: Lee ST, Weykamp CW, Lee YW, Kim JW, Ki CS. Effects of 7 hemoglobin variants on the measurement of glycohemoglobin by 1 analytical methods. Clin Chem 2007;53: J Lab Med Qual Assur 2016;38:

7 당화혈색소측정용 D-100 Hemoglobin Testing System 에대한검사성능평가이경훈 1,2 김성민 3 전선희 3 송상훈 1,2 박경운 1,3 송정한 1,3 1 서울대학교의과대학검사의학교실, 2 서울대학교병원진단검사의학과, 3 분당서울대학교병원진단검사의학과 배경 : 당화혈색소 (hemoglobin A1c, HbA1c) 는당뇨의진단및혈당조절을하는데널리쓰고있으며, 측정할수있는다양한방법이있다. 이온교환고성능액체크로마토그래피법을이용하는바이오래드사의 D-100 HbA1c 전자동측정기는높은처리율을가지고있다. 이연구에서는이새로운 HbA1c 측정기의성능평가를하였다. 방법 : Clinical Laboratory Standards Institute 가이드라인에따라서정밀도, 직선성, Variant II Turbo 와의상관성을평가하였다. 검체간교차오염률과질병관리본부국가표준진단의학검사실에서측정한 HbA1c 참값과차이를평가하였다. 우리나라에서자주발견되는헤모글로빈변이형이 HbA1c 결과값에주는영향을평가하였다. Excel 2010 (Microsoft Co., USA) 과 MedCalc ver (MedCalc Software bvba, Belgium) 프로그램을이용하여통계적분석을시행하였다. 결과 : 검사내정밀도의변이계수와검사기기내의변이계수는 National Glycohaemoglobin Standardization Program 및 international system of unit 을적용하였을때모든농도에서각각 1.08% 및 1.68% 미만이었고,.6%-15.% 범위에서우수한직선성을보였다 (R 2 =0.996). Variant II Turbo 와의상관계수도 로높았다. HbA1c 참값과차이는 95% 신뢰구간이 -3.3%-2.9% 이었다. 헤모글로빈변이형의영향을평가하였을때 hemoglobin Yamagata 외에는 HbA1c 값에크게영향을주지않았다. 결론 : D-100 system 은뛰어난정밀성, 직선성, Variant II Turbo 와우수한상관성을가지고있었으며, 참값과의일치도도높았다. 그러므로당뇨병을진단하거나치료및모니터링을할수있는충분한성능을보여주었다. (J Lab Med Qual Assur 2016;38:95-101) 교신저자 : 송정한우 )13620 경기도성남시분당구구미로 173 번길 82, 분당서울대학교병원진단검사의학과 Tel: 031) , Fax: 031) , songjhcp@snu.ac.kr J Lab Med Qual Assur 2016;38:

원저 Lab Med Online Vol. 3, No. 4: , October 임상화학 CAPILLARYS 2 FLEX Piercing HbA 1c 검사성능평가 Perfor

원저 Lab Med Online Vol. 3, No. 4: 221-22, October 2013 임상화학 CAPILLARYS 2 FLEX Piercing HbA 1c 검사성능평가 Performance Evaluation of the CAPILLARYS 2 FLEX Piercing Analyzer for HbA1c Determination 전용범 1 한민제 1