[State of the art and method validations: application to fidelity performance]. / Etat de l'art et validation de techniques: application aux performances de fidélité

Current events of clinical laboratories in France argue a lot about quality management. Setting up an assurance quality system can be realized in many approaches listed by increasing requirements: respect of reglementary Guide de bonne exécution des analyses (GBEA), BioQualite qualification, certification and at last accreditation. This last step corresponds to the recognition of the competence of the laboratory to execute specifics tasks. Validation of methods used in the laboratory is a key point when you realize an accreditation package. Fidelity (repetability and reproductibility) is one of the parameters to check in your lab for quantitative methods. These performances are validated in comparison with targets beforehand defined (according to biological variation or state of the art). This article reports fidelity performances obtained in 2000 and 2006 by the participants to ProBioQual internal quality controls. Considering these performances obtained in routine conditions, the different possible approaches to define acceptability limits were discussed.

[About the choice of acceptability limits in external quality assessment schemes]. / Réflexion sur le choix des limites acceptables dans les programmes d'évaluation externe de la qualité

Quality control schemes are a practical tool used in clinical laboratories and an essential element for any quality assurance process. In France, external quality assessment schemes (EQAS) can be mandatory (as national quality control organized by AFSSAPS) or voluntary as those suggested by French associations (ProBioQual, CTCB or Asqualab). These EQAS usually evaluate participants according to their performances: this ranking depends on acceptability limits which are here compared. Various examples based on ProBioQual's background illustrate difficulties to plan out analytical quality specifications. A comment is given about the best criteria (state of the art or biological variation mainly) to be considered to delimit analytic goals. This discussion includes approaches suggested by French committee on accreditation (Cofrac). All criteria could be criticized but it is important to compare oneself laboratory to peers and also to take account of biological variation.

[Improved reliability of critical data exchange between the biological laboratory and clinical centres. Type of data to transfer and identification of the most relevant tools]. / Fiabiliser les échanges d'informations précieuses entre le laboratoire et les services de soins. Types d'informations à échanger et choix des outils les plus fiables.

The reliability of the transfer of critical data between the biological laboratory and the clinical centres has to be reassessed. We have a legal constraint (GBEA and best practices) to make sure that the appropriate alert is issued for severe and urgent cases. In this paper, we present a study of the performances of data transfer tools, such as the comparison between regular phone and up to date available means like in line printers and network terminals. This paper focuses on the twenty-four hours duty in the hospital at the same level of reliability for alerts.

[Digoxin assay: results of 10 years of intralaboratory controls]. / Dosage de la digoxine: exploitation de 10 ans de contrôles interlaboratoires.

Serum digoxin measurement is often performed in medical laboratories. A professional association specialized in quality control, based in Lyon, has been organizing punctual controls of medication measurement for the past ten years. The results are analysed in term of intra and inter-technique precision, difference between methods and specificity in regard to endogenous or exogenous interfering substances. Methods have changed with a quasi disappearance of the methods used ten years ago (FPIA, EMIT) and introduction of new technologies on recent immunoanalysis automates. The results observed with the different instruments are similar. Reproductibility has not changed over ten years. Some difficulties remain in the measurement of low concentrations of digoxin. Many substances interfere in digoxin measurement : digoxigenine (inactive metabolite), endogenous digoxin-like immunoreactive factors, spironolacton, antidigoxin antibodies used for treatment of digitalic intoxications. These interferences depend on the method which is used, but it is essential to know them in order to interpret the results correctly.

French experience of quality assessment of quantitative urinary analysis.

Since 1986, clinical biochemists from the Rhône alpes area, in collaboration with a non-profit-making association (Pro Bio Qual), have been conducting an inter-laboratory quality assurance program for quantitative urine analysis. We investigated the precision and accuracy of individual methods and measurement systems routinely used in the monthly control for 13 analytes: albumin, alpha-amylase, calcium, chloride, creatinine, glucose, magnesium, phosphorus, potassium, total protein, sodium, urea, and urate. The number of laboratories participating in the program increased from 60 in 1986 to 277 in 1999. In 1986, the greatest inter-laboratory imprecision occurred in the assay of urinary total protein, because the commonly used sulfosalicylic acid turbidimetric methods displayed poor precision. Since 1989, the increasing use of pyrogallol red or Coomasie Blue dye colorimetric methods has improved inter-laboratory precision markedly. The acceptable precision and good practicability have encouraged the general use of the pyrogallol red method. As for albumin analysis, which uses a specific immunologic method, the precision was reasonably good. Quality assessment of chloride and sodium decreased when the method performance improved for other analytes (calcium, creatinine, glucose, magnesium, phosphorus, potassium, urea, urate). This program has helped the laboratories to improve the quality of quantitative urine analysis, particularly for total protein.

[Microdetermination to creatinine in biological fluids by kinetic methods. Comparison with continuous flow methods]. / Microdétermination de la créatinine dans les liquides biologiques par des méthodes cinétiques. Comparaison avec les méthodes en flux continu

The authors compare the performances of two micromethods of determination of creatinine in biological fluids, based on the measurement of the kinetics of the reaction of Jaffe: one was adapted to an analyser of the rate of reaction LKB 8 600, the other, a rapid analyser by centrifugation, the Centrifichem. Correlations with continuous flow techniques with dialysis are also reported.

[Advantages and constraints of recent multichannel selective analyzers]. / Avantages et contraintes des analyseurs sélectifs multiparamétriques récents.

The analyzers have been defined as follows: the analyzers are multiparametric (at least 5 parameters available simultaneously), selective, operating patient by patient, list price on May 1, 1984 below 1,000,000 francs, and distributed in France after January 1, 1981. This definition applies in particular to systems already widely used in French Laboratories, such as the Technicon RA 1000, and the Hitachi Boehringer 705. It also applies to systems distributed more recently, such as the Kone "Progress" and the Coultronics "Dacos". Based on data provided by the manufacturers (technical instructions, operating instructions, evaluation reports, etc) and data provided by a certain number of users, we have tried to appraise the advantages and constraints of these analyzers with respect to the following points: required environment (installation, electrical current, temperature), training of personnel, ease of adaptation, work organization, rhythms, operating cost, ease of use, verification of correct function, format of results, maintenance, etc. Lastly, we discuss the problem of the reliability of the results (precision, accuracy), in particular the problems of referencing in substrate assays, and the problems of accuracy in the measurement of enzymatic activities, considering the current norms of standardization.

[Improvement of result coherence in clinical enzymology: multicenter study of gamma-glutamyltransferase, alkaline phosphatase and amylase activities]. / Amélioration de la cohérence des résultats en enzymologie clinique: étude multicentrique concernant les activités gammaglutamyltransférase, phosphatase alcaline et amylase.

We report here on the results of a multicenter study of three enzyme activities (gamma-glutamyltransferase, alkaline phosphatase and amylase). For each activity, measurements were performed in two laboratories on different series of patients' specimens under routine conditions, at 30 and 37 degrees C, with techniques frequently used in France and with the IFCC reference method, when it exists. For each technique, precision was acceptable, but results differed considerably according to the technique used. The study also showed that for different techniques it is not possible to use a single transformation factor for activities between 30 and 37 degrees C. Patients' results determined by two techniques often showed a constant relationship. Groups of techniques that determined the same catalytic activity in patients' specimens were identified, whereas other techniques did not have this property. Several preparations, including reference materials produced by the Community Bureau of Reference (European Community, Brussels) and ten commercial secondary materials were tested for similar behaviour as compared to patients' samples. Results show the commutability of reference materials within a group of techniques indicating that they can be used as calibrators. This was seldom the case for the commercial secondary materials and we did not find any such material suitable for calibration of the three enzymatic activities. The present study demonstrates that with defined techniques and validated calibrators it is possible to reduce considerably differences between results obtained with different techniques at different temperatures and in different laboratories.

[Quality specifications and allowable standards for validation of methods used in clinical biochemistry]. / Analyses de biologie médicale: spécifications et normes d'acceptabilité à l'usage de la validation de techniques.

The purpose of this work is to provide, for a large number of analysis in the field of clinical biochemistry, appropriate criteria for the evaluation of the performance of in vitro diagnostic methods. Based on a first set of data established in 1986 and the experience cumulated by organisers in charge of internal and external quality assessment surveys, an expert group has proposed acceptable limits for a large list of analysis (n = 116). Data are reported and presented in tables divided into 7 chapters including: general biochemistry, enzymes, proteins, tumour markers, hormones, drugs (and toxic), and urinary analysis. For each analysis are given: analytical and reference ranges, three concentration levels for control specimens to be used during evaluations and the range of values within which they can be chosen, reproducibility and repeatability limits expressed as CV%. Maximal tolerable systematic error and inaccuracy are given for control and biological specimens and compared to those obtained using a reference or validated method. These data are essential for evaluations using the protocol designed by the SFBC and can serve as quality criteria for the choice and validation of in vitro diagnostic systems.

[Blood screening of benzodiazepines and tricyclic antidepressants. Results of 5 years of interhospital controls]. / Dépistage des benzodiazepines et des antidepresseurs tricycliques dans le serum. Bilan de cinq ans de contröles inter hospitaliers.

Pro bio Qual Association of Lyon have proposed a control which include the control for detecting benzodiazepines (BZD) and tricyclic antidepressants (ADT) since 2000. With this control, we have evaluated the specificity and the sensitivity of techniques used. We have tested the maprotiline reactivity too (tetracyclic antidepressant). We report results achieved with different immunoassay methods and their performances: specificity and sensitivity. The number of laboratories which realize these analyses is constant: these laboratories are hospital laboratories. Two methods are most common used: the fluorescence polarization immunoassay (FPIA) and the enzyme multiplied immunoassay test (EMIT). The evolution of these techniques does square with the evolution of chemistry analysers used in hospital laboratories. For the BZD, the specificity is good. For the ADT, carbamazepine at low concentration (5 mg/L) gives a positive result with FPIA Abbott assay; carbamazepine at high concentration (25 mg/L) gives a negative result with EMIT assay; phenothiazines give a positive response. For the BZD, the analyser Integra gives best results of sensitivity. Results of sensitivity obtained with the kit EMIT DAU are better than results obtained with the kit EMIT Serum. For the ADT, results of sensitivity obtained with FPIA Abbott assay seem better; the adaptation of EMIT assay on Integra analyser gives less good results. The reactivity for the BZD is very different: we can ignore a severe intoxication with alprazolam or lorazepam (response << cut off) but we can give a positive result for a therapeutic concentration of diazepam for example. With ADT, we haven't the same problem. But the reactivity for nortriptyline is less good than the reactivity for amitriptyline. So we should use a "cut off" concentration which corresponds to the best sensitivity and the best specificity.

The influence of bilirubin, haemolysis and turbidity on 20 analytical tests performed on automatic analysers. Results of an interlaboratory study.

The director of a laboratory has to be sure to give out reliable results for routine tests on automatic analysers regardless of the clinical context. However, he may find hyperbilirubinaemia in some circumstances, parenteral nutrition causing turbidity in others, and haemolysis occurring if sampling is difficult. For this reason, the Commission for Instrumentation of the Société Française de Biologie Clinique (SFBC) (president Alain Feuillu) decided to look into "visible" interferences--bilirubin, haemolysis and turbidity--and their effect on 20 major tests: 13 substrates/chemistries: albumin, calcium, cholesterol, creatinine, glucose, iron, magnesium, phosphorus, total bilirubin, total proteins, triacylglycerols, uric acid, urea, and 7 enzymatic activities: alkaline phosphatase, alanine aminotransferase, alpha-amylase, aspartate aminotransferase, creatine kinase, gamma-glutamyl transferase and lactate dehydrogenase measured on 15 automatic analysers representative of those found on the French market (Astra 8, AU 510, AU 5010, AU 5000, Chem 1, CX 7, Dax 72, Dimension, Ektachem, Hitachi 717, Hitachi 737, Hitachi 747, Monarch, Open 30, Paramax, Wako 30 R) and to see how much they affect the accuracy of results under routine conditions in the laboratory. The study was carried out following the SFBC protocol for the validation of techniques using spiked plasma pools with bilirubin, ditauro-bilirubin, haemoglobin (from haemolysate) and Intralipid (turbidity). Overall, the following results were obtained: haemolysis affects tests the most often (34.5% of cases); total bilirubin interferes in 21.7% of cases; direct bilirubin and turbidity seem to interfere less at around 17%. The different tests are not affected to the same extent; enzyme activity is hardly affected at all; on the other hand certain major tests are extremely sensitive, increasingly so as we go through the following: creatinine (interference of bilirubin), triacylglycerols (interference of bilirubin and haemoglobin), glucose (interference of bilirubin), cholesterol (interference of bilirubin), phosphorus (interference of bilirubin and haemoglobin), uric acid (interference of turbidity), iron (interference of haemoglobin and turbidity), total proteins (interference of bilirubin, haemoglobin and turbidity) and bilirubin (interference of haemoglobin and turbidity). Three categories of interferences can be found: interference by addition, chemical interference and spectral interference, and the results show that not only the choice of a method is important on the analyser, but also how this has been adapted. By looking into these conditions carefully, it is sometimes possible to find a reason for the problem and thereby a simple solution to correct it.(ABSTRACT TRUNCATED AT 400 WORDS)

Studies on circulating immunoreactive calcitonin in low birth weight infants during the first 48 hours of life.

Serum immunoreactive calcitonin (iCT), serum immunoreactive parathyroid hormone (iPTH) and serum Ca, Mg, P and total protein levels were determined sequentially at 5 given periods of time from 1 to 48 h of age in 16 low birth weight infants. Mean +/- SD serum Ca levels decreased from 8.99 +/- 0.79 mg/100 ml at time 1--2 h to 7.00 +/- 0.51 mg/100 ml at time 12--14 h; there was a small further decrease at time 22--26 h: 6.79 +/- 1.07 mg/100 ml. There was no significant change in serum Mg, P or total protein during the same periods of time. Serum iPTH levels increased steadily from time 1--2 h to time 44--48 h reaching above normal range values. Serum iCT levels were non detectable (less than 150 pg/ml) in 11 among 15 infants at time 1--2 h. A marked increase in mean +/- SD serum iCT levels was observed at time 12--14 h (1850 +/- 872 pg/ml) and time 22--26 h (1462 +/- 806 pg/ml) followed by a decrease at time 44--48 h. A negative correlation was found between serum iCT levels and respectively gestational age (p less than 0.01) and serum Ca levels (p less than 0.01) at time 22--26 h while serum iCT levels correlated positively with serum iPTH levels (p less than 0.05). Evidence obtained from this study indicates that a secretion of calcitonin takes place during the early neonatal period in low birth weight infants and that this secretion is a contributing factor of the socalled "early-type" neonatal hypocalcemia.

Usefulness of reference materials in calibration of enzyme activities.

alpha-Amylase, alkaline phosphatase and gamma-glutamyltransferase were studied in a multicentre evaluation. Analyses were performed on different patient samples. Each enzyme was assayed in two different laboratories at both 30 and 37 degrees C, with widely used reagent kits and with the IFCC reference method (if in existence). Results differed considerably according to the measurement procedure. Data also showed that it was not possible to employ a constant conversion factor for one enzyme and different techniques between 30 and 37 degrees C. Calibration with three reference materials extensively improved the intermethod consistency for most of the tested measurement procedures. It was possible to transfer accuracy from the method used for the certification of the reference material to routine procedures, by using the reference material as calibrator. Temperature did not seem to be a crucial variable for the implement of the enzyme calibrator approach.