| Title: |
Pilot studies in Mexican clinical laboratories . III. Strategy to evaluate quality of results ; Programa piloto en laboratorios clínicos mexicanos. III. Estrategia para evaluar la calidad de los resultados |
| Authors: |
Herrera, Manuel; Loria, Alvar; Terres, Arturo; Gonzalez Salayan, Marco Antonio; Hernández, Maria Antonieta |
| Source: |
Salud Pública de México; Vol. 36 No. 5 (1994); 484-491 ; Salud Pública de México; Vol. 36 Núm. 5 (1994); 484-491 ; 1606-7916 ; 0036-3634 |
| Publisher Information: |
Instituto Nacional de Salud Pública |
| Publication Year: |
1994 |
| Collection: |
Salud Pública de México (E-Journal) |
| Subject Terms: |
quality assessment; health care; clinical laboratoy; Mexico; control de calidad; sistemas de salud; laboratorio clínico; México |
| Description: |
Objective: to explore the validity of a simple and rapid strategy which discriminates precision and accuracy of assays done in sera. Participants: the same group of 18 laboratories. Methods: the sera and results of glucose, urea, creatinine and urates assayed in the routine work of a single day were collected by auditors of an agency of the Ministry of Health. Six laboratories used autoanalyzers and 12 manual kits in the assays. The sera were transported that same day (stored in ice) to a reference laboratory (REF LAB) to repeat the assays as soon as possible. The REF LAB used a spectrum analyzer which showed, throughout the seven months of the study, excellent precision in its internal program and good accuracy in the international EQA program of the World Health Organization. Strategy of analysis: precision and accuracy were independently evaluated with two correlates per analyte and participant: the coefficient of correlation of paired results (REF LAB/ participant) as a correlate of precision, and the mean of ratios of paired results (participant/REF LAB) as a correlate of accuracy. A correlation below 0. 70 and a mean ratio outside of 1 ¤ 0.15 were considered inadequate. Results: only 62 systems had information (10 either lacked reagents or had no samples that day). Imprecision was observed in only 8 per cent of the analytes (5/62) which is lower than the 30 per cent usually seen in Mexican EQA programs. On the contrary, the frequency of inaccuracy was high, i.e. the ratios were inadequate in 44 per cent of the systems (28/62). Conclusions: our strategy of evaluation was able to discriminate imprecision from inaccuracy. The low frequency of imprecision in the study suggests the possibility of an observer's effect on the observed (date and hour of the audit visit was known in advance). The high rate of inaccuracy highlights the convenience of having a larger local availability of reference materials and EQA programs with reliable control material. The search for indexes discriminating imprecision from ... |
| Document Type: |
article in journal/newspaper |
| File Description: |
application/pdf; text/html |
| Language: |
Spanish; Castilian |
| Relation: |
https://saludpublica.mx/index.php/spm/article/view/5785/6429; https://saludpublica.mx/index.php/spm/article/view/5785/6430; https://saludpublica.mx/index.php/spm/article/view/5785 |
| Availability: |
https://saludpublica.mx/index.php/spm/article/view/5785 |
| Accession Number: |
edsbas.9103432D |
| Database: |
BASE |