PLoS ONE (Jan 2011)

An evaluation of the performance and acceptability of three LED fluorescent microscopes in Zambia: lessons learnt for scale-up.

  • Eleanor R Turnbull,
  • Kaunda Kaunda,
  • Jennifer B Harris,
  • Nathan Kapata,
  • Mweemba W Muvwimi,
  • Annika Kruuner,
  • German Henostroza,
  • Stewart E Reid

DOI
https://doi.org/10.1371/journal.pone.0027125
Journal volume & issue
Vol. 6, no. 11
p. e27125

Abstract

Read online

The World Health Organization recommends the roll-out of light-emitting diode (LED) fluorescent microscopes (FM) as an alternative to light microscopes in resource-limited settings. We evaluated the acceptability and performance of three LED FMs after a short orientation among laboratory technicians from government health centers in Zambia. Sixteen technicians with varied light microscopy experience were oriented to FMs and divided into groups; each group read a different set of 40 slides on each LED FM (Primo Star iLED™, Lumin™, FluoLED™) and on a reference mercury-vapor FM (Olympus BX41TF). Slide reading times were recorded. An experienced FM technician examined each slide on the Olympus BX41TF. Sensitivity and specificity compared to TB culture were calculated. Misclassification compared to the experienced technician and inter-rater reliability between trainees was assessed. Trainees rated microscopes on technical aspects. Primo Star iLED™, FluoLED™ and Olympus BX41TF had comparable sensitivities (67%, 65% and 65% respectively), with the Lumin™ significantly worse (56%; p<0.05). Specificity was low for trainees on all microscopes (75.9%) compared to the experienced technician on Olympus BX41TF (100%). Primo Star iLED™ had significantly less misclassification (21.1% p<0.05) than FluoLED™ (26.5%) and Lumin™ (26.8%) and significantly higher inter-rater reliability (0.611; p<0.05), compared to FluoLED™ (0.523) and Lumin™ (0.492). Slide reading times for LED FMs were slower than the reference, but not significantly different from each other. Primo Star iLED™ rated highest in acceptability measures, followed by FluoLED™ then Lumin™. Primo Star iLED™ was consistently better than FluoLED™ and Lumin™, and performed comparably to the Olympus BX41TF in all analyses, except reading times. The Lumin™ compared least favorably and was thought unacceptable for use. Specificity and inter-rater reliability were low for all microscopes suggesting that a brief orientation was insufficient in this setting. These results provide important data for resource-limited settings to consider as they scale-up LED FMs.