Using LED Microscopy to Diagnose Tuberculosis
Author Information
Author(s): Maryline Bonnet, Laramie Gagnidze, Willie Githui, Philippe Jean Guérin, Laurence Bonte, Francis Varaine, Andrew Ramsay
Primary Institution: Epicentre, Paris, France
Hypothesis
Can LED-based fluorescence microscopy improve the diagnosis of tuberculosis compared to traditional Ziehl-Neelsen microscopy in a high HIV prevalence setting?
Conclusion
LED-based fluorescence microscopy did not increase sensitivity compared to Ziehl-Neelsen microscopy, but it offered faster reading times and was well accepted by technicians.
Supporting Evidence
- The detection yields of LED-FM and ZN microscopy were 20.3% and 20.6%, respectively.
- Sensitivity was 73.2% for LED-FM and 72% for ZN microscopy.
- Specificity was 96.7% for LED-FM and 95.9% for ZN microscopy.
- Inter-reader agreement was high with a kappa value of 0.9.
- Mean reading time with LED-FM was three times faster than ZN microscopy.
Takeaway
This study looked at a new way to check for tuberculosis using special lights. It was faster but didn't find more cases than the old method.
Methodology
A prospective study comparing LED fluorescence microscopy and Ziehl-Neelsen microscopy using sputum samples from suspected TB patients.
Potential Biases
Technicians had extensive experience with ZN microscopy, which may have influenced the results.
Limitations
The study could not stratify results by HIV status and not all patients produced a fourth specimen for culture.
Participant Demographics
{"mean_age":34.3,"gender_ratio":"1.6 (M:F)","past_TB_history":"20.6%"}
Statistical Information
P-Value
0.64
Confidence Interval
[18.2–22.5]
Statistical Significance
p=0.64
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website