A comparative study between Near-Infrared (NIR) spectrometer and High-Performance Liquid Chromatography (HPLC) on the sensitivity and specificity

• Published in: PloS one Vol. 20; no. 3; p. e0319523
• Main Authors: Maffioli, Elisa M., Anyakora, Chimezie
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 25.03.2025; Public Library of Science (PLoS)
• Subjects: Algorithms; Analgesics; Analgesics - analysis; Anti-Bacterial Agents - analysis; Antibiotics; Antihypertensives; Antimalarials - analysis; Antiparasitic agents; Artificial intelligence; Biology and Life Sciences; Chromatography; Chromatography, High Pressure Liquid - methods; Comparative studies; Computer and Information Sciences; Counterfeit Drugs - analysis; Counterfeiting; Drug stores; Engineering and Technology; FDA approval; High performance liquid chromatography; Human subjects; Humans; I.R. radiation; Infrared analysis; Infrared spectrometers; Libraries; Liquid chromatography; Machine learning; Medicine and Health Sciences; Near infrared radiation; Nigeria; People and Places; Pharmaceuticals; Pharmacists; Physical sciences; Portable equipment; Proprietary; Public health; Research and Analysis Methods; Sensitivity; Sensitivity analysis; Sensitivity and Specificity; Software; Spectroscopy, Near-Infrared - methods; Spectrum analysis; Vitamins; Nigeria
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0319523

It is estimated that 10.5% of medicines in low- and middle-income countries are substandard or falsified (SF), causing approximately 1 million deaths annually. Over the past two decades, there have been significant technological advancements in low-cost, portable screening devices to detect poor-quality medicines, which could be especially beneficial in these countries. The pharmaceutical market in Nigeria is valued at USD 4.5 billion and is growing at over 9% annually. However, SF medicines remain a major public health concern. We compared a novel Near-Infrared (NIR) Spectrometer with high-performance liquid chromatography (HPLC) by analyzing 246 drug samples purchased from retail pharmacies across the six geopolitical regions of Nigeria. We measured the sensitivity and specificity of a patented and Artificial Intelligence (AI) - powered handheld NIR spectrometer, which uses a proprietary machine-learning algorithm as well as hardware and software, across four categories of medicines: analgesics, antimalarials, antibiotics, and antihypertensives. Our findings reveal that the prevalence of SF medicines remains high, with 25% of samples failing the HPLC test. When tested with the NIR spectrometer, only a smaller subset of medicines—specifically analgesics—failed the test. Sensitivity and specificity for all medicines were 11% and 74%, respectively. For analgesics, the sensitivity was 37%, and the specificity was 47%. While these devices hold great potential, regulators should require more independent evaluations of various drug formulations before implementing them in real-world settings. Improving the sensitivity of these devices should be prioritized to ensure that no SF medicines reach patients.