Can portable devices really detect poor quality medicines?

A systematic review of 41 scientific studies evaluating the performance and abilities of portable devices to assess the quality of pharmaceutical products has been published today. The review revealed that although there is a wide range of devices available for medicines quality screening, there is also a lack of independent evaluation for those being used in the field.

Packets of medicine
i
istock

The devices ranged from small benchtop spectrometers to ‘lab-on-a-chip’ single-use devices, priced between US$10 to more than US$20 000. While most devices had the potential to distinguish genuine from falsified medicines, the review found that:

  • Only six devices had been field-tested
  • The average number of active pharmaceutical ingredients evaluated per device was only two when there are thousands of different medicines in use globally
  • There was very limited scientific evidence of their ability to accurately test liquid, capsule or topical formulations, or to distinguish between chiral molecules
  • Cost-effectiveness was not explored and little information was available on where these devices could fit into pharmaceutical supply chains
  • Most were complex devices:
    • Those that could test APIs (semi)-quantitatively required consumables and were destructive, except for spectroscopic devices
    • The 10 spectroscopic devices tested for their abilities to quantitate APIs required processing complex API-specific calibration models

Dr Céline Caillet, IDDO’s Medicine Quality group coordinator and pharmacist, said: “It’s great news that there are so many innovative portable devices available now to screen for poor quality medicines, but performance of these is key. Poor quality medicines have devastating consequences and more intensive research is needed to help national medicines regulatory authorities in choosing optimal medicine quality screening devices.”

 

Notes to editors:

Studies were evaluated across Embase, PubMed, Web of Science and SciFinder databases up to 30 April 2018.