Independent study · Police roadside screening · Oral fluid
Performance evaluation of on-site oral fluid drug screening devices in normal police procedure in Germany (Forensic Sci Int, 2014)
Police in North Rhine-Westphalia used three saliva screening devices on drivers for 11 months, and the results were compared with blood tests. The abstract reports results per drug and per device.
Read this first
P in the Cup does not perform the testing or certify these products. The figures below come from one source document and describe one study design. They are not a general accuracy rating, and a screen result is preliminary until a laboratory confirms it. See what accuracy figures mean.
Identity
Exactly what this record covers
| Field | As stated in the source |
|---|---|
| Citation | Musshoff F, Hokamp EG, Bott U, Madea B. Forensic Science International 2014;238:120-4. DOI 10.1016/j.forsciint.2014.02.005; PMID 24699311 |
| Devices | Rapid STAT (Mavand Solution GmbH); DrugWipe 5/5+ (Securetec Detektions-Systeme AG); Dräger DrugTest 5000 (Draeger Safety AG & Co. KGaA). Some drivers were also checked with an on-site urine device (DrugScreen, NAL von Minden). |
| Setting | Random oral fluid specimens from car drivers in normal police procedure, North Rhine-Westphalia, Germany |
| Size | 1,212 drivers over 11 months (the abstract writes "1.212"); per-device and per-drug counts are not given |
| Test method | On-site oral-fluid screening devices used by police; the abstract does not describe their chemistry or cutoffs |
| Reference method | Serum (blood) results, not laboratory analysis of oral fluid |
| Evidence type | Independent peer-reviewed study (not run by the manufacturers); a field evaluation, not a consumer-use study |
| Independence | Academic authors (Institute of Forensic Medicine, Bonn) with the North Rhine-Westphalia police; funding not stated in the abstract |
| What we read | The PubMed abstract only; the full text is behind a paywall |
| Evidence tier | A for the three named devices in roadside police use only |
| Last reviewed | 21 October 2026 |
Figures as stated
Sensitivity and specificity by drug
Copied from the abstract. Sensitivity is the share of drivers positive by blood test whom the device flagged; specificity is the share of blood-negative drivers it cleared.
| Drug | Measure | As stated in the abstract | Tier | Source |
|---|---|---|---|---|
| THC | Sensitivity | 71% (DrugWipe), 87% (Dräger), 91% (RapidSTAT) | A | Abstract, PubMed 24699311 |
| THC | Specificity | 29% (DrugWipe), 47% (DrugTest 5000); the abstract attributes this to low cutoff concentrations | A | Abstract, PubMed 24699311 |
| Opiates | Sensitivity | 95% (Dräger), 100% (DrugWipe, RapidSTAT) | A | Abstract, PubMed 24699311 |
| Amphetamine | Sensitivity | 84% (DrugTest 5000), 90% (RapidSTAT), 100% (DrugTest 5000), as printed; one of the two DrugTest 5000 figures is probably another device, but the abstract does not say which | A | Abstract, PubMed 24699311 |
| Methamphetamine | Sensitivity | 50% (DrugTest 5000), 100% (RapidSTAT) | A | Abstract, PubMed 24699311 |
| Cocaine | Sensitivity | 76% (DrugTest 5000), 100% (DrugWipe, RapidSTAT) | A | Abstract, PubMed 24699311 |
| Methadone | Sensitivity | 33-63%, with a low number of positives | A | Abstract, PubMed 24699311 |
| Benzodiazepines | Sensitivity | 0-33%, with a low number of positives | A | Abstract, PubMed 24699311 |
The authors conclude that oral fluid devices "still show a lack of sensitivity (methamphetamine, benzodiazepines) and specificity (THC)", and that THC specificity was poor "due to lowered cut-offs resulting in multiple false positive tests."
Limitations
- Roadside police screening of drivers, not home use; no consumer kit was evaluated.
- Compared with blood (serum), which can disagree with saliva for reasons unrelated to device accuracy.
- Abstract only: no per-drug sample counts, confidence intervals or device cutoffs, and one amphetamine figure appears to be attributed to the wrong device.
- Low THC specificity means many drivers flagged for THC had no THC in their blood test.
Sources
- PubMed abstract, PMID 24699311 (U.S. National Library of Medicine; opened and read 2026-10-21).
- Article page, DOI 10.1016/j.forsciint.2014.02.005 (Forensic Science International / Elsevier; opened and read 2026-10-21).
General education only, not medical, legal or purchasing advice. Found an error? See the corrections log.
Page last updated