Results - continued
Objective Measures of Drug Use
As noted earlier, 639 oral fluid samples were analyzed for drugs. Though 642 oral fluid samples were collected, 3 containers leaked during transport and sufficient volume to conduct analyses was available only for the remaining 639. Of these 639, there were also 394 blood samples available for analysis. Blood samples were provided by 406 subjects; however, 12 were unable to provide sufficient volume to permit analysis. Every subject who provided a blood sample also provided an oral fluid sample.
Drugs were detected in 96 cases:
In the subjects where blood only was positive for therapeutic drugs, the concentrations detected were all in the therapeutic range and would not be considered at a high enough level to cause impairment.
Drugs known to cause impairment were detected in many of the specimens and included marijuana, benzodiazepines (diazepam, alprazolam), carisoprodol, the narcotics (oxycodone, codeine, hydrocodone), methadone, tramadol, and cocaine.
The antidepressants/selective serotonin uptake inhibitors (SSRIs) such as fluoxetine and sertraline can cause impairment in rare circumstances where extremely high blood concentrations are measured. The data reviewed does not suggest that the concentrations found would have a significant effect on one’s ability to operate a motor vehicle. Fluoxetine and its metabolite norfluoxetine were found in four blood samples with corresponding oral fluid negative samples; and in four paired samples. Sertraline was found in five blood samples with correspondingly negative oral fluid specimens; and in five pairs of samples both were positive for the drug.
Benzodiazepines are known to cause impairment in traffic cases when present at high level. There were four cases in which alprazolam was reported, however, in two of the subjects, the concentrations were so low (1 ng/mL and 4 ng/mL) it is highly unlikely that the direct pharmacological effects of the drug would cause impairment. In the other two cases (27 ng/mL and 19 ng/mL in the presence of THC), while the concentrations were within the therapeutic concentration range, it should be noted that the desired/therapeutic effect of alprazolam is sedation - which would have a detrimental effect on driving a motor vehicle.
In both of these subjects, corresponding oral fluid analysis did not detect alprazolam and it is likely that the cross-reactivity on immunoassay tests for alprazolam is low. However, it is known that benzodiazepines do not appear in the oral fluid in very high quantity due to their high level of protein binding.
The most common benzodiazepine, diazepam (Valium™) and/or its metabolites nordiazepam, oxazepam and temazepam were detected in 4 blood samples and not in the corresponding oral fluid specimens. Diazepam and nordiazepam were found in one oral fluid sample, the corresponding blood sample was refused.
In one case, carisoprodol (Soma™) and its metabolite meprobamate were detected in both blood and oral fluid. Even at therapeutic concentrations, this may cause driving impairment as the desired effect is sedation.
The most prevalent drug detected was marijuana. There appeared to be a strong positive correlation between the oral fluid and blood tests, the only discrepancies (negative oral fluid and a positive blood) are from 10 cases where the inactive metabolites were detected in blood, and not the active tetrahydrocannabinol (THC).
A positive metabolite result (THCA) with a negative parent compound (THC) is consistent with remote use - in these cases a negative oral fluid would not miss an impaired driver.
THC or its metabolites were detected in 37 oral fluid cases and in 23 blood specimens. Thirteen had corresponding positive oral fluid samples; 10 contained only inactive THC metabolites, as described above.
In oral fluid, the active parent compound, tetrahydrocannabinol (THC) was detected in all cases where marijuana use was determined. In 22 samples where THC was detected in oral fluid, a blood collection was refused.
Cocaine or its metabolites were detected in 14 oral fluid samples and only 4 blood samples, hence oral fluid appears to be a better specimen type for the detection of recent cocaine use. Of the other 10 oral fluid positive samples, 5 had corresponding blood samples which were negative, and 5 subjects refused to give blood.
There was only one blood sample containing parent cocaine, whereas oral fluid detected 13 positives. The presence of parent drug, (cocaine) in either specimen indicated the presence of active drug in the system.
Methamphetamine, amphetamine, pseudoephedrine, phenylpropanolamine and phentermine are members of the same sympathomimetic stimulant group of drugs. They were detected in a small number of cases.
The other drugs detected (butalbital, amitryptiline) were only present in one or two cases, and given the reported concentrations, none at a concentration to cause impairment. Overall:
Table 30 presents each of the drug positive results for both oral fluid and blood, and Table 31 summarizes those results.
In Table 32, the overall results for drug positives are summarized by drug and sample type.
In Table 33, we summarize the combined results of the oral fluid and blood tests for drugs other than alcohol. Successful laboratory analyses were conducted on 639 drivers. Oral fluid was analyzed for all of these drivers and blood for 394 of them. All drivers who provided blood also provided oral fluid. Thus, 14 of these 639 drivers, or 2.2 percent tested positive for cocaine. The most frequently encountered drug was cannabinoids, with 47 or 7.4 percent of drivers having THC or a metabolite of that substance on board. One must bear in mind that metabolites of THC appear in blood well after the active phase of its potentially impairing effect. Cannabinoids were followed in frequency of appearance by two stimulants, cocaine and amphetamines, each of which presented 2.2 percent of the time in tested drivers. Benzodiazepines (Valium and its relatives) 1.6 percent, Sertraline (Zoloft) 1.6 percent, Fluoxetine (Prozac) 1.3 percent were the next most frequently encountered drugs. Opiates were encountered in 1.1 percent of the tested drivers. Other drugs appeared much less frequently. As indicated in Table 33, many drivers tested positive for more than one drug. Thus we provide as the last line of this table the number of drivers who tested positive for any drug, 96. This constitutes 15.0 percent of the total group who provided oral fluid or oral fluid and blood.
The results for ethanol are given in Tables 34 and 35. Some specimens correlate well between the breath, oral fluid, and blood samples, while others are difficult to explain.
Saliva has been shown to equilibrate rapidly with blood in terms of its alcohol content, being slightly higher on average than whole blood. Saliva/whole blood ethanol concentration ratios have been reported on average to be 1.08 for male subjects (n=48) within the first six hours after drinking (range: 0.84-1.36) (Jones, 1979). Other studies have reported similar ratios:
On the whole, the data appear to correlate well with the paired blood/breath, with some outliers. There were three cases where breath was negative, yet blood was positive at .03, .04 and .08 (oral fluid negative in all these cases).
These discrepancies could be related to possible collection issues particularly for oral fluid. In the early part of the study, the volatility of ethanol was not addressed. In the later location collections, the specimens were frozen as soon as possible after collection, and sent to the laboratory on dry ice. At the laboratory, improved sampling processes were implemented involving recapping and freezing the specimens as soon as sufficient sample volume was removed for testing. In this way, the loss of ethanol was minimized and correlation between the breath/blood and oral fluid improved as the study progressed.
When oral fluid is collected, the swab is placed into a transportation buffer; the device is capped, and sent to the laboratory for testing. The volume of the transportation buffer is 3 mL, therefore when one milliliter (1 mL) of saliva is placed into the tube, the total amount of sample volume for testing is 4 mL, and the drug and ethanol content in the sample is diluted by four. In the ethanol testing, it is possible when correcting for the dilution, a value below the limit of quantitation (LOQ) of the assay was multiplied by four, to give the final result. At the low end of the range (.005–.02), the correlation to the breath test was 17.3 percent; in the range above .02, the correlation to the breath test improved to 35/87 = 40.2 percent. This was also true with the blood samples, with 8.6 percent correlation at the low end of the testing range, improving to 19/87 (21.8%) above .02.
Table 34 presents, for those subjects who were positive for any alcohol in any matrix, the results for breath alcohol (%), oral fluid ethanol (%), and blood ethanol (%). Table 35 summarizes those data.
When breath was positive over the legal limit (.08), oral fluid was positive in 11 cases (11/21 = 52.3%); negative in 2 cases (9.5%) and 8 subjects refused to give a specimen (38%). Blood was positive in 4 cases (4/21 =19%) and subjects refused to give blood in all the other 17 cases.
The total refusal rate for blood was much higher than in oral fluid. From 110 subjects testing positively for alcohol via breath (>0.005), 68 (61.8%) refused to donate a blood sample (one blood sample was missing). In contrast, only 29 (26.4%) refused to give oral fluid.
Fourteen subjects were positive via all matrices (Figure 7).
Figure 7. Pattern of BAC Test Results by Biological Matrix
The collection rate for oral fluid was higher than blood. Initial problems with ethanol volatility during collection and storage have been addressed, and correlation between breath and oral fluid results should improve in future studies.