METHODS OF MEASURING MEDICATION USE/COMPLIANCE IN THE COMMUNITY DWELLING POPULATION OF OLDER PEOPLE
Biochemical measures of patient compliance consist of serum drug concentration and urine assay (blood and urine sampling) to measure drug presence and concentrations. Although this method is objective, it is not practical, convenient, or appropriate for most circumstances, and it is not always reliable. Serum drug concentrations detect compliance problems only for medications that have a long half-life; however, it does not account for individual differences in absorption and metabolism. Also, biochemical measures do not provide information about whether the patient took the proper amount of medication at the proper time. The costs of assays for the assessment of multidrug regimens can be prohibitive.
Behrensdorff and Steentoft (2002) conducted a roadside survey in Denmark to describe the prevalence of medicinal and illegal drugs in passenger vehicle drivers in a rural area. Police randomly stopped 1,000 drivers and requested that they provide on-the-spot saliva samples and complete a questionnaire once they returned to their homes. Drivers who were suspected of illegal driving (impaired driving, driving without a valid license) were not included in the stopped sample, as police wanted to follow their own procedures for illegal driving. Sample drivers participated voluntarily and anonymously; the refusal rate for saliva sampling was 1.9 percent and the response rate for the questionnaire study was 66 percent.
A total of 961 saliva samples were collected using the Cozart RapiScan System—a five-panel drug test designed to detect five main groups of frequently abused drugs: amphetamines, benzodiazepines, cocaine, cannabis, and opiates. Saliva screenings were performed partly by use of the Cozart RapiScan system and partly by a Cozart “Drugs of Abuse Microplate EIA” system. Both systems provide only a preliminary analytical test result (e.g., they detect the presence of a substance), and require a more specific alternative chemical method to obtain a confirmed analytical result (e.g., they do not provide a measure of the drug concentration). Saliva was chosen as the specimen for the study because it is easier, less invasive, and less time consuming for police to collect and handle than urine. In addition, the RapiScan saliva screen provided promising results in an earlier study in Scotland (Oliver and Seymour, 1998), and it was the only instrument available on the market that included five substances for screening. A drawback of the saliva screening systems used, however, is that neither is capable of tracing all types of benzodiazepines available on the Danish market. All samples were sent to the Department of Forensic Chemistry, Institute of Forensic Medicine, University of Copenhagen, for screening and confirmation by specific methods. After screening hundreds of samples using the RapiScan system, it was concluded that the test was not reliable and reproducible. Therefore, the RapiScan system was rejected and replaced by the Cozart “Drugs of Abuse Microplate EIA” system for screening of the 861 of the 961 samples with enough remaining saliva. Only the samples that screened positive were further analyzed using gas chromatography-mass spectrometry (GC-MS) for amphetamines, cannabis, cocaine/benzoylecgonine, and opiates. The benzodiazepines were confirmed by two methods: by gas chromatography with an electron capture detector, and by liquid chromatography with a diode array detector.
In the research conducted by Behrensdorff and Steentoft (2002), 7.1 percent of the sample (64 of the 896 stopped drivers who provided saliva) screened positive using the initial screen. In the confirmatory analysis, 2 percent (18 drivers) tested positive for one or more benzodiazepines or an illegal drug (amphetamines, cannabis, cocaine, or opiates, e.g., the drugs included in the screening method). The ages of the 18 drivers were not reported. In 0.7 percent (6) of the saliva samples, one or more benzodiazepines were detected: diazepam, bromazepam, alprazolam and oxazepam (all tranquilizers). The authors note that some of the frequently used benzodiazepines (in Denmark) would not be picked up by the screens used in this study, and therefore the number of samples that screened positive for benzodiazepines underestimates the prevalence of this drug in the population of rural drivers. The great difference in the number of positive saliva samples by the initial screening compared to the confirmatory analysis underscore the importance of conducting specific confirmation methods in combination with initial screening.