Quality Assurance of forensic investigations in toxicology and traffic safety

Mullen, Carrie (2014) Quality Assurance of forensic investigations in toxicology and traffic safety. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2014mullenphd.pdf] PDF
Download (6MB)
Printed Thesis Information: https://eleanor.lib.gla.ac.uk/record=b3084786


The work described in this thesis deals with three aspects of quality assurance in the field of forensic toxicology: proficiency testing schemes, validation of analytical methods for the piperazine group of abused drugs and validation of the police field impairment test, used at the roadside to test drivers for drug-induced impairment.
Proficiency Testing:
Long term reviews were performed for two forensic external quality assurance schemes. Rounds 30 (in 2007) to 48 (in 2012) of the UKAS-accredited commercial Quartz Forensic Blood Toxicology Proficiency Testing Scheme (PTS), and a ten year period from 1999 to 2009 of the freely-available United Nations Office on Drugs and Crime (UNODC) International Collaborative Exercises (ICE). Only limited ICE data could be made available as much of the original data had been stored on a database which had become obsolete, hence the data were only available as the original results forms provided to UNODC by the ICE participants. Data was entered to Microsoft Excel® spreadsheets and Microsoft Access® databases from the original forms for the years 1999, 2001 (2 rounds), 2003 (2 rounds) and 2005 (2 rounds), and summary data was extracted from the UNODC round reports for the years 2007, 2008 and 2009.
Four methods of scoring quantitative performance were reviewed and the most suitable, a z-score using an assigned ‘true’ value and a percentage of the true value as acceptable deviation, was applied to reanalyse the participants’ results and assess their performance. Methods of scoring proficiency which relied upon participants’ data to determine acceptable variation were found merely to describe the data rather than challenge participants on whether or not they were performing fit-for-purpose analyses.
Factors such as participation, analytes tested, participants’ methods of analysis and participants performance were summarised for each scheme before the performance of the two schemes, and that of their participants, were compared.
ICE tested more analytes per annum but from a smaller test menu than Quartz. This resulted in more repetitive testing and allowed for some trend analysis and performance monitoring. It was not possible to observe performance trends with Quartz due to the wide variety of analytes tested. The smaller array of potential analytes and more repetitive nature of ICE testing also meant that performance monitoring and detection of bias were easier to perform, and ICE was shown to be more effective as external quality assurance (EQA). Quartz provided a good educational resource as it incorporated the wide range of drugs which a forensic toxicology laboratory could realistically encounter. Following the review, however, it was recommended for QUARTZ that, to provide a safeguard against bias, more repetitive testing was required and this has now been adopted.
All piperazine analogues are now illegal in the UK, registered as Class C of the Misuse of Drugs Act (1971) and schedule 2, part III of the Misuse of Drugs Regulations (2001). Piperazines can elicit similar effects to some ATS and methods for their detection should be available in forensic toxicology laboratories.
In the present study, methods were developed for the detection of a range of piperazines in blood using LC-MS/MS (p-MeOPP, p-FPP, BZP, o-MeOPP, p-MPP and TFMPP) and GC-MS (p-FPP, BZP, TFMPP, p-MPP, o-MeOPP, m-CPP, p-MeOPP and p-CPP). Quality assurance required both methods to be validated. For all piperazine analytes accuracy was within ±15% (20% at low concentrations) and precision was within 15% (20% at low concentrations). For both methods LLOD of all analytes was 5 ng/ml of blood and upper limit of quantification was 2 µg/ml of blood. For the GC-MS method lower limits of quantification (LLOQs) were in the range 20 to 30 ng/ml of blood. For LC-MS/MS, LLOQs ranged from 50 to 60 ng/ml of blood, although quantification by the LC-MS/MS method was restricted by the lack of availability of appropriate internal standards. There were no apparent significant matrix effects and recovery by both methods was >60 % and, therefore, acceptable.
Short term stability of the piperazine analytes was investigated. Piperazines remain sufficiently stable when stored in the fridge for at least one week, and are stable through three freeze-thaw cycles. There was no detectable degradation when blood samples were left on the bench-top or when extracted ‘in-process’ samples were left in the autosampler for up to 72 hours.
The LC-MS/MS method could provide a readily applicable screening method. A small aliquot of a basic drug extract could be screened by LC-MS/MS for the presence of piperazines, leaving the majority of the extract for other analyses, for example, piperazines confirmation or amphetamines analysis. The GC-MS method was suitably validated to provide quantification but application to casework samples remains to be evaluated. It is recommended that piperazine testing be performed for all suspected MDMA or ‘club drug’ intoxication cases.
The Field Impairment Test:
The detection of drugged drivers primarily depends on the current method which is the driver field impairment test (FIT). FIT comprises measurement of pupil diameter and four physical tasks (the Romberg balance test, walk and turn test, one legged stand and finger to nose test) intended to simultaneously test comprehension, short term memory, balance and motor function. Despite FIT having ISO accreditation, it has been recognised that police officers lack confidence with the protocol and do not apply the test as often as is necessary. The main difficulty arises from the requirement to make a subjective judgement of impairment and officers lack confidence in their ability to do so.
FIT has never been fully validated. The present study was designed to meet the urgent requirement to develop FIT into an objective measurement, by determining what constitutes “normal” performance in FIT by unimpaired adults of different ages. FIT performance was recorded for 79 individuals, a statistically determined cohort size, confirmed by breath and oral fluid analysis not to be under the influence of impairing substances. Each error made during FIT, as defined by the FIT standard operating procedure, was recorded and collated in a Microsoft Excel® spreadsheet for analysis. It was found that the definition of ‘errors’ was too stringent as many which are required to be recorded are normal physiological or behavioural characteristics, such as body sway, and most subjects would be unable to complete the task without displaying them. A less stringent, evidence-based definition of “error” was developed which allowed statistically more significant analysis to be performed on the FIT results.
A statistically significant difference (P=0.00578) was shown to exist between the FIT performance of individuals under the age of forty years and those aged forty and over. Based on the principles of a PTS, robust mean and standard deviation were used to determine what constituted acceptable performance. Those in the younger age group could be considered impaired if the police officer witnessed more than seven errors, or, in the older age group, more than fifteen errors. Using these criteria the frequency of false positives, i.e. unimpaired drivers being assessed as impaired is estimated to be (less than 3%). Also, the ranges of errors observed in both groups was large and overlapped, such that it may be possible for an impaired person to appear unimpaired. This requires further investigation.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Keywords: piperazines, validation, forensic toxicology, field impairment testing, proficiency testing, drugs and driving
Subjects: R Medicine > RA Public aspects of medicine > RA1001 Forensic Medicine. Medical jurisprudence. Legal medicine
Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Medicine, Dentistry & Nursing
Supervisor's Name: Anderson, Dr. Robert A.
Date of Award: 2014
Depositing User: Ms Carrie Mullen
Unique ID: glathesis:2014-5465
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 16 Oct 2014 07:41
Last Modified: 18 Oct 2017 12:21
URI: https://theses.gla.ac.uk/id/eprint/5465

Actions (login required)

View Item View Item


Downloads per month over past year