« Previous | Contents | Next »
Listen
CHAPTER THREE: THE FIELD-BASED STUDY METHODS
3.1 As said previously, this second stage aimed to conduct an investigation into the impact that paraphernalia supply makes on sharing and health in the practice setting and compare this with non-supply. A number of objectives were set for this:
Objectives of the study
a) Perform a study comparing health and sharing practices of IDUs in a location where needle and syringe exchange plus paraphernalia is supplied, with the same measures taken in a location providing needle and syringe exchange only. Establish whether there are any differences and if so, whether these can be attributed to paraphernalia supply.
b) In the location where paraphernalia is supplied, establish participants' views on the paraphernalia supplied, including ease of use, self-reported nature of use, compatibility with the injection preparation process and its perceived impact on health and sharing.
c) In the location where paraphernalia is not supplied, establish participants' experiences of access to paraphernalia items needed, the items they use and perceptions on their impact on health and sharing.
d) Identify participants' suggestions and ideas for promotion of use of appropriate paraphernalia and ways to discourage sharing.
Design
3.2 A comparative cohort study with baseline and six month follow up comprising two groups of participants recruited from two needle exchange services in geographically distinct areas was planned, but due to difficulties in follow up, this became a cross sectional comparison study (i.e. no follow up). One group were receiving the intervention, which was the supply of injecting paraphernalia and the other received 'standard needle exchange care', which was no paraphernalia except swabs. The locations were separated by a distance of approximately 70 miles. This was to minimise the chance of those in the standard needle exchange group encountering IDUs in the intervention group and potentially obtaining paraphernalia.
3.3 This design method was chosen for pragmatic reasons after other possibilities were explored but deemed unsuitable, as discussed below.
3.4 Consideration was given as to whether it would be possible to use a randomised controlled design. The options were: to conduct the study at one location and randomise participants to intervention or no intervention; or, to conduct the study at many sites and randomise the sites to supply paraphernalia or not.
3.5 Randomisation at individual person level was considered inappropriate as injectors often present to needle exchanges with partners and peers. Cluster randomisation of social groups was considered. However, it was felt that in both designs it would not be possible to prevent 'contamination'. What is meant by this is that due to drug users often knowing other drug users and sometimes injecting in different groups, it would be impossible to prevent those receiving the paraphernalia (the intervention group) from passing it on to others who may be in the control group (i.e. no paraphernalia).
3.6 The possibility of cluster randomisation using a number of sites in different geographical regions was ruled out for two reasons: Firstly, lack of suitable areas. Most services in Scotland large enough to potentially provide the required number of participants had already begun supplying paraphernalia. It was a request of the funder that the study be performed in Scotland. Asking services to withdraw supply was considered inappropriate and initial discussions with providers confirmed that this would not be supported. In Dundee, the Harm Reduction Centre ( HRC) had been supplying paraphernalia for many years, with makeshift filter and tourniquet supply being available in the 1990's under local agreement. Further paraphernalia supply e.g. Stericups, was introduced by HRC subsequently, initially as part of a pilot trial. Distribution of paraphernalia became Tayside wide in 2003, and tourniquet supply stopped. Secondly, such a study design would have been very costly.
3.7 It was also recognised that the paraphernalia used by IDUs in either location could not be controlled. It could not be guaranteed that in the intervention location that participants would not on occasions use items other than those supplied. No accurate way of measuring paraphernalia used for every injecting episode could be identified, after the pilot phase (see later) attempted this but found reliability and recall to be problematic.
3.8 Hence the decision was taken to use a pragmatic study using distance as the main means of contamination prevention. Demographic data was collected from participants to allow the two groups to be assessed for comparability.
The intervention being studied
Original planned intervention
3.9 At study inception it was proposed to compare the supply of a safer injecting kit with non supply. However due to circumstances explained below this was adapted to become a comparison between paraphernalia already received with receiving only swabs. The supply of information and advice to needle exchange clients was unaltered from normal practice in both locations.
Changes to the original planned intervention
3.10 Ethical Approval of the study (3.26) was subject to the requirement that normal care was not withheld or any intervention was not withdrawn at the end of the study. On this basis, the study had to be 'naturalistic'. This meant that it had to compare practice in two locations and not seek to reduce supply in the control group or temporarily increase supply in the intervention group.
3.11 Aberdeen was identified as the only major Scottish city (i.e. large enough to potentially have enough injectors to undertake the study) where paraphernalia was not already being supplied, with the exception of swabs. Dundee was identified as a major Scottish city already supplying paraphernalia, having done so in some form since the 1990's.
3.12 It would have been preferred to have compared needles and syringe supply only with the supply of all items recommended from stage one, introduced into a population of drug users who had previously received no paraphernalia. However this was not possible. The study had to be more pragmatic in its design and measures for the following reasons:
3.13 Two appropriately sized geographically distinct Scottish sites not supplying any paraphernalia could not be found.
3.14 In the time between the protocol being first devised and stage two beginning (3 years), HRC in Dundee stopped supplying tourniquets (and later vitamin C) and were unable to introduce sterile water (see below). As said, Ethical Approval required that no change in 'normal' service was brought by the study.
3.15 Similarly the Aberdeen needle exchange providers could not be asked to stop supplying swabs. Therefore this meant that the non intervention group could access needles, syringes and swabs through the exchange scheme.
3.16 At the start of stage two, appropriate sterile water could not be sourced for supply in Dundee as they had planned. Although sterile water had been included in the paraphernalia laws of August 2003, it was still a prescription only medicine in January 2005 (stage 2 start). Steps had begun to agree a Patient Group Directive to allow supply, but an appropriate product was not available on the UK market at the time stage 2 began. In order to prevent sharing the ampoule had to be of small volume (and by law less than 2mls). It also had preferably to be plastic as opposed to glass to avoid the risk of cuts and blood spills from IDUs. Although an earlier agreement had been made with Exchange Supplies to donate their 1.4ml plastic ampoules of sterile water for the study, this could not be achieved as MHRA licensing issues meant they had to withdraw their plastic ampoules from the market 1. No appropriate alternative could be found. 5ml plastic ampoules could be sourced but this volume could promote sharing, was not legal and cost was prohibitive. Note 2ml glass ampoules are now marketed with syringe snappers, aimed at needle exchanges.
3.17 The launch of the Sterifilt prototype that had been used in the laboratory work onto the UK market was delayed. Therefore supply only began in Dundee once data collection had begun.
3.18 Vitamin C sachets were withdrawn from Dundee due to funding issues. The funding for this study was not sufficient to permit continued supply and this would have been against the ethical approval.
The factors under investigation
3.19 At project inception consideration was given to whether hepatitis C virus antibody presence should be measured. It was not considered possible to attribute HCV antibody status to paraphernalia access. HCV negative status would need to be confirmed prior to inclusion and followed up over time. It was considered that the size, duration and level of control necessary to draw any conclusions far exceeded a study of this size or funds. Instead self reported needle and syringe and paraphernalia sharing were investigated as an indicator of HCV risk taking behaviours.
3.20 Skin and soft tissue infections are associated with the sharing of equipment as discussed previously (1.8-1.10). Vascular complications may be aggravated by the use of harsh acids and the injection of particles, potentially due to lack of or ineffective filtering (1.12). The presence of skin and soft tissue infections and vascular complications in participants was recorded. This was based on self reporting and verified wherever possible by inspection by the researcher, who had received prior training in doing this. Sharing practices were also recorded. To add further understanding to the self reported use of paraphernalia and risk taking behaviours a sub group of participants took part in qualitative interviews to explore the use of paraphernalia in more depth. Data collection is described in 3.29.
Participant access to needle exchange supplies in both locations
3.21 Needle and syringe supply in both locations followed the Lord Advocate's Guidance for Scotland. This guidance does not specify paraphernalia quantities. In Dundee paraphernalia quantities supplied were as requested by the client.
3.22 Participants may have been users of several needle exchange outlets in their area (i.e. pharmacy based and agency based). In Dundee the pharmacies supplied paraphernalia, following the same Tayside-wide needle exchange protocol as the agencies. In Aberdeen the pharmacies supplied needle exchange packs that contained needles and syringes and swabs but no other paraphernalia items were supplied at this time.
Control of paraphernalia used
3.23 The paraphernalia used by the participants could not be controlled, in that it would not be possible to stop a participant from using something other than what was being studied. For example, in Dundee it would not be possible to prevent a participant who had access to sterile citric acid sachets from using lemon juice. The study collected data on all paraphernalia recently used by participants to try to capture information on actual items used.
The researchers and base locations
3.24 Data was collected by two part time (0.5 FTE) research officers. One was based at Drugs Action ( DA) in Aberdeen and the other was at The Harm Reduction Centre ( HRC) in Dundee. Data collection ran for 18 months. The researchers recruited participants mainly at the agency base locations. Participants were also recruited from outreach locations, and in Dundee from the Dundee Drug and AIDS Project, another agency in the Tayside needle exchange scheme. Attempts to recruit participants through pharmacy based exchanges were not very fruitful.
Procedures
3.25 Full details of the study procedures are given in the protocol. This and other study paperwork such as the Participant Information Sheet and Data Collection Tools can be provided from j.a.scott@bath.ac.uk A summary of the protocol is given below.
Ethical approval
3.26 Approval of the protocol was given by Tayside NHS Research Ethics Committee (reference 04/S1401/138). Site Specific Assessment ( SSA) approval was given by Grampian NHS Research Ethics Committee. Subsequent substantial amendments were given approval by the appropriate committee e.g. change of researcher.
Recruitment
Inclusion criteria
3.27 Mentally capacitated adults (16 years and over) who were current injecting drug users, willing to participate in the study.
Exclusion criteria
- Under 16 years of age
- Mental incapacity from known history or judgement of researcher/needle exchange staff
- Non injectors e.g. smokers of heroin
Publicity, approach and information for potential participants
3.28 Posters and leaflets, approved by the Ethics Committee, were used to raise awareness of the study to potential volunteers locally. Needle exchange workers in both agencies were briefed on the study by the principle investigator, who ran information sessions in both locations prior to agreement from the agencies to participate. Needle exchange workers highlighted the study to potential volunteers, which were all users of the needle exchanges that were considered to meet the inclusion criteria and not meet the exclusion criteria (this was further verified by the researchers). The needle exchange workers gave those who expressed an interest the leaflet which contains the participant information sheet ( PIS). The researchers also distributed the PIS and posters to pharmacies, homeless hostels, drop in centres and community centres. Potential volunteers were given time to consider participation then consent was sought by the researcher. Those who agreed gave written consent, except for the questionnaire where participation was deemed to indicate consent (see Measurements below).
Data collection
The study used three mixed methods of data collection as follows:
3.29 Quantitative information was collected on vascular health, presence and nature of injecting site injuries and self reported involvement in paraphernalia sharing. This was achieved using the following tools:
(1) Needle Exchange Short Questionnaire ( NESQ)
This was a brief questionnaire administered in both sites. It was completed with the participant by the researcher where possible or in some cases when the researcher was not available, by available needle exchange staff who had been trained by the researcher.
(2) Health assessment
This was always undertaken by the researcher, who received prior training. The participant was scheduled an appointment within the agency so objective health measures could be taken including blood pressure, measurement of extremities, heart rate, respiration rate and weight. Reminders were sent where possible by text message prior to the appointment.
(3) Semi-structured interview
Qualitative data was collected to facilitate understanding of the issues relating to paraphernalia from the participant's perspective. This was done using the third data collection tool. This interview gathered information on self reported injection preparation and administration practices, views on paraphernalia available and sharing practices. Ideas on how to reduce sharing amongst injectors were also sought. The researcher always undertook the interview. After consent was obtained, the interview was micro-tape recorded. The interview schedules asked participants to describe how they prepare and inject. Areas of risk were explored when highlighted. The schedules then varied between Aberdeen and Dundee. In Dundee views on the equipment supplied were sought, whereas in Aberdeen views were gathered on equipment used and thoughts on supply.
Study method and follow-up revisions
Original plan and post pilot modifications
3.30 During the pilot it was attempted to capture information on all injecting equipment used for every injection and detail each episode of sharing since the participant's previous visit to the needle exchange, by administering the NESQ (pilot version) at every visit. However it became obvious that this level of intensity would require greater researcher capacity than was available. The extent of time needed to collect the data at every visit and to contact and arrange follow ups detracted significantly from initial recruitment. Furthermore it would require significant incentives to retain participants. It was also initially planned to use all three data collection tools with every participant. However the pilot phase showed some potential volunteers were comfortable to complete the questionnaire but did not wish to undergo a health check or interview. Reasons given included lack of time and personal discomfort with this level of involvement. Therefore, since study funding was fixed, data collection was modified as described below to increase participation and make best use of available researcher time. The consequences of these modifications are discussed later.
Post pilot questionnaire and recruitment revisions
3.31 The questionnaire was revised to include collection of information on self reported injecting site injuries. Wherever possible these were verified by inspection by the researcher administering the questionnaire. Consent was separated into three distinct phases: Consent to take part in the questionnaire (verbal), consent to the health check (written) and lastly consent to the interview (written). This gave participants the option to take part in the questionnaire only, which collected core data required for quantitative analysis. This meant the health check became optional, which was not as originally desired, but it aided recruitment significantly. It was considered more important to gather in depth information in the interviews than interview every participant, so a sub-group of willing participants was considered appropriate.
3.32 The revised design utilised data collection twice (baseline and 6 months) instead of at every visit. An intensive follow up strategy was devised. Contact details were given by willing participants, including mobile telephone numbers and home phone numbers, or their address/place of residence in the absence of contact numbers. They were also asked to nominate as many 'stable contact points' as they wish, to aid follow up (Pickering, 2003), for example a non-drug using relative. Not all were able or willing to do this. Their needle exchange record was annotated to indicate recruitment and follow up dates flagged to needle exchange workers.
3.33 Supermarket vouchers (Dundee) or household items and toiletries (Aberdeen) to the value of £2 (per visit) were offered, following the pilot, as a small incentive and to thank participants for their involvement in each stage (Pickering, 2003 and Taylor (personal communication)). Incentives were guided by local preference expressed to the researcher. This value was considered low enough not be coercive but a small token to recognise involvement. It was also affordable within the budget, as incentives were not initially included. Consent was taken again at follow-up to check that the person has not changed their mind about participation.
Confidentiality and anonymity of data collected
3.34 The researchers signed up to and worked under the confidentiality policies of the host agencies. Only participant initials and date of birth were known to researchers and recorded on the questionnaires to code them. A participant study number was assigned and this was used to distinguish participants in the study database and on the tapes. Nominated contact phone numbers for follow-up were stored in a mobile phone under participant study number only (i.e. separate from initials and date of birth). Completed paper-based data was stored by the researchers in locked cabinets within the agencies. Data was transferred to Bath via recorded delivery mail and stored by the principle investigator in locked cabinets, as per ethical approval requirements. Data was entered in the database and tapes transcribed in Bath. All data was coded and password protected.
Data analysis
Quantitative data
3.35 Ten percent of the data entered was cross checked for quality assurance purposes. The pilot stage of the NESQ included validation of data gathered against data held in needle exchange records (n=20) and showed the reporting in the NESQ to be accurate. Quantitative results for the Dundee and Aberdeen groups were compared statistically ( SPSS v 14). Statistical guidance and analysis support was given by Dr G Taylor (medical statistician), Research & Development Support Unit, University of Bath. All data was assessed for normality and where appropriate was transformed. Otherwise it was analysed using non-parametric methods. The primary data set was formed from the initial collection data (baseline). Due to reduced follow up data, this was analysed where possible, including all available information. Analysis used a logistic regression model to analyse the prevalence rates adjusting for predicted confounding factors. Where appropriate Chi-squared tests were used. T tests were used if data was found to be normally distributed.
Qualitative data
3.36 The qualitative data was transcribed by the principle investigator and administrative staff at the University of Bath. Transcription checks were performed by the principle investigator. Transcripts were subject to grounded theory thematic analysis to identify key themes that emerged in response to each area of investigation. Progressive focusing was then used in order to group themes. This allowed key questions raised by the quantitative data to be answered, as well as providing additional understanding and information. Quotes given under the results section have been selected from interviewees where it is felt that they captured the meaning of the theme well or they provided particularly vivid insight.
Researcher competence assurance
3.37 Competence assessment of the researchers was undertaken, mapped against the identified competencies for the post listed in the job description. Training gaps were identified and addressed prior to data collection and if any new gaps were identified as part of the ongoing study governance process.
Summary of study method
Figure 14 below summarises the study process and methods used diagrammatically.
Figure 14: Summary of the methods used in stage two of the study
« Previous | Contents | Next »