Medication safety in community pharmacy: a qualitative study of the sociotechnical context

Background While much research has been conducted on medication safety, few of these studies have addressed primary care, despite the high volume of prescribing and dispensing of medicines that occurs in this setting. Those studies that have examined primary care dispensing emphasised the need to understand the role of sociotechnical factors (that is, the interactions between people, tasks, equipment and organisational structures) in promoting or preventing medication incidents. The aim of this study was to identify sociotechnical factors that community pharmacy staff encounter in practice, and suggest how these factors might impact on medication safety. Methods Sixty-seven practitioners, working in the North West of England, took part in ten focus groups on risk management in community pharmacy. The data obtained from these groups was subjected to a qualitative analysis to identify recurrent themes pertaining to sociotechnical aspects of medication safety. Results The findings indicated several characteristics of participants' work settings that were potentially related to medication safety. These were broadly classified as relationships involving the pharmacist, demands on the pharmacist and management and governance of pharmacists. Conclusion It is recommended that the issues raised in this study be considered in future work examining medication safety in primary care.


Background
Medication safety has long been recognised as a key issue within the broader patient safety agenda [1][2][3]. A number of studies have shown that medication error is relatively common and identified a range of contributory factors occurring at the individual (for example, fatigue and training) and organisational (for example, staffing, organisational climate and system design) levels of analysis [4][5][6].
Much of this research has been conducted in secondary care settings, with relatively few studies taking place in primary care, and fewer still focusing on the dispensing of medicines in community pharmacy. However, a large proportion of medication prescribing and dispensing occurs in primary care settings [3]. The British National Health Service [7] reports that approximately 11,000 community pharmacies existed in England and Wales, dispensing approximately 785 million prescription items each year. While there is no population-level estimate for the incidence of adverse events in primary care dispensing, a prospective self-report study in a sample of 35 British community pharmacies found an incidence rate of 22 near misses and four errors for every 10,000 items dispensed [8]. A similar error rate (0.26%) was found in an observational study of a single pharmacy in the United States [9]. In both cases, errors were attributed to a range of factors, including misidentification of drugs, workload, distractions and dispensing against product labels created during the process rather than against the prescription itself. Meanwhile, Peterson, Wu & Bergin [10] note concerns among Australian community pharmacists that an increase in workload is creating more opportunities for dispensing errors to occur, and Szeinbach, Seoane-Vazquez, Parekh & Herderick [11] found a relationship between volume of prescriptions and frequency of dispensing errors in a self-report survey of US community pharmacists.
These studies highlight the need to address what have been termed "sociotechnical" factors in medication safety. Sociotechnical factors are those concerning the relationship between the technical, psychological and social elements of a work system [12]. Their relevance to community pharmacy is illustrated by Ashcroft, Morecroft, Parker & Noyce [13], who note the influence of both technical elements (such as workload and staffing) and social elements (attitudes towards incident reporting and organisational learning) on the proliferation of medication incidents. During their investigation, Ashcroft et al. collected qualitative data from a sample of community pharmacy staff. Their analysis of this data, though, focused only on reporting and learning from medication incidents, their aim being to develop a safety culture assessment tool [13,14]. The aim of the present study is to extend Ashcroft et al.'s initial work by identifying sociotechnical factors that participants felt were related to medication safety in general -that is, the causation, detection and prevention of medication incidents as well as their reporting.

Method
The authors carried out a secondary analysis of the qualitative data collected for Ashcroft et al.'s [13] study; the original aim of which was to explore participants' views about risk management culture in community pharmacy. The data was collected from ten focus groups conducted in the North West of England between December 2003 and April 2004. These focus groups involved a total of 67 participants, all working in community pharmacy. Participants were recruited on a purposive basis, having taken part in a previous questionnaire study by the authors and consented to be contacted about further research. The participants represented a range of roles and locations across the region, as shown in Table 1. Ethical approval for the research was granted by the University of Manchester Senate Committee.
Each focus group lasted for approximately two hours and was moderated by an experienced qualitative researcher. The topic guide was semi-structured, and included the following themes: • What are the attitudes and views of pharmacists and support staff about patient safety in community pharmacy? Note: "Locum" is the UK term for sessional pharmacists. "Pre-registration" is the UK term for trainee pharmacists. "Support staff" is a collective term for counter assistants and non-pharmacist dispensers.
• What is the prevailing culture with respect to patient safety in community pharmacy?
• How does the prevailing culture affect how incidents are handled?
• What types of incidents are reported/not reported in community pharmacy?
• What factors facilitate and inhibit reporting?
As well as these themes, other discussion topics were allowed to emerge spontaneously during each focus group. All focus group discussions were tape recorded and fully transcribed for subsequent analysis.
Two researchers (DLP and DMA) identified themes relevant to the sociotechnical aspects of community pharmacy using template analysis [15,16]. Template analysis is an inductive process that involves the analyst reading through the data and creating a "template" consisting of the general themes that emerge from this reading. The template is then modified and extended through successive readings until it provides sufficient coverage of the data. In the current study, the researchers began by reading the entire dataset and identifying the general issues that participants raised in relation to medication safety. These issues became the superordinate themes. They were used as a framework to categorise the data from each focus group, following which each theme was developed by identifying subordinate themes in the data assigned to it. Version 7 of NVivo [17] was used to document the analysis, which continued until the analysts had identified as many themes as possible from the data. As DLP and DMA worked separately to identify superordinate themes, reliability of the analysis was established by comparing the two set of themes for consistency, with the final template incorporating both sets; this being a common approach to establishing reliability in qualitative studies that use a realist epistemological framework [18]. The final template, consisting of the superordinate and subordinate themes, was reviewed by PRN and DP to confirm the relevance of the themes to medication safety.

Results and discussion
Across all of the focus groups, the factors thought to affect medication safety were summarised using three superordinate themes: relationships involving the pharmacist; demands on the pharmacist; managing pharmacy work. These themes, and their subordinate themes, are listed in Table 2.

Relationships involving the pharmacist
The first theme -relationships involving the pharmacistreflected the interdependencies and interactions between the pharmacist and other people. These include peers, other health care professionals outside the pharmacy, and service users.

Peers
Pharmacists highlighted the role of their professional peers in maintaining safe practice. As with other professional groups, collective norms of practice may emerge amongst pharmacists. However, while some of these norms may encourage safe practice, others may discourage it.
[ The effect of professional norms on safety-related behaviour has been noted in previous studies; for example, anaesthetists' approaches to the use of guidelines are influenced to some extent by how they expect their peers to act [19,20]. Hence one of the ways to achieve medication safety is to encourage pharmacists to develop group norms that support safe behaviour.
Other comments made by participants referred to perceived differences between agency-supplied (locum) pharmacists and permanent (employed) pharmacists with regard to involvement in pharmacy activities. The locum pharmacist talks of his disconnection from the day-to-day activity of the pharmacy; this can lead to pharmacists (whether employed full-time, part-time or locum) being confronted with an ongoing or complex situation of which they have only partial knowledge. This may become more of a problem as pharmacies extend their opening hours, increasing the need for shift handovers between both pharmacists and support staff [21]. In any case, there is perhaps a need for more documentation of cases for the purposes of information sharing.
Other health care professionals Some participants commented on the relationship between pharmacist and prescriber. In particular, issues related to the partnership working that occurs between the two, for example the pharmacist agreeing to dispense an incomplete prescription on the understanding that the prescriber will issue an amended version later. However, there was also a fear on the part of pharmacists that they would be blamed for medication incidents that actually originated with the prescriber.
When a prescription comes into us we're relying on the fact that the GP has made the correct diagnosis and has prescribed a dose that is safe for that patient. [ These excerpts point to a tension within the pharmacistprescriber relationship. On the one hand, pharmacists are keen to protect their dispensing from substandard prescribing, particularly if they may be held culpable for any problems that result. On the other hand, pharmacists are dependent on prescribers for their business, and so may be reluctant to disrupt their working relationship. Bradley et al. [22] have noted that trust in each others' capabilities is fundamental to collaborative working between pharmacists and GPs. It could be hypothesised that the extent to which both parties are willing to collaborate over It is interesting to consider where the power lies in the pharmacist-customer relationship. From one point of view, the pharmacist has a basis for power, as it is he or she rather than the customer who has insight about performance and safety issues in the pharmacy, as well as being the one who is responsible for managing any interactions with the customer. However, some of the comments made suggest that the customer has some influence over the pharmacist; the latter perceives pressure to turn around medicines quickly for the customer. This may be of particular relevance to medication safety given that one source of high workload could be the perceived or actual productivity demands from customers, as well as interruptions to the pharmacist [10].  [23] argues that, from a safety engineering perspective, it would be difficult to ensure that patients provided a consistent and reliable safety check, given the diversity of patients and a possible unpreparedness to take responsibility for safety of their treatment. However, Entwistle [24] argues that, from a medical ethics perspective, it would be beneficial to involve patients in the final check, even if it is impractical to place complete reliance on them to do the check themselves, as this can enhance carer-patient communication and provide an opportunity for patient education. The extent to which patients should be involved in medicine checking in community pharmacy, and how much pharmacists want them to be involved, is a question that may warrant further investigation.

Demands on the pharmacist
As well as interacting with different people and groups, the community pharmacist operates within a number of constraints. These arise both from business needs and from the legal and regulatory context of practice.

Commercial
In the UK, community pharmacy differs from many other healthcare services in that it operates as a private business, rather than being government owned. This means that a key concern for community pharmacists is the cost of their services and the financial impact of any decisions made in the course of their work. Commercial concerns could potentially influence the attitude of pharmacists to service provision.
We The first two quotes concern the integration between community pharmacies and the not-for-profit sections of the health service. The pharmacist in Group 4 expresses a reluctance to report errors to the body that is responsible for funding the pharmacy, while the pharmacist in Group 8 suggests that community pharmacists have more to lose than do other healthcare providers by reporting others' bad practice. Hence, the relationship between pharmacists and the safety management functions of the health commissioning organisation may be influenced by the pharmacists' need to protect their business. The third comment refers to a perceived conflict between the need to meet commercial targets and the need to ensure adequate resources -in this case, staff. Shortage of staff, though, may have adverse consequences for safety in the face of increasing workload in pharmacies [4,10].

Corporate
In the UK, community pharmacies come in a number of forms -some operate as independent stores; others as small chains consisting of several stores; yet others as larger regional or national chains, often integrated with a general department store or supermarket. As organisations in general vary in terms of their culture and hence operating practices [25], so do these different types of pharmacy. In general, larger chains were seen as having a more centralised approach, in which pharmacists at the various chain locations were expected to take direction from senior managers based in a headquarters.

I think the protocols that [chains] have as well tend to be stricter [than in independents], and they won't let you bend from the protocols [Employed pharmacist, Group 6]
No, because they've got so many branches to cover, they've got to put it down as a must do, rather than a, well, we'll get round it type of thing [Locum pharmacist, Group  As the first two quotes suggest, the centralised approach of pharmacy chains carries some benefits for safety management. First, by establishing a managerial hierarchy and formally defining standards of practice, it provides a form of quality assurance. Second, the organisation may have mechanisms for sharing resources and knowledge. However, the third quote surmises that differences in morale and communication between staff may be found across the different types of pharmacy. Whether this is the case is not clear from the data in this study; however, Ashcroft & Parker [26] suggest that staff commitment and communication (as well as resourcing) are amongst the factors associated with safety in pharmacies. Hence, it would be of value to evaluate the effect of organisational structures and practices on safety.

Legal and regulatory
Pharmacy practice in the United Kingdom is governed both by legislation such as the Medicines Act, and by the regulations of the pharmacy professional body, which at the time of writing is the Royal Pharmaceutical Society of Great Britain. Hence there are statutory requirements by which pharmacists need to abide. As autonomous practitioners they are very sensitive to the risk of disciplinary action or litigation should a patient be harmed. This emerged as a key driver for safe practice. There was a general disinclination on the part of participants to incur the wrath of the law or the regulator. However, some participants described situations in which pharmacists may feel they should "bend" the law, suggesting a direct conflict between doing "the right thing" and staying within legal boundaries. Interestingly, the last quote suggests that the regulator may also play a role in arbitrating between the two imperatives. Benson, Cribb and Barber's study of pharmacy practitioners [27] also described situations in which ethical responsibilities and legal responsibilities were felt by the respondent to be in conflict; however, the relative frequency of such situations is not fully clear. Another issue to consider is which course of action (legal or ethical) is felt by pharmacists to be most closely aligned with safe practice.
A further potential role that participants identified for the regulator is to negotiate with pharmacy employers and drug manufacturers in order to ensure adequate resources for safe practice. We

Management and governance
In order to manage the work of pharmacists, both their employers and their professional regulator may make use of a range of methods. These include governance processes such as incident reporting, regulating practice through the use of protocols, and the physical design of workplaces and tools. All of these can influence the work of pharmacists as shown in the following examples.
Blame culture vs learning culture Some of the participants recognised the benefits of developing a culture in which incidents were openly discussed and lessons shared and acted upon. However, many participants felt that a barrier to adopting such practices is a tendency to attribute blame to individuals unnecessarily. The effect of blame culture on safety management in healthcare settings has been discussed in depth in previous literature [28]. Here, it will suffice to say that the comments made by participants in this study are consistent with the findings of previous studies; it would appear that sensitivities about being seen to blame colleagues, or being the subject of blame oneself, may provide a barrier to exploring and learning from adverse events in pharmacy practice [29].

Formal vs informal approaches
While there are usually formal governance processesmainly based on reporting and audit -some participants alluded to the alternative use of informal approaches. For example, a pharmacy store manager may, having found out about a dispensing error, discuss it with the people involved in the first instance, only resorting to formal reporting if the matter cannot be resolved informally. Also, some participants noted that their level of trust in governance processes depends on who is administering them. There's no point in being proactive to a system or management or a body which is itself being reactive and disciplinary, because that defeats the point of you being proactive in the first place [Pharmacy owner, Group 10] The data suggested that pharmacists will engage with any governance system -whether formal or informal -to the extent that it supports development of practice rather than sanctioning individuals or sites. This may reflect the presence of a pharmacist community of practice -that is, a social group that fosters collective learning and norms of practice [30]. Communities of practice tend to be informal and peer-led, rather than management-led, and so healthcare professionals may prefer to engage with them than with formal governance structures [31]. The implication is that those responsible for managing pharmacists should be aware of the existence of communities of practice and work with them, rather than against them, for mutual benefit.

Protocols
Protocols were recognised to be important in principle, as they set minimum standards for practice. However, some participants felt that an over-reliance on protocols ran counter to their belief in being autonomous professionals. The participants explained that there can sometimes be a conflict between what a protocol dictates that the pharmacist should do and what the pharmacist judges to be the right thing to do in the situation. It may be the case that a number of factors play into the decision whether or not to follow a protocol -for example, the nature of the protocol, the experience of the pharmacist and the nature of the situation at hand [19,20]. Nevertheless, deviations from protocols have been implicated in adverse events in other healthcare domains [32], and so there is a need to consider the relevance of pharmacy protocols to medication safety.

Work design
A key feature of the work environment that was reported to influence pharmacists' work is its physical manifestation -that is, the layout of the workspace and the equipment that is found within it. New technologies -such as decision support or automation -may change the nature of the pharmacist's task activity and demands. Szeinbach et al. [11] and Rapport, Doel & Jerzembek [33] have observed that the topography of community pharmacists' work areas can either facilitate or interfere with their practice. For example, Rapport et al. note that a wellorganised workspace engenders a sense of professionalism and reduces the likelihood of dispensing error, while layouts that make pharmacists too accessible to customers can invite ill-timed interruptions to consultations or dispensing. Respondents to Szeinbach et al.'s survey suggested that the configuration of the pharmacy (that is, its shape and the presence of a drive-through window) could affect the efficiency and accuracy of dispensing; automation, though, was generally felt not to have a negative impact on efficiency and accuracy. As suggested by the quotes here, participants in the current study held similar views about layout to those expressed in these previous studies. The fourth quote refers to the impact of automated dispensing; the pharmacist's role -much like other operators working with closed-loop control systems, such as pilots and anaesthetists -becomes one of monitor and decision-maker rather than hands-on assembler. This change may or may not be welcome to particular pharmacists.

Study design: strengths and weaknesses
The use of semi-structured focus groups provided the researchers with a large volume of rich data about the experiences of community pharmacy staff. The openended nature of the data collection allowed the participants to raise other issues that might not have been considered by the researchers. However, the findings may be limited in that the sample was confined to community pharmacy staff working in the North West of England. While the researchers expect that the findings apply to pharmacies elsewhere in the United Kingdom, if not internationally, it is possible that the study has highlighted some issues that are specific to the geographical location in which it was conducted. The sample consisted predominantly of locum pharmacists. While a 2002 workforce survey also indicated a high proportion of locums in comparison to other types of pharmacists [34], it is possible that the views of locums have been given undue predominance in the current study. Support staff and technicians had relatively little representation in the sample. Also, there is a time lag of five years between the focus groups and the publication of the current paper. While the authors would suggest that the issues raised in this paper remain of relevance, it is acknowledged that they predate more recent developments in pharmacy practice such as extended services (such as supplementary prescribing) and registered pharmacy technicians (who, unlike many support staff at the time of the present study, have nationally-recognised formal training and professional accreditation), as well as the forthcoming transfer of pharmacist regulation to an independent council. Any of these developments may change the views of pharmacists about sociotechnical influences on safety; alternatively, the changes may be viewed from the perspectives described here. In any case, what cannot be inferred from this study is the predictive or causative relationship between any of the factors indicated and measures of quality or safety. Such a link should ideally be a focus for future research. One way of studying this may be to use the factors discussed here as the basis for a questionnaire, the data from which could be subjected to factor and regression analysis.

Conclusion
While medication safety might be viewed in terms of the dispensing process itself, the focus group data from community pharmacy staff indicate various social and organisational factors that also have a potential impact. Some of these issues are common to both hospital and community dispensing. However, others are peculiar to community pharmacy, in particular the strongly commercial nature of services in this setting. Either way, it is important to examine their origin, nature and influence on safety.
Publish with Bio Med Central and every scientist can read your work free of charge