This review of the literature has revealed a number of themes and data to inform understanding of OTC medicine abuse, However, what is perhaps most apparent is the extent of the omissions
in the extant literature, particularly as they relate to the lack of:
- qualitative methods that may be appropriate for exploring individual perspectives;
- reliable quantitative data in some countries;
- fully evaluated or implemented interventions;
- data relating to Internet supplies; and
- consensus over definitional terms.
These concerns are now considered in turn, before a number of specific suggestions for further research and policy involvement are proposed.
The various definitions described previously have a number of implications for research and understanding in this area. First, whilst they can positively reflect a range of different types of societal medicine use, they may also lead to confusion, particularly if, like some studies did, there are not accurate and consistent attempts to distinguish between them. This may be further complicated by the origins of these terms, with some such as “dependency” and “abuse” being associated with a clinical or diagnostic perspective (American Psychiatric Association, 2000
), “addiction” carrying a societal broader interpretation and “misuse” being associated with pharmacy studies particularly. This reflects enduring debates about and changes to terminology in the wider addiction literature, including the WHO’s adoption of “dependence” over “addiction” (World Health Organisation, 1964
) nearly half a century ago, to recent debates about these terms in the DSM-IV and proposed DSM-V (Dean & Rud, 1984
; O’Brien et al., 2006
). Underscoring this definitional variation are also fundamental issues about stigma, identify and also agency. The use of the term “dependency” and not “addiction” has been argued to have occurred due to issues of stigma of the latter (Dean & Rud, 1984
; Erickson & Wilcox, 2006
; Reay, 2009
) as well as the issue of an “addict” or “spoilt” identity (Goffman, 1990
; McIntosh & Mckeganey, 2000
). In terms of agency, it is interesting to reflect on the distinction between misuse and abuse in some of the extant literature, since this appears to recognise a difference between intentionally experimenting with a medicine (to elicit a different effect) and abusing it, and unintentionally deviating from standard use (taking at different dose or indication) and therefore misusing it. Whether these can be adequately mapped onto additional concerns about the loss of control in addiction, as argued by Reith (2004
), for example, are additional issues. One further omission is the absence of any reference to pseudo-addiction in the OTC literature identified in this review. Pseudo-addiction has been defined as the under-treatment of pain (Bell & Salmon, 2009
), which may lead to symptoms that are similar to dependency and which reveal a potentially even more complex area.
Methodologically, quantitative approaches have dominated, illustrated by the use of cross-sectional descriptive survey designs, often using self-completion postal surveys of pharmacist participants in UK studies. Response rates appear to have varied significantly using this approach, and whilst Matheson et al. (2002
) reported very good response rates across two surveys using a prepaid envelope and two reminder letters, and Hughes et al. (1999b) received responses from just under half of pharmacists sampled using two mailings. These studies reflect a trend to using pharmacists proxies and hence obtaining data that reported on pharmacists’ perceptions of the problem and the profile of those they considered to be affected, which as Orriols et al. (2009
) noted is “much too subjective to obtain reliable qualitative and quantitative data”. Although not explicitly noted by the researchers, this may reflect a belief that those who are abusing or misusing OTC medicines may be a hard-to-reach or covert
) group and hence using pharmacist proxies is perhaps perceived as being more appropriate. However, several study designs have involved sampling those suspected of abusing/misusing OTC medicines, either via pharmacies (Phelan & Akram, 2002
; Orriols et al., 2009
), at targeted venues such as gyms (Kanayama, et al., 2001
) or by post (Sproule et al., 1999
). Although these represent less subjective accounts of the problem, they have resulted in poor response rates except in the study by Orriols et al. (2009
), who argued that allowing purchasers to complete a questionnaire away from the pharmacy and return it via post, as compared to completing it in the pharmacy, meant those who were abusing or misusing could complete the forms anonymously. However, Orriols et al. (2009
) were disappointed by the poor level of pharmacy participation, which may be related to the need for the pharmacies involved to undertake the administration of the questionnaires, as was identified in other studies (Wazaify et al., 2006
Of particular note is that qualitative methods have been neglected and only one identified study used focus groups (Björnsdóttir et al., 2009
) and one which reported the use of semi-structured interviews (Mattoo et al., 1997
) presented detailed statistical data and the absence of qualitative data suggested this was a structured survey design. Nielsen et al. (2010
) used qualitative interviews and reported a range of different types of abuse of codeine, as well as barriers to treatment, illustrating the unique data that this method can generate. Adopting such methods may reveal further insights that could help understanding of the contested definitional issues raised above, as well as providing more than the proxy summaries of those perceived to be affected, as offered by some pharmacist-participant studies.
The use of secondary data sources, such as those in various US reports (Substance Abuse and Mental Health Services Administration, 2004
) and using details of patients attending drug treatment centres in South Africa (Myers et al., 2003
), for example, offers potentially more robust statistical information on the extent of the problem. However, such data are not unproblematic and in the case of some US data, for example, prescription and OTC medicines were often reported together.
Linked to the source of this last type of secondary data is any evaluation or indeed thorough detail of treatment options for those affected by OTC medicine abuse. Empirical studies have identified a range of often pragmatic solutions, but evidence-based interventions and attendant evaluations are a clear omission in this field.
Finally, the emergence of new forms of medicine supply, such as via the Internet, in what Fox et al. (2005
) termed the “second moment” of “e-pharmacy” has not been studied, despite being recognised as a potential threat (McBride et al., 2003
). Such developments may not only stretch the metonymic accuracy of the term OTC, but also require a redefinition of what such supplies involve, as such supplies transcend national boundaries and attendant regulation in many cases (Bessell et al., 2003
) and may challenge the international patterns identified.
In relation to policy, this review confirms that there is a problem in a number of countries but concerns about what is being investigated – whether this is misuse, abuse, dependency, addiction or pseudo-addiction – coupled with a lack of systematic data on the scale of the problem make appropriate and proportionate policy-based interventions difficult to consider. There exists a tension between making OTC medicines available to individuals to increase their access to medicines and enabling them to self-manage conditions and accepting that there is some degree of risk of such products being misused or abused, with potentially serious consequences for some. Raising awareness of potential problems of OTC medicines, as the recent response in the United Kingdom has illustrated in terms of making purchasers aware of the possibility of addiction, would appear a prudent response. But whilst this may arguably warn those using products for the first time, for those with an existing problem, more support may be needed in the clinical pathway.