Discussion

The present study is the first in the country to consider psychotropic medication use in the context of mental disorders. The rate of 14.9% for 12-month use of any psychotropic medication in the general population is at the higher end of reported rates from other countries. NI has the third highest utilisation rate of the participating WMH survey countries, ranking behind France (19.2%) and Spain (15.5%). The consumption rate is also higher than the European average of 12.3% reported in the European Study of the Epidemiology of Mental Disorders (ESEMeD; Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004). However, it is important to consider the difference in the years of studies. Our study was conducted between 2004 and 2008, with the ESEMeD conducted between 2001 and 2003. Given that several studies suggest an increase in the use of these medications over the years, this may account for our higher rate. Then again analysis of our earliest data suggests a rate of 16.5%, higher than five of the six countries in the ESEMeD. In addition, the NI 12-month prevalence rate for any mental health disorder is one of the highest worldwide (Bunting et al. Reference Bunting, Murphy, O'Neill and Ferry2013a ), so it is perhaps not surprising that psychotropic consumption is equally high. Legislation surrounding prescription charges is an obvious potential explanation for differences between countries. Health care in NI is delivered by the UK National Health Service and consultations are free at point of delivery. The majority of prescriptions issued in the province at the time of the study were without charge. The results need to be considered in light of this context. As might be expected, a meta-analysis by Gibson et al. (Reference Gibson, Ozminkowski and Goetzel2005) found a negative relationship between charges and consumption. Other explanations as to the varying rates between countries include provision and use of treatment, awareness of disorders and cultural beliefs (McManus et al. Reference McManus, Mant, Mitchell, Montgomery, Marley and Auland2000; Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004).

Over one-third (38.5%) of respondents who met the criteria for a 12-month disorder had taken any psychotropic medication. This is higher than the other rates (19.6% in Israel (Grinshpoon et al. Reference Grinshpoon, Marom, Weizman and Ponizovsky2007); and 32.6% in Europe (Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004)). NI also has a higher proportion of individuals who met the criteria for any mood disorder in the previous 12 months who had taken any psychotropic medication (50 v. 21.1% in Israel) (Grinshpoon et al. Reference Grinshpoon, Marom, Weizman and Ponizovsky2007), although this is in line with the European rate of 45.6% (Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004). These trends may represent a greater preference for the use of psychotropic medications in NI by those with a mental health disorder and/or doctors. Nevertheless, these figures still suggest possible under treatment. Almost two-thirds (61.5%) received no psychotropic medication in the past 12 months despite meeting the criteria for a disorder in the same period. Half of those with a mood disorder received no medication. While some of these individuals may be availing of treatments other than medication, Bunting et al. (Reference Bunting, Murphy, O'Neill and Ferry2013a ) reported that just 40% of individuals meeting the criteria for a mental disorder in the past 12 months received any treatment. Rates of medication use for other disorders, however, were higher. Of those who met the criteria for any anxiety disorder in the previous 12 months, 47.5% had taken any psychotropic compared with 25% in Israel (Grinshpoon et al. Reference Grinshpoon, Marom, Weizman and Ponizovsky2007) and 32.4% in Europe (Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004). In addition, 66.9% of those who met the criteria for 12-month Generalised Anxiety Disorder had taken any psychotropic compared with 18.6% in Israel (Grinshpoon et al. Reference Grinshpoon, Marom, Weizman and Ponizovsky2007) and 25.5% in Europe (Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004).

The rate of 9.4% for 12-month use of antidepressants is one of the highest, if not the highest, rate of consumption among the WMH survey countries. Another study carried out in the USA in the years similar to the present study found a rate of 10.8% (Pratt et al. Reference Pratt, Brody and Gu2011), however this includes those aged 12 and over (the present study interviewed those 18 and over), and used past month prevalence as opposed to our 12-month prevalence. The rate of 9.4% in the present study corroborates the rates of 9.1% found in NI in 2006/2007 and 12% in 2010/2011 (National Advisory Committee on Drugs, 2012). The high rate found in the present study is perhaps not surprising given the 9.6% 12-month prevalence rate for mood disorders in the country, second only to the USA rate of 9.7% (Bunting et al. Reference Bunting, Murphy, O'Neill and Ferry2013a ). It should also be noted, however, that one in 20 (4.8%) of those with no disorder in the general population had consumed an antidepressant in the previous 12 months. Moreover, almost one in ten (8.6%) of those who had not met the criteria for any 12-month disorder indicated they had consumed any psychotropic medication. Some disorders (such as sleep disorders) were not assessed in the current study. In addition, those who were subthreshold cases were not included in this figure of 8.6%. Both of these factors may have contributed to an overestimate of unneeded use. Nevertheless, inclusion of those with subthreshold disorders led to a decrease of less than one percentage point to 8%. Therefore it is possible these results correspond to true unneeded use of medication, such as individuals continuing to take medication despite their state improving. Alternatively, health professionals may be providing medication to individuals not meeting the criteria for relevant disorders. Future research should attempt to ascertain whether it is patients' preferences for medication or doctors' inclinations to prescribe them that are influencing these rates.

The discovery that those previously married are more likely than married individuals to take psychotropics has been previously reported (Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004; Grinshpoon et al. Reference Grinshpoon, Marom, Weizman and Ponizovsky2007). Possible explanations for this include higher odds ratios of mental disorders among this group and factors relating to social support (Lindström & Rosvall, Reference Lindström and Rosvall2012). A positive association between age and psychotropic use has also been consistently found worldwide (Cooperstock & Parnell, Reference Cooperstock and Parnell1982). In the present study, however, it was those aged 50–64 years old, rather than the oldest group of participants (65+), who were more likely to have consumed psychotropic medication. The fact that those in care homes were not interviewed may have led to an underestimation of usage in the oldest age group, and higher odds ratios in those aged 50–64 years old, particularly given that recent evidence suggests a higher rate of psychotropic prescribing among this over 65 age group in NI care homes (Maguire et al. Reference Maguire, Hughes, Cardwell and O'Reilly2013). Results by Bunting et al. (Reference Bunting, Murphy, O'Neill and Ferry2013a ) also show that those aged 50–64 years old in the province have the highest proportion of any 12-month anxiety disorder and second highest proportion of any 12-month mood disorder. Another predictor of use was trauma – those who had experienced conflict-related traumatic events, as well as those who had experienced non-conflict-related events, were more likely than individuals who had experienced no traumatic events to have taken psychotropic medication. Almost one in five (19.6%) of those who had experienced conflict-related events had used a psychotropic medication in the past 12 months. The fact that 60.6% of the population have experienced a traumatic event and 39% have experienced a conflict-related traumatic event (Bunting et al. Reference Bunting, Ferry, Murphy, O'Neill and Bolton2013b ), further highlights the importance of providing treatments for those who have experienced mental health difficulties as a consequence of trauma. Policy makers should also ensure that those affected by the conflict are offered appropriate mental health treatments. Even after controlling for presence of a mental disorder, those who had experienced conflict-related trauma were over 1.5 times more likely than those who had experienced no trauma to use antidepressants. This supports the contention by Tomlinson (Reference Tomlinson2012) that some of those most affected by the conflict may use antidepressants as a means of dealing with the transition from conflict to peace. While there was a significant association between use of any medication and gender, when the other variables were accounted for using logistic regression analysis, the odds ratios were non-significant. However, females were almost two times more likely than males to have taken antidepressants, and this finding was significant. The greater likelihood of females to consume psychotropics is a longstanding and established international finding (Cooperstock, Reference Cooperstock1976; Paulose-Ram et al. Reference Paulose-Ram, Jonas, Orwig and Safran2004; Rodrigues et al. Reference Rodrigues, Facchini and Lima2006; Chien et al. Reference Chien, Bih, Lin, Chou, Lee, Lee and Chou2011). Possible explanations include the greater likelihood of females to disclose mental health problems and be prescribed psychotropic medication (Women's Health Council, 2005; Tedstone Doherty & Kartalova-O'Doherty, Reference Tedstone Doherty and Kartalova-O'Doherty2010), and the historical role of women as the family caregiver and healer and the stress arising from their various other roles (Ingram Fogel & Fugate Woods, Reference Ingram Fogel and Fugate Woods2008).

A number of limitations of the research should be considered. The prevalence rates may be an underestimation of true use for a number of reasons. Firstly, the NISHS did not survey individuals in institutions such as prisons, hospitals or care homes. These settings have been previously found to have higher rates of prescribing and/or use than the general population for certain classes of psychotropics (Elger et al. Reference Elger, Geohring, Revaz and Morabia2002; Gasquet et al. Reference Gasquet, Medioni, Lellouch and Guelfi2002; Maguire et al. Reference Maguire, Hughes, Cardwell and O'Reilly2013). While survey weights were calculated and applied to account for response bias, some non-response bias may still remain. However, the response rate of 68.4% is similar to that of other psychotropic medication surveys (Alonso et al. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha, Bryson, de Girolamo, Graaf, Demyttenaere, Gasquet, Haro, Katz, Kessler, Kovess, Lépine, Ormel, Polidori, Russo, Vilagut, Almansa, Arbabzadeh-Bouchez, Autonell, Bernal, Buist-Bouwman, Codony, Domingo-Salvany, Ferrer, Joo, Martínez-Alonso, Matschinger, Mazzi, Morgan, Morosini, Palacín, Romera, Taub and Vollebergh2004; Goldney & Bain, Reference Goldney and Bain2006; Grinshpoon et al. Reference Grinshpoon, Marom, Weizman and Ponizovsky2007). A further limitation of the study was that participants were not asked detailed questions about their medication such as duration of use, dosage and reasons for use. As such, where medications may have had a variety of uses they were classified according to their typical use rather than their actual use for each individual. This is particularly important given the high amount of ‘off-label’ use of these medications (Baldwin & Kosky, Reference Baldwin and Kosky2007). Where medications had a psychoactive use (e.g., sedative) and another use (e.g., antihistamine), it was assumed that respondents understood from the question wording that we were asking about the psychoactive use(s) of the medications and not their alternative use(s). This lack of information also means that conclusions cannot be drawn about the appropriateness or efficacy of these medications. Furthermore, this study did not consider over-the-counter medications. Another drawback of the study was the use of self-report data. While it may be thought to be more accurate with regards to actual use than prescription data, an objective measure of consumption was not included in the study. While the use of a 12-month recall period may also be viewed as a disadvantage of this study, memory bias was minimised using techniques which have been previously suggested to aid recall such as a showcard/booklet and asking participants to examine their medication inventories (Strom, Reference Strom2006). Another strength of the study was that it was part of the WMH Survey Initiative meaning that studies in other countries used the same methods and therefore cross-national comparisons were available.

As previously mentioned, the present study is the first to examine the rate of psychotropic medication use, in those with and without mental health disorders, in the NI population. The rate of general psychotropic medication consumption and particularly antidepressants appears to be among the highest worldwide. While some of this prevalence may be accounted for by time elapsed between studies, comparison with more recent studies still remains high. More detailed analysis shows that use among those with mental disorders is low, with the majority having not received psychotropic medication. In addition, there is a sizeable minority identified as having no disorder who consumed medication, particularly antidepressants. These findings have implications for prescribing practice. It is recommended that prescribing policies promote the need for prescribers to adopt more detailed diagnostic assessment to ensure that those who meet the criteria for a disorder are offered appropriate medication. Individuals who request psychotropic medication should also be carefully assessed to prevent inappropriate prescribing. Future research should examine the doctor–patient relationship in more detail to pinpoint potential causes of under treatment or inappropriate prescribing.