International treatment and good practice guidelines recommend that before prescribing benzodiazepines to any patient, clinicians should first exclude any physical illness and determine the benefit-to-risk ratio.¹ The guidelines also recommend that benzodiazepines may be prescribed as a short-term treatment for severe insomnia^2,3 to counter any side effects of co-prescribed medication¹ while waiting for the full effect of other treatment modalities to occur^3,4,5 and to alleviate withdrawal symptoms in patients with substance misuse.^1,6,7 Benzodiazepines should not be used routinely in patients with anxiety disorders,^2,5,7 except for symptomatic relief of severe anxiety or panic attacks as they can provide rapid relief. They should be prescribed for only 2 weeks and thereafter reviewed.¹ The South African Society of Psychiatrists’ treatment guidelines for psychiatric disorders that includes the use of benzodiazepines⁸ are similar to other international treatment guidelines.

Worldwide, variable rates in the prescribing of benzodiazepines have been reported.^{9,10,11,12,13,14} Maust, Blow, and Lin¹⁵ reported that, in the United States (US), 30.6 million adults (12.6%) used benzodiazepines in 2016, although Agarwal and Landon¹⁶ reported a slightly lower rate of 7.4% in 2019. The European study of the epidemiology of mental disorders reported a prescribing rate of 18.4% across Europe,¹² while in South Africa, a study comparing prescribing patterns in Gauteng and the Northern Cape found rates of 7.9% – 8.5% and 9% – 9.5%, respectively.¹⁷ The variation in published rates by different countries may be attributed, at least in part, to different methodological approaches and clinical settings for example, lower rates may be attributed to possible underreporting by the participants because of stigma, while rates in tertiary outpatient psychiatry clinics may differ from that in community district clinics, because of less specialised staff in community clinics.

Despite not being recommended as first-line treatment by all guidelines,^1,2,4,17 benzodiazepines in actual clinical practice tend to be prescribed for a broad range of psychiatric conditions, including anxiety, mood, psychotic, personality, and sleep disorders,^10,18,19,20 and in patients with comorbid medical illnesses.^19,20,21,22 They are also frequently prescribed for older patients,^23,24,25,26 females,^27,28,29 patients with lower education and socioeconomic status,²⁴ and those with perceived low levels of support.^24,30 Although the guidelines recommend that benzodiazepines be prescribed for short periods of 2 weeks initially and then reviewed, several studies have reported a high rate of prescribing benzodiazepines for long-term periods.^{18,19,31,32,33} The World Health Organization (WHO) defined long-term benzodiazepine use as exceeding 180 days,³⁴ while others have suggested between 4 and 12 weeks.³³ The WHO definition is the most frequently used in research and was the definition used in this study.

Inappropriate and long-term prescribing of benzodiazepines by healthcare professionals is one of the leading causes of iatrogenic benzodiazepine use disorders.³¹ The adverse health effects of benzodiazepine use disorder include psychomotor impairment, impairment in judgement and dexterity, forgetfulness, confusion, irritability, aggression, and paradoxical disinhibition.³⁵ Tolerance, dependence, and withdrawal symptoms may also occur.^35,36 Long-term benzodiazepine use may be appropriate in cases such as terminally ill or severely handicapped patients, where the benefits outweigh the risks and side effects.^1,37

There is a paucity of research in the South African public health sector, especially in district clinics, where anecdotally, long-term benzodiazepine prescribing is a common practice, hence the need for this study. Establishing the prescribing patterns of benzodiazepines at the clinical level may help to detect and prevent long-term use and subsequent adverse health consequences. Such findings would also help district managers develop systems for monitoring benzodiazepine prescribing patterns and ensuring that the recommendations of approved guidelines are correctly implemented.

The study aimed to determine the rate of long-term prescribing of benzodiazepine and to determine any associations between demographic and clinical variables and this prescribing pattern at community psychiatry clinics in Johannesburg health district.

The study was a descriptive retrospective and cross-sectional record review. Of the 27 community psychiatry clinics in the Johannesburg health district, five were randomly selected. The names of all 27 clinics were written on pieces of paper and placed in a hat. From the hat, five pieces of paper were randomly drawn, which made up the study sites. The clinics included in the study were Discoverers, Eldorado Park Ext 8, Crosby, Zola, and Hillbrow.

Approximately one out of every four patients (23.8%) in the Johannesburg Health District community psychiatric clinics were prescribed a benzodiazepine. While similar high prescribing rates of 18.7% – 24.0% have been reported,^10,32,38 lower rates of 3.8% – 12.6% have also been reported.^13,14,15 The variation in reported prescribing rates could be because of differences in the methodology employed, such as study design, settings, and inclusion and exclusion criteria. The high rates in the studies of Ghosh-Nodia and Ahuja³⁸ and Haw and Stubbs¹¹ may have been as a result of much larger sample sizes, whereas in the studies performed by Haw and Stubbs¹¹ and Johnson et al.³², the high prescribing rates may have been because the prescription was initiated during hospital admission and continued in the community clinics where the studies were conducted soon after discharge. The high prescribing rates in this study may in part be because of clinicians’ lack of knowledge of benzodiazepine prescribing guidelines,³⁹ a lack of capacity and time for the re-evaluation of benefits and risks associated with benzodiazepines,⁴⁰ and failure to provide patient psychoeducation on non-pharmacological interventions to manage symptoms.¹⁵ In addition, there may have been inadequate monitoring of prescriptions by pharmacists, who are required to serve as gatekeepers.⁴¹ In South Africa, benzodiazepines are classified as schedule five (S5) drugs by the Medicines and Related Substances Control Act No. 101 of 1965, which serves as a control measure to prevent inappropriate prescribing and limit misuse of these drugs.⁴² However, these control measures have not been entirely effective.²¹

Benzodiazepines are divided into four groups based on their elimination half-life. They include ultra-short acting (midazolam, triazolam), short-acting (oxazepam, temazepam), intermediate-acting (alprazolam, bromazepam, lorazepam), and long-acting (clonazepam, diazepam, nitrazepam).⁶ The most frequently prescribed benzodiazepine (96.7%) by the clinicians in this study was clonazepam. Singh and Oosthuizen¹⁸ also reported that clonazepam was the most prescribed benzodiazepine in their patients, followed by diazepam. However, Summers, Schutte and Summers⁴³ reported a preference for the longer-acting benzodiazepines in their public sector study. Generally, in the South African public healthcare sector, clinicians’ choice depends on what is available on the country’s essential medicine list and in the central pharmacy stocks at the time of prescribing.^21,44 During this study, the benzodiazepines available on the essential medicine list were diazepam, lorazepam, oxazepam, and clonazepam. However, as there are no limitations in the South African private health sector; there is a preference for intermediate-acting benzodiazepines such as alprazolam and bromazepam.^14,29 Other countries have also shown variation in the choice of benzodiazepine prescribed between the public and private sectors. Lebanon, Ramadaan, Sheik-Taha, and Deep¹⁰ reported that the most used intermediate-acting benzodiazepines were alprazolam (34.6%) and bromazepam (33.6%). It is unclear what guides and informs the choice of one benzodiazepine over the other in public and private healthcare sectors of different countries. The mean dose of clonazepam was like that of other studies^10,20 and within the manufacturers’ recommendations.⁴⁵ There was also no progressive increase in the benzodiazepine dosage, as might be expected with the development of tolerance.⁴⁶

This study found that in almost all the patients (90%), the benzodiazepines were prescribed for a long-term period (exceeding 180 days). This rate was considerably higher than the range of 10% – 19% reported in most other studies.^18,32,33 However, it must be noticed that the Johnson et al.³² study had a small sample size, and the benzodiazepines were initiated during admission to a hospital, with only those still on benzodiazepines at community clinics soon after discharge included in their study. The Bernard et al.¹⁸ review was an observational study at primary healthcare clinics with a larger sample size and included only adult patients with anxiety disorders. Although higher rates of between 60% and 70% have been reported,^12,47 the Lagnaoui et al.¹² study included only patients with mood and anxiety disorders, while the Valenstein et al.⁴⁷ study included only depressed patients on antidepressants. Notwithstanding the wide variation in published long-term benzodiazepine prescribing rates, which may be attributed to differing methodologies, the rate found in this study was still high.

It is therefore relevant to determine what factors may be associated with this long-term benzodiazepine prescribing habit. Several studies have reported that long-term prescribing of benzodiazepines has been more common for female rather than male patients.^47,48,49,50 Females tend to have better help-seeking behaviours than males, which may account for them consulting doctors more often and being on medication for more extended periods.⁵¹ However, Zandstra et al.³⁴ and Sjostedt et al.²⁴ reported no gender differences to support this. Contrary to the aforesaid, although not statistically significant, this study found a slight preponderance of males. Franken et al.⁵² also reported most males among their participants. It is likely that this study population, at the outset, consisted of more males. This male preponderance may be attributed to the observation that males tend to have medication non-compliance and poor social reintegration after discharge,⁵² which necessitates regular and ongoing follow-up at community clinics.

This study also found that approximately half of the patients on long-term benzodiazepine use (43.3%) were older than 50 years of age, like findings in the French study of the long-term prevalence of benzodiazepine use.⁵³ Patients of advancing age generally have comorbid medical conditions that may warrant benzodiazepine initiation and possible subsequent prolonged use.^54,55 Chronic medical conditions in the older population have been associated with increasing psychosocial stressors that might precipitate or exacerbate anxiety and depression.^16,56 However, more than half of the patients in the current sample (56.7%) had no medical comorbidities to support this argument. Gerlarch et al.⁵⁶ reported that low quality of sleep in the elderly is also a predictor for long-term benzodiazepine use. However, this was not a clear indication for prescribing benzodiazepine to the patients in this study. Although this study did not show statistically significant associations between age and long-term prescribing of benzodiazepines, the studies that have demonstrated this have focussed primarily on individuals older than 60 years with chronic diseases with depressive, anxiety, or insomnia disorders.^12,22,28

Being single with perceived poor social support has also been associated with long-term benzodiazepine use because of an increased likelihood of anxiety and related disorders.^9,30,48,52 Similarly, a lower socioeconomic status because of a lower level of education and unemployment may add to the anxiety of low social support and perpetuate long-term benzodiazepine use.^9,21,30 The settings of these studies were similar to our research. Many participants who were prescribed benzodiazepine for long-term use in this study were also either unmarried, divorced, or widowed, had a level of matric or below education, or were unemployed. This may be because they were generally from a relatively poor socioeconomic community with a lower level of education, unemployment, and lower income or disability grants.⁵³

The origins of substance use disorders are complex and multifactorial, and there is evidence of a correlation with poor mental health, poverty, low levels of education, unemployment, a lack of social amenities, and social exclusion.⁵⁷ The increasing level of discomfort relating to anxiety may lead to prolonged periods of individuals being prescribed and misusing benzodiazepines. Benzodiazepines may be taken inappropriately to treat the undesirable effects of substance abuse, enhance or augment other drugs’ euphorigenic effects, or induce intoxication.⁵⁸ Several studies have reported that comorbid substance use is also associated with long-term benzodiazepine use.^9,20,22 A South African study⁵⁹ reported a high prevalence rate of substance use disorders in psychiatric patients; however, no studies reported an association between comorbid substance use and prolonged benzodiazepine use. Our research found that most patients (80%) had no comorbid substance use disorder, similar to Aragaines et al.⁵³ This may be because of the patients’ under-reporting, poor documentation by the doctors, or small sample size.

Despite it not being recommended by approved guidelines, common mental disorders such as depression, anxiety, and sleep disorders are often associated with long-term benzodiazepine use.^9,26 Surprisingly, this study found that most of the patients who were prescribed long-term benzodiazepines had bipolar and psychotic disorders. Similar findings were reported by Johnson et al.³² This may be because most of the patients at these clinics have had psychiatric admissions for severe mental illnesses and are down-referred for long-term follow-up.⁵²

Only three patients on long-term benzodiazepines had documented reasons for a benzodiazepine being prescribed. It is possible that there were clinical grounds for prescribing a benzodiazepine; however, because of poor record-keeping, the prescribers did not document it. Agarwal and Landon¹⁶ also reported poor record-keeping. It may be that clinicians fail to record the indications for the use of these drugs because they are not familiar with accepted prescribing guidelines because of limited training or a lack of confidence in correctly prescribing these drugs.^39,60

The high rates of prescribing benzodiazepine and the long duration of prescribing found in this study are of concern. In addition, this study found that there were poor records of the indications for prescribing benzodiazepine. It would imply that the recommended accepted prescribing guidelines are not being adhered to. There is a high risk of developing dependence and the emergence of serious adverse effects especially in the elderly. Apart from impacting the patient’s quality of life, there is a risk of litigation against the Department of Health with financial implications. It demands that the pharmaceutical industry and mental health trainers ensure that regular training on guidelines is carried out and that proper clinical monitoring systems be established.