Towards Medicines Reuse: A Narrative Review of the Different Therapeutic Classes and Dosage Forms of Medication Waste in Different Countries

Background: Medicines reuse, the idea of re-dispensing returned medicines to others following quality control, is yet to be implemented in the UK. This practice is potentially a sustainable way of dealing with returned medicines, which are otherwise classed as medication waste and destroyed. To inch towards medicines reuse, it is important to know more about the different therapeutic classes and dosage forms that make up medication waste. For example, it is helpful to know if medicines being returned are mostly solid-dosage forms and thus have the potential to be reused or are from therapeutic classes that would make medicines reuse cost-effective. Little is known about the therapeutic classes and the dosage forms of wasted medicines. This study aimed to narratively review and report findings from the international literature on the different therapeutic classes and the dosage forms of medicines that are returned by patients to community pharmacies, hospitals, general practitioners’ clinics, or collected through waste campaigns. Studies based on surveys without physically returning medicines were also included where relevant. Methods: A comprehensive electronic search of databases, including PubMed and Google Scholar, was carried out over one month in 2017 and updated by 5 November 2020, using a combination of carefully created keywords. Results: Forty-five studies published in English between 2002 and 2020, comprising data from 26 countries were included and reviewed. Oral solid dosage forms (mostly tablets) were the commonly reported dosage form of all wasted medicines in 14 studies out of the 22 studies (64%) that described the dosage form, with percentages ranging from 40.6% to 95.6% of all wasted medicines. Although there was variability among the levels of medication waste reported in different countries, findings from the UK and Ethiopia were relatively consistent; in these, medicines for the cardiovascular system and anti-infective medicines, respectively, were the most common therapeutic classes for medication waste. Conclusion: This narrative review provides insights about the different therapeutic classes and dosage forms of medication waste either returned by patients, collected through waste campaigns, or indicated in survey responses. The findings could help policy makers understand the potential implications of treating most unused medicines as medication waste and whether therefore pursuing a medicines reuse scheme could be environmentally or financially logical. The quality and the safety of these returned medicines using criteria related to the storage conditions (such as heat and humidity), physical shape (such as being sealed, unopened, unused, and in blister packaging), and tampering are other important considerations for a medicines reuse scheme.


Introduction
Waste can be referred to as any substance or object the holder discards, intends to discard, or is required to discard [1]. The World Health Organisation (WHO) defines pharmaceutical or medication waste as "expired, unused, spilt, and contaminated pharmaceutical products, drugs, vaccines and sera" [2]. Medication waste is a growing problem in the UK and different parts of the world in terms of its negative impact on governmental expenditures, the environment, and human health [3][4][5][6][7]. Waste associated with prescribed medicines cost the National Health Service (NHS) in England an estimated £300 million a year in 2009, £110 million of which related to medicines returned to community pharmacies for disposal [6]. However, the financial cost is only one part of the medication waste burden. The negative impact on the environment is also significant with one reason for finding pharmaceuticals in the water environment [7] being the improper disposal of medication waste [8,9]. The presence of medication waste in the environment can modify the physiological function of living creatures and has been linked to the possible emergence of antibiotic-resistant bacteria such as vancomycin-resistant enterococci and beta-lactam-hydrolysing Enterobacteriaceae [10], as well as the feminising effects of endocrine deactivating compounds such as ethinyl estradiol [11]. The risk to human health is not limited to pollution and contamination of the drinking water, as there is also a risk when others in the home consume unused medicines that have been stockpiled but ought to have been dealt with safely. For example, patients might self-medicate for a new illness with medication previously prescribed for a different illness, causing harm through misdiagnosis or mistreatment [12]; there might be accidental poisoning if children use stockpiled medicines; and medicine abuse might occur where the medicines are controlled or have addictive properties [13].
The causes of medication waste are divided into preventable (e.g., patient stockpiles medicines, overprescribing, or repeat dispensing of unwanted medicines), non-preventable (e.g., death of a patient, or a change in the prescription meaning the previous medicines are no longer required) and non-adherence behaviours [1,6,14]. Therefore, prevention is one way to reduce medication waste. Preventing waste is in fact the top option according to the Waste Hierarchy, which is a grading framework that ranks waste management options according to what is best for the environment, with "prepare for reuse", "recycle", "other recovery", and "disposal" following "prevention" in decreasing preference order [15]. Many interventions have been attempted to prevent medication waste, but these have not always been effective, as the most common causes of medication waste are actually non-preventable [14]. Medicines reuse-the idea of re-dispensing returned medicines to others following quality control-is an underexplored concept in the UK but could help reduce medication waste regardless of the cause. What is more, qualitative studies have previously analysed intentions and actions towards the reusing of medication waste, reporting a possible future for the idea [16][17][18][19]. Numerous factors influence the practicalities of such an idea, including the prior storage conditions, as well as the therapeutic classes and the dosages forms of medicines considered to be waste but which might then be reused. Knowing information about the different therapeutic classes and dosage forms of medication waste creates some understanding of which medicines might potentially be up for reuse. For example, it is helpful to know if medicines being returned are mostly solid-dosage forms (thus having the potential to be reused), or liquids, injectables, etc., and whether these medicines are over the counter (cheaper/not critical to NHS costs) or other therapeutic classes that could be more relevant in terms of environmental sustainability or cost-effectiveness.
Despite a thorough literature review on the causes of medication waste [6,14,[20][21][22][23], the financial [4,6,20,[24][25][26][27][28] and environmental impact of medication waste [7,10,11], medicine disposal practices [8,9,22,24,[28][29][30][31][32][33][34], and management strategies of medication waste [6,14], only some studies have reported the type and therapeutic classes and dosage forms of unused or returned medication waste, and none have brought the information together in a focused review [6,23,28]. This study aimed to narratively review and report findings from the literature about the different therapeutic classes and the dosage forms of medication waste that are returned by patients to community pharmacies, hospitals, general practitioners' clinics, or collected through waste campaigns in different countries around the world. Results from studies based on surveys (without the physical return of medicines) were also included to take account of relevant data collected via this alternative method.

Materials and Methods
A search of electronic databases was carried out over one month in 2017 and updated in 2020 ending on 5 November 2020 to identify reports and studies published in English detailing therapeutic classes and dosage forms of medication waste. Electronic databases searched comprised PubMed/Medline, Cochrane library, Grey literature (open grey and British library), National Audit Office (NAO), and National Institute for Health and Care Excellence (NICE) evidence. The bibliographies of retrieved references were also searched.
The search activity used combinations of a list of terms that included the following: types of unused medicines OR classes of unused medicines OR dosage forms of unused medicines OR types of medicine waste OR classes of medicine waste OR dosage forms of medicine waste OR types of unused drugs OR classes of unused drugs OR dosage forms of unused drugs OR types of drug waste OR classes of drug waste OR dosage forms of drug waste.
The inclusion criteria aimed to select studies published in English that reported the therapeutic classes and dosage forms of returned medication waste, either dispensed following a prescription or purchased over the counter (OTC), or a medicine sample that had expired (or had no clear expiry date) or was never fully consumed (or not used at all). Studies describing medical waste, medical device waste, and/or clinical tissue waste were excluded.
Study selection was completed by two researchers (H.A. and N.P.) using a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow of identification, screening, eligibility, and inclusion [35] (Figure 1). At first, 3390 candidate studies were identified; then, 18 duplicates were removed. All study titles and abstracts of the remaining 3372 studies were screened, with 3311 studies removed, resulting in 61 potentially eligible studies. After a thorough full-text review of the 61 studies to assess for eligibility, 45 studies published between 2002 and 2020 were included in this narrative review. Data obtained from the retrieved studies described demographic information of the participants, the types and dosage forms of medication waste, study settings and sample size, and the time/duration of the collection of the returned medicines (varying from 4 weeks up to 12 months).

Results
The search yielded 3390 candidate studies. A total of forty-five studies published between 2002 and 2020 and comprising data from 26 different countries from around the world (Australia, Austria, Egypt, Ghana, India, Jordan, Kuwait, Malaysia, Mexico, New Zealand, Oman, Qatar, Saudi Arabia, Spain, Taiwan, Tanzania, Thailand, United Arab Emirates, United Kingdom, United States of America, China, Malta, Indonesia, Iraq, Nigeria, and Ethiopia) were included and reviewed. In some of these studies, medication waste was returned by patients to community pharmacies, general practitioners' clinics, hospitals or sometimes collected via medicine take-back and medicine waste campaigns. However, twenty nine (the majority) studies used a survey to collect information about the therapeutic classes and dosage forms of medication waste by asking participants for information without physically collecting the waste: six studies from Ethiopia [36][37][38][39][40][41], three from India [42][43][44], two from Malaysia [24,45], two from the USA [34,46], two from Jordan [47,48], two from Egypt [26,49], two from Thailand [20,50], one from Qatar [51], one from China [52], one from Iraq [53], one from

Results
The search yielded 3390 candidate studies. A total of forty-five studies published between 2002 and 2020 and comprising data from 26 different countries from around the world (Australia, Austria, Egypt, Ghana, India, Jordan, Kuwait, Malaysia, Mexico, New Zealand, Oman, Qatar, Saudi Arabia, Spain, Taiwan, Tanzania, Thailand, United Arab Emirates, United Kingdom, United States of America, China, Malta, Indonesia, Iraq, Nigeria, and Ethiopia) were included and reviewed. In some of these studies, medication waste was returned by patients to community pharmacies, general practitioners' clinics, hospitals or sometimes collected via medicine take-back and medicine waste campaigns. However, twenty nine (the majority) studies used a survey to collect information about the therapeutic classes and dosage forms of medication waste by asking participants for information without physically collecting the waste: six studies from Ethiopia [36][37][38][39][40][41], three from India [42][43][44], two from Malaysia [24,45], two from the USA [34,46], two from Jordan [47,48], two from Egypt [26,49], two from Thailand [20,50], one from Qatar [51], one from China [52], one from Iraq [53], one from Indonesia [54], one from Nigeria [55], one from Spain [56], one from Saudi Arabia [27], one from Tanzania [57], one from Malta [58], and one from Ghana [59]. The methodologies used and the targeted populations are summarised in Appendix A Table A1.

Studies' Samples
The studies' samples were reported in different ways. Most studies reported sample size as the number of medication waste items returned or collected. Other studies reported the sample as per weight (kg), per bag, packs, or containers of the collected returned medication waste. The sample for survey-based studies was reported as the number of completed questionnaires collected or the number of participants surveyed. For more details about the sample of the studies included, please refer to Appendix A Table A1.

Demographics of the Participants
Gender was not reported in the majority of the studies (Appendix A Table A1). Fifteen studies (36% of the retrieved studies) described the gender of the participants, and it was not apparent that there is a gender difference associated with the presence/reporting of medication waste. For example, more women took part in seven of the studies [20,38,45,52,54,56,57] and more men took part in eight of the studies [36,37,39,40,42,49,55,59]. In the study from Egypt [49], the number of people who returned their medication waste happened to be more male than female and one study from Malaysia [24] recruited female students only.
Age of participants was described in 23 studies out of 45 studies (51%) (Appendix A Table A1). Participants' age profile varied in these studies and was up to 81 years. Twelve studies of the 23 studies (25%) found an apparent relationship between the mean number of returned medicinal items per patient and their age. Here, the majority of medication waste was reported to be from participants with the age ranges of 60-80 years [21,31,32,49,57]. Two studies [43,59] had more data relating to participants in the age range 20-40, but this was an artefact of the study designs, focussing on students who are likely to be in that younger age range. It is not possible to conclude that the age range of 60-80 years was associated with more medication waste as, additionally, age data was absent from half of the studies (49%).

Dosage Forms of Returned Medication Waste
Dosage forms were investigated in 22 out of the 45 studies (49%) on medication waste (Appendix A Table A2). Dosage forms included a range of oral solid dosage forms (tablets, capsules, granules, powders, and lozenges), liquids (syrups, injections, eye drops, suspensions, emulsions, and lotions), semisolids (ointments, creams, gel, paste and suppositories), and other items such as inhalers, sprays, patches, strips, and chewing gum. Oral dosage forms were the most commonly reported formulation in fourteen studies out of 22 (64%) with percentages ranging from 40.6% to 95.6% of all medication waste. Moreover, tablets were reported to be the most common of the oral dosage forms.
One study from Oman (60) reported that during handling of the dosage forms, most of them appeared in a suitable condition for reuse and were still in their original container. However, some had changed in colour, consistency, and odour and therefore were deemed not to be suitable for reuse. Results from a UK study [36] were consistent with the Oman study [60] in which many of the returned medication waste items were reported to be in a condition suitable for reuse as assessed by a pharmacist. These were the only two studies that directly commented on whether the medication waste returned was potentially suitable for reuse.

Therapeutic Category of the Returned Medication Waste
Except for two studies [31,57] in which only prescribed medicines were included in the authors' analysis, the majority of the studies include both prescribed and OTC medicines. Moreover, only three studies [25,26,61] included medicinal samples in addition to prescribed and OTC medicines.
The majority of the studies (42 out of the total 45) reported the therapeutic category of the medication waste, and these were included in the current analysis (Appendix A Table A2). The remaining three studies reported the medication waste individually by generic or brand name and were therefore excluded from the current analysis.
The therapeutic categorisation systems used for reporting the findings were not the same in all studies. Seven studies used the British National Formulary (BNF) categories [6,26,49,60,[62][63][64]. Seven studies used the Anatomical Therapeutic Chemical Coding (ATC) of the WHO [33,36,48,56,58,65,66]. Other ways of therapeutic categorisation included national codings such as the Saudi National Formulary (SNF) [27], Chinese Pharmacopoeia [52], and the Monthly Index of Medical Specialities online (MIMS) [20]. The remaining studies used disease and class of medicine classification such as diabetes/anti-diabetic. A breadth of therapeutic categories reported included cardiovascular system (CVS), central nervous system (CNS), alimentary tract/gastrointestinal tract (GIT), respiratory system, musculoskeletal system and joint disease, analgesics and antipyretics, non-steroidal anti-inflammatory drugs (NSAIDs), endocrine system, malignant disease and anticancer medicines, nutrition and blood, vitamins and minerals, gynaecology and medicines for urinary tract infection (UTI), antibiotics, medicines for ear, nose, and oropharynx, and skin medicines.
The different therapeutic categorisation systems used in reporting medication waste (sometimes in studies completed in the same country) make the interpretation of results difficult. For example, two studies, one from India [42], and one from the USA [25], combined analgesics with nonsteroidal anti-inflammatory drugs (NSAIDs) into one therapeutic category, while five studies from India [43], the USA [34,46], Mexico [61], and Thailand [50] described analgesics and antipyretics as one category and musculoskeletal and joint disease medicines as another category. In addition, the number of studies that investigated medication waste by therapeutic categorisation was more likely to be from a small number of countries. For example, seventeen studies out of forty-two (40%) were from four countries: the UK [6,13,62,64], Ethiopia [36][37][38][39][40][41], New Zealand [21,31,65,67], and the USA [25,34,46]. This makes reporting of the results by the number of studies less representative of the international literature.
Therefore, in order to synthesise the results for this narrative review, all the different therapeutic categories were re-classified according to the BNF categorisation system and then represented by country ( Figure 2). For example, NSAIDs were re-classified under musculoskeletal system medicines (BNF Chapter 10), analgesic and antipyretics were re-classified under CNS medicines (BNF Chapter 4), and alimentary tract system medicines were re-classified under gastrointestinal system medicines (BNF Chapter 1). In addition, in countries where more than one report was found, such as the Ethiopia, UK, New Zealand, Jordan, and Egypt, the sum of all returns of medication waste was calculated and reported by country. Figure 2 shows the results of the common therapeutic categories of medication waste reported by country and after re-classification according to the BNF categorisation system. In the UK, CVS medicines were the most common therapeutic class of medication waste, with CNS medicines being the second most common therapeutic class. Other therapeutic categories of medication waste, such as gastrointestinal and respiratory medicines, were also reported but less commonly in the UK. Similar results to the UK were reported from countries such as Australia, Austria, Mexico, and Oman where CVS medicines were the most common therapeutic class of medication waste. Moreover, in Mexico, Australia, and Austria, musculoskeletal system medicines were also common and the second most reported category.
In New Zealand, CNS medicines were the most common therapeutic class of medication waste. Other therapeutic categories such as gastrointestinal, cardiovascular, and musculoskeletal system medicines (with diclofenac sodium and ibuprofen reported to be commonly returned as waste) were also reported in studies from New Zealand but less than CNS medicines. In Jordan and Qatar, results were similar to New Zealand, where CNS medicines were the most common therapeutic class of medication waste. In Jordan and Qatar, paracetamol was the most common individual tablet considered as waste. In addition, in Jordan, gastrointestinal medicines were reported as the second most common therapeutic class of medication waste followed by anti-infective medicines. In Qatar, anti-infective medicines were reported as the second most common therapeutic class of medication waste. Other therapeutic categories of medication waste such as musculoskeletal system medicines were reported in Jordan and Qatar but less commonly.
In Spain, both the gastrointestinal system and CNS medicines were the most common therapeutic classes of medication waste. In Taiwan, gastrointestinal system and CVS were the most common therapeutic classes of medication waste. While in Saudi Arabia, both the respiratory system and CNS medicines were the most common therapeutic classes of medication waste.
In Ethiopia, Egypt, and Tanzania, anti-infective medicines were the most common therapeutic class of medication waste. The CNS medicines (in Ethiopia), and CVS medicines (in Egypt and Tanzania) were reported as the second most common therapeutic class of medication waste. Other therapeutic categories of medication waste such as musculoskeletal and gastrointestinal system medicines were reported in Ethiopia, Egypt, and Tanzania, but less so.
Studies from the USA, Thailand, India, and Indonesia showed that musculoskeletal system medicines were the most common therapeutic class of medication waste in these countries. Finally, in Malaysia, vitamins and minerals were the most common therapeutic category of medication waste.

Discussion
Despite the extensive literature on medication waste, no literature review to date had explicitly focused on the therapeutic classes and the dosage forms of medication waste items. This narrative review addresses that gap. The principal finding is that CVS (certainly in the UK) and anti-infective (certainly in some African countries) medicines are reported as some of the most commonly returned/accumulated medication waste category. Arguably, knowing the therapeutic category of medication waste is as crucial as the quantity of the returned medication waste in terms of environmental and financial potential for medicines reuse. This is because medicines in certain therapeutic categories cost more. For example, one UK study [63] reported the volume of waste relating to respiratory system medicines to be about half (8%) that reported for CNS medicines (19%), but the cost of the medicines in the respiratory group was about the same as those in the CNS category. Thus, knowing the therapeutic categorization of medication waste makes it easier to judge where medicines reuse might be financially logical. It is of course essential to quality assure the safety of any returned medicines using criteria related to the storage conditions (such as heat and humidity), physical shape (such as being sealed, unopened, unused, and within blister packs), and tampering. Two studies conducted in Oman [60] and the UK [63] directly commented on whether the medication waste returned was potentially suitable for reuse. These studies reported that during handling of the dosage forms, most of the returned medicines appeared in a suitable condition for reuse and were still in their original container, with only a few having changed in colour, consistency, and odour; thus, these were deemed unsuitable for reuse. Findings of these studies are also important considering that unused medicines from the so-called developed world are sometimes sent for reuse to so-called developing countries.
In the UK, CVS medicines were the most common therapeutic class of medication waste. A possible explanation is that CVS medicines are one of the commonly prescribed medicines in the UK, comprising approximately 20% of all the medicines prescribed because of the prevalence of cardiovascular disease. Moreover, CVS medicines are one of the commonly amended classes of medicines because of frequent changes in doses and drugs necessitated by guidelines [62]. In Ethiopia, Egypt, and Tanzania, anti-infective medicines were the most common therapeutic class of medication waste. This is possibly because antibiotics have been available without a prescription in these countries, where also it is common for people not to complete their course of antibiotic treatment when their symptoms resolve. In New Zealand and Jordan, CNS medicines were the most common therapeutic class of medication waste with paracetamol as the most common individual tablet returned as waste. The potential explanation here is that analgesics (with paracetamol reported to be the most common) are frequently used for the self-medication of headaches, which is a commonplace discomfort. Similarly, in the USA, Thailand, and India, the musculoskeletal system medicines were the most common therapeutic class of medication waste with NSAIDs being the most common group reported in these countries, again reflecting their common usage. In the study from Malaysia, vitamins and minerals were reported as the most common therapeutic category of medication waste, but this is likely an artefact of the methods, which only sampled female students.
This narrative review synthesised information about the most commonly found medication waste products from different countries around the world. However, the results should be interpreted carefully. First, the findings apply to medication waste that was returned by patients only or reported in surveys and does not take into account the substantial amount of medication waste likely to be disposed of into household garbage or via the sink or simply kept stockpiled unreported at home [65]. Second, the small sample size and the small number of returns of medication waste in the majority of the studies made these studies less likely to be representative of a global picture. Third, the CNS classification of paracetamol as the most commonly reported item as waste raised the percentages of waste from the CNS therapeutic class compared to other therapeutic classes such as musculoskeletal, alimentary tract, and respiratory systems. Paracetamol is considered cheap, and one may argue that it is not worthwhile to set up a medicines reuse system if this is the largest category of returned medicines in any one country. Fourth, the quality of the studies included in this narrative review was not checked because of the disparity of methods and a lack of specific criteria, which could affect the results reported (as none of the papers found were excluded) and could be seen as another limitation of this study. Finally, the results of this narrative review cannot be generalised. For example, results from Ethiopia, Egypt, and Tanzania of having antibiotics as the most common therapeutic category of medication waste cannot be generalised (although reported to be the commonest along with CVS medicines) to other countries where antibiotics are only available with a prescription such as in the UK or the USA. In addition, results from the two Malaysian studies, which reported that vitamins and minerals were the most common therapeutic category of medication waste, is impossible to generalise to the larger population, as one study was only completed with female Malaysian students (no males). The other Malaysian study was also completed by more females than males.
This narrative review has other limitations that should also be acknowledged. First, it included results from reports, theses, audits, and the grey literature, but there is always a risk that some studies were not included as a result of not performing a thorough enough systematic search. Second, the search strategy was restricted to studies that were published in the English language only and so could have missed other valuable research. Third, the reasons behind the accumulation of the returned medication waste from each therapeutic category were not clearly evidenced in all the studies. Some studies provided possible explanations that may apply only to the country from which data were obtained, and therefore, it may not be appropriate to generalise these explanations. Finally, information about what motivates people to return their medication waste and if they returned a certain type of medication waste over others were not investigated and remain unknown.
This review is the first to provide narrative information about the different therapeutic categories and dosages forms of returned medication waste from different countries around the world. Pooling information about the different therapeutic classes and dosage forms of medication waste can help increase knowledge about medicines that are returned unused and or otherwise classed as wasted medicines, so that extrapolations can be made about the costs of waste and whether it is worthwhile to reuse these medicines. For example, paracetamol is considered cheap, and one may argue that it is not worthwhile to set up a medicines reuse system if this is the largest category of returned medicines in any one country. In addition, oral solid dosage forms are more likely to be suitable candidates for reuse compared to other dosage forms such as liquids or injectables; therefore, it is useful to know where this is the most commonly returned formulation. However, a pharmacist's hand inspection of such medicines would not be sufficient to address concerns about the quality and the safety of returned medicines. For example, there would also need to be additional checks in place for storage conditions (e.g., under excessive heat and humidity), and physical characteristics (such as being sealed, unopened, unused, and being in a blister) which could indicate (non-)tampering, degradation or contamination, in addition to the visual indicators. These concerns would need to be addressed before medicines reuse becomes a reality [23,68], and one way to do this would be through the application of technology [69,70].
This narrative review identified a large number of studies from the literature that investigated the different therapeutic classes and the dosage forms of medication waste returned by patients to healthcare settings, and through waste campaigns, as well as information obtained from survey responses. Although there was variability among the levels of medication waste reported in different countries, findings from some countries such as the UK and Ethiopia were relatively consistent and appeared to reflect the local usage of these medicines. This suggests that medication waste categories might be proportional to the prevalence of medicines in each specific country, which remains to be tested in future studies. Future studies that focus on assessing the quality and the safety of retuned medicines, and exploring public and healthcare providers' perception about medicines reuse should also be performed before medicines reuse becomes a reality. For example, contained sites such as long-term care facilities or hospice care settings where the medications are presumably stored correctly might be more capable of reusing medicines and could be a realistic site for trialling medicines reuse in the future.

Conclusions
Findings from this narrative review provide insight about the different dosages forms and the therapeutic classes of medication waste, which can be used to support future medicines reuse-related research and explorations. It appears that the therapeutic categories of medication waste are reflecting prevalence of usage, inviting policy makers in each country to reflect on whether medicines reuse could be cost-effective in their own settings. Any medicines reuse scheme would still need to consider quality and safety checking of returned medicines over and above the pharmacists' visual checks.  Unused medicine were returned to 114 pharmacies in 5 primary care trusts.

Appendix A
In total, 8626 items were reported as returned with 7500 of the returned items identified and coded for analysis.
Not studied. Cross-sectional, observational two phases study was conducted using a convenience sample in Southern California. In Phase I, a web-based survey was conducted at one health sciences institution; and in Phase II, a paper-based survey at drug take back events.
Web and paper-based survey.     Take back collection program.

Sample of 200 households in
Kuwait received an educational letter and special plastic bags in which to place unwanted medicines to be collected by the municipality. A second convenience sample of an additional 14 households in Kuwait received the same educational letter together with a face-to-face interview and assistance in collecting unwanted medicines.
Not studied.

2013
Aditya [43] India; dental hospital in North India Descriptive cross-sectional survey of dental students based on a structured questionnaire format) was carried out in a teaching dental hospital in North India.  Survey study.
A total of 800 houses, 400 each from urban and rural areas, were included for the study, which was based on the prevalence of self-medication as per a previous study done in India.
The participants above the age of 18 years, capable of giving information of medicine use within the family (the heads of the households or their spouses or any adult capable of delivering required information) were interviewed for the study.
The presence of any healthcare professional amongst the family members in a visited house was excluded in order to avoid biased answers.

2009
Ali et al. [24] Malaysia; Universiti Sains A prospective descriptive, cross-sectional survey was conducted from February to June 2005 in the Universiti Sains, Malaysia.

Survey study.
A total of 481 single female respondents were targeted for a questionnaire-based survey on randomly sampled students. A total of 1724 different types of medicines were found with an average number of 4 medicines found per student.
Respondent were only females ages varied from 19 to 54 years old. 89.2% (n = 429) of the students were categorised in the 19-24 years age category, while 8.7% (n = 42) were aged between 25 and 30 years old. The remaining 2.1% (n = 10) were aged between 31 and 54 years. 2020 Hassali and Shakeel [45] Selangor, Malaysia    Survey study.
A total of 1034 participants were enrolled in the study. A multi-stage sampling method was employed to select households.
Overall, 1000 (97%) households responded to the interview, among them 504 urban and 496 were rural. The median family size of the households was 5; just above half (52%) of the households had at least five family members. Only 7% of the surveyed households had a health professional as a family member.

2017
Teni et al. [36] Gondar town, northwestern Ethiopia A cross-sectional household survey was conducted from 5 April to 6 May 2015. In the study, 809 households were surveyed from four sub-cities selected through multi-stage sampling with 771 included in the final analysis.
Survey study.
In the study, 809 households were surveyed from four sub-cities selected through multi-stage sampling with 771 included in the final analysis.
Of the participants of the study that represented their respective households, upwards of three quarters (76.3%) and two-fifths (40.9%) were female and those in the age group of 18 to 29 years, respectively. Nearly three-fourths (73.3%) followed Orthodox Christianity, and almost all (90.3) were Amhara in their ethnic identity. Survey study.
A total of 4 health facilities were included in the study. During the 1 month of the study period, 56 types of medications were found unused at the health facilities.
Three of the heads were male and one was a female. All of them were BSc nurses with a work experience of a minimum of 4.6 and a maximum of 8 years. All the pharmacy heads were male and degree holders with a minimum experience of 4 years and maximum experience of 8 years. A total of 3 store women and 1 store man were interviewed. All the store men/women were diploma holders with a work experience of a minimum of 4 years and a maximum of 8 years. 2020 Gudeta and Assefa [39] Jimma city, Ethiopia A facility-based descriptive cross-sectional study was conducted among private practitioners in retail outlets of Jimma city from 20 November to 19   Awi zone, Amhara regional state, Northwestern Ethiopia A community-based cross-sectional study was conducted through interviews with representatives of households.
Survey study.
A total of 23 kebeles (the smallest an administrative unit in Ethiopia) (2 urban and 21 rural kebeles) from four woredas were selected using a multi-stage sampling technique. A total of 507 households were included in the study.
The majority of the study participants, 368 (72.6%), were female. The mean age of the study participants was 40 years, and the majority were between the ages of 30 and 65 (67.9%) 2020 Yimenu et al. [38] Awi zone, Amhara regional state, Northwestern Ethiopia A community-based cross-sectional study was conducted through interviews with representatives of households.
Survey study.
A total of 23 kebeles (the smallest an administrative unit in Ethiopia) (2 urban and 21 rural kebeles) from four woredas were selected using a multi-stage sampling technique. A total of 507 households were included in the study.  The predominant therapeutic group was drugs affecting the nervous system, but individually, diclofenac sodium and ibuprofen were the most returned medications, respectively. Nervous system drugs: 19%. Alimentary tract and metabolism: 13%. Cardiovascular system: 12%. Musculoskeletal system: 11%. Respiratory system and allergies, and miscellaneous: 8%. Blood and blood-forming organs: 7%, Dermatological and anti-infective: 7%. Genitourinary: 3%, Hormones: 3%.
Oral solid forms (tablets and capsules) were counted. Liquid medications were quantified by the amount left in the original container, semisolid preparations were estimated as a proportion of original container. Inhalers were recorded as either full, half-full, or empty. Anything almost empty was excluded from the analysis.
The chosen sample of the total returned unused medicine was around 20%, which maybe not representative of the whole sample. Oral solid forms (tablets and capsules) with tablets as the most common returned dosage form.
Oral liquid forms. Cream and ointment.
Unable to describe respondent demographic information.
Not studied.
Unable to describe respondent demographic information. In addition, due the different policies for collection and disposal of medicines, the majority of unused medicines were disposed into landfills and water system, which may mean that the returned amount may be underestimate of the extent of unused medicines.
Tablet or capsule, oral liquid, cream or ointment, and inhalers.
Tablet or capsule, oral liquid, cream or ointment, and inhalers.
The author reported that this study did not attempt to estimate the quantities of unused medicines at patient's home; as a result, it is more likely that the unused medicines from primary care was underestimated.
It is an audit report with a result from Cumbria/northwest of England, which may not representative of the whole United Kingdom and may underestimate the extent of unused medicines.
Not studied.
Unable to describe the respondent demographic information clearly.
In Phase II, 17% of medications for chronic conditions (hypertension, diabetes, cholesterol, heart disease) and 8.3% for mental health problems (antidepressants/antipsychotic/anti-convulsant) were commonly reported as unused. 7% painkillers and 4% electrolytes and dietary supplements.
Use of a web-based survey may limit the accessibility of this study to people without computer and Internet access at home, which may to some extent underestimate the extent of unused medicines. Unable to describe respondent demographic information.
Unable to describe respondent demographic information.
Not studied.
Unable to describe respondent demographic information.  Al-Azzam et al. [47] Jordan; North of Jordan particularly Irbid Central nervous system drugs were found to be the most common, followed by anti-infective agents. The most common individual medications found were amoxicillin, paracetamol, metronidazole, antihistamines, hypoglycaemic medications, and adult cold medications.
Central  The majority of the drugs stored (n = 58; 21%) in the participating homes were analgesics. Nonsteroidal anti-inflammatory drugs were the second most commonly stored drugs, representing 16% of all the drugs.
Not reported.
There was potential for selection and social desirability bias as a result of the strategy of using the telephone to conduct an interview. In addition, interviews were conducted during working hours, which could run the risk of excluding highly educated young subjects.
Due to the small sample size, the results of this exploratory study should be considered with caution.
Not studied.
Unable to describe respondent demographic information. As it included only medicines with SQUH labels, others were missed, and this may underestimate the extent of unused medicines.

2004
Wongpoowarak et al. [20] Thailand; Songkhla Musculoskeletal system drugs were the most common pharmacological group of returned drugs. The medications were pharmacologically classified using MIMS Thailand, which is a standard reference source. This study was a snapshot study, as the studied population was one of 14 provinces in southern Thailand. A total of 2208 drug items found in household surveys were classified into 5 groups of the mostly found drugs. These were 343 non-opioid analgesics and antipyretic drugs, 188 antacids, anti-reflux agents and anti-ulcer, 180 nonsteroidal anti-inflammatory drugs (NSAIDs), 127 antihistamine and anti-allergic and 119 anti-diabetic drugs. Top 5 of the most found rarely or unused drugs, classified as leftover medicines, were NSAIDs (49 items), penicillin (38 items), G.I.T. regulators, and antiflatulents (36 items). Of the total of 2208 drug items found in household, 82 items (3.7%) and 45 items (2.0%) of drugs were already expired and deteriorated, respectively.
Not studied.
Unable to describe respondent demographic information.
Not studied.
This study did not estimate the quantities of unused medicines in patient's home. As result, it is likely that it may underestimate the extent of unused medicines in the community. Oral solids drugs were the most common wasted dosage form 40.6% followed by injections 9.2%, with very few topicals preparations.
Since only hospital-prescribed medicines was included, others may be missed, which may underestimate the extent of unused medicines.

2007
Abahussain et al. [33] Kuwait; Kuwait city No medicines were collected from the 200 households participating in the municipal collection program The second intervention yielded 123 medicines from 14 homes; the most common class of unwanted medicines were drugs for respiratory system. Unwanted medications were classified according to the ATC WHO classification. A third of all unwanted medicines were for the respiratory system (38% of these were cough and cold preparations, 25% nasal preparations). 12% of the medicines were for the musculoskeletal system (53% oral NSAIDs) or were dermatologicals (33% topical antibiotics).  68.5% (n = 1181) of the medications were in the form of tablets and pills while capsules constituted 14.6% (n = 252) of the overall amount. 5% (n = 87) syrups and suspensions while 4.9% (n = 84) were creams and ointments. Less than 1.0% (n = 5) consisted of inhalers, with 0.2% (n = 4) suppositories of the overall total.
Sampling of only female students made it impossible to generalise the results to the whole student population in the campus.
Not studied.
The sample size of the study was small to depict a clear picture of the entire Selangor population; hence, the findings of the current study are not generalisable to all of Malaysia.  Huang et al. [52] Six provinces in North, Central, and Southern regions of China Cold medication (86.1%) was the most common category of medicines kept in households. Specifically, the following were the major classes of medicines found in the households: gastrointestinal medicines (27.0%), pain medications (22.9%), vitamins (20.6%), antibiotics (19.0%), external painkillers (16.5%), and external anti-inflammatory antidotes (15.4%).
Not studied. Not reported.

2019
Vella and West [58] Maltese village, Malta The most common class of disposed medications was that pertaining to the alimentary tract (24.6%), closely followed by medicines belonging to the respiratory group (23.8%). 10.5% of the unused disposed medications were from the musculoskeletal group, which includes medications such as nonsteroidal anti-inflammatory drugs (NSAIDs), and supplements, such as glucosamine. The medications with the lowest return rate were anti-neoplastic and immunomodulating agents (0.7%), followed by anti-parasitic medications (0.2%).
Solid dosage forms were counted manually, liquid dosage forms were measured using a calibrated measuring cylinder, dermatological preparations were measured using kitchen weighing scales, and inhalers that had a counter were recorded as per value available on the counter. Unused inhalers without a counter, eye drops, ear drops, nasal drops, and nasal and oral sprays were not quantified as effective entries, as their quantities could not be safely determined.
This study excluded some dosage forms whilst quantifying and costing waste, such as eye drops, inhalers, and nasal sprays. Therefore, the actual cost of waste presented in this study is an underestimate. 2020 Insani et al. [54] Bandung, Indonesia NSAIDs were the most common medicines left unused (n = 372) followed by vitamins and nutritional supplements (n = 215) and antibiotics (n = 171).
Not reported.
This study was conducted in one region in Bandung (small sample size); thus, its generalisation for the Indonesian population is limited. In addition, the predictors associated with disposal practice were not identified.
Not reported.
This study was based on the self-reported presence of medicines in respondents' residence. Therefore, it is possible that the medicines were under-reported or some names of unidentified medicines were wrongly reported. In addition, the sample size was small.

2015
Wondimu et al. [41] Tigray Region, Northern Ethiopia The most common classes of drugs found in the households were analgesics (29%) and antibiotics (25%). Generally, more than half (62%) of the medications were used for ongoing treatment.
One of the study limitations was the cross-sectional design employed, which might be affected by temporal relationship establishment with some variables and could not provide much more substantial evidence of causality, unlike a longitudinal design.
Of the total 553 medicines stored, more than three quarters (80.8%) were of solid dosage forms. Liquid dosage forms were (16.6%) and semisolids were (2.5%).
The study did not include the rural parts of Gondar Town. The small sample size makes the findings not representative of the pattern of household medicine storage practice in those areas.
Not reported.
Health centres and private health facilities were not included in the study, and thus, the results may have been slightly different if those facilities were included. The sample size was small. In addition, the current study was conducted among private practitioners. Thus, prospective researchers may consider both private and public professionals for their comparative study.
Not reported.
The small sample size and the cross-sectional nature of the study design prevent us from drawing causal inferences about the relationship between the chosen covariates and outcome variables over a period.
Not reported.
The small sample size and not including the health centres and private health facilities were limitations to this study. Thus, the results may be slightly different if those facilities were included.