Prevalence, patterns and associated factors of self-medication among older adults in Ghana

Abstract

Abstract: Health literacy has pushed the frontiers of self-care in recent decades, particularly among older adults. However, not much is known about self-medication for minor ailments and chronic conditions among older adults in developing countries. This paper addresses this knowledge gap by examining the prevalence, patterns, and associated factors of self-medication among older adults (aged 50 years and more) in Ghana. The study employed a cross-sectional survey within four communities in the Bosomtwe District, Ghana. Using the purposive, snowballing, and convenient techniques, 384 older adults were involved in the study. Multivariate logistic regressions were used to estimate the factors associated with self-medication. Up to 54.4% of the respondents had self-medicated at least 3 months before the survey, while 60.2% self-medicated at least once every month. Open markets (43%) were the dominant drug sources, while quick relief (26.6%) and cheap cost (22.7%) were the reasons for self-medication. Females (AOR: 2.626, CI: 0.015–0.030, p = 0.022), tertiary graduates (AOR: 1.825, CI: 0.007–0.045, p = 0.045) and uninsured older adults (AOR: 3.293, CI: 0.015–0.026, p = 0.007) were more likely to self-medicate. Furthermore, those with no past illness (AOR: 0.036, CI: 0.007–0.033, p = 0.013) and without non-communicable disease (AOR: 0.008, CI: 0.001–0.012, p = 0.020) were less likely to self-medicate. Policy recommendations targeting the educated, uninsured, female older adults as well as stiffening the sale of medicines without prescription will be necessary for reducing the prevalence of self-medication among older adults.

Keywords:

• Self-medication
• older adults
• drugs
• Ghana
• medicine

1. Background

Self-medication—the utilization of drugs unprescribed by a qualified medical practitioner (Jafari et al., 2015)—has been on the rise among different age groups (Bamgboye et al., 2006; Gutema et al., 2011; Sarahroodi et al., 2012). This practice takes several forms, some of which include the consumption of herbal or chemical medications, previously prescribed medicine for similar cases, using additional medicine at home, or not consuming medicine entirely (Karimy et al., 2011). In this context, the term ”self-medication” is used to imply the intake of drugs, unprescribed by a qualified health practitioner, for minor ailments such as cold or cough, headache, back discomfort, indigestion, and nasal congestion, as well as for chronic conditions like diabetes, cardiovascular and cerebrovascular diseases (Lei et al., 2018). Data from several countries including Ghana have shown an increasing rate of self-medication and medicine consumption, with the prevalence of self-medication in Ethiopia, Nigeria, and Nepal being reported to be 43.2%, 73%, and 59%, respectively (Bamgboye et al., 2006; Gutema et al., 2011; Sarahroodi et al., 2012). According to Van den Boom et al. (2008), in Ghana and Nigeria, the commonest means by which people cope with diseases is through self-medication to the extent that for every two Ghanaians, one practice self-medication in times of illness. However, when all age groups are considered, the elderly use more unprescribed medicine drugs than other age groups (Sarahroodi et al., 2012).

One of the several reasons for the prevalence of drug use and self-medication among the elderly is the higher vulnerability of older adults to many ailments (Sarahroodi et al., 2012). A large number of the elderly population suffers from diseases such as cardiovascular disease, diabetes, and cancer. Also, the risk of multiple chronic diseases is common in this age group and can lead to increased use of medications (Meranius & Hammar, 2016; Vali et al., 2012). On the other hand, age-related changes in the human body’s reaction to drugs and vice versa make drug-related issues in the elderly more intricate than in other age clusters (Vali et al., 2012), and this needs to be addressed with major attention (Milte et al., 2014; Ten Bruggencate et al., 2018).

In many African countries, informal pensions and other social welfare schemes are practically nonexistent. In cases where they do exist, inadequate resources have contributed to minimal packages for basic healthcare services for the elderly (Alli & Maharaj, 2013). Consequently, older people are confronted with a lot of health challenges without sufficient social security systems or well-functioning alternative care systems (Alli & Maharaj, 2013; Morgan et al., 2022a). With an inability to generate adequate financial resources, older people are deprived of the opportunity to meet their basic health needs (Waweru et al., 2003). Not surprisingly, many older people are deterred from seeking the necessary and appropriate medical care (Sun & Smith, 2017), thereby, forcing them to resort to self-medication, which has been noted to have serious social health and economic implications (Bennadi, 2013; Chang & Trivedi, 2003; Petermans & Olivier, 2017). According to Machado-Alba et al. (2014), indiscriminate self-medication is considered a public health problem that entails disadvantages including diminished clinical effectiveness, an increase in treatment durations, and a prolongation of recovery. Furthermore, the practice carries several potential risks, such as incorrect self-diagnosis, delays in seeking medical help when necessary, rare but severe adverse reactions, unsafe drug interactions, incorrect administration, incorrect dosage, incorrect therapy selection, masking of severe disease, and abuse and dependence risks (Ruiz, 2010). It is therefore worth understanding the determinants of the practice, especially among older people.

However, ageing and health studies in Africa are not without limiting factors (Alli and Maharaj, 2013; UNFPA, 2008). One of the major limitations of studying the health of older people in Africa, according to Alli and Maharaj (2013), is a lack of research on ageing. Similarly, many African countries, including Ghana, lack national-scale data on the ageing population that would aid in making informed decisions (UNFPA, 2008). Conterminously, while several types of research on self-medication have been conducted in Ghana (Donkor et al., 2012; Gbagbo & Nkrumah, 2020), the majority of these studies have not focused on older people, so there is limited information on older people's self-medication practices and associated factors (Biritwum et al., 2013; Gyasi & Phillips, 2020; Saeed et al., 2016). This limited research and data mean that ageing is poorly understood, and as a result, resources are not well channeled to meet the needs and wants of older people. The development of evidence-based data and planning for healthcare for such a needy, often dependent, and vulnerable age group is increasingly attracting attention for present-day social and public health strategies, especially in the advancement of Universal Health Coverage as a response (UHC; Gyasi & Phillips, 2020).

Focusing on the health and well-being of the elderly is important for informing health policy and planning (Alli & Maharaj, 2013) particularly as demographic trends in Africa and particularly Ghana reveal that there has been a considerable rise in the population aged 50 and over, propelled by decreasing rates of fertility and mortality (Kowal et al., 2010). Again, it is particularly important in Ghana due to widespread spatial disparities that exist in healthcare provision, with urban areas well-served and home to the majority of the nation’s hospitals, clinics, and pharmacies, while modern healthcare is frequently unavailable in rural areas (Peprah et al., 2020). Additionally, despite the government’s constant efforts at expanding individuals’ access to healthcare and the range of services, data from Ghana’s 2021 population and housing census reveals that only 68.6% of the population is insured under the National Health Insurance Scheme (NHIS) or private health insurance plans, with more females having insurance coverage than males. This is particularly worrying as financial barriers to healthcare utilization constitute a major hindrance to formal healthcare use in Ghana (Mills et al., 2012; Oduro Oduro Appiah et al., 2020). Furthermore, in Ghana, medications are easily accessible, with or without a prescription, which raises the risk of abuse and the emergence of unintended occurrences (Aboagye & Kyei, 2014; Atinga et al., 2018).

Following the definition of older adults as persons aged 50 years and above (Kowal et al., 2010; World Health Organization, 2011), this study investigates the prevalence and predictors (demographic, socioeconomic, health-related, and health behaviour variables) of self-medication among older adults in the Bosomtwe District of Ghana. Self-medication will either directly or indirectly result in the misuse of medications if it is not supported by credible medical knowledge. It is beneficial to develop educational initiatives to alter older adults’ perspectives about self-medication. How is this possible? It is achievable when the context of self-medication is understood, in relation to the predictive factors. This will enable both the health and educational administrations to take action and close the gap in health care caused by self-medication practices among older adults in Ghana.

2. Profile of the study area

The Bosomtwe District is one of the Ashanti Region’s forty-three (43) Districts and one of Ghana’s 260 Metropolitan, Municipal, and District Assemblies (MMDAs). It is bordered to the north by Kumasi Metropolis, to the east by Ejisu Municipal, to the south by Bekwai Municipal and Bosome-Freho District, and the west by Atwima Kwanwoma District. The District’s population in 2010 was 93,910 people, with 44,793 males (47.7%) and 49,117 females (52.3%). The District is mostly rural (69.8%) (Ghana Statistical Service, 2014). The District’s health care system is comprised of sixteen public and private health facilities (three hospitals, three health centres, five clinics, two maternity homes, four Community Health Planning Service (CHPS) compounds, and a training institution). These health facilities, however, are insufficient to serve the District efficiently due to the District’s expanding population and increased demand for healthcare services (Ghana Statistical Service, 2014).

3. Methods

3.1. Study design

The descriptive cross-sectional survey method was used. The research method proved beneficial in assessing the prevalence and predictors of behaviour (Burbridge, 1999; Hemed, 2017; Levin, 2006; Setia, 2016).

3.2. Unit of analysis

Following Gyasi et al. (2020), World Health Organization (2011) and Kowal et al. (2010), the study defines the study’s unit of analysis (older adults) as persons aged 50 years or over. As a result, those under the age of 50 were barred from participating in the study.

3.3. Sample size and sampling technique

A total of 384 older individuals were sampled from four settlements in the Bosomtwe District (Kuntanase, Boneso, Jachie, and Homabenase) using the purposive, snowballing, and convenient sampling approaches. The sample size was calculated using the single proportion formula without continuity correction: n = Z² *P(1-P)/d² (Naing et al., 2006) with the following assumptions: the proportion (P) was assumed to be 0.5 in the absence of past studies, the Z statistic for a 95% confidence level of confidence (Z = 1.96), and the degree of precision (d = 0.05 in a proportion of 1).

3.4. Data collection procedure

The data was gathered via a questionnaire administered to the study participant by the researchers. The survey question was broken into four sections. The first section contained information on the respondents’ socio-demographic characteristics (age, gender, ethnicity, level of education, and socio-economic status, among others). Part two covered health-related variables, including chronic Non-Communicable Disease (NCD) diagnoses, past illness records, and self-rated health. The third section dealt with health behaviour (smoking, alcohol intake, vegetable intake, and exercise).

The final section of the questionnaire collects information on self-medication, with specificity on the record of practice, frequency of practice, source of drugs, and reasons for self-medication. A total of 384 questionnaires were distributed face-to-face to older adults using purposive, snowballing, and convenient sampling techniques (Bhattacherjee, 2012; Creswell, 2012; Denscombe, 2010; Gravetter & Forzano, 2015; Northrop, 1999; Saunders et al., 2009). The study’s data was collected between 9 September 2020, and 28 October 2020. To demonstrate their desire to participate in the study, all respondents orally agreed to participate in the study.

3.5. Measures

The outcome variable—self-medication, was measured as a dichotomous variable signifying either “no self-medication” or “self-medication” in the three months preceding the survey. This was employed to reduce recall bias, and inherent retrospective studies (Colombo et al., 2020). Self-medication is described as the selection and use of non-prescription medications by persons acting on their initiative to address self-identified diseases or symptoms (Heidarnia, 2011; Jafari et al., 2015; Kassie et al., 2018). The variables that predict: The predictor factors were examined in three categories: sociodemographic, health-related, and health behaviour. Gender (1 = male, 2 = female), age in years (1 = 50–59, 2 = 60–69, 3 = 70–79, 4 = 80 and above), ethnicity (1 = Akan, 2 = non-Akan), religious group (1 = Christianity, 2 = non-Christian), level of education (1 = no formal education, 2 = basic school education, 3 = high school education, 4 = tertiary education), socioeconomic status (1 = extremely poor, 2 = quite 3 = not very well off, 4 = quite well off) and enrolment in the NHIS (1 = Yes, 0 = No). Being sick in the last three months before the data collection or past sickness records (1 = Yes, 0 = No), being diagnosed with chronic non-communicable diseases (NCDs) (1 = Yes, 0 = No), and perceived health status or self-reported health or self-related health (1 = Good health status, 2 = Poor health status) were health-related factors. Formal healthcare utilization (1 = Yes, 0 = No), alcohol drinking (1 = Yes, 0 = No), and smoking (1 = Yes, 0 = No) were health behaviour factors.

3.6. Data analysis

Using SPSS software (version 23.0), data was processed and statistical analyses were carried out. To define the research sample’s background characteristics, descriptive statistics were performed. Descriptive statistics were utilized to describe the prevalence of self-medication, the sources of self-medication medications, and the motives for self-medication. In addition, sequential logistic regression models were created to assess the factors related to self-medication. In this regard, four distinct models have been created. Model 1 was made up of demographic factors. Model 2 included socioeconomic factors as well as all variables from Model 1. Model 3 included all factors from Model 2 as well as health-related variables. Model 4 included all of the factors from Model 3 as well as health behaviour variables. At a significant level of 0.05 or less, odds ratios (ORs) with 95% confidence intervals (CIs) were provided.

4. Results

4.1. Background information of the respondents

The background information of the participants is presented in Table 1. Approximately 66% of the respondents were females. In terms of age, 45.8% of the respondents were aged between 60–69 years. Respondents from the Akan ethnic group were the most predominant among the participants representing a total of 69%. Approximately 77.1% of the respondents profess the Christian faith. Basic school education (41.9%) is the dominant educational attainment of the respondents. Up to 40.4% of the participants were quite poor (they were within the lowest 25% of Ghanaians). More than three-fourths (75.8%) of the participants are economically active, and 73.4% of them have enrolled in the national health insurance scheme (NHIS).

Table 1. Demographic and socioeconomic characteristics of the respondents

Variable	Categories	Count (384)	Percentage (%)
Gender	Male	129	33.6
	Female	255	66.4
Age	50–59 years	115	29.9
	60–69 years	176	45.8
	70–79 years	53	13.8
	80 years and above	40	10.4
Ethnicity	Akan	265	69.0
	Non-Akan	119	31.0
Religion	Christianity	296	77.1
	Non-Christian	88	22.9
Level of education	No formal education	78	20.3
	Basic school education	161	41.9
	High school education	114	29.7
	College/tertiary	31	8.1
Socioeconomic status	Extremely poor	65	16.9
	Quite poor	155	40.4
	Not very well off	110	28.6
	Quite well off	54	14.1
Employment status	Economically active	291	75.8
	Inactive	93	24.2
NHIS enrolment	Yes	282	73.4
	No	102	26.6

4.2. Prevalence of self-medication among older adults

Table 2 describes the prevalence and patterns of self-medication among the participants. Results showed that 54.4% of the respondents have self-medicated at least 3 months before the conduct of the study. We further established that 60.2% of the participants self-medicate at least once every month. Open markets (43%) are the dominant source from which the participants obtained drugs or medicines which they took without prescriptions. Quick relief from pain (26.6%) and the cheap cost of drugs (22.7%) were the dominant reasons for self-medication. In terms of the conditions for which they self-medicate, body pains (28.7%), headache (21.1%), and cold (19.1%) were the dominant ailments.

Table 2. Self-medication rates among older adults

Variable	Categories	Count (384)	Percentage (%)
Participants have self-medicated in the last 3 months	Yes	209	54.4
	No	175	45.6
How frequently do the participants self-medicate	Every week	13	3.4
	Every two weeks	43	11.2
	Once every month	231	60.2
	Every three months	97	25.3
Where the un-prescribed drug was obtained	Chemical stores (shops)	95	24.7
	Open markets	165	43.0
	Sharing medicines with relatives	76	19.8
	Leftover medicines at home	48	12.5
Reasons for self-medication	Cheap cost of drugs	87	22.7
	Quick relief from pains	102	26.6
	Availability of drugs	69	17.9
	Waiting times at health facilities	56	14.6
	Non-seriousness of the illness	70	18.2
Conditions for which they self-medicate	Cold	40	19.1
	Headache	44	21.1
	Body pains	60	28.7
	Indigestion	20	9.57
	Nasal congestion	18	8.61
	Diabetes	15	7.18
	Cardiovascular	10	4.78
	Stroke	2	0.96

4.3. Predictors of self-medication among older adults

In Table 3, four multivariate logistic regression analyses were performed to ascertain factors associated with self-medication among older adults in Ghana. The variables in each model were demographic (Model 1), socio-economic (Model 2), and health-related (Model 3) as well as health behaviour (Model 4). In Model 1, female older adults (AOR: 2.694, CI: 0.031–0.047, p = 0.026) and older adults who are 80 years and above (AOR: 1.459, CI: 0.011–0.036, p = 0.003) were significantly more likely to self-medicate compared with their respective counterparts. In Model 2, older adults who are not insured (have not registered with the NHIS) were 3.066 times significantly more likely to self-medicate compared with older adults who are insured (AOR: 3.066, CI: 0.022–0.048, p = 0.031). The introduction of socio-economic variables specifically NHIS enrolment dissipated the association between gender and age on the one hand and self-medication on the other hand. In Model 3, health-related variables were added to Model 2. Results indicated that older adults with no past illness (3 months before the survey) (AOR: 0.088, CI: 0.017–0.039, p = 0.043) and with no non-communicable disease (AOR: 0.159, CI: 0.002–0.049, p = 0.027) were significantly less likely to self-medicate compared with their respective counterparts. In the Full Model, females (AOR: 2.626, CI: 0.015–0.030, p = 0.022), older adults with tertiary education (AOR: 1.825, CI: 0.007–0.045, p = 0.045) and those who have not insured (AOR: 3.293, CI: 0.015–0.026, p = 0.007) were significantly more likely to self-medicate compared with their respective counterparts. Furthermore, older adults with no past illness (AOR: 0.036, CI: 0.007–0.033, p = 0.013) and with no non-communicable disease (AOR: 0.008, CI: 0.001–0.012, p = 0.020) were significantly less likely to self-medicate compared with their respective counterparts.

Table 3. Multivariate logistic regression results on the determinants of self-medication among older adults

	Model 1		Model 2		Model 3		Full Model
Variable	AOR	95% C.I	AOR	95% C.I	AOR	95% C.I	AOR	95% C.I
Demographic
Gender ^a
Female	2.694*	(0.031–0.047)	2.780	(0.969–7.974)	2.386	(0.719–7.920)	2.626*	(0.015–0.030)
Age (years) ^b
60–69 years	1.047	(0.244–3.075)	0.438	(0.103–1.860)	0.444	(0.081–2.438)	0.382	(0.030–4.791)
70–79 years	2.234	(0.140–3.005)	0.419	(0.076–2.308)	0.273	(0.036–2.094)	0.072	(0.005–0.999)
80 years and above	1.459*	(0.011–0.036)	0.500	(0.121–2.061)	0.419	(0.081–2.17)	0.836	(0.071–9.778)
Ethnicity ^c
Akan	1.254	(0.001–0.015)	0.013	(0.001–0.017)	0.0021	(0.0025–0.0023)	0.022	(0.001–0.0024)
Religious group ^d
Non-Christian	0.760	(0.162–1.364)	0.807	(0.238–2.735)	0.437	(0.115–1.659)	0.619	(0.097–3.961)
Marital status ^e
Married	2.572	(0.761–6.810)	2.050	(0.645–6.519)	1.950	(0.559–6.802)	2.876	(0.580–14.266)
Socio-economic
Education ^f
Basic School			1.687	(0.252–1.877)	1.387	(0.116–1.292)	1.441	(0.025–0.787)
High School			1.868	(0.199–3.794)	1.442	(0.075–2.588)	1.500	(0.017–2.374)
Tertiary education			1.672	(0.442–2.215)	1.722	(0.654–1.617)	1.825*	(0.007–0.045)
Socio-economic status ^g
Quite poor			1.343	(0.492–3.661)	1.364	(0.421–4.415)	2.275	(0.458–11.300
Not very well off			1.532	(0.396–5.932)	1.744	(0.366–8.309)	3.134	(0.416–23.599
Quite well off			1.641	(0.276–9.742)	1.652	(0.245–11.145)	2.768	(0.309–24.814)
Uninsured			3.066*	(0.022–0.048)	0.205	(0.025–1.660)	3.293*	(0.015–0.026)
No Family support			3.253	(0.065–0.655)	1.175	(0.028–0.552)	2.237	(0.010–0.600)
Health-related
No Past illness (3 months)					0.088*	(0.017–0.039)	0.036*	(0.007–0.033)
Poor health status					0.530	(0.179–1.568)	0.627	(0.135–2.909)
No NCDs					0.159*	(0.002–0.049)	0.008*	(0.001–0.012)
No Disability					2.303	(0.445–11.925)	22.151	(1.661–295.417)
Health behaviour
Non-physical activity							1.542	(0.554–14.168)
Non-alcohol users							0.457	(0.103–2.555)
Non-tobacco smokers							0.312	(0.895–120.698)
Non-fruit intake							1.535	(0.051–3.740)
Non-vegetable intake							1.202	(0.362–10.20)
Model fitting information -2Log Likelihood		143.543		123.064		112.875		83.234
Hosmer-Lemeshow χ^2 (significance)		1.354(0.865)		9.986(0.266)		10.645(0.224)		12.567(0.234)
Nagelkerke R^2		0.089		0.569		0.422		0.569

*p ˂ 0.05.

^aMale is the reference category for the gender variable

^b50–59 years is the reference category for the age variable

^cNon-Akan is the reference category for the ethnicity variable

^dChristianity is the reference category for religious variable

^e Single is the reference category for the marital status variable

^f No formal education is the reference category for the education variable

^gExtremely poor is the reference category for the socio-economic status variable

Model 1 = Demographic variables; Model 2 = All variables in Model 1 plus socio-economic variables; Model 3 = All variables in Model 2 plus health-related variables; Model 4 = All variables in Model 3 plus health behaviour variables

CI = Confidence Interval; OR = Odd Ratio; AOR = Adjusted Odd Ratio

5. Discussion

The study examined the prevalence, patterns and associated factors of self-medication among older adults in Ghana. The study revealed that 54.4% of the participants have self-medicated at least 3 months before the survey. We further established that 60.2% of the participants self-medicate at least once every month. This rate is relatively higher than that reported among older adults in Spain (7.8%) and Poland (49.4%; Brandão et al., 2020), Spain (13.3 %; Carmona-Torres et al., 2018) and from 14 countries in Europe (26.3%; Brandão et al., 2020). The relatively low prevalence rate in these studies could be attributed to the conceptualization of older adults as persons 65 years or above while our study adopted 50 years and above. Interestingly, the paper reported prevalence rates that are lower than those reported in Egypt (73%; Ghazawy, 2017), Malaysia (62.7%—63.5%; Hassali et al., 2011; Mok et al., 2021), Iran (83%; Jafari et al., 2015; Karimy et al., 2011; Sarahroodi et al., 2012) and Nepal (59%; Gutema et al., 2011). That notwithstanding, the results reveal a high prevalence of self-medication among the participants. Whereas self-medication in an appropriate manner can reduce costs and allow health professionals to focus on more serious health problems (Galato et al., 2009), the prevalence rate of 54.4% is high and requires policy initiatives and behavioural changes to address it. These should include education and awareness programmes on the effects of self-medication and strict enforcement of regulations on the sale of drugs on the open markets, in chemical shops and pharmacies without prescriptions. These measures will help reduce self-medication and adverse health events associated with self-medication (Hassali et al., 2011; Petermans & Olivier, 2017) among older adults.

The findings reveal that open markets and chemical shops or pharmacies were the places most of the participants obtained the un-prescribed drugs. This corroborates past evidence (Alghanim, 2011; Donkor et al., 2012; Hassali et al., 2011; Jafari et al., 2015; Sharif et al., 2012; Zafar et al., 2008) where non-prescription drugs were reported to be easily available to clients on the open market and in pharmacies. Weak regulatory regimes as far as the sale of drugs are concerned is the promoting factor for the illegal sale of drugs on the open market and in chemical shops. Strengthening regulatory, supervisory and sanctioning regimes could prove instrumental in clamping down on the illegal sale of drugs on the open market and thereby reduce the untold stories of adverse health events associated with self-medication (Cobbold & Morgan, 2022; Morgan et al., 2022b; Petermans & Olivier, 2017). Furthermore, sharing medicines with relatives and the use of leftover medicines at home were also discovered as ways by which the participants obtained drugs without a prescription. Ghazawy (2017) noted that obtaining medications from relatives or friends constitutes a significant source of drugs for older adults who self-medicate. Other scholars (Alghanim, 2011; Sharif et al., 2012; Zafar et al., 2008) also identified the use of leftover medicine as a prevalent source of drugs for self-medication purposes. This indicates the need for change in the perception and practices towards the safe use of medicines. Interventions at different levels are required to reduce the frequency of medication misuse.

The reasons for self-medication, the cheap cost of drugs, quick relief from pains and non-seriousness of the illness were the dominant factors identified. This lends credence to past research where the cheap cost of drugs, quick relief from pains and non-seriousness of the illness (Ghazawy, 2017; Jafari et al., 2015; Lawan et al., 2013; Sallam et al., 2009; Swetha & Usha, 2016) were reported to motivate self-medication. Rational beings maximize utility and minimize costs. As such, the lower cost of drugs in the open market motivates self-medication as against healthcare use. Nonetheless, untold adverse health effects (Petermans & Olivier, 2017) are associated with the practice, which when added to the cost will make it more expensive than formal healthcare use. Health education is required to make older adults aware of the negative repercussions associated with self-medication and the need to treat every health problem with the needed urgency. Aside from these factors, easy accessibility of drugs and long waiting times at health facilities were also identified as reasons for self-medication, thereby corroborating past research (Donkor et al., 2012; Gbagbo & Nkrumah, 2021; Ghazawy, 2017; Lawan et al., 2013; Mensah et al., 2019; Nkrumah et al., 2019; Swetha & Usha, 2016). With chemical shops and pharmaceutical stores dotted all over the country, in addition to the sale of drugs on the open markets, access to drugs without prescriptions is easy. It is therefore more convenient to self-medicate with over the counter drugs than visit a health facility. Interventionist (improvement in health service delivery and enforcing regulations against the inappropriate sale of drugs) and reactionary measures (sanctioning offenders) are needed to reduce misuse of medicines or drugs among older adults.

The study revealed that gender, educational attainment, health insurance enrolment, past illness (3 months before the survey) and diagnosis with non-communicable chronic disease(s) were associated with self-medication among the older adults in Ghana. Specifically, the study has established that female participants were more likely to self-medicate compared with their male counterparts. Past research (Brandão et al., 2020; Jafari et al., 2015) have also reported higher rates of self-medication among females. The high tendency of females to self-medicate could be explained by their high healthcare need as a result of perceived poor health status (Balbuena et al., 2009; Loyola Filho et al., 2005; Nielsen et al., 2003). Unfortunately, poverty in Africa is feminised because female-headed households lack access to, and control over resources, such as land, as likened to their male counterparts (Matsa, 2011). This reduces their potential to shoulder healthcare costs (Agrigoroaei et al., 2017; Aguila et al., 2016; Agyemang-Duah et al., 2019; Appiah et al., 2020; Gyasi et al., 2019), especially for female older adults who are often non-economically active. With most females having a higher tendency to be within the poorer socio-economic groups, their potential to afford healthcare services is hampered, as such they become more likely to depend on alternative treatments such as self-medication. Gender issues should be factored into self-medication campaigns and awareness programs.

Regarding the level of education, participants with tertiary education were found to have higher odds of self-medication. This corroborates previous research in which higher educational attainment was found to increase the odds of self-medication (Brandão et al., 2020; Jafari et al., 2015; Carrasco-Garrido et al., 2014; Sarahroodi et al., 2012; Delaney et al., 2011; Karimy et al., 2011; Fuentes & Villa, 2008). That said, educational attainment may have different influences on self-medication depending on where the participant lives, as seen in studies conducted in different geographic regions (Coelho Filho et al., 2004; Loyola Filho et al., 2005; Miralles & Kimberlin, 1998; Stoehr et al., 1997). A possible explanation for the greater level of self-medication among the educated elderly may be that people with higher educational levels may feel more confident in their active search for self-diagnosis and/or self-medication as a form of self-care (Kumar et al., 2013; Vidyavati et al., 2016). Self-medication campaigns should focus on older adults with higher education.

Furthermore, participants who were not insured were significantly more likely to self-medicate compared with their counterparts who have insured. The results imply that health insurance provides a financial buffer that promotes formal healthcare use as against self-medication. The finding however contradicts that of Widayati et al. (2011) who found that Indonesians who had access to healthcare (having insurance) were more likely to self-medicate than those who reported no health insurance coverage. Again, the findings contradict evidence from Ghazawy (2017) and Sarahroodi et al. (2012), who reported no statistically significant association between self-medication and having or not having medical (health) insurance. Based on the predictive capacity of health insurance, there is a need for older adults to be encouraged to enrol in the NHIS since it reduces the tendency to self-medicate. Evidence abounds in the literature that financial and operational barriers (inaccessibility of NHIS offices, delays in issuance of membership cards, internet challenges and poor quality of services under the health protection policy) hinder NHIS enrolment and membership renewal among older adults in Ghana (Morgan et al., 2022a; Quartey et al., 2023; Salari et al., 2019). Consequently, it is recommended that these barriers be addressed through effective policies (i.e further decentralization of NHIS operations, revising the exemption age for older adults from 70 years to 60 years, to coincide with the retirement age in Ghana and improving the quality of health services in general and particularly for services provided under the NHIS).

The paper found that older adults with no past illness and with no non-communicable disease were significantly less likely to self-medicate compared with their respective counterparts. This implies that perceived good health status was associated with a lower likelihood to self-medicate. Evidence abounds that people with poor health conditions are more likely to self-medicate (Conner, 2010; Mekonnen et al., 2018), and this tendency is higher when the people are far from health facilities (Arikpo et al., 2010; Awad et al., 2006). This implies that self-rated health ultimately motivates people to self-medicate. Studies by Carrasco‐Garrido et al. (2010) among persons aged 16 years and over, and Brandão et al. (2020) among older adults, identified poor health status as a factor that drives the decision of pregnant women to self-medicate. Based on the finding, people awareness campaigns should focus on whipping the desire of people with underlying health conditions to seek formal healthcare as against self-medication. This will avert tendencies of episodic adverse health outcomes (Petermans & Olivier, 2017).

6. Implications for policy and practice

This study forms part of a step in research and data contribution on self-medication among older adults in Ghana. The practice of self-medication is prevalent among older adults in Ghana and can be observed at various socio-demographic levels. First, evidence from this study can help policymakers in Ghana to develop legislation on self-medication practices. Specifically, legislation should cover the importation and sale of drugs both prescription and non-prescription drugs. While the sale of non-prescription drugs on the open market is illegal, the rule has been lackadaisically implemented over years. The findings of the study necessitate the rigid enforcement of such laws to safeguard the public, most especially older adults from having unbridled access to drugs. Second, the data from the study could help policymakers develop an effective self-medication risks communication system that addresses the various socio-demographic levels. For instance, gender, level of education and other socio-demographic factors should be considered in the development of self-medication risk communications. Beyond this, the risk communication should cover the prevalence of the self-care practice among the various and the health effects it may have on them. This should be done through community engagements and advertisements, where persons who self-medicate suffer certain consequences. A supplement of first-hand information or accounts of victims of unregulated self-medication could also be used in such risk communications to further buttress the point. Third, the findings bring to bear the cracks (cost, long waiting times and overall patient satisfaction) within health service delivery in Ghana. Frameworks can be developed based on this to improve overall healthcare delivery. Finally, the paper reveals the importance of the NHIS as a health protection scheme that cushions older against healthcare expenditure; thereby necessitating efforts to promote enrolment and continuous renewal of membership among older adults (to guarantee formal healthcare use).

7. Limitations of the study

This study has three major limitations, which we recognize. First and foremost, the study was confined to only one district in Ghana, limiting the study’s generalizability and representativeness. The addition of other districts in Ghana might have broadened the geographical reach of the study, allowing the study findings to be generalized. Nonetheless, the findings may be generalized to older individuals who share comparable socioeconomic, demographic, and health characteristics as the research participants. To that aim, a nationally representative survey on self-medication among Ghanaian older persons is required to offer national data and evidence to advise and guide policy. The use of a quantitative method is the second drawback of this study. Due to the quantitative technique, the researchers were unable to obtain in-depth perspectives from study participants on self-medication. Again, the researchers were unable to demonstrate a causal link among the study participants due to the use of a cross-sectional methodology. Due to the cross-sectional nature of the study, there was also the possibility of social desirability or memory bias. There is a potential that individuals would over-report or under-report their use of self-medication. Finally, the distribution of the sample differs from the population of the district. For instance, while the proportion of females in the district was 52.3% as against 47.3% of the males, the study reports a female proportion of 66.4% as against 33.6% for males.

8. Conclusion

The findings suggest the need to include gender, educational attainment, health insurance enrolment, past illness and diagnosis with non-communicable chronic disease (perceived health status) in the formulation of self-medication policies in Ghana. More importantly, the study findings necessitate urgent public education on the implications of self-medication on the health of older adults and the intensification of regulatory regimes against the illegal sale of drugs on the open market and in pharmacies.

Impact of findings on practice statements

• Steps must be taken to increase awareness of the ills of uncontrolled self-medication.

• Regulatory regimes on drugs must be strengthened.

Sanctions must be applied to people who sell drugs without a prescription.

Abbreviation

CHPS: Community Health Planning Service

GHS-ERC: Ghana Health Service Ethics Review Committee

MMDAs: Metropolitan, Municipal, and District Assemblies

NCD: Non-Communicable Disease

NHIS: National Health insurance Scheme

SPSS: Statistical Package for Service Solution

UHC: Universal Health Coverage

UNFPA: United Nations Population Fund

Authors’ contributions

AKM was involved in the design of the study, data collection and analysis, drafting, reviewing and final preparation of the paper. AWA designed the instrument for data collection and performed data organization and interpretation and editing. SUN was involved in the analysis and critical review of the manuscript. All the authors read and approved the final version of the paper for publication.

Availability of data and materials

Data available from the corresponding author upon reasonable request

Ethics approval and consent to participate

All procedures performed in this study involving human participants were per the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. A retrospective exemption approval was granted by the Ghana Health Service Ethics Review Committee (GHS-ERC) according to the Standard Operating Procedures 2015. As approved by the Ethics Board, informed consent was obtained from the participants by agreeing orally to participate in the study. As the dignity, safety and well-being of the interviewees were a matter of primary concern to the researchers, participation in the study was strictly voluntary, and no identifying or sensitive information was recorded.

Acknowledgements

The authors are grateful for the voluntary participation of all participants in this study and their cooperation.

Disclosure statement

The authors declare that there exist no financial or personal relations that may have incongruously influenced them in writing this paper.

Additional information

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or for-profit sectors