Effects of Demographic Characteristics and Consumer Behavior in the selection of Retail Pharmacies and Over-the-Counter Medicine

A cross-sectional study was conducted between February and March of 2016 in a convenience sample of consumers recruited in the metropolitan area of Thessaloniki. A total of 314 consumers were surveyed (approximately 0.1% of the target population) via anonymous, face-to-face interviews using a structured written questionnaire. The interviews were typically conducted during weekdays from 9 am to 3 pm and from 6 pm to 9 pm in public spaces (shopping malls, public transportation stations, etc.) and written consent was obtained from all participants. The sample was stratified by age and employment status, per the results of the 2011 Population and Housing Census for this administrative region (ELSTAT, 2014). The differences in the distribution of age and employment status between the study sample and the target population were minor; there was a notable difference in regard to gender, but this is consistent with the fact that women are more likely to go shopping and tend to make more frequent OTC purchases. Ten questionnaires were excluded due to invalid or incomplete answers and four due to uniform questionnaire scores. The final sample comprised 300 participants: 185 (61.7%) female and 115 (38.3%) male respondents.

The main instrument used for data collection was a structured questionnaire with close-ended, multiple-choice questions. The instrument was developed by the first author following an in-depth review of the literature concerning consumer behaviour as it pertains to the selection of community pharmacies (Arneson et al., 1989; Boström, 2011; Merks et al., 2014; Villako & Raal, 2007; Wirth et al., 2010) and the purchase of OTC medicines (Boström, 2011; Gyaneshwari, 2015; Lodorfos et al., 2006; Paddison & Olsen, 2008; Wazaify et al., 2005). The content and face validity of the questionnaire were assessed by an independent panel of three specialists in pharmacy management and organisation. Three questions were revised for ease of comprehension and clarity and further assessed by a convenience sample of nine adults recruited at a shopping mall in Thessaloniki. The pilot phase responses were not included in the study results. The final version of the instrument comprised of three sections (available upon request from the first author).

The first section collected the demographic information of the participants (gender, age, educational level, employment status, personal income). The second section contained questions regarding the consumers’ choice of pharmacy. The participants were asked to answer multiple-choice, close-ended questions regarding their tendency to make purchases in a single pharmacy and their desired relationship to the pharmacy staff and to rate, on a five-level Likert-type scale, their agreement with statements regarding the degree that certain pharmacy factors (location, staff, product range, additional services, membership programmes, anonymity/confidentiality, atmosphere, opening hours) affect their selection of pharmacy. The third section concerned the selection of OTC medicine. It contained multiple-choice questions assessing the confidence, spontaneity and specificity of the consumer's selection. Finally, the participants were asked to indicate the degree to which each of eight factors (the product's country of origin, manufacturing company, packaging, experience of previous use, advertisement, pharmacist's opinion, family or friends’ opinion, price) affected their purchase of OTC medicines.

The relationships between the demographic characteristics of the participants and their choice of pharmacy, as well as their purchase preferences of OTC medicine (research questions 1 and 2) were assessed using the chi-square test followed by Cramer's V, one-way analysis of variance (ANOVA) with Tukey's post hoc test and Spearman's rho correlation coefficient, where appropriate. Effect sizes for ANOVA were calculated with eta squared (η²), where 0.01 represents a small effect size, 0.06 a medium effect size and 0.14 a large effect size. Spearman's correlation coefficients were computed to identify the inter-correlations among factors influencing the choice of pharmacy and OTC products (research question 3; Table 1). All tests were two tailed, and the significance level was set at 5%. Statistical analyses were performed using the IBM Statistical Package for Social Science statistics software, version 21.

The relationships among the factors influencing the choice of pharmacy and the factors influencing the purchase of OTC medicine (research question 3) were examined. Table 1 presents the Spearman's inter-correlations among the factors influencing the choice of pharmacy. All values were positive and ranged from very low to moderate. The highest correlations were observed between the following pairs of factors: membership programme and additional services that are provided, rho (300) = 0.46, p < 0.01; the store's atmosphere and anonymity/confidentiality, rho (300) = 0.41, p < 0.01; the store's atmosphere and pharmacy staff, rho (300) = 0.40, p < 0.01; the store's atmosphere and the product range, rho (300) = 0.39, p < 0.01; and the additional services and the product range, rho (300) = 0.40, p < 0.01.

The correlations among factors influencing the purchase of OTC medications are shown in Table 1. The strongest correlation was found between the product's country of origin and the manufacturing company, rho (300) = 0.77, p < 0.01. The packaging was found to moderately correlate with both the product's country of origin, rho (300) = 0.36, p < 0.01, and the manufacturing company, rho (300) = 0.38, p < 0.01. Moreover, family's/friends’ opinion was also found to moderately correlate with the product's advertisement, rho (300) = 0.38, p < 0.01.

While examining the differences in the choice of pharmacy and the factors influencing this choice in relation to the demographic characteristics of the participants (research question 1), statistically significant relationships emerged for their educational level, occupation and age and are reported in detail below.

The relationship between the participants’ educational level and answers to the question, ‘I tend to make my purchases in a single pharmacy’, was found to be statistically significant (χ² (10) = 24.90, p = 0.006, Cramer's V = 0.29) (Table 2). Those who have completed primary school or lower secondary (gymnasium) reported that they always favour making their purchases in a single pharmacy (73.3% and 46.7%, respectively). On the contrary, respondents with higher completed levels of education tend to do so as well, but only most of the time (45.5%–56%).

The relationship between occupation and answers to the question, ‘I tend to make my purchases in a single pharmacy’, was found to be statistically significant (χ² (12) = 68.41, p < 0.001, Cramer's V = 0.338) (Table 2). A significant percentage of students (40%), unemployed people (44.1%) and part-time workers (40%) tend to make their purchases in various pharmacies. This is in contrast to the frequency of retired people who always made their purchases in a single pharmacy (61.9%).

A significant relationship was also observed between the participants’ occupation and their reported relationship with the pharmacy staff (χ² (12) = 47.08, p < 0.001, Cramer's V = 0.281). Students strongly preferred their relationship to be formal (80.0%), while others mostly wanted it to be familiar (Table 2).

The importance of the pharmacy staff was found to be significantly different across educational levels (F (5, 296) = 2.76, p = 0.019, η² = 0.02). Tukey's post hoc test revealed that the pharmacy staff is significantly more important to consumers who have completed gymnasium (lower secondary) (4.55) and attained a doctorate (4.50) than to those who have undergraduate-level (3.95) education (Table 3).

Results also revealed statistically significant differences in the importance of the pharmacy staff across occupations (F (6, 298) = 7.41, p < 0.001, η² = 0.05). Post hoc tests revealed that retired consumers (4.63) and homemakers (4.33) consider the pharmacy staff as a significantly more important factor than students (3.60), unemployed (3.63), part-time workers (3.90), full-time workers (4.18) and self-employed respondents (3.83) (Table 3).

Spearman's rho correlations were calculated between age and the eight factors affecting the choice of pharmacy. Low to moderate positive correlations were observed between age and the pharmacy staff, rho(299) = 0.32, p < 0.01, and between age and additional services that are provided, rho(298) = 0.19, p < 0.01. Therefore, older participants tend to rate these factors as significantly more important than younger participants.

This section summarises the purchase preferences of OTC medicines and the factors influencing this purchasing behaviour in relation to the demographic characteristics of the participants (research question 2).

A significant relationship was observed between the participants’ educational level and purchase patterns of OTC medicine (χ² (10) = 22.90, p = 0.011, Cramer's V = 0.197). A vast majority of primary school graduates (80%) reported that they know exactly what they need when they purchase OTC medicine, in comparison with respondents of other educational levels who are evenly split between those who know ‘exactly’ and those who know ‘approximately’ what they need (Table 2).

A significant relationship was found between gender and unscheduled purchases of OTC medicine (χ² (2) = 6.32, p = 0.042, Cramer's V = 0.146). As shown in Table 2, most female respondents tend to sometimes make unscheduled purchases of OTCs (52.5%), whereas most male respondents tend to never do so (55.3%).

There is a marginal statistically significant relationship between occupation and answers to the question, ‘When I am in a pharmacy, I tend to make unscheduled purchases of OTC medicine’ (χ² (12) = 21.25, p = 0.047, Cramer's V = 0.189). More specifically, the majority of homemakers (59.5%) and part-time workers (60.0%) reported that they sometimes make unscheduled purchases of OTC medicine, whereas the majority of retired people (56.6%) claimed that they never tend to do so (Table 2).

The importance attributed to the product's advertisement, as a factor that potentially influences the purchase of OTC medicine (Table 3), was statistically significantly different across occupations (F (6, 299) = 2.15, p = 0.048, η² = 0.01). Tukey's post hoc test revealed that this factor is considered significantly more important by full-time workers (3.28) than self-employed participants (2.33).

In addition, the results of ANOVA indicated significant differences in the importance given to the pharmacist's opinion across different occupation levels (F (6, 298) = 2.93, p = 0.009, η² = 0.02). Tukey's post hoc tests revealed that homemakers (4.52) and retired respondents (4.47) consider this factor as significantly more important than students (3.93), self-employed (4.00) and unemployed respondents (4.05) (Table 3).

The importance of the product's country of origin was found to be significantly different across levels of monthly personal income (F (5, 297) = 2.78, p = 0.018, η² = 0.02). Tukey's post hoc tests showed that the product's country of origin is significantly more important to respondents with an income of 801–1200 € (3.93) than to those with an income below 400 € (3.39) (Table 4).

Furthermore, significant difference was found in the importance attributed to the product's manufacturing company across income levels (F (5, 296) = 4.03, p = 0.001, η² = 0.04). Tukey's post hoc tests indicated that respondents with a higher income (more than 801 €) tend to consider the medicine's manufacturing company as more important than those with a lower income (less than 800 €) (Table 4).

Finally, the results of ANOVA suggested the existence of significant differences in the importance attributed to the pharmacist's opinion across income levels (F (5, 298) = 2.41, p = 0.036, η² = 0.02). Tukey's post hoc tests revealed that respondents with an income between 400 and 1200 € consider this factor as more important than those with an income between 1201 and 1600 € (Table 4).

Moreover, a low negative correlation was found between age and the product's advertisement (rho (300) = −0.14, p < 0.05) and a positive correlation between age and the pharmacist's opinion (rho (299) = 0.22, p < 0.01).

The results presented indicate that in the urban area of Thessaloniki, Greece, pharmacy customer behaviour regarding the choice of pharmacy is influenced by demographic factors, particularly age, occupation and education. In agreement with our findings, in the pharmacy literature, pharmacy customer behaviour has been reported to be greatly influenced by demographic factors (Castaldo et al., 2016; Kouwenberg & Jaiboon, 2019; Rabbanee et al., 2015; Shiyanbola et al., 2016).

A recent study suggests that a large portion of the urban residents of Thessaloniki have established customer loyalty to a specific community pharmacy (77%), as well as a preference for a familiar relationship with the pharmacy staff (69%) (Kevrekidis et al., 2018). The results of the present study suggest that, among these consumers, retired consumers and consumers of low educational level particularly display a significantly higher level of customer loyalty to a particular pharmacy.

These findings are compatible with past literature indicating high pharmacy patronage among elderly customers (Rabbanee et al., 2015). This, in turn, may suggest that the relationship between the pharmacist and elderly customers is trusting. The customer–pharmacist interpersonal relationship has been reported to be critical in the retail pharmacy sector (Antunes et al., 2015; Athavale, et al., 2015; Castaldo et al., 2016; Minarikova et al., 2016b), especially with regard to elderly customers (Shiyanbola et al., 2016).

The observation that students strongly preferred for their relationship with the pharmacy staff to be formal, as opposed to the remaining respondents, could be attributed to a lack of interest for a familiar relationship due to their age difference with the staff or the fact that these younger consumers tend to be healthier. By not having to manage as many chronic conditions, the members of this demographic segment are less likely to constitute regular pharmacy customers or to have a strong motivation to establish a good rapport with the pharmacist.

The pharmacy's location, opening hours, pharmacy staff and anonymity/confidentiality have previously been reported to be among the most important factors that influence the choice of pharmacy of residents of the metropolitan area of Thessaloniki, whereas the availability of a membership programme, the provision of additional services, the product range and the store's atmosphere are among the least important (Kevrekidis et al., 2018). The results of the present study indicate that the importance of the pharmacy staff's quality for the choice of pharmacy depends on the demographic characteristics of consumers in Thessaloniki; in particular, this criterion appeared to be the most important for elderly consumers, consumers of low and high educational level, as well as retired consumers and homemakers.

The above findings, which seem to be incongruous with regard to the educational level, could possibly be interpreted as consumers of different demographic characteristics emphasising different aspects of the quality of the pharmacy staff. This explanation is supported by the findings of earlier studies, which illustrated that consumers of different demographic categories favour different aspects of the pharmacy staff quality (Shiyanbola et al., 2016; Villako & Raal, 2007). For instance, older adults identify a quality pharmacy as one in which the pharmacist provides patient-centred care (Shiyanbola et al., 2016). Quick service has been found to be valued by younger customers, with that appraisal decreasing with age (Villako & Raal; 2007).

The observation that the perceived importance of additional services that are offered by the pharmacist appears to increase with the consumers’ age could be related to the fact that the elderly might have a growing demand for pharmaceutical services due to age-related health challenges (Francis et al., 2005; Pelicano-Romano et al., 2015).

The perceived importance of the pharmacy's atmosphere was found to intercorrelate with that of the provision of anonymity and confidentiality, the pharmacy staff and the available range of products. The pharmacy staff (El Hajj et al., 2011; Merks et al., 2014; Villako & Raal, 2007; Wirth et al., 2010) and the provision of a good range of products and services (El Hajj et al., 2011; Villako & Raal, 2007) have been independently shown to be among the primary factors of pharmacy choice, and both functional experience and hedonic experience factors have been found to positively correlate with consumer satisfaction and with loyalty and purchases (Gavilan et al., 2014). Castaldo et al. (2016) have also found that equivalent retail-level levers, such as the store's environment, the product assortment and communication, indirectly influence trust in pharmacies. Therefore, these may simply be factors that are consistently valued by consumers who seek pharmacies that can inspire their loyalty. Alternatively, the pharmacy's layout and atmosphere could be set up in such a way as to make finding sought products or products of interest more convenient (Emmett et al., 2006), achieve privacy (Hattingh et al., 2015) and encourage communication (De Vera et al., 2018).

An inter-correlation between the provision of additional services and a membership programme was also found. Both features could reward health-conscious regular customers, the former in terms of continuous lifestyle management and the latter in terms of incentivising repeat purchases, possibly of self-medication products, both of which can be broadly tied to self-care.

Kevrekidis et al. (2018) investigated the consumers’ preferences concerning the selection of OTC medicines in the metropolitan area of Thessaloniki and noted the reported tendency of about half of the respondents to know exactly what they need when purchasing OTC medicines and of most of them to ‘never’ or only ‘sometimes’ make unscheduled OTC purchases, as well as frequently favouring a specific product from a certain OTC medicine category. The results of the present study, in accordance with preceding information (Paddison & Olsen, 2008), indicate that the demographic characteristics of these consumers influence their perception of OTC products.

In particular, the results presented indicate that the respondents who had only completed primary school and retired participants were more decisive consumers than those of higher educational levels, with their vast majority claiming that they know exactly what product they need during their shopping. When taken at face value, this decisiveness appears to be counterintuitive and incompatible with the fact that educated people are more knowledgeable and competent in making OTC purchase choices as well as less likely to trust their physicians (Figueiras et al, 2000), as well as with literature that indicates that a higher level of education is a predictor of direct product requests (Collins et al., 2020). Nevertheless, as these patients with a higher education are more likely to self-medicate when suffering from an acute disorder (Figueiras et al., 2000), they may also simply be more willing to experiment with their purchases while engaging in self-care behaviours and being more open to a wider variety of OTC products, including ones with which they have no prior experience.

While participants tend not to make completely compulsive OTC purchases (Kevrekidis et al., 2018), women were more likely to do so, as they reported a higher likelihood of making unscheduled purchases of OTC medicines. This is in line with the fact that women are more self-care oriented (Papakosta et al., 2014), and that they use more types of medicines (Zadoroznyj & Svarstad, 1990). Additionally, self-medication is more prevalent among them (Figueiras et al, 2000; Ya-Ning, 2006). While there is little consensus in the literature about why it is so, there may be biomedical, social (Zadoroznyj & Svarstad, 1990) and cultural elements (Figueiras et al., 2000) that factor into this tendency.

According to Kevrekidis et al. (2018), the most influential factors in the purchase of OTC medicines for the residents of the urban area of Thessaloniki are experience of previous use and the pharmacist's opinion, followed by product price, the product's country of origin and manufacturer; on the contrary, the packaging, the product's advertisement and the opinion of friends or family were reported as the least important factors in an OTC purchase. The results of the present study indicate that respondents with higher incomes tend to place more importance on a product's country of origin and manufacturing company than those in lower brackets. It would appear that by judging such non-medical product features, they are looking for a more premium and reputable product, or at the very least take more care in finding a ‘proper’ one. The country of origin has been reported to drive purchase decisions (Kauppinen-Räisänen et al., 2012; Sanyal & Datta, 2011; Temechewu, 2020). A domestic company would especially act as a risk reduction cue (Pecotich & Ward, 2007). A study conducted in Finland showed that between two equally priced choices, consumers would choose the domestic product (Honkanen, 2013). Moreover, well-known brands associated with high perceived quality can also act as risk reduction cues in OTC drugs (Aufegger et al., 2021).

In addition, the product's country of origin, manufacturing company and packaging appeared to be interrelated attributes of OTC product choice. A drug's package identifies the country and company by displaying their information and can also serve as a means for the aforementioned search for a reputable product, with the consumers essentially ‘judging a book by its cover’.

The pharmacist's opinion, which is a medical criterion, carried more weight with low-to-middle income brackets, instead. One aspect of this might be the role that pharmacists play by recommending cheaper products to price-sensitive consumers (Ricks & Mardanov, 2012). Nevertheless, homemakers and retirees appeared to value the pharmacist's opinion more than students, self-employed and unemployed consumers. Considering the latter, the association between occupation and the importance of the pharmacist's opinion could not be exclusively linked to price-sensitive consumers. Homemakers and retirees are largely made up of older consumers, and therefore, the previously described relationship of trust between pharmacists and older consumers, who have a reported preference towards patient-centred care (Shiyanbola et al, 2016), could play a role in rendering homemakers and retired consumers more receptive towards the pharmacist's opinion. Self-employed workers were also more likely to discount the importance of the product's advertisement, another external source that could impact the information search stage of the buyer decision process.

The product's advertisement moderately correlated with the opinion of family and friends. This could be the case of laymen, without previous experience with a certain product, seeking information from such sources, which may not come from caregivers, but are nonetheless easily accessible. The desired impact of a direct-to-consumer advertisement of an OTC is dependent on whether the ad was informative (Spake & Joseph, 2007). These advertisements could function by supporting the word of mouth that has already been received by family and friends. For interviewees of Paddison and Olsen (2008), the consumer's interpersonal communication with family and friends was vital due to its credibility and empathy. In fact, some preferred certain painkillers due to a sense of family tradition. For them, the effect of advertisements was that of reinforcement rather than conversion to a different brand.

Loyalty appears to develop through time and interaction with the pharmacy staff themselves, as consumers tended to appreciate them and trust the pharmacist's opinion more, the older they get. With age comes increased morbidity, an increasing tendency to visit the pharmacy as well as increased importance of additional services; consumers undergo point-of-care testing (such as glucose, cholesterol and blood pressure measurements). A previous study has found that younger community pharmacy customers (<40 y.o.) were more likely to exhibit information-seeking behaviour prior to visits to the pharmacy (Burghle et al., 2020). This cohort may rely less on the pharmacy staff's input into self-care decisions or the choice of OTC products, and therefore value it less than older customers. There was a low negative correlation between respondent age and the product's advertisement. The younger the people were, the more they relied on advertisements. A survey by Paddison and Olsen (2008) found that younger consumers are more likely to recount adverts.