Impact of this study on practice

  • Our findings confirm previous research that consumers seek out and use a variety of sources, of varying quality, for information on their medicines

  • Our findings suggest that consumers do not accept this variety of information sources passively, rather consumers actively evaluate the quality of their information sources

  • Our findings highlight that consumers evaluate the information sources they use on the basis of the perceived knowledge and trustworthiness of the source, and the ability for interaction

Introduction

There is growing consensus that consumers should be provided with the opportunity to be more involved in their own healthcare. This is marked by a shift from the concept of consumer compliance to that of adherence and concordance, with the consumer adopting an autonomous rather than a passive role [1–3]. Shared models of decision-making are proposed [4, 5], with an informed, and therefore empowered, consumer taking an active part in decisions about their medical management. A challenging, but crucial, aspect in achieving shared decision-making between consumers and healthcare professionals is acknowledging and managing the uncertainty evident in medical knowledge [6]. Inseparable from this movement towards shared decision-making is recognition of the ‘democratisation’ of medical knowledge, that is, a change in the traditionally paternalistic power relation between the ‘sick patient’ and healthcare professional as holder and arbiter of medical knowledge [7, 8].

The challenges for consumers to be more involved in their healthcare are prompted further by the vast array of information available to help consumers decide whether, and how, to use a medicine [7, 8] and the deregulation of an increasing number of medicines in many countries (i.e. medicines available without need for consultation with a prescriber). Such shifts in focus raise questions for how healthcare professionals can best assist patients to utilise these information sources optimally to ensure the quality use of medicines (QUM). In Australia, QUM is one of the central objectives of the National Medicines Policy 2000, which states that medicines should be used judiciously, appropriately, safely and efficaciously. To achieve QUM, healthcare professionals need to know not only which information sources consumers actually use but also how they evaluate the reliability of these sources. This study aims to provide some information towards achieving this goal.

Sources consumers utilise include healthcare professionals, traditional media (television, newspapers, magazines), the Internet, specific consumer-directed information (package inserts or consumer medicine information leaflets (CMIs), consumer support groups and telephone support), and the knowledge and experience of family and friends [9–15]. The most frequently used resource identified in the above studies is a healthcare professional (usually doctors and pharmacists) although there is increasing use of the internet [12], and also a trend for reliance on the ‘media’ for health/medicine information [9, 12–14]. Concern has been expressed regarding the reliability of some of these information sources, in particular, the media and internet [16, 17].

There is limited literature available on how consumers perceive the reliability of the information sources they use. Consumers’ perceptions of information sources have been qualified in terms of satisfaction [10, 13], usefulness and ease of use [11], saliency and credibility [15], or have focused on specific information sources with regard to their perceived reliability/utility [8, 18–21]. The methods used to elicit consumer perceptions on information sources have included case-studies [8], phone surveys [9–11, 13], semi-structured interviews [12], and attitude questionnaires with additional survey methodologies [19]. Although these are recognised as valid methodologies, such techniques often constrain the information that consumers can provide to investigator-generated and framed questions. Consumer responses may have more relevance if researchers know how people understand and frame a particular concept, making it easier to measure or determine consumer perceptions of the concept [22]. There is a need to utilise alternative methods that are sensitive to individual differences in consumer perceptions and provide insight into why these differences exist.

In reply to these challenges, Frewer et al. [23] has suggested the use of the repertory grid technique in health-related research. Originally developed in the area of personality theory [24], the repertory grid technique has been used in eliciting consumer perceptions and attitudes to genetic engineering [25], the food industry [26], complementary medicines [27], and the treatment of angina [28]. The advantage of using the technique in this study includes having highly individualised perceptions of the information sources that can be analysed both quantitatively and qualitatively.

The aim of this study was to use the repertory grid methodology to (i) establish the range of medicine information sources that consumers use and (ii) determine consumers’ perceptions on the reliability of these medicines information sources (i.e. the information sources consumers personally utilise).

Methods

The study used a structured interview based on the repertory grid technique. All participants were consumers who attended one of three community pharmacies—two located in Brisbane, Australia and the other just outside the Brisbane metropolitan area—over a period of 6 weeks (February–March 2004). Approval for the study was obtained from the ethics committee at the School of Pharmacy, The University of Queensland.

Sample population

The sample population was a convenience sample of consumers attending one of the three community pharmacies. Participants excluded were those who could not consent to the study, were less than 18 years of age, or those who refused to participate. All participants provided written informed consent prior to taking part in the interview process. All interviews were conducted by a final year pharmacy student, within the community pharmacy out of earshot from the dispensary or front counter. Demographic information considered pertinent to the analysis was collected as part of the interview.

Interview procedure

The interview was divided into three parts and was generally 15  min, although in some instances lasted up to 25 min.

Identifying information sources

Participants were asked whether they had ever purchased a medicine from a supermarket, a pharmacy (over the counter (OTC)) or received a medicine on prescription from a doctor (Supermarket medicines are limited in type available and pack size. Medicines available in a pharmacy consist of a wider variety of more potent medicines and larger pack sizes of medicines available in supermarkets). Participants who answered ‘yes’ to the above were asked to specify the medicine (or if multiple items were identified to give one example). Participants were then asked to identify three information sources that they would consult to find out more about each type of medicine. The first three information sources the participant named for each type of medicine was recorded. Some participants identified the same information sources while other participants identified differing information sources for the different medicine types. Each participant could identify a maximum of nine information sources. The information sources identified were placed in one of seven categories as detailed in Table 1. If the participant identified two or less information sources, the interview was terminated at this stage.

Table 1 Categories of identified information sources
Full size table

Eliciting consumer perceptions of the information sources

Participant perceptions of their individually identified information sources were determined by comparing information sources both within and between the seven categories. Each information source identified by the participant was written down on a separate card. The participant was then presented with groups of three information sources (triads) on the individual cards and asked to state which two of the three sources were more reliable. They were then asked the reason why they thought the two were more reliable (termed the emergent pole) [24]. The reason why the participant felt the third source was less reliable (the implicit pole) was also asked. The responses obtained—the emergent and implicit poles—formed the participant’s personal association or construct for that triad. This method to elicit constructs using triads is termed the ‘minimum context card’ form for eliciting constructs in the repertory grid technique [29]. An example of a triad and resulting construct is shown in Fig. 1.

Fig. 1
figure 1

Eliciting constructs and rating information sources—an example from a single participant

Full size image

If the participant only generated a small number of information sources then they were presented with pairs of information sources (dyads). The participant was asked if they thought the information sources were equally reliable and if so the reason for this. If the dyad was not equally reliable, the participant was asked why they felt one was more reliable than the other and the construct poles generated by the participant were again recorded.

The combinations of information sources allowed opportunity for all the information sources identified by the participant to be included in the triads (or dyads). This process was repeated for further sets of triads (or dyads) until the participant was not generating any new constructs or ten triads (or dyads) had been presented to them.

Rating information sources

Each construct pole generated by the participants was put onto a 10-point scale where ‘1’ was the emergent pole and ‘10’ was the implicit pole (Fig. 1). The participant was asked to rate each information source on these scales and the scores were then recorded in a rating matrix (grid) (Fig. 1). If the participant felt that a construct was not applicable to a particular information source, then this was noted on the matrix instead of a score.

Analysis

The participants demographic information, medicines identified, information sources identified and constructs elicited were analysed using standard descriptive statistics.

The participant-generated constructs were put into categories based on the meanings that they represented using generic content analysis undertaken by all three authors [30]. Where discrepancies existed in placing a construct into a category, mutual consensus was reached after discussion. To explore the participant’s perceptions of the reliability of their information sources, the median scores for these against the constructs were determined.

Results

A total of 110 consumers were interviewed of which 73 (66%) were female. The median age (IQR) of consumers was 55.5 years (42, 67). The remainder of the demographic information is shown in Table 2.

Table 2 Consumer demographics
Full size table

All the consumers had received a medicine on a prescription from their doctor (either currently or in the past), 86 (78%) had purchased a medicine from a supermarket and 99 (90%) from a pharmacy. The consumers identified 648 information sources out of a possible 990 for their named medicines. The most frequent information source identified was the doctor, 91 (83%), written sources, 90 (82%) and the pharmacist, 86 (78%) (Fig. 2).

Fig. 2
figure 2

Information sources identified by consumers

Full size image

The number of consumers who could not identify any information sources they would use decreased with the increasing restrictions on medicine availability; supermarket 15 (13.6%); pharmacy 4 (3.6%); and prescription 3 (2.7%).

The information sources identified changed dependant upon the type of medicine (Fig. 3).

Fig. 3
figure 3

Elicited information sources compared to the of medicine

Full size image

From the 648 named information sources, the pharmacist was identified as the predominant source of information for OTC medicines (32%), the doctor for prescription medicines (37%) and the pharmacist and doctor equal proportions (19% versus 17%) for supermarket medicines. Breaking down written information into CMI/package insert and pack/box, the reliance on the CMI increases (0.6% to 0.5% to 7.5%) as the medicine type changes from supermarket to OTC to prescription and this is reversed for reliance on the pack (15.8% to 11.9% to 4.2%).

Twenty-two (20%) of the 110 consumers did not complete the final section of the interview; 15 consumers (13.6%) identified 2 or less information sources, 5 consumers (4.6%) stopped their interviews due to time constraints, and 2 consumers (1.8%) thought that all of their information sources were equally reliable. The remaining 88 consumers (80%) generated a total of 299 constructs (median of 3.4, range 1–7) that were categorised under 16 main themes (Table 3).

Table 3 Categorisation of construct themes
Full size table

Knowledge, training and trustworthiness were the constructs most frequently elicited by consumers as a measure of reliability of the information sources they used. These key constructs need to be interpreted in light of which source is being considered. For example, while the knowledge and training of the doctor and pharmacist is clear, for the internet or print media, knowledge and training need to be understood as the consumer’s perceived knowledge and training of the producer of these information sources. The median values for each information source, when rated against each construct, were calculated (Table 4).

Table 4 Median scores of information sources on top eight themed constructs
Full size table

These values give an indication of differences, participants may have had in the way they perceived the reliability of their information sources. In Table 4, a median score of less than 5 suggests that the perception of the information source is towards the named construct. A score greater than 5 suggests that the perception is not towards this construct, but to some other theme. Finally, a median score of 5 suggests that the particular information source does not comfortably lie at one end or the other for that construct. For example, the low median scores for the doctor and pharmacist suggest that they are perceived to have good knowledge, be trained, trustworthy, experienced, provide up-to-date and specific information and can provide the opportunity for more interaction compared with the other information sources.

Discussion

The use of the repertory grid technique to interview consumers visiting community pharmacies was successful in providing a method of identifying the information sources that consumers use and also how the reliability of these information sources are perceived relative to each other.

In this study, consumers identified a number of information sources that they access to find out information about their medicines. The doctor and pharmacist were identified in the top three information sources, 83% and 78%, respectively, and this is comparable with other studies [10, 11, 13]. Given the sample population, this may be expected as all consumers had received a medicine on prescription and can be regarded as frequent visitors to both the doctor and a pharmacy with a median number of 3 and 6 visits in 6 months, respectively. However written information was the second main information source identified (82%) and this is different from previous studies where written information did not figure as highly [12, 13, 19]. This may be accounted for due to the different approaches used in these studies, which included asking only for two sources [13, 19], or providing a list for the participant to choose from [12]. This is a potential strength of the repertory grid technique in that the information sources elicited were provided by the consumers as opposed to a list from the researcher. Thus the information sources are meaningful to the consumer who provided them and the fact that they named a source, strongly suggests that they have experienced accessing that source. The importance of this is further enhanced when considering that the subsequent perceptions of these sources are then based on experience as opposed to an ‘artificial’ response to a source that they have never used. This is a key concept in the application of the repertory grid that the choice of elements (in this example information sources) is meaningful to the individual and in the context of the subject under study [24].

The 88 consumers who were able to generate constructs, produced a median of 3.4 (range 1–7). In a previous study using the repertory grid in angina, participants generated a median of seven constructs (range 5–12) [28]. The low median number may reflect that these are the main constructs used by individuals in their perception of the reliability of their information sources and the other constructs are considered less important.

Consumers contrasted the reliability of their information sources in terms of the key constructs of perceived knowledge, training and trustworthiness. When comparing the doctor and pharmacist to the other information sources that they used, consumers judge the health professionals as more reliable on the basis of their knowledge, training, trustworthiness and experience. Similar reasons have been cited elsewhere for patients wishing to use the doctor and pharmacist as their primary information source for medicines [10]. The internet was contrasted with alternative information sources as being seen as coming from a knowledgeable source but less trustworthy and up- to-date. This perception has been reported in other studies where the internet was ranked as the fourth most distrusted information source [11], and also acknowledged as potentially an unreliable source of information [31]. This may be explained by difficulties experienced by consumers in determining the trustworthiness of the source [15]. Written information in comparison with the internet was regarded as coming from a more trustworthy and knowledgeable source. This also has been reported previously where credibility and accuracy were considered important by consumers for written medicines information [32]. In the report by Bunn, media was also regarded as less accurate and credible when compared with other information sources [32]. This was a perception reflected by the consumers in our study where media was regarded as coming from neither a trustworthy nor knowledgeable source. These latter two findings suggest that although consumers access the internet and the media as sources of information on medicines, they do acknowledge that there are limitations on the reliability of these sources. The repertory grid technique provided an excellent tool for eliciting key discretionary themes that consumers use to evaluate the reliability of their information sources.

There are limitations that must be considered in interpreting the results of this study. The convenience sample may not be reflective of all consumers who seek information on medicines. The consumers who did not complete the interview or could not generate constructs may have very different perceptions of their medicine sources although often they would claim that they regarded them to be equally reliable. The fact that the interviews were conducted within the community pharmacy may have resulted in a more favourable response to the pharmacist as a reliable source of medicine information.

Conclusion

The repertory grid technique was successful in identifying the information sources consumers accessed to find out about their medicines and also in identifying the perception of these sources in terms of their reliability. The repertory grid provided a different approach to identifying consumer perceptions without introducing interviewer generated themes that may not have been meaningful to the participant. Thus this technique could offer a valuable tool for future research into consumer preferences for their treatment options or establishing perceptions towards medications and subsequent compliance with these.