Creating opportunities for patient participation in managing medications across transitions of care through formal and informal modes of communication

Abstract Background Communicating about medications across transitions of care is important in older patients who frequently move between health care settings. While there is increasing interest in understanding patient communication across transitions of care, little is known about older patients' involvement in formal and informal modes of communication regarding managing medications. Objective The aim of this paper was to explore how older patients participated in managing their medications across transitions of care through formal and informal modes of communication. Methods The study was conducted across two metropolitan hospitals: an acute hospital and a geriatric rehabilitation hospital in metropolitan Melbourne, Australia. A focused ethnographic design was used involving semi‐structured interviews (n = 50), observations (203 h) and individual interviews or focus groups (n = 25). Following thematic analysis, data were analysed using Fairclough's Critical Discourse Analysis. Results Data analysis revealed two major discursive practices, which comprised of an interplay between formal and informal communication and environmental influences on formal and informal communication. Self‐created patient notes were used by older patients to initiate informal discussion with health professionals about medication decisions, which challenged traditional unequal power relations between health professionals and patients. Formal prompts on electronic medication administration records facilitated the continuous information discourse about patients' medications across transitions of care and encouraged health professionals to seek out older patients' preferences through informal bedside interactions. Environmental influences on communication comprised health professionals' physical movements across private and public spaces in the ward, their distance from older patients at the bedside and utilization of the computer systems during patient encounters. Conclusion Older patients' self‐created medication notes enabled them to take on a more active role in formal and informal medication communication across transitions of care. Older patients and family members did not have continuous access to information about medication changes during their hospital stay and systems often failed to address older patients' key concerns about their medications, which hindered their active involvement in formal and informal communication. Patient or Public Contribution Older adults, family members and health professionals volunteered to be interviewed and observed.

documentation of patients' progress, treatment goals and medication regimens in medical records. Informal modes of communication refer to opportunistic, unplanned, spontaneous, unstructured interactions that can take place at any time in different locations such as at the bedside, in corridors and in office spaces, and also involve self-initiated communication aids including handwritten notes or reminder scribbles. 6 Older patients' participation in medication communication can differ depending on the structure and the context of communication events that they are involved in as well as the severity of their cognitive impairment during hospitalization. 7,8 Although involving older patients with cognitive impairment in formal interactions is challenging for health professionals, a few strategies have been found to be helpful for patients with mild cognitive impairment such as simplifying decisions, holding the discussions in a quiet environment, using teach-back methods and using printed tools including decision aids. 9 Previous studies also indicated lack of involvement of older patients in formal communication encounters during ward rounds where discourses were shaped by organizational culture, clinician behaviours and entrenched power relations between patient-provider relationships. 3,10-13 For example, tensions between patient-centred discourses and organizational discourses in medical or nursing education have been reported where senior doctors focused more on meeting the educational needs of junior medical staff or where nursing educators focused on memorizing isolated facts about medications than involving patients in decision-making processes. 14,15 Studies of formal ward rounds have revealed inconsistent engagement with patients, patients' experiences of health professional dominance and the exercise of hierarchical power by controlling medication decisions as well as the spaces where communication occurred. 10,12 Furthermore, organizational factors, such as competing demands of health professionals, staff workloads, temporal and spatial challenges, 16 or inaccurate information transfer between health professionals and settings have been identified as hindrances to collaborative conversations between health professionals and patients during formal encounters upon patient admission and at hospital discharge. 17 Temporal challenges included lack of availability of patients or family members at a time that coincided with health professionals' availability, whereas environmental and spatial challenges included barriers to communication due to distance of health professionals from patients' bedside. 18 Notably, there has been a lack of focus on informal interactions between patients and health professionals concerning medication information 19 and on patients' proactive communication about their needs and goals as they move across settings. 16

| Study design
A focused ethnographic design using multiple methods was adopted to explore communication processes between older patients, family members and health professionals in relation to managing medications across transitions of care.

| Setting and study participants
This study took place in two metropolitan hospitals in Melbourne, Australia: a 500-bed tertiary teaching hospital and a 150-bed geriatric rehabilitation facility. Data collection was conducted in a general medicine ward and two medical wards specialized in infectious diseases and general respiratory in the teaching hospital, and in three aged care and two rehabilitation wards in the geriatric rehabilitation facility.
Participants comprised older patients, family members and health professionals. Inclusion criteria for patients were those aged 65 years or older on admission. We chose this age cut-off because 65+ has been conventionally used to designate an older person in the literature. 20,21 Older patients with severe cognitive impairment were excluded. Researchers screened all potential patients with nurse unit managers of designated wards to make sure that they were cognitively capable of consenting before approaching them at the bedside. Every older patient had a Glasgow Coma Scale (GCS) score documented by ward nurses in the observation charts within patients' medical records. This score was used to complement other means of determining patient eligibility to participate, which included discussions with the nurse-in-charge, patients and family members.
Older patients who had a GCS score lower than 14 were not included in the study. Inclusion criteria for family members were any individuals who were associated with older patients admitted to hospital. Inclusion criteria for health professionals were registered nurses, pharmacists and doctors employed for at least 1 week at the study settings.
Two researchers from nursing and pharmacy backgrounds undertook data collection. In consultation with nurse unit managers and bedside nurses, researchers initially viewed patient lists of each study ward to determine the potential eligibility of patients. After providing written and verbal information about the project at the bedside, researchers obtained informed consent from older patients.
Purposive sampling was adopted to ensure that the older patients represented various age categories including youngest-old (65-74), middle-old (74-84) and oldest-old (85 years or over). Older patients were approached at the bedside at a time that was convenient for them. We approached 44 older patients in acute care settings and 34 patients in subacute care settings. Older patients were given unlimited opportunity to decide whether they wished to take part.
Of those who were approached in acute care settings, 26 patients accepted and 18 patients declined, whereas, 24 patients accepted and 10 patients declined in subacute care settings. Health professionals were also recruited using purposive sampling taking into consideration their discipline, level and the length of their experience to ensure a variety of professional backgrounds. Purposive sampling was conducted by identifying specific characteristics in older patients and health professionals and ensuring that all characteristics were covered before the recruitment. During the recruitment of participants, information about the study was provided through participant information sheets and a verbal set script summarizing the information that was included in the participant information sheet.
Participants were provided sufficient time to read the study information and consent form and ask questions. The time given to patients to consider whether or not to take part ranged from a few minutes to a couple of days depending on patients' availability to read the participant information and consent forms.

| Data analysis
Initially, verbatim transcriptions of data were read and reread to increase familiarity. First, to explore key characteristics of issues, the research team coded a selection of transcriptions. The first selection comprised five different transcriptions and all researchers independently coded those transcriptions inductively. After all codes were compared and contrasted and any discrepancies were resolved amongst the research team, a thematic coding framework was constructed to be applied to the remaining transcripts. All transcriptions and the coding framework were transferred into NVivo 12 (QSR Melbourne), and the remaining data were coded by three researchers independently. Codes were clustered into larger groupings and into themes. 22 Data analysis was discussed at fortnightly research team meetings.
The Critical Discourse Analysis (CDA) developed by Fairclough was used to conduct a complementary discourse analysis of the data. According to a Faircloughian approach to CDA, discourse is associated with language use as a form of social practice. CDA is the synthesis of theoretical positions that provide opportunities to conduct a critical examination of the relations between power inequalities and language, discourse and social contexts and language and ideology. Given that a focused ethnographic design comprises exploring social phenomena and shared experiences within a specific culture, using CDA was a suitable method of analysis that allowed the researchers to explore overarching ideologies, roles, social relations, unspoken rules of discourses relating to whose voices were dominant or marginalized and common-sense assumptions affecting health professionals' everyday social practices. The CDA approach helped us to explore the data at three different dimensions including the text level, the discursive practice level and the social practice level. 23 The text level involved examining the structure of the communication encounters, the aspects of the language that people used and the content that was prioritized. At this level, we also analysed language devices including the use of humour, modality, back-channelling and hedging according to their broad definitions. The discursive practice level explored how discourses were produced in various ways in specific social contexts. At this level, researchers focused on exploring power relations between participants, roles and professional status of participants, the way they were positioned in the interaction and times and places that were relevant to conversations.  24,25 Three researchers analysed the data independently according to the codebook and discrepancies were discussed and resolved during regular research team meetings. Researchers used pseudonyms when transcribing and analysing the data to maintain the confidentiality. Verbal and written consent was obtained from patients to conduct individual interviews and focus groups and to collect the data from medical records. Informed consent was also obtained for the use of audiovisual media such as photos. For observations, researchers attempted to obtain written consent from health professionals and older patients. If, after explaining the project, the older patient wished to participate, but was unable to read the participant information and consent form due to visual problems, general tiredness and lack of desire to read, or was unable sign the consent form due to poor hand grip as a result of frailty, a process of verbal consent was followed thereafter. This process of obtaining verbal consent was approved by the ethics committees. Any identifiable information, such as patients' names or name of the hospital, was removed at the time of data collection. The confidentiality of data collected from medical records was ensured by associating these data with the patient pseudonyms instead of using real names during data collection processes. The data collected from medical records were stored on an online-secured storage drive of the University. To identify eligible patients from patient lists, researchers obtained permission from head medical consultants and nurse unit managers of each study unit to access these patient lists. This process was also approved by the ethics committees.

| RESULTS
In all, 50 older patients participated in semi-structured interviews. In addition, 25 focus groups were conducted across both hospitals. If  Table 1. Two major discursive practices were interpreted

| Interplay between informal and formal communication
Communication via formal and informal modes flowed backwards and forwards in health care settings, which created opportunities for older patients' participation in decision-making processes. Self-created patient notes and the prompts on medical records enabled patient involvement in medication changes across transitions of care. In the following observation, the nurse was administering morning medications and a 70-year-old patient was unwilling to take aperient Movicol ® (macrogol) since he believed that the timing was inappropriate. He created notes to remind himself to discuss the timing of medication doses with the doctors ( Table 2).
The patient was unhappy about the lack of availability of doctors since there was no opportunity to ask for the timing of administration of Movicol ® to be moved from morning to night.
The patient was also frustrated by his inability to remember to relay his complaint during ward rounds, so had taken to writing his questions on a piece of paper as a reminder. The nurse supported the patient writing reminder notes, emphasizing the challenge of recalling the medication-related questions during ward rounds since the flow of the conversation was usually steered by doctors.
When questioned at a later time, the patient confirmed that his notes helped him to ask questions and provide suggestions about changing the administration time of his laxative, which indicated that the older patient's interest was eventually being served since he positioned himself as an active agent in urging doctors to change the administration time of the medication. Self-created patient notes prompted the patient to initiate informal discussion to impact medication decisions, which eventually played a role in  F I G U R E 1 A template demonstrating patients' self-medication administration on electronic medication administration record.  These field notes in  Pharmacist: Atorvastatin, which is also known as Lipitor ® .

| Environmental influences impacting the flow of informal and formal medication communication
Wife: Lipitor ® . Yeah I know Lipitor ® .
Pharmacist: And he was on 40 milligrams, now after as someone has had a stroke and also had a heart attack, we try to maximise their statin.
Pharmacist: 80. And we would do that because not just for cholesterol and lipids which is what it's traditionally used for, but also there might be a site on that blood vessel which has a plaque which may have contributed to the stroke. So this helps stabilise that plaque to prevent further strokes. And there's good evidence to show that it does that. I know- Wife: That's right.
Pharmacist: And yeah I know in the media there's a bit of-Wife: Bit of negative?
Pharmacist: But not-I think that's more to do with primary prevention but this is for secondary prevention and there is a lot of goodstrong evidence to support that.
Wife: That it helps.

T A B L E 4 Field notes from an observation of ward rounds
'The team included the registrar, the resident and the intern doctor stand in the ward corridor in front of computer to review the medications for a 79 years-old patient on the eMAR. While standing in the corridor, the resident and registrar discuss the patient's inhaler technique at home and they conclude that the patient has poor technique. The resident suggests Handihaler ® and summarises the medication changes to the registrar. Doctors move on the patient's room to talk about the patient's inhaler (Symbicort ®budesonide and formoterol fumarate dihydrate). The registrar tells the patient that they will put her back on her normal dose that the patient used to take at home. The registrar informs the patient: "We'll put you back on what you were on before" and she changes the topic and starts talking about her bowels, explains the change of her laxatives to PRN (as needed) and that she can ask for it if she needs it. The ward round team leaves the patient's room, move to stand in the ward corridor outside the room and have further discussions about the patient's inhaler and aspirin'.  (Table 5).
Although mobile computers provided nurses with easy access to medication details while in the bedside area, information about medications was seldom communicated to older patients themselves.  their general practitioners in the community 27,28 and health professionals at the hospital. However, patients also found that some health professionals did not really look at these notes even though patients carried these with them across care transitions. 27 Our research has further identified that self-created notes not only functioned as aide-memoires but also empowered older patients to initiate opportunistic discussions with health professionals.
However, these opportunistic discussions only occurred when health professionals came to the bedside and made themselves available to listen to older patients' concerns. As indicated by Liu et al.,14 collaborative discourse between older patients and health profes- Older patients' involvement in the medication communication was influenced by health professionals' spatial behaviour and the utilization of the computers during their formal and informal patient encounters.
As previously highlighted, doctors' flexibility of moving across different spaces when making important decisions about medications created a hindrance to older patients' potential input, which contributed to maintenance of medical dominance in decision-making processes. 35 Although computers on wheels provided health professionals with an Therefore, simplified information about medication changes should be made available for older patients and family members through userfriendly tablet computers or patient portals at the bedside so that they can track any changes following their hospitalization and also access written information simultaneously during ward rounds or nurses' medication rounds.
This study has limitations. Potential observer effects may have occurred with participating health professionals when they were communicating with patients. However, observer effects were mitigated by building rapport with health professionals and continuous observations over several weeks in a single ward before moving to the next ward. Additionally, the study was conducted at two metropolitan hospitals, and therefore, the findings may not be transferable to individuals located in regional or rural areas. More nurses than doctors and pharmacists participated in observations within this study, as they were the health professionals most commonly present at the bedside.

| CONCLUSIONS
This study illustrated the importance of older patients' self-created medication notes not only as an aide-memoire but also as a tool

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.