[Posted as supplied by author]

Details of the research process, interview schedules and observation grid

1. Details of the research process

1.1. Data analysis

All interviews were taped and transcribed verbatim. All transcripts were entered into QSR NVivoÒ (version 1.2) for support in coding and analysis. Analysis was performed using the principles of grounded theory, and using an inductive approach combining biomedical and sociological perspectives. This balanced approach was felt to be the best approach, considering that the main objective of the study was to contribute to a better understanding of medicines use in the elderly.

AS (a clinical pharmacist) and VL (a sociologist) started the analysis with open inductive coding, through an intense line-by-line reading of transcripts of interviews with two doctors, one nurse and one pharmacist. Data collection and analysis continued simultaneously, and observations took place while the analysis of interviews with HCPs was in process. Finer coding was applied as more data were coded and analysed. Throughout the analysis, AS and VL wrote memos to store ideas, insights, and interpretations. This was essential in the development of analytical ideas.

The provisional coding scheme generated by AS and VL was discussed with other members of the research team, after independent reading of a sample of transcripts. Some codes were refined, and researchers moved to axial coding, looking for relationships between categories, and then to selective coding. Researchers searched for elements in the data that seemed to contradict the emerging explanation of the phenomena under study. Analysis was enhanced by constant comparison between interview and observation data, looking for deviant cases, and by comparison with available research in this field.

Findings were further validated by sending all participants (HCPs) a summary of the results and asking them to report any disagreement. None disagreed on the categories as they were presented, but the majority added comments on the perceived relative importance in everyday practice (eg by saying that it occurred inconstantly, or that in contrast it was a very important factor). We did not supply them to patients because the findings as they were presented were obtained several months after focus groups were conducted, and it seemed difficult to ask patients’ views on something that happened several months before.

Besides the inductive analysis described above, researchers had defined in advance descriptive codes to facilitate text retrieval and comparisons during analysis. These codes were: step of the medication use process (prescribing, patient counselling, information transfer); perceived appropriateness (appropriate, difficult-inappropriate).

AS coded all transcripts, using inductive and descriptive categories. A second pharmacist, not involved in the rest of the study, coded two transcripts to check for reliability. Calculated Cohen’s kappas were 0.87, 0.60 and 0.54 for nodes relating to activity, perceived appropriateness, and influencing factor respectively. Since the third value was not satisfactory, these subnodes were redefined and all material was subsequently recoded.

The data from focus groups were used to compare patients’ accounts to providers’ account. The data set did not enable researchers to perform an inductive approach on this.

The quotes presented in the article were chosen for representativeness, and further checked by external reviewers for relevance and understanding.

1.2. Roles of the researchers

None of the persons collecting the data were involved in patient care, to limit the risk of researcher-respondent interaction. One member of the research team (AS) was involved in data collection (all interviews with HCPs and one set of observations) and data analysis. Other persons involved in data analysis were not involved in data collection and vice-versa, because it was not feasible or unsuitable. All members of the research team were involved in data analysis, and they represented different disciplines (geriatrics, pharmacy, pharmacology and social sciences). This helped to prevent personal or disciplinary biases of a single researcher from excessively influencing the findings.

1.3. Process for informed consent

For semi-structured interviews with HCPs, each potential participant was contacted by telephone or by electronic mail. The researcher introduced herself as a pharmacist doing a research project whose aim was to collect HCPs’ views on the use of medicines in their practice. All agreed to participate (i.e. we had a 100% acceptance rate), and an interview time and date was arranged. They all received a letter providing further details about the project, assuring them of confidentiality and asking for written consent (see supplemental file). No financial incentive was offered.

For focus groups with patients, after identification of eligible participants with the medical team, AS had a first contact with each patient in his/her room. She presented herself as a person doing research on the use of medicines in elderly patients, and explained that we were interested in their views on treatment changes and information received. She also explained the modalities of the group discussion. Each patient received a personalised letter with additional details but no agreement was sought at that time. AS came back the following day to ask whether they agreed to participate. The acceptance rate was 95%. No financial incentive was offered.

Anonymity was assured to all participants in the information letter. No names of persons or institutions were transcribed and participants were identified by numbers only. When using quotes for illustration of the results, researchers removed each piece of talk that could lead to an identification of the person or institution.

2. Interview guides and examples of interview questions

Note to readers: all questionnaires were developed and piloted in the French language. Interview guides and questions have been translated for illustration purposes only.

2.1. Semi-structured interviews with health care professionals

A. Interview guide

1° Introduction

• Personal experience in geriatrics, and roles/responsibilities with regard to medicines use
• Definition of appropriate medicines use

2° Detailed questioning on the following domains of medicines use

• prescription and follow-up
• patient counselling
• transfer of information to the general practitioner

For each domain, discuss the following:

• description of activities
• perceived quality in practice; difficulties encountered (encourage examples)*
• reasons for difficulties/ inappropriateness
• approaches for optimisation

3° Miscellaneous (years of experience in geriatrics, years of practice under current position, postgraduate training in geriatrics)

B. Examples of interview questions (*)

• [After defining appropriate medicines use] What is the situation like in your daily practice? (positive, negative perceptions)
• Which difficulties do you face in the prescription of medicines (/follow-up) for elderly patients admitted on your unit?
• Many scientific publications have looked at the use of medicines in elderly patients, and some of them highlighted that medicines use is often suboptimal. What do you think? What does it mean to you?
• Do you think that medicines tend to be misused (/overused, /underused) ? Overused? Misused? Can you give examples from your practice ?
• What are the reasons for this? Why does this happen? Can you explain the reasons why this happens?

2.2. Focus groups with inpatients

A. Interview guide

1° Introduction (opening questions)

• Roundtable; names and origin

2° Treatment changes

• Description (introductory question)
• Positive experience, advantages
• Difficulties, discomfort, annoyance
• Optimisation (suggestions)

3° Information received

• Description
• Satisfaction / Difficulties
• Optimisation (suggestions)

4° Discharge

• Fear, difficulties
• Optimisation

5° Ending questions

B. Examples of interview questions (key questions)

Let’s talk about the changes that have been made during your stay here  with regard to your medicines.

• What makes you happy about it?
• What makes you unhappy? What bothers you?
• Which difficulties have you encountered?

Tell me about the last time somebody told you something about your medicines

• What did this person tell you? Are you satisfied? If not, why?
• What else would you like to know about your medicines?

 

3. Observation grid

A. Instructions

• Objectives and instructions for taking notes
• Types of notes to be taken (observed events, documents, introspection, informant interviewing)
• Activities to take part during the observation period, and types of data to access

B. Observation grid – main categories

• Prescribing / treatment changes *
• Follow-up for safety and efficacy
• Patient counselling
• Discharge
• Interactions between members of the multidisciplinary team in relation to medicines use
• Comments by patients/caregivers/family with regard to their medicines
• Interactions between the observer and HCPs working on the unit
• At the end of the observation period: global perception of the quality of medicines use; changes in views over time; perceived bias due to the presence of the observer (pharmacist)

C. Observation grid - example of subcategory (*)

• Treatment changes performed: examples – when, who, what, rationale, attitudes, documentation, patient information
• Events illustrating difficulties to ensure optimal medicines use (…)
• Sources of information used to answer questions on medicines (…)
• Events illustrating patient influence (…)

Online poll

Find out more >>