Table 1.

The dataset

Authors (year)	Content
Schofield et al. (2011)23	Context	Semi‐structured interviews with 61 patients recruited from GP practices in three areas in the UK. Eligible patients invited via a letter from their GP. All participants had been prescribed antidepressants against depression and/or anxiety for at least a year. Purposive sampling is used to reflect the population of users of antidepressants in terms of age, sex, ethnicity and socio‐economic background. Most participants had experienced several episodes of depression at the time of the interview.
	Findings	At the time of initiation of treatment, medicines were seen as a short‐term measure to get better at a time when they really needed it. Many participants had reservations against antidepressants when first consulting their GP, due to concerns about using medicines in the long‐term, fear of addiction and negative views held by themselves or others. Using and experimenting with medicines over time makes patients experts on their own conditions and treatments. This expertise gives the patient more control over decisions regarding management of their illness.
Conrad (1985)10	Context	In‐depth interviews with 80 people with epilepsy about the meanings of medications in everyday life and why medicines are taken or not, carried out as a part of a larger project about living with epilepsy. All participants were or had been using medications against the condition. Recruitment was done via community channels. Participants are described as between 14 and 54 years old, mostly lower middle class and coming from urban areas in the United States. Interviews were held independent of medical and institutional settings.
	Findings	Peoples' practices regarding medicines are related to aspects of control in two directions: taking control over symptoms or worries about the disease, or being controlled by a disease that forces usage of medicines. A continuous evaluation of the medicines is undertaken, and patients may stop taking them if no specific effect is perceived. The author refers to a general view in society that it is better to try to achieve health goals without medicines, and reports reflections of this view in the participants' accounts.
Dolovich et al. (2008)24	Context	The study aims to investigate expectations and influences thereon to find out if expectations have an impact on how medicines are used. Purposive sampling and recruitment through community and health‐care channels in Canada were used. The 18 participants represent different ages, living conditions and types of medicines used. Semi‐structured interviews were conducted around the medicines the participant considered most important and analysed using grounded theory.
	Findings	Expectations are most clearly expressed by patients who want a clearly defined outcome in a condition that affects their daily life. Participants are realistic in what they want medicines to achieve and use different sources of information to adapt and confirm their expectations. The number of medications used every day was described as an issue; either as burdensome for an individual or in general terms.
Unson et al. (2003)29	Context	With a stated aim to increase adherence, patients' beliefs about osteoporosis (OP) medication and medicines in general are assessed. Focus groups based on ethnicity was recruited via senior centres and housing estates in deprived areas in the United States. A convenience sample of 55 women aged 60 years or older participated. Most participants were on prescribed medication, but not for OP. Authors suggest that what is handled as a dichotomous question (treatment or not) in medicine is a more complex decision for patients, where heuristics, moral aspects and power relations are at play.
	Findings	Side‐effects are considered serious and on a short‐term basis; ‘they can be worse than the disease itself’ and ‘25% protection without side‐effects is better than 50% with’. Participants use both their own experiences and those shared by others in decisions about medication. Participants expressed scepticism regarding prescribers' knowledge about how medicines affect specific patients and saw prescribing partly as doctors' experimentation.
Granger et al. (2013)	Context	Mixed methods study aiming at exploring theoretical linkages between symptom experience over time and the meaning of medication adherence. Ten patients with chronic heart failure completed questionnaires measuring beliefs, behaviours, symptoms and satisfaction and were interviewed about the meaning associated with medicines. Patients were recruited by research nurses during an admission to a US university hospital with exacerbation of their condition.
	Findings	Experience of symptoms influences the meaning attributed to medicines. Medicines are described in positive ways as tools to care for oneself, but also with negative notions of being inevitable if wanting to avoid death. Patients were questioning how a large number of medicines everyday can be helpful.
Mazor et al. (2010)25	Context	Telephone interviews with women ≥65 years that fulfilled WHO criteria for osteoporosis recruited from a multispecialty practice in Massachusetts, United States. History of dispensed prescriptions was used to classify participants into three groups of equal sizes: not using medication, started but discontinued medication and currently on medication. The study links core beliefs about medicines to patients' views on perceived need, safety and efficacy of a medication.
	Findings	The need and usefulness of medication is described as related to age, but with different conclusions: old age and brittle bones make medicines necessary, or offset the efficacy of medicines. Patients that make use of peers' knowledge or experience of the medication are sceptical to treatment. Connects participants' views on medicines with their core beliefs about health and illness.
Nair et al. (2007)26	Context	Semi‐structured interviews seeking to investigate patients' experiences with risk–benefit assessment when making decisions about treatment for type II diabetes. The 18 Canadian patients used different types of treatment and were recruited through community and health‐care channels. Both purposeful and theoretical sampling was used to ensure inclusion of patients that found treatment easy as well as difficult. The interpretation of the interviews was validated in a focus group session towards the end of the analysis process.
	Findings	Avoidance of medication may be based on the impression that one is already taking too many medicines. Duration of illness and experimentation with medicines influences the understanding of the effects of disease and the treatment. Patients develop a personalized understanding of the value of a treatment, and this forms a basis for the decision on whether to use a medicine.
Smith et al. (2000)35	Context	Experiences, concerns and willingness to participate in decision making about medicines were explored and compared between patients with the three conditions. Group interviews were arranged via voluntary organizations for each condition in the UK.
	Findings	Patients with schizophrenia had experienced medicines as a way for doctors to modify their behaviour in order to be acceptable in society. Patients wishing to leave decisions about medicines to doctors still wanted to be informed about positive and negative effects so that they could monitor their treatment. Information about medicines is used to participate in decisions and challenge doctors.
Cranney et al. (1998)30	Context	In an investigation of barriers to implementation of guidelines for hypertension treatment, UK healthy elderly patients' and GPs' perceptions of risks and benefits were addressed. Participants recruited via a GP practice (75 patients) and on a training course (121 GPs). Attitudes to risk with untreated hypertension and ideas about benefit from prescribed medicine were assessed with questionnaires accompanied by visual aids during semi‐structured interviews.
	Findings	Most patients overestimate both the risk with untreated hypertension and the benefit from preventive treatment and accepted treatment based on trusting their doctor. When provided with information about the clinically proven relative risk reduction, fewer patients accept treatment and more patients mention the risk of side‐effects. Communicated ideas about benefit with treatment are rather vague and based on conceptions of prevention being necessary or helpful in general.
Leaman and Jackson (2002)31	Context	Questionnaire completed by 216 patients from a single GP practice in the UK. A random sample of patients, stratified for age and gender, were asked to state the level of benefit requested for acceptance of a first, second and third medicine for treatment of hypertension. Benefit was represented with fixed levels of NNT(5). Hypothetical scenarios explaining the consequences of a myocardial infarction and some practical aspects of the treatment accompanied the questionnaire, and respondents were asked to answer with only these aspects in mind.
	Findings	Patients request a much higher level of benefit than what has been clinically proven. Authors mention patients' lack of rationality in making decisions about medicines, leaving decisions to doctors and altered circumstances when facing a real rather than a hypothetical situation as explanations for why so many patients are on treatment for hypertension despite the results in the study.
Fuller et al. (2004)32	Context	Older people's attitudes to stroke prevention were examined by presenting probabilities of risks and benefits with warfarin treatment. People aged 66–97 years answered questionnaires about hypothetical scenarios of risk reduction and practical aspects of treatment. The 81 participants were recruited via an elderly medicine outpatient clinic at a large university hospital in the UK.
	Findings	Participants expressed wishes for a guaranteed number of years of disease‐free survival in order to engage with medicine‐taking. Experiences of disease in the family and personal health beliefs are influential on the acceptance of treatment.
Hux and Naylor (1995)33	Context	Data on benefit of lipid‐lowering medication from a large clinical study were presented in different formats (relative and absolute risk reduction, NNT, average and stratified survival) to 100 participants aged 35–65 recruited from an outpatient setting in Canada. Treatments were presented as free of charge, without side‐effects and suggested by a doctor in hypothetical scenarios. Participants' preferences and their stated certainty about the decision were recorded to investigate how the format of benefit data influences decisions about treatment.
	Findings	Relative risk reduction generated the highest acceptance for treatment, followed by absolute risk reduction. Stratified survival data were preferred over average numbers.
Arkell et al. (2013)34	Context	Experiences, attitudes and expectations about information given prior to starting anti‐TNF therapy were assessed in focus group interviews with ten rheumatoid arthritis (RA) patients in the UK. All participants were currently on treatment and purposively sampled to represent different ages, disease duration and activity and anti‐TNF agent used. Data were analysed with a phenomenological approach.
	Findings	Patients described a willingness to face increased cancer risk due to treatment if sustained relief from RA symptoms could be achieved. Fear of disease symptoms and long‐term effects influenced the desire to start and stay on anti‐TNF medication; side‐effects were considered secondary.
Gale et al. (2012)36	Context	Meta‐ethnography of qualitative literature on usage of medication for prevention of cardiovascular disease, seeking to explore variations in behaviour and implications for practice.
	Findings	The context for patients' decisions about medicines goes beyond the clinical setting – social interactions in the personal community, other people's experiences and various sources of information are influential. Doctors are trusted sources for information about medicines; information from academia, pharmaceutical companies and media is less trusted.
Sale et al. (2011)28	Context	A phenomenological study conducted to investigate patients' experiences with the decision to take OP medication after sustaining a fracture. Participants aged over 65 years who had had a fracture in the last five years and were at high risk for having another one were recruited via an OP screening programme in Canada. Two‐thirds of the 21 patients were currently taking OP medication. Cost for medication was covered by a local drug plan for all participants.
	Findings	A decision to use or not to use medicines is often not permanent; patients report they have changed their mind about medicines or might do it later. The decision to start OP medication is often based on trusting the prescriber's recommendation. Discussing the decision about medicines with friends or family or searching information elsewhere often resulted in not accepting treatment.
Adams et al. (1997)37	Context	Attitudes of patients with asthma to prophylactic medication are explored with a patient‐centred perspective. In‐depth interviews were carried out with 30 participants recruited from a GP practice in Wales. Participants represented different ages, social backgrounds and duration of asthma.
	Findings	Using medication every day is closely linked to the idea of having a disease. Those patients that accepted the treatment as part of living with asthma still disliked using medicines every day. Negative views on steroids, associated with illicit use, were overcome by the fact that they are needed by the patient and prescribed by a professional.
Stack et al. (2008)38	Context	Patients' beliefs about multiple medicines are addressed in interviews with 19 patients diagnosed with cardiovascular disease and type II diabetes. Authors acknowledge that usage of many medicines is associated with poor adherence and self‐management in patients. Recruitment was done via two urban GP practices in the UK.
	Findings	Diabetes medicines are seen as necessary, whereas medicines for the management of cardiovascular risk, especially lipid‐lowering agents, are given lower status. The patients that describe a perceived risk for cardiovascular events have experienced heart attacks or strokes among family or friends.
Marshall et al. (2006)39	Context	Relations between level of cardiovascular risk, acceptance of treatment and demographic characteristics were investigated quantitatively. Patients without diagnosed cardiovascular disease from GP practices in the UK were invited to participate in coronary risk screening and a research study. Preferences regarding treatment in hypothetical scenarios were recorded from the 181 participants before the screening, and a second interview was conducted afterwards to see whether patients changed their minds when told about their own risk.
	Findings	Patients' preferences are stable over time but vary between individuals. A difference in acceptance of treatment between patients from different social classes is suggested to be related to ideas about moral obligation to use medicines.