Attitudinal predictors of older peoples’ and caregivers’ desire to deprescribe in hospital

Potentially inappropriate medicines (PIMs) offer more risks than benefits and are associated with adverse outcomes including morbidity, hospitalisation and mortality [1]. A prospective study across six European hospitals reported between 34.7 and 77.3% of older people were prescribed a PIM on admission [2]. There is therefore a need to review medicines to determine suitability for discontinuation, a process termed ‘deprescribing’ [3]. While the prevalence of PIMs on hospital admission is high [2], deprescribing practice is limited and reactive in response to iatrogenic harm rather than proactive to prevent future harm [4]. Given that deprescribing requires an accurate medication history and monitoring to observe response to medication withdrawal [5], an admission to hospital where these two activities are routine, provides an opportunity to develop a deprescribing intervention.

Prescribing should be based on a partnership as the prescriber is the disease expert and the patient is the expert on their illness [6]. It is therefore unsurprising that patient engagement in decision-making is a proposed essential component of deprescribing [5]. Consultations between practitioners and patients are an opportunity to determine whether deprescribing is appropriate, agree strategies for ongoing monitoring and establish the patient’s desire to try deprescribing [5].

Trials across multiple settings report up to half of older patients decline deprescribing interventions [7‐10]. Exploration of predictors for this lack of desire to deprescribe has focussed on external characteristics such as age, gender and number of prescribed medications. A recent retrospective analysis of hospital electronic medical records reported that all external characteristics analysed, including PIM prevalence, number of medicines at admission and comorbidities had no effect on patients’ willingness to accept deprescribing [10]. It is unsurprising that these characteristics do not predict desire to deprescribe as there is a substantial body of evidence in the field of behavioural science confirming that a key predictor of behaviour is attitude towards the behaviour, which is poorly predicted by external characteristics [11‐13]. Furthermore, external demographic characteristics cannot be changed thus contribute little to guiding physicians or those developing deprescribing interventions. Identification of attitudinal predictors of desire to deprescribe may therefore provide modifiable targets for an intervention targeting patients’ and caregiver’ behaviour. Such attitudinal predictors are likely to be related to patients’ barriers and enablers to deprescribing, which have been characterised in a variety of settings [14]. Patient reported barriers include perceived continued need for medicines, fear of withdrawal effects and previous bad experiences. Conversely, experience or fear of side effects, absence of observable benefits and reduction in pill burden are patient reported enablers. The extent to which these potentially modifiable attitudinal factors predict desire to deprescribe remains unknown.

Informal caregivers such as family members are increasingly involved in medication decision-making. For patients that are unable to participate in these decisions, such as those living with cognitive impairment, caregivers frequently assume sole responsibility [15, 16]. Furthermore, caregivers influence engagement with deprescribing by physicians and patients who are able to participate in decision-making [14, 17]. Despite the wide ranging influence exerted by caregivers on the deprescribing processes, their level of engagement with and attitudinal factors influencing desire to deprescribe are unknown.

Older peoples’ attitudes towards deprescribing have been described using the Patients’ Attitudes Towards Deprescribing questionnaire (PATD) [18‐21]. The majority of respondents in the reported studies indicate being satisfied with existing medication whilst incongruously, over 90% also indicate willingness to accept deprescribing proposed by a doctor. This high level of willingness contrasts the significant proportion of participants in deprescribing trials declining deprescribing propositions [7‐10]. This gap between reported willingness and observed behaviour requires further exploration. Given that willingness to accept deprescribing proposed by a doctor has demonstrated limited variation in responses, this may not provide the best data for explaining this gap.

The revised Australian-validated Patients’ Attitudes Towards Deprescribing questionnaire (rPATD) explores factors that influence desire to deprescribe not captured by the PATD. The rPATD items are grouped into the four factors of burden of taking medication, appropriateness of medication (perceived harms and benefits), concerns about stopping the medication and level of involvement in making decisions about medicines. The appropriateness factor provides the new item “I would like to try stopping one of my medicines to see how I feel without it”. This item provides an indication of the patient’s attitude towards their prescribed medication by indicating their desire to try stopping a medicine. Furthermore, given that a significant proportion of previously reported deprescribing trials have been pharmacist-led [22], this item may provide better data for explaining the gap between reported willingness to accept deprescribing proposed by a doctor and observed declining of deprescribing propositions.

This hospital-based study aimed to describe the likely desire of patients and caregivers to be involved in medicine decision-making and, identify any attitudinal predictors of desire to try stopping a medicine.

Engagement of patients and caregivers is a core component of deprescribing, yet a substantial proportion indicated limited desire to be involved in medication decision-making. Furthermore, the low desire to try stopping a medicine is in agreement with the substantial proportions of participants declining deprescribing in the trial environment [7‐10]. However, patients and caregivers overwhelmingly report agreement with deprescribing proposed by a doctor. Practitioners should not therefore dismiss deprescribing opportunities due to patients and caregivers choosing to be less involved in decision-making. The three diagnostic indicators for establishing desire to try stopping a medicine are perceptions of the number and necessity of medicines and, a desire for dose reduction. These may also assist physicians with targeting relevant attitudinal predictors during deprescribing discussions.

Given similarities between the two English-speaking nations, minimal adaptations to the Australian rPATD were required before UK use. The item exploring burden of paying for medication was adapted to reflect the UK context. Whilst the sample size estimation was based on PATD data, the observed variation in responses to the global items was comparable, yielding confidence intervals equal to or narrower than predicted. The high consent rates afford some confidence in the generalisability of the findings to the populations of the hospitals at which the research was conducted. The presence of a researcher to support patients self-completing the rPATD may have biased responses. However, similarities with the caregiver rPATD responses indicate that researcher presence is unlikely to have unduly influenced the findings.

Half of potentially eligible patients were excluded due to inability to participate in medication decision-making. Inclusion of caregivers therefore provides representation of this previously under-researched population [18‐20, 26]. The patient participant population is comparable to previous PATD studies [18‐20, 26] and to a pan European study evaluating older people’s hospital admissions [2]. The caregiver population was comparable with a US study exploring treatment preferences of caregivers involved in medication decision-making [15]. These similarities indicate that the study findings may be generalisable beyond the two hospital study sites.

Similar to previous patient PATD responses in the outpatient clinic, acute hospital and care home settings, the global rPATD items in the present study demonstrated little variation, with the majority of respondents agreeing with deprescribing proposed by a doctor whilst also being satisfied with current medicines [18‐20, 26]. There was, however, greater variation in responses to the items relating to patients’ and caregivers’ desire to be involved in medicine decision-making. This agrees with the existing literature in relation to some older people expressing preference for a passive role in decision-making [27‐31] and may also be true of caregivers, who were similarly older in age [32].

The appropriateness factor demonstrated greatest divergence between patients and caregivers. The majority of patients perceived their medicines were appropriate, whereas caregivers were ambivalent. This may be due to caregivers feeling that they lack understanding of their care recipient’s treatments [27].

Whilst there is qualitative literature indicating that deprescribing generates concerns for patients [17], the majority of patient respondents indicated that they did not hold concerns about stopping medication. This may be due to differences between actively inviting people to generate potential concerns versus inviting an opinion on specific concerns as in the present study [33]. Caregiver responses were similar to patients’, however resistance to deprescribing long-standing medication was conveyed but did not predict lack desire to try stopping a medicine. Physicians report reluctance to propose deprescribing for fear of patients perceiving this as withdrawal of care [14]; the present study suggests neither patients nor caregivers hold this view.

Some caution should be applied to this message, as whilst the majority of respondents agreed with deprescribing proposed by a doctor, they also reported content with existing medication. This potentially reflects a desire to conform, which may lead to agreement with a doctor’s recommendation to deprescribe despite concerns [31] and reluctance to report adverse outcomes such as return of symptoms [30].

The reported preference for a passive role in medication decision-making by older people in the literature [27] was expressed by some patients and caregivers in their responses to items in the involvement factor. Whilst items relating to the passive behaviour of knowledge acquisition regarding prescribed medicines attracted high agreement, the item relating to liking to be involved in decisions about medicines was lower.

The attitudinal predictors of desire to try stopping a medicine for both patients and caregivers are perceived necessity and a desire for dose reduction. As both items are from the appropriateness factor, this may represent a limitation of using an appropriateness item as the primary outcome. However, this could also suggest that attitude towards the appropriateness of medication is the most suitable target for a behaviour change intervention. Additionally, the predictive ability of the burden item regarding taking too many medicines for patients and not for caregivers suggests that a patients’ perceived burden of medicine taking is an important indicator of their desire to try stopping a medicine.

As the target behaviour is deprescribing and a key predictor of deprescribing is attitude towards deprescribing [11‐13], the three attitudinal predictors are potential intervention targets. The finding that perceived medication necessity and a desire for dose reduction are predictors of both patients’ and caregivers’ desire to try stopping a medicine may offer efficiencies for intervention design. Behaviour change techniques offer an evidence-based approach to modifying attitudes towards a behaviour. For example, a practitioner may identify that a patient is prescribed an inappropriate medicine who is ambivalent to deprescribing. The present study indicates that one or more of three attitudinal predictors of desire to try stopping a medicine may alter this ambivalence. For example, the patient’s perception that they are not taking too many medicines can be targeted with the evidence-based behaviour change technique ‘information about emotional consequences’ [34, 35]. This theory-based approach to changing patients’ attitude towards deprescribing has been reported in the EMPOWER trial, which includes the behaviour change technique ‘information about health consequences’ [36].

Patients and caregivers overwhelmingly report agreement with deprescribing proposed by a doctor yet vary in the extent to which they want to be involved in medicine decision-making. Practitioners should not therefore dismiss deprescribing opportunities due to patients and caregivers choosing to be less involved in decision-making. Three attitudinal predictors of reported desire to try stopping a medicine provide modifiable targets for developing a hospital deprescribing intervention targeting patients’ and caregivers’ behaviour. Future work should focus on identifying and testing evidence based behaviour change techniques targeting these attitudinal predictors for inclusion in deprescribing interventions.