Plugging the patient evidence gap: what patients with swallowing disorders post-stroke say about thickened liquids*

Participants and sampling

A purposive sample of 16 community-dwelling individuals who had experienced a stroke and been prescribed TL was obtained in the Republic of Ireland (ROI). Participants were recruited from four local stroke support clubs across the ROI via club organizers who facilitated contact between researchers and participants. Club organizers were contacted directly from a list of national stroke groups. Information sessions were held at individual clubs from which participants were recruited into the study. Inclusion criteria specified that participants were community-dwelling persons who had a stroke that resulted in a swallowing disorder and were prescribed TL for a period of not less than 1 week during their stay in hospital. Participants also needed to be able to communicate their experiences of TL. A principal standard for qualitative sample size is to collect data until saturation (Bernard 2006) and direction for a representative sample size was guided by recommendations for powering Delphi consensus, which suggests a minimum of 15 respondents (Hsu and Sandford 2007). Sixteen people were recruited to participate in the study. Data from two participants were not included as one presented with significant receptive and expressive communication difficulties. The second was excluded as the participant's carer contributed most to the interview process, thus not solely reflecting the participant's experience. Data were collected until questions were thoroughly explored through the 14 interviews with no new insights being achieved and it was clear that theoretical saturation had been achieved.

Data collection

The method of data collection and generation were individual semi-structured interviews. The researchers contacted club organizers to obtain permission to attend the weekly stroke club to explain the study, provide aphasia-friendly study information sheets and recruit participants. The interviews lasted between 15 and 30 min and were carried out during the time allocated for the stroke club meetings. They were audio-recorded with field notes taken by an observing researcher. Topic guide questions were based on a comprehensive literature review and related to perceptions of experience of the intervention, the decision-making process and acceptability of the intervention. Communication impairment effecting either or both receptive and expressive language is common in individuals post-stroke with 35% of individuals having symptoms of aphasia at hospital discharge (Dickey et al. 2010). An interview protocol was developed to account for communication difficulties which acknowledged that communication may be difficult and the interviewer may help by clarifying participant responses, interpreting words, vocalizations and gestures, giving sufficient time to respond and by use of communication aids. Due consideration was given to issues of informed consent, anonymity and confidentiality. Ethical approval was granted by the University Faculty Research Ethics Committee.

Data analysis

Data were transcribed, anonymised and analysed using inductive thematic analysis (Braun and Clarke 2006). An independent researcher observed the first set of interviews to provide feedback on interviewing technique. Data were transcribed and analysed independently by two researchers and compared to enhance inter-rater reliability. Coding occurred in six phases and was verified by discussion between analysts at each phase. An additional 40% random sample of the transcripts was double-checked by third and fourth researchers on the team—each reviewing 20% of scripts. Where disagreement arose in coding, consensus was reached via discussion between two to three members of the research team. Triangulation of interpretations was achieved by comparing data within and across interviews and the field notes.

Theme 1: Uncertainty

Participants’ perceptions and experiences of decision-making regarding the implementation of TL are listed in table 2. Thirteen participants reported limited participation regarding treatment implementation, a stage when they may have been very ill. They considered it an automatic process in which they described little active participation. Within this overarching theme were two subthemes (table 2) which reflect participants’ confusion and ambiguity regarding the TL treatment.

Thirteen participants reported not being involved in the decision to commence TL with many participants assuming that TL was routine treatment for people who had a stroke. Because of this perception, participants did not tend to question the decision. ‘It was just one of those things … like when you were young, your mum would give you medicine, you couldn't really ask questions … you'd just take [it]’ (J). Alongside this assumption was a vague understanding of who was involved in implementing the intervention. Collectively the hospital, nurses, physiotherapists, doctors, SLTs and occupational therapists were the varying healthcare professionals who were perceived as primarily responsible for prescribing the treatment. This lack of clarity was also characterized by a more general uncertainty. ‘I don't know who it was’ (G).

The perceived lack of sharing in the decision-making process and vagueness regarding who prescribed the treatment was further underpinned by an incomplete understanding of why the treatment was prescribed by some participants. Four participants believed they never had a swallowing difficulty which in turn led to confusion regarding the rationale for being prescribed TL. ‘I didn't feel like I needed them’ (K). Four participants identified a perceived lack of communication which one participant felt might be explained by the hospital context. ‘Well, most of the hospital are busy and they don't have time to be explain everything’ (J). Two participants observed that their acute clinical condition at the time might account for their reduced understanding rather than resulting from information not being communicated to them. The use of TL especially in the initial post-stroke stages was therefore underpinned by at best a vagueness about the intervention and participants’ own condition. It emphasizes what participants felt to be a passive role in their treatment.

Theme 2: An unpleasant experience

Participants’ experiences of physically taking TL are noted in table 3. Individuals typically recounted the experience as a disagreeable one and in contrast to the uncertainty associated with treatment decision-making, memories of drinking TL were vivid. Participants used evocative images to describe their experiences and recount how taking TL impacted on their fluid intake, quality of life and how they modified the treatment in response.

Thirteen of the 14 participants indicated their dissatisfaction with the product and resultant drinks. This was reflected in the pejorative terms used to describe thickened drinks which ranged from ‘awful’ (J) to ‘vile’ (K). The response to the taste of thickened drinks included perceptions that the taste of different drinks was neutralized by the intervention. ‘That certainly took whatever flavour was in anything, it took the flavour of it’ (K). With the exception of one individual, participants who mentioned the altered consistency resulting from the use of TL did so unfavourably, often referencing visually unappealing substances. Changes in the consistency of drinks from pre-stroke experiences did not tally with expectations ‘When you're drinking a cup of tea, you expect it to be liquid … not something you can nearly chew’ (B).

In addition to taste and consistency changes, three participants who were involved in preparing thickened drinks noted problems in using the thickener resulting, for example, in drinks that were ‘Sometimes … too thick’ (E). Strong negative memories of taking TL remained with one participant who had a stroke a number of years previously and physically recoiled when shown the product. Drinking TL was largely an unpleasant sensory experience which adversely affected participants’ enjoyment of fluids and resulted in unpleasant recollections.

Participants also commented upon other negative consequences associated with the treatment, in particular fluid intake restrictions. Such consequences along with the unpleasantness associated with using TL often resulted in active cessation of the treatment irrespective of whether participants were on it long- or short-term. ‘I asked them could I stop’ (K). Although not everyone discontinued treatment against recommendations, two additional participants reported not adhering strictly to individualized directions for using TL by modifying their treatment regime. This theme illustrates that participants were able to recall what impact taking TL had on them and often acted to modify or eliminate its effects. They did not like the sensory aspects of the treatment and reported that it negatively influenced their enjoyment of drinking and fluid intake.

Theme 3: A trade-off

Theme 3 reflects on more positive interpretations of the treatment. Participants’ reflections on the overall experience and benefits of taking TL are itemized in table 4.

Eleven participants concluded that although the TL treatment was distasteful, they could ultimately understand the purpose of it. This was essentially to promote their swallowing and avoid excessive coughing, which would in due course progress their recovery. There was a trade-off in understanding this in retrospect whilst recalling the unpleasantness of the treatment. On reflection, participants could see that there were reasons for being prescribed TL: ‘It was serving its purpose’ (D). There were varying degrees of understanding demonstrated with some participants assuming it helped because they were prescribed it but without stating clearly why. Three participants referenced potential improvements in their recovery generally. Six were able to be more specific relating the treatment to the prevention of respiratory consequences. ‘It makes sure that the tablets and the foods are gone down properly’ (F). Two other participants recognized that the stroke had affected their ability to swallow although this tended to be stated in general terms. This interpretation of TL as serving a purpose whether specifically explicated or not may explain the continued use of the treatment by some despite the general dissatisfaction with it. Despite their dislike of taking TL themselves, half the participants saw benefits and would therefore recommend it to others in a similar situation. Two participants reported it to be a good experience, while others constructed their recommendations to use it in terms of lack of choice. ‘[It is] not an experience you would choose, if you had a choice, but there's no choice really’ (D). Three participants gave clear recommendations to try the treatment with two reporting that it provides the PWD with confidence when drinking. One participant recommended against the treatment: ‘Well I don't think I would recommended them no, if you could do without them, do without them’ (E), while another would not advise a fellow patient either way: ‘I suppose I wouldn't advise her, it's very hard to make that decision’ (C). While some participants would therefore recommend the intervention, this is not typically an enthusiastic endorsement, reflecting more a needs-must scenario.

Treatment burden

The use of TL appears to constitute what can be considered a burden on the individuals prescribed this treatment. The TL treatment burden can be postulated to result from a number of factors, some which may be remediable. These include patients being in the acute stage of their illness and having a genuine dislike of the treatment. Further, Gallacher et al. (2011) describe the work of developing an understanding of a treatment and inadequate communication between health professionals as a treatment burden for patients, both of which are reflected by participants in this study. Individuals may be attempting to ease the treatment burden by not using TL as prescribed, by either making modifications to or terminating it. Treatment fidelity is an acknowledged issue for PWD using TL (Garcia et al. 2005) and such problems with adherence have been described as a financial burden in healthcare (Vermeire et al. 2001) resulting in known unfavourable patient outcomes (Kaizer et al. 2012). Specifically, non-use or alteration of TL as prescribed may result in a number of adverse individual outcomes including dehydration resulting from reduced fluid intake, aspiration and aspiration pneumonia. These translate into additional costs for the healthcare system (Altman et al. 2010, Martino et al. 2005). If clinicians are to continue prescribing TL, then additional efforts are required to reduce the treatment burden for patients and improve treatment outcomes. Given the busyness of hospital environments the acute context may be less than ideal for monitoring and ensuring proper implementation of TL. However, policy changes, collaborative working, new approaches to patient education and individual and contextual adaptations may help improve treatment use and reduce aspiration risks and dehydration (Rosenvinge and Starke 2005). Meeting the personal beverage preferences of individual patients (Simmons et al. 2001) for example, may be a route to improved use of TL. More frequent TL offerings have also been found to significantly increase fluid intake (Simmons et al. 2001). Given the known struggle of PWD on TL to meet adequate hydration levels and the potential satiety effect of TL (Cichero 2013, Sharpe et al. 2007), online software applications likewise have the potential to reduce treatment burden, as they provide treatment reminders and real-time access to product information, dosage calculators and mixing instructions for specific products.

Treatment options

It is clear the TL is a treatment that is firmly embedded in SLT practice (Jones et al. 2017, McCurtin and Healy 2017, Smithard et al. 2013); the evidence also suggests that dysphagia treatment choices both generally and specifically for TL are underpinned by clinical experience among SLTs (Garcia et al. 2005, McCurtin and Clifford 2015). As it is unlikely that significant changes in management strategies for aspiration are imminent, improvements in thickener products are required which respond to sensory issues identified by patients. Encouragingly, the commercial market is showing evidence of addressing issues relating to poor palatability, suppressed or altered taste, grainy texture (Matta et al. 2006) and appearance. The traditional starch-based thickeners which can be undesirable due to their ‘starchy’ flavour, granular texture and lump formation (Cichero 2013) are now accompanied by newer gum-based thickeners that have a smoother texture, are less grainy, more stable in maintaining their viscosity and tend to be preferred by patients given their palatability. This in turn may improve treatment adherence and hydration (Penney 2013).

Alongside improving thickening agents, the profession also has a responsibility actively to consider and trial other treatment options for managing aspiration, both alongside TL and in isolation from it. These include traditional techniques in current SLT repertoires such as postural changes and less utilized techniques such as carbonation for which there is some evidence that it reduces penetration/aspiration into the airways (Bülow et al. 2003). A number of participants in this study also reported on issues with flavour when using TL, and thickened liquids have been found to have low flavour release (Hollowood et al. 2002, Matta et al. 2006). Such patient evidence should encourage clinicians to consider flavour enhancements to facilitate taste when prescribing TL. Other alternatives are water protocols, a supplemental approach to TL that allow patients with dysphagia to consume unthickened water between meals. While evidence is also limited for this intervention, Carlaw et al. (2012) report that studies demonstrate lower rates of dehydration and chest infection. In essence, given the patient evidence, it is worth considering adopting a broader approach to managing aspiration which encompasses a variety of treatment options. This is especially important in the context of what Smithard (2016) describes as promising research for alternative treatments.

Patient involvement and ownership

The participants in this study present a picture of limited involvement in the decision to use TL and in their understanding of its function particularly in the initial stages of their illness. This may reflect a broad-brush approach to managing swallowing disorders by using TL and/or the acute nature of their illness. It suggests that communication with patients regarding the treatment occurs at a less-than-optimal time and that patient education may be insufficient to meet their needs. It may be worth reconsidering the delivery of education to aid patient comprehension at this pivotal time not limited to but including alternative strategies such as peer counselling. Given SLTs’ understanding of communication deficits and of swallowing disorders, SLTs are well placed to address this issue.

A standard response to the participants’ experiences of lack involvement in the decision-making around TL might be to discuss promoting patients’ active roles in their own healthcare (Barello et al. 2012, Mockford et al. 2011). While it is well known that patient involvement in decision-making remains poor generally (Gunnarsson et al. 2014, Watts-Pappas et al. 2008), it is difficult to see how models such as shared decision-making might be effectively operationalized for post-stroke patients with dysphagia in the acute stages. This may be especially so given that older people, who are representative of people with strokes, have a tendency to be less involved in their treatment decisions and to seek less information from their clinicians (Deber et al. 2007). Despite this, there is evidence that healthcare professionals tend to rate patient involvement highly (Hamann et al. 2012), and while there remain questions about how best to address the uncertainty reported by patients regarding this treatment, it is important to tackle it in order to ensure both treatment fidelity and improved patient experiences. Irrespective, more attention needs to be paid to educating patients in the use of TL, engaging them in the ongoing decision-making and in problem-resolution regarding this treatment.

Limitations and clinical implications

This study, in common with other qualitative work, is bound by the context and culture in which it was conducted and, hence, reflects the experiences of the particular participants interviewed. The participants involved were asked to recall their memories of the post-stroke period which may have been affected by their acute and continuing clinical condition. They may not have been able to participate fully in their treatment decisions at certain points given the nature of their illness. The findings from this study relate to a small number of individuals, most of whom had short-term experiences with the intervention and some who were a number of years post-stroke. This may have impacted or biased recollections and the findings should be viewed in this context. Future research should consider the experiences of long-term users of TL. Further details such as type of thickener prescribed, consistency prescribed and independence in drink preparation were not available and might impact on perceptions of the intervention.

Individuals who are in the acute stages post-stroke are obviously not in a position to be active participants in their treatment decision-making or communication exchanges regarding same. This of course impacts on SLTs’ abilities to operationalize shared decision-making. Irrespective, the data suggest that patients are confused about the treatment and there remain opportunities for health professionals to consider how to improve patient education regarding TL.

The participants in this study did not appear to associate the TL treatment with the SLT profession specifically. This has implications for the profession both in terms of public knowledge of the SLT role but also potentially for treatment success. Patients who dislike the treatment, for example, might not know who to discuss their concerns with, which may contribute to their early treatment cessation. Improved communication regarding the SLT role is necessary.

Fully informed decision-making utilizes a number of forms of evidence including that of patients. Such evidence should be taken into account by clinicians when making treatment decisions. Fully informed patients should also have access to evidence which represents EBP in all its forms, i.e., research, practice, patient and contextual evidence. This study contributes evidence to that process.

More consideration needs to be given by both clinicians and researchers to alternative or complementary treatments targeting aspiration.