Introduction

Delirium is a syndrome, characterised by a disturbance in attention and awareness that develops over a short period and fluctuates over time,¹ with altered arousal also an important factor.² Delirium has a high prevalence in hospitalised patients and a range of delirium screening tools have been developed to improve detection.^3,4 Screening tools should be validated⁵ with psychometric properties preserved in clinical settings.⁶ The SQiD is a validated screening tool that has well established negative predictive value, specificity (89%) and sensitivity (44%),⁷ and has become widely used and encouraged in routine clinical care.⁸ It has been popular with clinicians, for example in the United Kingdom, where one in three units demonstrated use.⁹

We hypothesise that the popularity of the SQiD may relate to its simplicity and face validity. The SQiD comprises a single question Do you feel that [patient’s name] has been more confused lately? asked of a carer/relative/friend (hereafter referred to as the informant) closely involved in the patient’s care, as part of routine clinical care.¹⁰ The SQiD was shown to be specific, moderately sensitive, and easy to integrate into routine care.⁷ The SQiD performed better than other tools available at the time of its inception and has been suggested as an appropriate tool to guide whether clinical review is needed.¹¹

Although validated and used clinically,⁷ not much is known about how the SQiD works. Qualitative methods may help answer this question, as they enable in-depth exploration of a phenomenon.¹² The primary objective of the study is to derive themes that emerge in clinician-informant conversations prompted by the SQiD, to assist in determining in what way the SQiD screening tool for delirium may function.

Methods

The study setting was the haematology/oncology ward at a large university-affiliated acute hospital in Sydney, Australia. A nursing in-service on delirium and screening tools, including the SQiD, was conducted on the ward. A brief outline of the SQiD talks study was provided, including the aim and consent processes. Nurses were provided with verbal and written instructions about obtaining assent from both the patient and informant to participate in the interview in the context of limited initial disclosure, followed by full consent (see below). Nurses were given a printed prompt for their use of the SQiD question Do you feel that [patient’s name] has been more confused lately? and requested to ask the question of informants of inpatients. There was no additional formal training on its use. Nurses were neither coached nor discouraged from discussions with informants, and guided only with just answer how you normally would if a relative/carer asks you questions. Due to the binary nature of the SQiD answer and the iterative discussion process between clinician and informant, a non-negative response was considered SQiD positive.

Patients and their nominated informant were recruited to the study if they were adults able to communicate in English and provide informed consent. Patients and their informant were approached by nursing staff opportunistically when an informant was present at the bedside. As per the instructions provided to nursing staff, potential participants were invited to participate in a clinical communication study with nurses and advised about what this entailed (recording the conversation and accessing electronic medical records [EMR] for clinical details). Full details about the study were delayed until after the recording took place. Assent (agreement) to participate in the recording was essential. If assent to record was given, the nurse then audio-recorded the SQiD and subsequent conversation, in the patient’s presence.

After the recording was made, the study purpose and rationale were explained to the patient and informant by a research team member, to enable them to provide informed consent to participate. If the patient assented to participate but was unable to provide informed consent, their ‘person responsible’ was approached to consent on their behalf. Limited delayed full disclosure of the nature of the study was necessary to mitigate bias in the patient and informant response, caused by education regarding delirium that would occur during full informed consent.⁷

Basic demographic data (age, sex), and documentation of delirium and/or dementia diagnoses in the patient’s EMR were recorded.

For consenting participants, audio files were transcribed verbatim, checked, and de-identified. A record was made of whether the SQiD interview was positive or negative for delirium, and the SQiD outcome was compared to contemporaneous clinician documentation of delirium on the patient’s EMR. No research standard was applied to determine true positives or negatives for delirium as this was not a validation study. Participants were planned to be recruited until data saturation (i.e., when no new themes emerged with analysis of further transcripts).¹³

Interview transcripts were analysed using thematic analysis (Braun and Clarke (2006), within a grounded theory framework (Glaser & Strauss, 1967) (Box 1), by two investigators (NH and AW).

Results

Characteristics of patients and their informants

Recruitment was from November 2022 to April 2023. Twenty-nine patients and their informants were recruited, the range in interview duration was 7-211 seconds (mean 66.1 seconds) (Table 1). A positive SQiD was observed in 15/29 interviews. Six of 29 had a diagnosis of delirium recorded in the EMR, while five of these six had a positive SQiD outcome. None of the patients had a recorded diagnosis of dementia. One patient had documentation of cognitive impairment. Twenty-eight participants were able to provide informed consent, one consent was provided by the ‘person responsible’. No patient withdrew consent following enrolment.

Thematic analysis

Three themes emerged in the thematic analysis: recognition of “confusion”, operational factors, and the SQiD outcome. Illustrative quotations for the themes and subthemes are provided in Table 2.

(i) Recognition of confusion

This theme reflected both nursing staff and informant concepts of confusion (Table 2). Nursing staff varied in their elaboration of aspects of confusion after the SQiD prompt. Informants recognised confusion by symptoms associated with syndromic delirium, such as fluctuation in mental state, change in cognition and functional abilities, and hallucinations (Table 2). The opposite polarity (i.e., the patient was not confused) also emerged in some instances where there was actual clinical evidence of confusion. Some informants seemed to normalise confusion as an expected, age-related change.

(ii)Operational factors

This theme reflected practical aspects of administering the SQiD. Patients and informants commented on the ease and brevity of the tool (Table 2). Another was the most appropriate informant choice, whereby informants suggested who among the patient’s family/friends might be best placed to answer the SQiD.

The patient’s presence during the SQiD conversation had various impacts. Sometimes the patient provided useful information, but on other occasions appeared to inhibit information gathering. Some informants appeared to be uncomfortable disclosing information in front of the patient, leading to normalising or minimising potential symptoms of delirium.

(iii)The SQiD outcome

The interpretation of the SQiD outcome was not straightforward. There were occasions where an initial negative response to the SQiD became a positive result through further discussion with an inquisitive nurse. On other occasions, the SQiD result was unclear despite discussion (Table 2).

The outcome of informants raising concerns about possible delirium with treating teams was infrequently discussed. However, there was a range of responses from dismissal to action (Table 2).

Overarching themes

Three overarching themes emerged: clinician investment/confidence, communication techniques, and knowledge of delirium and other cognitive disorders.

Illustrative quotes are presented in Table 3.

(i)Clinician investment/confidence

Immersion in the SQiD transcripts revealed that some nurses were more engaged or confident in following on from the SQiD prompt compared with others. Following an initial negative response to the SQiD, nurses varied from immediately ending the discussion to asking follow-up questions. In some cases a negative response was revised to a positive result via an iterative discussion between informant and nurse.

(ii) Communication techniques

A variety of communication techniques were used by nurses, influencing the richness of information gathered and outcome. These techniques included active listening, checking and encouragement to enhance engagement. This intersected with the nurse’s curiosity and persistence that characterised some interviews, resulting in clarifying and sometimes even completely revising responses.

A nuanced interviewer tended to demonstrate a combination of open and closed questions, conveyed empathy and validation during the conversation, or used humour to defuse tension.

Some nurses followed up on the initial SQiD by paraphrasing it or exploring specific features of delirium, which elicited different or more detailed responses from the informant.

Nurse: Yeah and it took her like longer to reply and like…

Informant: Definitely

Nurse: …and like a few times to clarify what the meaning was?

Informant: Definitely, yeah

Nurse: Was that the only thing that you noticed, or was there anything else, like any behaviour, or anything like that?"(daughter, interview 5)

SQiD conversations where the patient actively contributed to discussions were more challenging to interpret (see Table 2). In these instances, nursing staff used communication techniques to defuse defensiveness when it occurred, including humour, or appeared to collude with the patient to maintain rapport.⁷

(iii) Knowledge of delirium and other cognitive disorders

The nurse’s knowledge of delirium was evident in follow-up questions asked. Often, probing questions centred on orientation, rather than the full range of symptoms comprising syndromic delirium (e.g. inattention, fluctuation).

Nurse: Okay, that’s good. So he’s like oriented where he is…who you guys are?

Although the SQiD conversation was focused on identifying recent changes in cognition, there were occasions where this may have been conflated with an underlying longer-term cognitive disorder, like dementia, which was not distinguished from syndromic delirium by the nurse.

Discussion

Few publications utilise qualitative methods to better understand the functionality of delirium screening tools.^19,20 This study focused on the SQiD, and how it works as a delirium screening tool. Emergent themes of’recognition of “confusion”, operational factors, and the SQiD outcome were identified. The results confirm previously hypothesised strengths of the SQiD, including ease of administration and brevity .

New insights were derived about what nurses and patients understand by the term confusion, with various symptoms and signs, some more specific to delirium than others.^21,22 Interestingly, although inattention has long been in the diagnostic criteria for delirium, this was not specifically asked in any follow-up questions after the SQiD, and disorientation was overemphasised. It has been proposed that the term confused is colloquially understood.²³ However, the word confusion has been problematic in delirium research, due to a lack of consistency in definition, and as different specialities utilise different terms when referring to delirium, resulting in parallel bodies of literature. The use of the term delirium rather than confusion is encouraged and provides a more precise diagnosis,^24–26 but may not be meaningful to informants without education.

The premise of the SQiD is that clinicians could lead the informant from a colloquial understanding of confusion, via nuanced questioning, towards an indication of delirium presence or absence.¹⁰ This is reliant on clinician understanding of the components of syndromic delirium. No tool can be both ultra-brief and operationally didactic in supporting delirium detection, therefore a combination of approaches, integrated with education may prove useful. A short tool like the SQiD may encourage participation and engage carers, whereas a more structured validated tool like the 4AT may assist clinician training.

It also emerged that judging the SQiD outcome can be difficult. While some nurses accepted a negative response to the SQiD at face value, others were more curious and went on to explore symptoms of delirium, sometimes eliciting a different, potentially more accurate result. The SQiD may engage curiosity and interest in detecting delirium in some clinicians, while others apply it with less engagement.

The overarching themes derived illustrate the challenges of embedding delirium screening in clinical practice, and opportunities for enhancing the efficacy of the SQiD by optimising the information gathered from informants. Clinician knowledge about delirium appears to be an important factor in achieving a useful SQiD. This study shows that clinician understanding of delirium cannot be assumed. Nurses with a more thorough understanding of the nature of delirium and its differentiation from other cognitive disorders, may be able to probe for delirium symptoms in a more targeted manner and identify longer-term cognitive impairment which may predispose to delirium.^27,28

Delirium management is an important consequence of detection.²⁹ However, informants only infrequently shared their observations of confusion with treating teams. When they did share their observations, information appeared to be dismissed rather than a lever for management. The need to link screening results with action has been previously identified,^26,30 and is relevant to effective implementation of the SQiD. Approaches to trigger action include using a delirium monitoring tool after screening,³¹ adding a scoring system to screening that is linked to action,³⁰ and embedding workplace processes that ensure staff respond appropriately to a positive screening test.³²

Communication techniques emerged as important in helping or hindering the SQiD conversation. The use of validation, encouragement, paraphrasing, and humour, facilitated data gathering from informants, echoing the observations of others.^33,34 The study identified the potential consequences of the patient listening and contributing to the SQiD discussion, hitherto an unexplored factor in SQiD administration. Patients could endorse or elaborate informant responses, but also inhibit the SQiD conversation, through distraction, instigating separate conversations, or silencing the informant. These observations suggest there may be value in conducting SQiD conversations away from the bedside, or finding ways to support open discussion.

This overarching theme of communication was closely linked with investment in the SQiD tool. Curiosity and/or investment in engaging in discussion moved the interaction beyond the single question to activate effective delirium screening. This is in contrast with tools like the 4AT, which although giving the clinician less flexibility in administration, provide measurable observations and cognitive tests with patients themselves.³⁵ Knowledge of delirium is associated with increased interest and engagement with the syndrome and its detection.^21,36 This study suggests potential value in providing delirium education to those administering the SQiD.

Limitations

Although no new themes emerged in the final interview, we may not have reached data saturation. As a pragmatic time-limited clinical study with minimal funding, we relied on staff who were undertaking the study additional to their regular clinical duties during a period of COVID-related staff shortages, leading to challenges in participant recruitment. Additionally, as analysis was based on transcripts only, important concurrent nonverbal communication may have been missed. Nursing staff administering the tool were not blinded to the study aims. This, along with recording the informant interview, may have affected participation and discussion in various ways, e.g., by encouraging an enhanced clinical interaction that was not reflective of usual practice.

Conclusion

This is the first qualitative study examining how the SQiD works in delirium screening. We gained insights into how nurses and informants conceptualise confusion, operational aspects of the tool, and the nuances of interpreting the screening result. Overarching themes situated the tool within real-world clinical contexts, observing challenges of clinician investment and engagement in delirium screening, variable knowledge bases and communication skills. The findings suggest a need for further qualitative analysis of the SQiD in clinical settings, particularly how effectiveness may be improved by harnessing the function of the SQiD in leveraging communication between clinicians and families/carers to encourage clinician curiosity and interest in delirium detection.

Disclaimer

The views canvassed in the following article reflect the beliefs of the three authors and not the institutions they represent.

Acknowledgments

The authors thank Dr Alexandra Thoms and Ms Yu Ting Chan, along with the nursing staff of the study ward, for their assistance in data collection.

Author Contributions

The authors confirm contribution to the paper as follows:

Conceptualisation: MS, AW; data curation: NH; formal analysis: NH, AW; funding acquisition: NH, AW; investigation: NH; methodology: NH, AW, MS; project administration: NH, AW; supervision: MS, AW; writing– original draft: NH, AW; writing – review and editing: NH, MS, AW.

Ethics statement

Approval for this research was gained from the Concord Hospital Human Research Ethics Committee (protocol no. 2022/ETH01450).

Funding sources

This research was supported by the Beverley Raphael New Investigator Grant, 2022, Royal Australian and New Zealand College of Psychiatrists.

Declaration of Interests

MS was the original developer of the SQiD tool.