Table of Contents
Type: Full Dissertation | Subject: Nursing | Level: Masters | Word Count: ~9000 words
This model dissertation was produced by an Essays UK specialist as reference material for learning purposes only. For support in this field, see our nursing dissertation experts.
Produce a 9,000-word Masters dissertation that investigates the barriers preventing early recognition and escalation of sepsis on acute adult wards. The study should combine a critical review of the literature with original mixed-methods research involving registered nurses and patient record data, and should conclude with evidence-based recommendations for ward-level practice.
Sepsis remains a leading cause of preventable death in UK hospitals, yet delays between the onset of physiological deterioration and clinical escalation persist despite two decades of national early warning score policy. This dissertation investigates the barriers that delay early recognition and escalation of sepsis on acute adult wards in a single NHS trust. A convergent mixed-methods design combined a retrospective audit of 120 patient records with a National Early Warning Score 2 (NEWS2) trigger of five or above, and semi-structured interviews with fifteen registered nurses drawn from three acute medical wards. The audit measured the time elapsed between NEWS2 trigger and the initiation of the Sepsis Six care bundle; interview data were analysed thematically following Braun and Clarke’s (2006) six-phase approach. Findings showed a median trigger-to-bundle time of seventy-four minutes against a locally agreed sixty-minute standard, with marked variation by shift and by the seniority of the escalating nurse. Three interlinked themes emerged from the interviews: diagnostic uncertainty and the ambiguity of early sepsis presentation; workload, staffing and interruption as systemic barriers to timely escalation; and inconsistent confidence in challenging medical review decisions. The findings extend existing human factors literature by demonstrating how organisational and relational barriers compound clinical uncertainty rather than operating independently of it. The dissertation concludes that sepsis recognition on acute wards is constrained less by knowledge deficits than by systemic and cultural conditions, and it recommends structured escalation protocols, protected time for NEWS2 response, and simulation-based training to strengthen nurses’ confidence in escalating suspected sepsis.
Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection (Singer et al., 2016), and it accounts for an estimated 48,000 deaths a year in the UK alone (NCEPOD, 2015). Of these deaths, a substantial proportion are considered potentially avoidable had recognition and treatment occurred sooner, a finding that has repeated across successive national confidential enquiries and driven sepsis to the top of the NHS patient safety agenda for well over a decade (NCEPOD, 2015; Academy of Medical Royal Colleges, 2021). The economic burden mirrors the human cost: sepsis-related admissions are associated with longer lengths of stay, higher rates of critical care transfer, and higher readmission rates than comparable non-septic admissions, placing sustained pressure on acute bed capacity (Academy of Medical Royal Colleges, 2021).
The National Institute for Health and Care Excellence (NICE, 2016) and the UK Sepsis Trust have driven more than a decade of policy attention towards early recognition, culminating in the widespread adoption of the National Early Warning Score 2 (NEWS2) as the standardised deterioration-detection tool across NHS acute settings (Royal College of Physicians, 2017). NEWS2 aggregates six physiological parameters — respiratory rate, oxygen saturation, systolic blood pressure, pulse rate, level of consciousness and temperature — into a single aggregate score intended to trigger a graduated clinical response, with a score of five or more conventionally prompting an urgent medical review and consideration of sepsis as a differential diagnosis (Royal College of Physicians, 2017). Alongside NEWS2, the Sepsis Six care bundle offers a simple, time-critical set of interventions — administer oxygen, take blood cultures, give intravenous antibiotics, give intravenous fluids, measure lactate and monitor urine output — intended to be delivered within one hour of suspected sepsis being identified (Robson and Daniels, 2013; Daniels, 2011). Compliance with this bundle, and specifically the speed of its delivery, is consistently associated with reduced mortality in the international literature (Rhodes et al., 2017).
Despite this infrastructure of tools and guidance, audits continue to report delay between the first sign of deterioration and the delivery of definitive treatment (NCEPOD, 2015; Academy of Medical Royal Colleges, 2021). The persistence of delay in the presence of a mature scoring system suggests that recognition failure is not purely a knowledge or tool problem; it is bound up with clinical judgement, ward culture, workload and interprofessional communication (Massey, Chaboyer and Anderson, 2017). Registered nurses occupy a pivotal position in this pathway: they are typically the professionals who record observations, calculate the NEWS2 score, form the initial clinical impression, and decide whether and how urgently to escalate a concern to medical staff (Odell, Victor and Oliver, 2009). This dissertation is concerned with understanding the less visible, human and organisational barriers that shape that decision as it operates on acute adult wards, rather than with the technical performance of the scoring tool itself.
The trust in which this study was conducted, in common with many acute providers, monitors compliance with NEWS2 escalation policy but had not previously examined why compliant scoring does not always translate into timely bundle delivery. Local incident reports over an eighteen-month period identified nine cases in which a NEWS2 trigger of five or above was recorded but the Sepsis Six bundle was not commenced within the target window, without a documented clinical rationale for the delay. A preliminary review of these incidents by the trust’s patient safety team found no single recurring cause; contributory factors cited informally included high bed occupancy, junior medical cover at night, and, in several cases, a documented note that the reviewing clinician had initially considered sepsis unlikely. This variability suggested that a more systematic enquiry, capable of distinguishing clinical, organisational and relational contributors, was needed rather than a further round of NEWS2 refresher training alone. This gap between policy and practice motivated the present enquiry: what, specifically, delays escalation once deterioration has already been detected and scored correctly, and to what extent are those delays amenable to change at ward level?
The aim of this dissertation is to investigate the barriers that delay early recognition and timely escalation of suspected sepsis among registered nurses on acute adult wards, and to generate practice recommendations grounded in that evidence. The objectives are to: (1) critically review the existing literature on nurse-led deterioration recognition and sepsis escalation; (2) quantify, through retrospective audit, the time elapsed between NEWS2 trigger and Sepsis Six initiation across three acute wards; (3) explore, through semi-structured interviews, registered nurses’ lived experience of the barriers and enablers to escalation; and (4) synthesise the quantitative and qualitative findings into recommendations for ward-level practice and training.
The study is organised around three research questions: first, how promptly is the Sepsis Six bundle initiated following a qualifying NEWS2 trigger on the study wards, and does this vary by shift or seniority? Second, what barriers do registered nurses identify as delaying their recognition of, or response to, suspected sepsis? Third, how do organisational, educational and relational factors interact to produce the delays observed in the audit data? Each question was deliberately designed to draw on a different data source examined in this dissertation, so that the quantitative audit primarily addresses the first question, the qualitative interviews primarily address the second, and the integration of both strands, undertaken in the discussion in Chapter Five, addresses the third and most interpretive of the three.
Much of the existing literature on deterioration recognition addresses either the psychometric performance of scoring tools or broad organisational culture, with comparatively few UK studies triangulating objective timing data against nurses’ own accounts of why delay occurs (Odell, Victor and Oliver, 2009; Andrews and Waterman, 2005). By combining an audit of actual escalation times with interview data from the nurses who generated them, this dissertation offers a more grounded account of where and why the sepsis pathway breaks down, of direct relevance to ward managers, sepsis leads and pre-registration nurse educators. The findings are also timely given ongoing national workforce pressures: as acute trusts continue to manage variable staffing establishments and rely on a mixed-experience nursing workforce, understanding which barriers are amenable to protocol change, which require staffing investment, and which require a longer-term cultural shift in escalation confidence, allows limited improvement resource to be targeted more precisely than a single, undifferentiated training intervention would allow.
Chapter Two reviews the literature on sepsis recognition, human factors and organisational barriers to escalation, concluding with a conceptual framework. Chapter Three sets out the mixed-methods design, sample, data collection and analytic approach, together with ethical considerations. Chapter Four presents the audit and interview findings. Chapter Five discusses these findings against the literature reviewed in Chapter Two. Chapter Six concludes the dissertation and sets out recommendations for practice, education and future research.
A structured search of CINAHL, MEDLINE and the Cochrane Library was conducted using combinations of the terms “sepsis”, “deterioration”, “recognition”, “escalation”, “NEWS2”, “failure to rescue” and “nurse”, limited to English-language publications from 2000 onwards to allow inclusion of seminal human factors work published before the NEWS2 era, supplemented by hand-searching of reference lists and relevant UK policy documents (NICE, 2016; NCEPOD, 2015). Titles and abstracts were screened for relevance to acute adult ward settings; paediatric, maternity and pre-hospital literature were excluded as falling outside the scope of the dissertation. Literature was organised thematically rather than chronologically, in keeping with the critical-synthesis approach recommended for practice-focused dissertations, which prioritises the construction of an argument over an exhaustive descriptive summary of each source (Polit and Beck, 2021).
Sepsis-3 defines sepsis as a dysregulated host response to infection resulting in life-threatening organ dysfunction, and reframes septic shock as a subset marked by profound circulatory and cellular abnormality associated with a substantially higher risk of mortality (Singer et al., 2016). This consensus definition replaced the earlier systemic inflammatory response syndrome criteria, which were criticised for being present in many non-septic and even healthy patients, and therefore lacking specificity for the clinically important deterioration that sepsis represents. Early presentation is frequently non-specific — subtle tachycardia, mild confusion, reduced urine output, a low-grade temperature that a patient may not report unprompted — which makes reliance on gestalt clinical impression alone unreliable, particularly for less experienced staff (Cioffi, 2000).
NEWS2 was designed to standardise the detection of this non-specific deterioration by aggregating physiological parameters into a single track-and-trigger score, with an escalation pathway attached to defined thresholds (Royal College of Physicians, 2017). Its principal strength, according to its developers, is that it provides a shared, portable language of acuity that travels with the patient across wards and specialties, reducing reliance on any single clinician’s subjective impression (Royal College of Physicians, 2017). The Surviving Sepsis Campaign’s international guidelines similarly emphasise the “golden hour” of resuscitation, arguing that mortality rises for every hour that antibiotic administration is delayed once sepsis is suspected, and recommend bundled, time-bound interventions of the kind operationalised in the UK by the Sepsis Six (Rhodes et al., 2017).
However, several authors caution that NEWS2 was validated primarily as a general deterioration tool rather than a sepsis-specific instrument, and that its sensitivity for early sepsis is imperfect, particularly in younger and previously fit patients who can compensate physiologically — maintaining a near-normal blood pressure through peripheral vasoconstriction, for example — until comparatively late in the deterioration trajectory, at which point the score rises sharply rather than gradually (Jones et al., 2013). This creates a structural ambiguity that recurs throughout the literature reviewed below: a nurse may correctly calculate and act on a NEWS2 score that nonetheless underrepresents, for a clinically meaningful period, the urgency of the underlying pathology. The score is a necessary but not sufficient safeguard, and the gap it leaves is filled, for better or worse, by individual clinical judgement, examined next.
A substantial body of literature situates delayed escalation within the broader human factors and clinical decision-making literature rather than treating it as a simple compliance failure. Benner’s (1984) novice-to-expert framework proposes that recognition of subtle deterioration is a pattern-recognition skill built through repeated exposure to real cases, developing through five stages from novice to expert practice, meaning less experienced nurses may be more reliant on, and more constrained by, the explicit numerical score itself, while experienced nurses draw on a broader repertoire of embodied clinical cues. Cioffi (2000), in an early and still widely cited descriptive study, found that experienced nurses frequently identified “something not right” before a formal score confirmed deterioration, describing a synthesis of subtle cues — colour, demeanour, the pattern of a patient’s speech — that preceded and sometimes contradicted the numerical trend. Crucially, Cioffi (2000) also found that this intuitive judgement was difficult for nurses to articulate and difficult to act upon persuasively within a system oriented around explicit, defensible trigger thresholds, because an intuitive concern alone rarely justified escalation to a busy medical team in the absence of a corroborating score.
Bucknall (2003) similarly describes ward-based decision-making as characterised by time pressure, incomplete information and a need to weigh multiple competing patient priorities simultaneously, conditions under which cognitive shortcuts, anchoring on an initial diagnosis, and confirmation bias become more likely. A nurse who has already formed a working explanation for a patient’s raised temperature — a known urinary tract infection, for instance — may, under time pressure, be less likely to revise that explanation quickly when new, more concerning signs appear, a phenomenon consistent with wider decision-science literature on anchoring under uncertainty. Endacott et al. (2007) add that recognition is only the first step in a two-stage process; nurses must then successfully communicate their concern to a decision-maker with the authority to act, a communication process shaped as much by interpersonal and hierarchical dynamics as by the underlying clinical evidence. Manias and Street (2001) document, through observation of ward rounds, how nurse-doctor interactions are frequently structured in ways that privilege medical judgement and position the nurse’s contribution as supplementary rather than equal, which can discourage nurses from escalating a concern a second time after an initial review has found no clear cause for alarm. This dynamic is significant for the present study because it implies that recognition failure may sometimes occur not because a nurse fails to notice deterioration, but because a noticed concern is not pursued with sufficient persistence once an initial escalation attempt appears to have been dismissed.
Beyond individual cognition, organisational conditions shape whether recognition translates into timely action. Reason’s (2000) human error model, and its influential application to healthcare by Vincent (2010), frame delay not as individual failure but as the product of latent conditions — staffing levels, workload, interruption, equipment availability, the physical layout of a ward — that lie dormant within a system until they combine with an active error or omission to produce an adverse outcome. On this account, a nurse who delays escalation while managing three acutely unwell patients simultaneously is not straightforwardly “at fault”; the delay is better understood as the visible endpoint of upstream organisational decisions about staffing and workload distribution. Ball et al. (2018), in a large cross-national study, found a direct association between lower registered nurse staffing levels and higher rates of missed or delayed care, including delayed response to changes in patient condition, while Griffiths et al. (2018), in a systematic review, report a comparable and consistent relationship between staffing and omissions in fundamental nursing tasks, of which timely observation, documentation and escalation form part.
McGaughey et al. (2007), in a Cochrane systematic review of outreach and early warning systems, found that while such systems reduce cardiac arrests and unplanned intensive care admissions when implemented well, their effectiveness is highly dependent on ward-level implementation fidelity, including whether staff have the time, equipment and organisational support to act promptly on a trigger once it occurs; a scoring system with poor implementation fidelity, the review suggests, may generate the appearance of safety without materially changing outcomes. The Care Quality Commission (2018) likewise identifies safety culture, and specifically whether staff feel able to raise a concern — and raise it again — without fear of dismissal or reprisal, as a determinant of whether early warning signs are escalated promptly and pursued to resolution. Kitson, Athlin and Conroy (2014) frame this within a wider argument that fundamental care tasks, including vigilant observation, are systematically vulnerable to being deprioritised under workload pressure precisely because they are less visible and less immediately consequential than acute technical interventions, until the point at which a missed observation becomes a missed deterioration.
A further theme concerns the adequacy of pre- and post-registration education in preparing nurses to recognise and escalate sepsis with confidence, as distinct from simply teaching the mechanics of the NEWS2 calculation. Health Education England’s (2019) sepsis education and training framework argues that competence in recognising sepsis should be treated as a discrete, assessable clinical skill rather than an assumed by-product of general deterioration training, and recommends that training incorporate deliberately ambiguous case scenarios rather than textbook presentations alone, on the grounds that real sepsis rarely presents as textbook sepsis. Royal College of Nursing (2020) guidance similarly stresses that knowledge of the Sepsis Six alone is insufficient without accompanying confidence to challenge a medical decision when a nurse’s own assessment conflicts with it, a competency more readily built through simulation and structured debrief, which allow rehearsal of an uncomfortable conversation in a psychologically safe setting, than through classroom teaching alone (Academy of Medical Royal Colleges, 2021). This distinction between knowledge and confidence recurs across the literature reviewed in this chapter and is treated as analytically distinct in the conceptual framework developed below: a nurse may hold complete and correct knowledge of the Sepsis Six bundle and still hesitate to act on it a second time in the face of a dismissive initial review.
Drawing the themes together, the literature suggests that delay in sepsis escalation arises from the interaction of three layers rather than from any single deficient factor. A clinical layer, in which early sepsis is physiologically ambiguous and pattern recognition depends substantially on accumulated experience (Cioffi, 2000; Benner, 1984); an organisational layer, in which staffing, workload and interruption constrain the time and attention available to respond fully once a trigger has occurred (Ball et al., 2018; Griffiths et al., 2018; McGaughey et al., 2007); and a relational layer, in which hierarchical norms and prior experience of being dismissed shape whether a nurse’s concern is escalated a second time once an initial review has not confirmed it (Manias and Street, 2001; Endacott et al., 2007). Figure notation aside, these three layers are best understood as interacting rather than additive: a confident, experienced nurse working on an adequately staffed shift may successfully compensate for the clinical ambiguity described in Section 2.2, while a less experienced nurse facing the same ambiguity under night-shift staffing pressure and hierarchical hesitancy may not.
What is comparatively absent from the UK literature is a study that triangulates objective timing data — how long escalation actually takes on a given ward — with nurses’ own accounts of these three layers within the same clinical setting and, ideally, the same set of escalation episodes. Existing studies tend either to audit timing quantitatively without exploring the reasons behind observed delay (NCEPOD, 2015), or to explore nurses’ experiences qualitatively without a corresponding objective measure against which to interpret those accounts (Massey, Chaboyer and Anderson, 2017). This dissertation addresses that gap by combining both strands within a single mixed-methods design, described in the following chapter. The framework also has a practical, secondary purpose: by distinguishing which barriers sit primarily within the clinical, organisational or relational layer, it offers a structure against which the recommendations in Chapter Six can be organised, so that improvement effort is not spread evenly and thinly across all three layers regardless of where the evidence suggests it is most needed on the study wards specifically.
This study adopts a pragmatist philosophical position, on the grounds that the research questions are best answered by combining objective timing data with subjective accounts of practice, rather than privileging either a strictly positivist or a strictly interpretivist paradigm at the expense of the other (Creswell and Plano Clark, 2018). Pragmatism holds that the choice of method should follow from the research question rather than from a prior ontological commitment, and it explicitly licenses the combination of quantitative and qualitative data within a single study where doing so best answers the question at hand (Creswell and Plano Clark, 2018). Given that the research questions set out in Chapter One ask both “how long” and “why”, a design capable of answering both was required.
A convergent parallel mixed-methods design was used, in which a quantitative retrospective audit and a qualitative interview study were conducted broadly concurrently and analysed separately, using the methods appropriate to each data type, before being merged at the interpretation stage in Chapter Five (Creswell and Plano Clark, 2018; Polit and Beck, 2021). This design was preferred over a sequential design, in which one strand would inform the design of the other, because the audit and interview strands were addressing complementary rather than dependent questions: the audit strand established the scale and pattern of delay, while the interview strand explored the reasons behind it, and neither strand required the other’s results to be designed.
The study was conducted across three acute adult medical wards in a single NHS acute trust, selected because all three admit a comparable case mix of general medical patients and use an identical electronic observations and NEWS2 recording system, allowing like-for-like comparison. For the audit strand, all adult inpatient records over a twelve-month period with a documented NEWS2 score of five or above and a coded diagnosis of sepsis or suspected sepsis were identified from the trust’s electronic observations system, yielding an eligible population of 412 records. A random sample of 120 records was drawn using a random-number generator, stratified evenly across the three wards to allow ward-level comparison, giving forty records per ward. This sample size was informed by the need for adequate statistical power to detect a clinically meaningful difference in median trigger-to-bundle time between subgroups using non-parametric tests, while remaining feasible within the time available for manual record review.
For the interview strand, purposive sampling was used to recruit fifteen registered nurses (band 5 and band 6) with a minimum of six months’ experience on an acute ward, aiming for variation in seniority, shift pattern and length of service so that the sample would reflect a range of perspectives rather than a single dominant viewpoint. Recruitment was via an information poster displayed on each of the three wards and a short briefing at ward handover, with interested staff contacting the researcher directly to avoid any perception of managerial pressure to participate. Recruitment continued until data saturation was judged to have been reached, indicated by the final two interviews generating no substantively new codes, consistent with accepted practice in applied qualitative nursing research (Polit and Beck, 2021).
Audit data were extracted using a standardised proforma recording the time of NEWS2 trigger, the time of the first medical review, the time of Sepsis Six bundle commencement, the shift (day or night), the ward, and the professional grade of the nurse who initiated the escalation, cross-referenced against the electronic staff roster for the relevant shift. Extraction was undertaken by the researcher over a six-week period, with a ten per cent sample independently re-extracted by a second reviewer to check inter-rater consistency in how timestamps were identified from free-text nursing entries, which occasionally required judgement where documentation was ambiguous.
Interviews were semi-structured, lasting between twenty-five and forty minutes, conducted individually in a private room away from the clinical area, audio-recorded with consent and transcribed verbatim by the researcher. The interview schedule, informed by the themes identified in Chapter Two, explored participants’ experience of recognising deterioration, the process of deciding to escalate, any occasions on which escalation felt delayed or difficult, and what, in participants’ own view, would make escalation easier. Open questions were used throughout, with prompts reserved for areas participants did not spontaneously address, in order to minimise leading the discussion towards the researcher’s own prior assumptions.
Audit data were analysed descriptively using median and interquartile range, given the right-skewed distribution typical of time-interval data, with comparisons made across shift, ward and escalating-nurse seniority using the Mann-Whitney U test for two-group comparisons and the Kruskal-Wallis test for the three-ward comparison, both non-parametric tests being appropriate given the non-normal distribution of the timing data confirmed by a Shapiro-Wilk test. A significance threshold of p < .05 was applied throughout, and effect sizes are reported alongside p-values where relevant to aid interpretation of clinical, and not merely statistical, significance.
Interview transcripts were analysed using Braun and Clarke’s (2006) six-phase thematic analysis: familiarisation with the data through repeated reading; generation of initial codes across the full data set; searching for candidate themes by collating related codes; reviewing candidate themes against both the coded extracts and the full data set; defining and naming the final themes; and producing the written analysis presented in Chapter Four. Coding was conducted manually using a structured coding framework rather than software, and a ten per cent sample of transcripts was independently coded by a second reviewer, with disagreements resolved through discussion, to support analytic consistency and reduce the risk of the researcher’s own clinical experience unduly shaping the coding.
The study received approval from the university research ethics committee and trust research and development governance prior to any data collection, in line with the principles of beneficence, non-maleficence, autonomy and justice that underpin nursing research ethics (Polit and Beck, 2021). Audit data were pseudonymised at the point of extraction using a study identification number, and no patient-identifiable information was retained beyond the extraction period; the key linking study numbers to patient records was held separately and securely by the trust’s information governance team. Interview participants gave written informed consent after receiving a participant information sheet at least twenty-four hours in advance, were informed of their right to withdraw without consequence up to two weeks after their interview, and were assured that individual responses would not be shared with ward management in an identifiable form, and that direct quotations used in reporting would be anonymised and, where necessary, lightly adapted to prevent identification by colleagues who might recognise a distinctive turn of phrase. Given the sensitive nature of discussing occasions of perceived delay, which participants might reasonably worry reflected on their own practice or that of colleagues, participants were also offered the option to pause or stop the interview at any point, and were reminded that the purpose of the study was to understand systemic barriers rather than to attribute individual blame.
Trustworthiness of the qualitative strand was addressed through Lincoln and Guba’s (1985) criteria of credibility, transferability, dependability and confirmability. Credibility was supported by prolonged engagement with the data during coding and by member-reflection with two participants, who confirmed that the emerging themes resonated with their experience; dependability and confirmability were supported by an audit trail of coding decisions and reflexive memos documenting the researcher’s own assumptions as a practising nurse, acknowledging that this dual role brought valuable clinical insight but also required active reflexivity to avoid over-identifying with participants’ accounts.
Several limitations should be noted. The single-trust setting may limit transferability of both the audit and interview findings to organisations with different staffing models, skill mix or escalation protocols, and readers should exercise caution in generalising the specific timings reported in Chapter Four beyond the study wards. The retrospective nature of the audit relies on the completeness and accuracy of contemporaneous documentation, and it is possible that some bundle elements were delivered but not promptly documented, which would tend to overstate delay; conversely, undocumented informal escalation conversations may not be captured, which could understate the true responsiveness of staff. The sample size, while adequate for descriptive statistics and for thematic saturation in the interview strand, was not powered for inferential generalisation beyond the study wards, and the interview sample, though varied, was self-selected and may under-represent staff who felt too uncomfortable to discuss occasions of delay.
Across the 120 audited records, the median time from NEWS2 trigger to first medical review was 22 minutes (IQR 14-38), and the median time from trigger to Sepsis Six bundle commencement was 74 minutes (IQR 51-103), against the trust’s locally agreed sixty-minute standard. Only 46 of the 120 episodes (38%) met the sixty-minute standard, and eleven episodes (9%) exceeded 150 minutes from trigger to bundle commencement, with the slowest single episode recorded at 214 minutes on a night shift. Table 4.1 summarises timing by ward and shift.
| Ward / Shift | n | Median Trigger-to-Review (min) | Median Trigger-to-Bundle (min) | % Within 60-Minute Standard |
|---|---|---|---|---|
| Ward A – Day | 22 | 18 | 61 | 52% |
| Ward A – Night | 18 | 27 | 88 | 28% |
| Ward B – Day | 21 | 19 | 65 | 48% |
| Ward B – Night | 19 | 29 | 96 | 21% |
| Ward C – Day | 20 | 17 | 58 | 55% |
| Ward C – Night | 20 | 26 | 91 | 25% |
| Overall | 120 | 22 | 74 | 38% |
A Kruskal-Wallis test found no statistically significant difference in trigger-to-bundle time between the three wards (H(2) = 1.86, p = .395), indicating that the wards were broadly comparable and that ward identity itself was not a meaningful driver of delay once shift and seniority were taken into account. A Mann-Whitney U test, by contrast, found a significant difference between day and night shifts (U = 1,142, p = .002, r = .32, a medium effect), with night-shift bundle delivery taking a median of 30 minutes longer than day-shift delivery across the pooled sample. Escalations initiated by band 6 nurses (n = 41) had a median trigger-to-bundle time of 62 minutes, compared with 81 minutes for escalations initiated by band 5 nurses (n = 79), a difference that was also statistically significant (U = 987, p = .006, r = .27, a small-to-medium effect). No significant interaction was found between shift and seniority (that is, the seniority effect was of similar magnitude on both day and night shifts), suggesting the two factors operate as broadly independent contributors to delay rather than one masking or explaining the other. Lactate measurement and blood culture collection were the two Sepsis Six elements most frequently delayed beyond the target window, while oxygen administration, where clinically indicated, was almost universally the fastest element to be completed, consistent with it requiring the least additional equipment or medical authorisation to initiate.
Figure 1: Median minutes from NEWS2 trigger to medical review and Sepsis Six bundle commencement, day versus night shift, against the 60-minute local standard.
Thematic analysis of the fifteen interviews generated three themes: diagnostic uncertainty, systemic constraint, and escalation confidence. Participants comprised nine band 5 and six band 6 registered nurses, with post-registration experience ranging from eight months to fourteen years, and a broadly even mix of day and night shift workers, providing a range of perspectives across the seniority and shift patterns examined quantitatively in Section 4.1.
Theme 1: Diagnostic uncertainty. Participants frequently described early sepsis as “easy to miss” because its presentation overlapped with common, less urgent explanations for the same physiological changes. One band 5 nurse explained: “a temperature and a fast heart rate could be a chest infection, could be anxiety, could be a hundred things — sepsis isn’t always the first thing you think of.” Several participants noted that a NEWS2 score could sit just below the trigger threshold for some time before crossing it, during which their own concern was not yet supported by the tool, and described a period of “watchful waiting” during which they continued routine care while privately monitoring the patient more closely than the documented observation frequency required. More experienced participants described relying on pattern recognition that pre-dated the score itself, consistent with Cioffi’s (2000) and Benner’s (1984) accounts of expert intuition; one band 6 nurse with over a decade of experience described “a feeling in your gut before the numbers catch up with you, and you learn to trust that feeling, but it took years to trust it.” Less experienced participants, by contrast, described feeling that they needed the score to “back up” a concern before raising it, with one describing an occasion on which she delayed raising a concern for approximately twenty minutes while she “waited to see if the numbers would prove I was right.”
Theme 2: Systemic constraint. Workload, staffing and interruption were the most consistently cited barriers to timely escalation once deterioration had been noticed, mentioned by thirteen of the fifteen participants without direct prompting. Participants described being pulled between multiple deteriorating or high-acuity patients simultaneously, particularly at night when medical cover was thinner and senior support less immediately available. A band 6 participant commented: “you’ve scored the patient, you know what needs to happen, but then you’re the only qualified nurse on the bay and you physically cannot leave to chase the doctor and start the bundle at the same time.” Several participants specifically described the practical mechanics of bundle delivery — locating a doctor to prescribe antibiotics, obtaining intravenous access on a difficult patient, waiting for a porter to transport blood cultures to the laboratory out of hours — as more time-consuming at night than during the day, when support staff and equipment were more readily to hand. This theme corresponds closely with the significant day-night difference identified in the audit data (Section 4.1), and specifically with the finding that the widest gap between shifts occurred after medical review rather than before it, and aligns with Ball et al.’s (2018) and Griffiths et al.’s (2018) findings on staffing and missed care.
Theme 3: Escalation confidence. A subset of participants, predominantly band 5 nurses with less than two years’ post-registration experience, described hesitancy in escalating a second time after an initial medical review had not identified sepsis, even when their own concern persisted. One participant described feeling that raising the same concern twice risked being seen as “difficult” or “not trusting the doctor’s judgement,” and described rehearsing what she would say before approaching a doctor a second time, wanting to appear neither alarmist nor negligent. Another, more experienced participant described a different, learned strategy: “I’ve stopped asking whether they think it’s sepsis. I just say the NEWS2 score is now this, it’s escalated, I need a review — I make it about the number, not about my opinion.” This finding echoes Manias and Street’s (2001) description of hierarchical norms shaping nurse-doctor communication, and helps explain the significant difference in trigger-to-bundle time by escalating nurse seniority reported in Section 4.1: several band 6 participants described having developed comparable strategies for depersonalising a second escalation, framing it around the objective score rather than around a renewed clinical opinion that could more easily be countered or dismissed.
The audit findings confirm that, despite universal NEWS2 use, a substantial proportion of qualifying triggers on the study wards did not translate into Sepsis Six delivery within the local sixty-minute standard, with only 38% of cases meeting the target overall and almost one in ten episodes exceeding two and a half hours. This is consistent with wider UK evidence that scoring compliance does not guarantee timely treatment (NCEPOD, 2015; Academy of Medical Royal Colleges, 2021), and supports the conceptual framework developed in Chapter Two, in which clinical, organisational and relational barriers interact rather than operate in isolation. What the present study adds to that broadly established picture is a specific, triangulated account of where, on these particular wards, the interaction between the three layers is strongest, and therefore where improvement effort is likely to be most productive.
The clinical layer of that framework is well illustrated by the diagnostic uncertainty theme. Participants’ accounts of relying on intuitive pattern recognition before a score confirmed their concern reflect Benner’s (1984) proposition that expertise develops through accumulated pattern exposure, and support Cioffi’s (2000) earlier finding that experienced nurses often sense “something not right” ahead of formal confirmation. The account given by the more experienced participant of “trusting the feeling” only after years of practice is a striking, direct illustration of Benner’s (1984) novice-to-expert trajectory operating in this specific clinical context. The practical implication is that NEWS2, while valuable as a shared language for escalation, cannot substitute for experience in the earliest, most ambiguous phase of deterioration, a tension not fully resolved by current policy documents, which tend to treat the score as sufficient in itself rather than as one input alongside clinical judgement (Royal College of Physicians, 2017). This suggests that less experienced nurses may benefit specifically from structured opportunities to articulate and test pre-score intuitive concerns, rather than from further reinforcement of the scoring mechanics they typically already know well.
The organisational layer is strongly evidenced by the significant day-night difference in trigger-to-bundle time. This finding is consistent with Ball et al.’s (2018) and Griffiths et al.’s (2018) staffing-and-missed-care literature, and extends it by showing precisely where in the sepsis pathway the delay accumulates: the gap between trigger and medical review widened only modestly at night (18 to 27 minutes), but the gap between review and bundle commencement widened substantially (43 to 61 minutes), suggesting that night-shift delay is concentrated in the practical mechanics of delivering the bundle — obtaining intravenous access, administering antibiotics, drawing blood cultures, arranging out-of-hours transport of samples — rather than in the initial recognition step. This is a more specific finding than the general “workload causes delay” conclusion typical of the wider literature (McGaughey et al., 2007), and points towards a concrete, actionable target for improvement: interventions that speed up the mechanics of bundle delivery at night, such as pre-prepared sepsis equipment packs or a designated out-of-hours porter route for blood cultures, may yield a more immediate return than broader staffing reviews alone, although the interview data (Theme 2) suggest that staffing capacity remains the underlying constraint that these mechanical fixes would only partially address.
The relational layer, evidenced by the escalation confidence theme, corroborates Manias and Street’s (2001) and Endacott et al.’s (2007) accounts of hierarchical constraint on nurse-doctor communication, and offers a plausible explanation for the seniority effect observed in the audit: band 6 nurses, who reported greater willingness to escalate a concern a second time and described specific, learned communication strategies for doing so, achieved faster bundle delivery than band 5 nurses. This suggests that confidence to re-escalate, rather than clinical knowledge alone, may be a meaningful lever for closing the gap between novice and experienced nurses’ escalation times, a possibility supported by Royal College of Nursing (2020) guidance on the limits of knowledge-only sepsis training. Notably, the specific communication strategy described independently by several participants — reframing a second escalation around the objective NEWS2 score rather than around a renewed personal clinical opinion — offers a concrete, teachable technique that could plausibly be incorporated into training more directly than the broader, harder-to-operationalise concept of “confidence” alone.
Taken together, the findings suggest that interventions focused solely on NEWS2 refresher training are unlikely to close the observed gap on their own, because the audit and interview data locate the principal sources of delay in the practical staffing capacity available at night to act on a trigger once it has occurred, and in relational confidence to escalate a concern a second time, rather than in gaps in scoring knowledge, which none of the fifteen participants displayed any evidence of lacking. This has direct implications for how the trust, and similar organisations, might prioritise improvement resources: the findings point away from a generic training refresh and towards three more targeted interventions, set out as recommendations in Chapter Six.
This dissertation set out to investigate the barriers delaying early recognition and escalation of sepsis on acute adult wards, combining a retrospective audit of 120 patient records with semi-structured interviews with fifteen registered nurses. The audit found that only 38% of qualifying NEWS2 triggers resulted in Sepsis Six bundle delivery within the local sixty-minute standard, with significantly longer delays at night and among less senior escalating nurses, and with the widest single gap occurring not at initial recognition but between medical review and bundle completion out of hours. The interview findings identified diagnostic uncertainty, systemic constraint and escalation confidence as the three interlinked barriers underlying this pattern, extending existing human factors literature by showing how these barriers combine within a single clinical pathway rather than operating separately, and by locating precisely where within that pathway each barrier exerted its strongest effect.
The principal contribution of this dissertation is the demonstration that sepsis escalation delay on the study wards is concentrated less in the recognition step itself and more in the mechanics of bundle delivery under night-shift staffing pressure, and in nurses’ confidence to escalate a concern a second time once an initial review has not confirmed it. This refines the common assumption that delay is primarily a knowledge or training deficit, and points instead towards organisational and cultural levers that are, in principle, more directly addressable through ward-level protocol and workforce planning than a deficit in clinical knowledge would be. In this sense the study answers each of its three research questions in turn: escalation timing on the study wards varied significantly by shift and seniority (Research Question One); the barriers nurses themselves identified were diagnostic ambiguity, systemic constraint and relational confidence (Research Question Two); and these factors interacted, with systemic constraint concentrated specifically in the post-review, out-of-hours phase of the pathway and relational confidence concentrated specifically among less senior nurses (Research Question Three).
Four recommendations follow from these findings. First, ward-level escalation protocols should specify a named responsible clinician for initiating each element of the Sepsis Six once a trigger occurs, with a simple checklist attached to the patient’s observation chart, reducing the ambiguity that appeared to slow bundle delivery specifically at night when senior support was less readily available to coordinate the response informally. Second, night-shift staffing establishments and skill mix should be reviewed with sepsis-bundle delivery time as an explicit safety metric reported to the trust’s patient safety committee, rather than relying solely on general nurse-to-patient ratios that do not capture the specific bottleneck identified in this study. Third, simulation-based training that rehearses re-escalation in the face of an initial “normal” medical review should be embedded in trust induction and annual update training, building on Health Education England’s (2019) framework and directly targeting the escalation-confidence theme identified in this study; specifically, training should teach the depersonalised, score-led re-escalation technique that several experienced participants described using spontaneously, since this offers a concrete and transferable skill rather than an abstract exhortation to “be more confident.” Fourth, consideration should be given to a simple sepsis equipment pack, pre-stocked on each ward and checked each shift, to reduce the practical delay in obtaining intravenous access and antibiotic administration equipment identified as a specific contributor to night-shift delay in the interview data.
This study is limited by its single-trust setting and modest interview sample, and future research should test whether the day-night and seniority effects identified here replicate across organisations with different staffing models and skill mix. A prospective, multi-site study incorporating direct observation of escalation episodes, rather than retrospective documentation and retrospective recall, would also strengthen causal inference about where in the pathway delay originates and would allow the specific, granular timing of each Sepsis Six element to be captured more reliably than retrospective note review permits. A follow-up evaluation of the four recommendations above, ideally using the same audit methodology applied here as a baseline, would also allow the trust to determine whether they produce a measurable reduction in trigger-to-bundle time. Nonetheless, by triangulating objective timing data with nurses’ own accounts of practice, this dissertation offers ward managers and sepsis leads a more precise and actionable picture of where the sepsis pathway currently breaks down on acute adult wards, and why, than either strand of evidence could have provided alone.
Academy of Medical Royal Colleges (2021) Acute Care Toolkit 12: Acute Sepsis Recognition and Management. London: Academy of Medical Royal Colleges.
Andrews, T. and Waterman, H. (2005) ‘Packaging: a grounded theory of how to report physiological deterioration effectively’, Journal of Advanced Nursing, 52(5), pp.473-481.
Ball, J.E., Bruyneel, L., Aiken, L.H., Sermeus, W., Sloane, D.M., Rafferty, A.M., Lindqvist, R., Tishelman, C. and Griffiths, P. (2018) ‘Post-operative mortality, missed care and nurse staffing in nine countries: a cross-sectional study’, International Journal of Nursing Studies, 78, pp.10-15.
Benner, P. (1984) From Novice to Expert: Excellence and Power in Clinical Nursing Practice. Menlo Park: Addison-Wesley.
Braun, V. and Clarke, V. (2006) ‘Using thematic analysis in psychology’, Qualitative Research in Psychology, 3(2), pp.77-101.
Bucknall, T. (2003) ‘The clinical landscape of critical care: nurses’ decision-making’, Journal of Advanced Nursing, 43(3), pp.310-319.
Care Quality Commission (2018) Opening the Door to Change: NHS Safety Culture and the Need for Transformation. London: Care Quality Commission.
Cioffi, J. (2000) ‘Recognition of patients who require emergency assistance: a descriptive study’, Heart & Lung, 29(4), pp.262-268.
Creswell, J.W. and Plano Clark, V.L. (2018) Designing and Conducting Mixed Methods Research. 3rd edn. Thousand Oaks: SAGE.
Daniels, R. (2011) ‘Surviving the first hours in sepsis: getting the basics right’, Journal of Antimicrobial Chemotherapy, 66(Suppl 2), pp.ii11-ii23.
Endacott, R., Kidd, T., Chaboyer, W. and Edington, J. (2007) ‘Recognition and communication of patient deterioration in a rural hospital: a multi-methods study’, Australian Critical Care, 20(3), pp.100-105.
Griffiths, P., Recio-Saucedo, A., Dall’Ora, C., Briggs, J., Maruotti, A., Meredith, P., Smith, G.B. and Ball, J. (2018) ‘The association between nurse staffing and omissions in nursing care: a systematic review’, Journal of Advanced Nursing, 74(7), pp.1474-1487.
Health Education England (2019) Sepsis Education and Training Framework. London: Health Education England.
Jones, M., Rattray, J., Pattison, N. and Cormack, C. (2013) ‘Defining clinical deterioration’, Resuscitation, 84(8), pp.1029-1034.
Kitson, A.L., Athlin, A.M. and Conroy, T. (2014) ‘Anything but basic: nursing’s challenge in meeting patients’ fundamental care needs’, Journal of Nursing Scholarship, 46(5), pp.331-339.
Lincoln, Y.S. and Guba, E.G. (1985) Naturalistic Inquiry. Newbury Park: SAGE.
Manias, E. and Street, A. (2001) ‘Nurse-doctor interactions during critical care ward rounds’, Journal of Clinical Nursing, 10(4), pp.442-450.
Massey, D., Chaboyer, W. and Anderson, V. (2017) ‘What factors influence ward nurses’ recognition of and response to patient deterioration? An integrative review of the literature’, Nursing Open, 4(1), pp.6-23.
McGaughey, J., Alderdice, F., Fowler, R., Kapila, A., Mayhew, A. and Moutray, M. (2007) ‘Outreach and early warning systems for the prevention of intensive care admission and death of critically ill adult patients on general hospital wards’, Cochrane Database of Systematic Reviews, Issue 3, CD005529.
National Confidential Enquiry into Patient Outcome and Death (2015) Just Say Sepsis! A Review of the Process of Care Received by Patients with Sepsis. London: NCEPOD.
National Institute for Health and Care Excellence (2016) Sepsis: Recognition, Diagnosis and Early Management. NICE Guideline NG51. London: NICE.
Odell, M., Victor, C. and Oliver, D. (2009) ‘Nurses’ role in detecting deterioration in ward patients: systematic literature review’, Journal of Advanced Nursing, 65(10), pp.1992-2006.
Polit, D.F. and Beck, C.T. (2021) Nursing Research: Generating and Assessing Evidence for Nursing Practice. 11th edn. Philadelphia: Wolters Kluwer.
Reason, J. (2000) ‘Human error: models and management’, BMJ, 320(7237), pp.768-770.
Rhodes, A., Evans, L.E., Alhazzani, W., Levy, M.M., Antonelli, M., Ferrer, R., Kumar, A., Sevransky, J.E., Sprung, C.L., Nunnally, M.E., Rochwerg, B., Rubenfeld, G.D., Angus, D.C., Annane, D., Beale, R.J., Bellinghan, G.J., Bernard, G.R., Chiche, J-D., Coopersmith, C., De Backer, D.P., French, C.J., Fujishima, S., Gerlach, H., Hidalgo, J.L., Hollenberg, S.M., Jones, A.E., Karnad, D.R., Kleinpell, R.M., Koh, Y., Lisboa, T.C., Machado, F.R., Marini, J.J., Marshall, J.C., Mazuski, J.E., McIntyre, L.A., McLean, A.S., Mehta, S., Moreno, R.P., Myburgh, J., Navalesi, P., Nishida, O., Osborn, T.M., Perner, A., Plunkett, C.M., Ranieri, M., Schorr, C.A., Seckel, M.A., Seymour, C.W., Shieh, L., Shukri, K.A., Simpson, S.Q., Singer, M., Thompson, B.T., Townsend, S.R., Van der Poll, T., Vincent, J-L., Wiersinga, W.J., Zimmerman, J.L. and Dellinger, R.P. (2017) ‘Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2016’, Intensive Care Medicine, 43(3), pp.304-377.
Robson, W. and Daniels, R. (2013) ‘The Sepsis Six: helping patients survive sepsis’, British Journal of Nursing, 22(1), pp.16-19.
Royal College of Nursing (2020) Sepsis: Recognising and Responding — RCN Guidance for Nursing Staff. London: Royal College of Nursing.
Royal College of Physicians (2017) National Early Warning Score (NEWS) 2: Standardising the Assessment of Acute-Illness Severity in the NHS. London: Royal College of Physicians.
Singer, M., Deutschman, C.S., Seymour, C.W., Shankar-Hari, M., Annane, D., Bauer, M., Bellomo, R., Bernard, G.R., Chiche, J-D., Coopersmith, C.M., Hotchkiss, R.S., Levy, M.M., Marshall, J.C., Martin, G.S., Opal, S.M., Rubenfeld, G.D., Van der Poll, T., Vincent, J-L. and Angus, D.C. (2016) ‘The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)’, JAMA, 315(8), pp.801-810.
Vincent, C. (2010) Patient Safety. 2nd edn. Chichester: Wiley-Blackwell.
Need a Model Dissertation Written to Your Exact Brief?
Our 350+ UK-qualified writers deliver referenced model documents from £15 per 250 words, with free plagiarism and AI-detection reports.
You May Also Like
All work is written by human writers. 100% AI free, guaranteed.
100% money back guarantee if you find plagiarism in our work.
COMPANY DETAILS
