Nursing care left undone in community settings: Results from a UK cross-sectional survey

1 INTRODUCTION

Set against the background of a global crisis within the health care workforce (Britnell, 2019), concerns have been expressed for some time about what constitutes the correct number and training level of nurses required to promote safe and effective patient care. While debates around these concerns are far from new (e.g., McKenna, 1995), high-quality empirical work has helped to develop an evidence base around such debates over the last two decades. Research in California (Aiken, Clarke, Sloane, Sochalski, & Silber, 2002; Aiken et al., 2010) has demonstrated a clear association between nurse-to-patient ratio and patient mortality and failure to rescue (deaths following complications). The European-led RN4CAST study replicated and expanded this work and confirmed that across all 12 European countries that participated, hospitals with better nurse staffing (a greater proportion of professional nurses) had better outcomes for patients and for nurses (Aiken et al., 2017).

While the main patient outcome measures in much of this work have been related to mortality rates, there has been interest shown in other indicators such as patient safety, adverse events and care quality in relation to discussions around nursing ratios and skill mix (the proportions of different levels of nurse, including the level of qualifications, expertise and experience, available for patient care during a nursing shift; Jacob, McKenna, & D'Amore, 2015). Linked to this interest in broader patient outcomes has been the impact of nurse staffing levels and skill mix on ‘missed care’ or ‘care left undone’ with the former being defined as ‘any aspect of required care that is omitted either in part or in whole or delayed’ (Kalisch & Williams, 2009). For example, research across 46 NHS hospitals in England (Ball, Murrells, Rafferty, Morrow, & Griffiths, 2014) highlighted that 86% of nurses reported care being left undone on their last shift and that the majority of this related to aspects of relational care (such as comforting and talking with patients) and health education. Furthermore, this work noted that the nurse-to-patient ratio was significantly associated with the incidence of ‘missed care’ and suggested that care not being delivered could be the mechanism through which low nurse staffing levels adversely affect quality and safety. This mechanism has been acknowledged and detailed in some of the conceptual models associated with missed care such as the Process of Care and Outcomes Model (Lucero, Lake, & Aiken, 2009). Building on this work, Ball et al. (2018) then demonstrated that missed nursing care was associated with increased odds of patients dying following common surgical procedures, further supporting the view that missed care mediates the relationship between registered nurse staffing levels and patient mortality. In addition, a systematic review of the impact of nursing care missed on patient outcomes showed decreased patient satisfaction, medication errors, urinary tract infections, falls, pressure ulcers and readmissions could all be associated with missed care—though the three studies reviewed to consider the link between missed care and mortality found no clear association (Recio-Saucedo et al., 2018).

It is important to note that many countries have roles located between health care assistants and registered nurses, such as licensed practical nurses and the new nursing associate role in England. These staff may also have an impact on the prevalence of missed care, though it is not within the scope of this paper to consider this. Similarly, the perspective of the patient on unmet needs (another way of considering missed care) and the relationship that this has to staffing levels is of key importance, but is also beyond the remit of this paper.

There are therefore many studies that provide robust evidence about required staffing levels for delivering safe and effective care and on the relationship of nurse staffing and skill mix to missed care. However, the majority of this empirical work has been done within acute, hospital care settings. Far less research has considered nurse staffing levels, missed care and service user outcomes in community and primary care settings, with one recent study stating that it is ‘the first to consider the concept of missed care in community based nursing’ (Phelan, McCarthy, & Adams, 2018a, 2018b, p. 626). Certainly, work on missed care in community nursing contexts lags far behind that in the acute sector. Indeed, a recent systematic review of missed care in the community context (Sworn & Booth, 2019) identified only five empirical studies (reported in eight publications), undertaken in Ireland (Phelan & McCarthy, 2016; Phelan et al., 2018a, 2018b), Australia (Blackman, Henderson, Willis, & Toffoli, 2015; Henderson, Willis, Blackman, Toffoli, & Verrall, 2016; Henderson, Willis, Xiao, & Blackman, 2017; Willis et al., 2016) and the United States (Nelson & Flynn, 2015). Collectively, this small body of research provides early insight into the prevalence of, types of and reasons for missed care in community nursing and care home settings. As with acute care sector findings, staffing levels and patient acuity stood out as major reasons given for the occurrence of missed care. However, none of these studies looked directly at the relationship between registered nurse staffing levels and the prevalence of missed care.

This paper therefore aims to contribute to the emerging evidence base on missed care in primary and community care settings by reporting findings from a UK-based survey administered by the Royal College of Nursing in 2017. Specifically, we focus on data relating to a single question on care left undone within the community care, primary care and care home setting. We consider prevalence of care left undone and its relationship to levels of registered nursing staff within these settings.

2 METHODS

This paper presents secondary data analysis from a cross-sectional Staffing Survey conducted in the UK by the Royal College of Nursing (RCN, 2017). Data were collected for this survey between 14 May and 30 May 2017.

3 FINDINGS

4 DISCUSSION

In line with previous studies from Ireland (Phelan & McCarthy, 2016) and Australia (Henderson et al., 2017), this study notes a prevalence of nursing care left undone reported by registered nurses within the UK community care and care home sector that is sufficiently high to warrant concern. The reported rates of community nursing care left undone in the current study appear lower than those reported in the community context in Ireland (Phelan & McCarthy, 2016) where the rates for many aspects of care ran at over 50%. However, the current study relies on a single question measure, whereas previous community studies have used the MISSCARE survey tool (Kalisch & Williams, 2009), or amended versions of this, which report on missed care in relation to diverse specific aspects of care (e.g., assessment, promotion of skin integrity, health education, documentation). It is possible that the single generic question asked in the current study does not help with the recall of care left undone which measures such as the MISSCARE tool, using specific cases, might elicit. In addition, a review by Jones, Hamilton, and Murry (2015) has shown that reports of unfinished care increase with longer periods of recall. Given the current study asked about the last shift worked, this could also partly account for the comparatively lower rates of prevalence. We believe these factors, alongside possible response bias (discussed below in ‘Limitations’), may have led to an under-reporting of the prevalence of care left undone in the current study.

There was a relatively low prevalence of care left undone within the UK primary care nursing context. We could find no other UK or international studies to compare this prevalence to. It is possible that patient appointment systems, particularly clearly defined appointment times, might partly account for this lower rate of care left undone compared to other community care nursing contexts. The greater focus on specific tasks (rather than relational care) in primary care nursing, and the incorporation of time for patient education into the allocated length of time for appointments could also be important factors here, given that relational care and health education are the most frequently reported types of care left undone (Ball et al., 2014).

Importantly, the current study looked specifically at the relationship between the proportion of registered nurses (actual number/expected number) on duty and the prevalence of care left undone. We could identify no previous UK or international work that has looked directly at this relationship in the community care, primary care or care home context. Our study demonstrates a significant relationship with lower proportions of registered nurses than anticipated leading to higher prevalence of care left undone across these settings. As noted earlier, work in the acute sector has suggested that nursing care left undone might be the mechanism through which low nurse staffing levels adversely affect the quality and safety of patients (Ball et al., 2014; Kalisch, Tschannen, & Lee, 2011). While much mandatory and legislative work on safe staffing numbers in the four countries of the UK has focused on acute care settings, we would suggest that ensuring adequate registered nurse numbers in community settings also requires attention. This is a concern if high-quality, safe and effective care is to be maintained in community nursing. It is particularly important in a policy context, such as that in the UK, where there is a shift towards more health and social care, and therefore concomitant nursing care, in community settings. We also now know that frailty and multi-morbidity are prevalent in the community across the life course, especially for those older adults living at home (Hanlon et al., 2018). This brings with it significant population need, dependent upon the provision of skilled nursing care, at a time when health care providers are also dealing with the challenges of severe shortages in the community nursing workforce (Charles, Ham, Baird, Alderwick, & Bennett, 2018). Little research evidence is currently available concerning what constitutes appropriate levels of registered nurses in various community contexts, and this makes it difficult to proceed with mandating or legislating for safe staffing within the community setting. Jackson, Wright, and Martin (2016) highlight both the lack of research considering safe staffing in community nursing contexts and the difficulty in establishing generalizable recommendations on numbers alongside the prevailing complexity across health care setting and environments. Further, research in this area would help community nurse managers rationalize, articulate and justify their staffing requirements.

Also of note in the current study is that the care home sector reported far more shifts with the expected full complement of registered nurses than the community and primary care sector. We suggest this may be related to mandatory staffing requirements for at least the vast majority of the sector. It is mandatory for one RN to be on duty in a nursing home, thus generating the high rates of full complement seen in the current study. Importantly, despite this high planned level of registered nurses, the prevalence of care left undone in the care home sector remained high, suggesting that current planned levels of registered nurses are often not sufficient in providing high-quality, safe and effective care in the care home setting.

Finally, we noted a lower prevalence of care left undone on night shifts compared with day shifts in the care home sector. However, in line with Willis et al.'s (2016) study on missed care in residential care, this difference was not statistically significant. One possible explanation for this difference would be the reduced number of patient tasks required on a care home night shift leaving more time for the often-missed relational care.

5 LIMITATIONS

As with almost all studies on missed care and care left undone (Jones et al., 2015), our study relies on nurses self-reporting. Such reporting is open to subjective judgement, not least in terms of what level and type of care are required and therefore what an individual nurse may consider ‘care left undone’. The use of a single, generic question to capture care left undone in this study could further increase the level of subjectivity in data responses, by being more prone to subjective judgement than using a validated tool incorporating a battery of questions related to specific care tasks such as the MISSCARE tool. This single construct measure could also miss some of the important nuances related to differing activities of care. Furthermore, it does not tell us which aspects of care are left undone, by whom and within what context within the community nursing environment. In addition, for nurses working relatively autonomously, such as district nurses, assessing ‘care left undone’ may be less difficult than would be the case for those respondents working in larger teams, such as in care homes. This is because those working in larger teams may be unaware of the nursing actions taken by others.

We did not use regression modelling to assess whether other variables (such as patient volume and acuity, availability of other ancillary and professional health staff, issues of geography such as rural settings and time required to travel) influenced the relationship between the proportion of registered nurses and the level of care left undone. Future research in the community nurse context could consider such modelling, though accounting for variables in the community nursing context is perhaps even more complex than in the acute care setting which is rather more bounded.

While the sample size in this study offers a strong basis for generalizability, no data were examined in relation to whether the sample demographics (gender, ethnicity, etc.), years of experience post-qualification or the numbers of those with specialist qualifications were representative of the wider community nursing workforce. Caution should therefore be exercised in this regard.

Finally, the team has identified issues with data quality. A significant proportion of respondents chose to state that they neither ‘agreed nor disagreed’ with the statement ‘Due to the lack of time, I had to leave necessary care undone’. This is likely indicative of a response bias (Furnham, 1986) with nurses being reluctant to self-report this issue for fear it reflects badly on them. There is therefore the potential that ‘care left undone’ is under-reported in this paper. Due to recall bias, we also cannot verify the accuracy of calculations of staffing ratios as they were based on participants’ recall data. We know from previous research (Althubaiti, 2016) that recall bias is a systematic error that occurs when participants do not remember previous events or experiences accurately, and this could be the case here.

6 CONCLUSION

Community nursing staff reported a high prevalence of care left undone on their last shift, whereas primary care nurses reported a lower, but still clinically important, prevalence. Less than 40% of community nursing shifts had the planned number of registered nurses present, and the prevalence of care left undone increased as the proportion of registered nurses dropped below planned numbers.

In the care home sector, despite over 80% of shifts reporting the planned number of registered nurses, there remained a high prevalence of care left undone. This was the case across both day and night shifts, raising questions about whether the planned number of registered nurses is sufficient for providing safe and effective care in the care home sector.

For community nurse managers to make effective, evidence-based safe staffing decisions, more research is required to examine which aspects of care are left undone, and for whom, in various community contexts. Caseload management tools for calculating staffing levels in community nursing could be useful here (Jackson et al., 2016) and could be implemented, while the levels of care left undone were simultaneously measured to gauge their impact. In addition, such research should be more nuanced, accounting for a wider variety of variables such as patient acuity, skill mix and the geography of locations, among others. Policy consideration should be directed at current staffing levels in the community context and whether these are adequate even when shifts have the fully planned numbers of staff. This need is especially critical in the care home sector where even shifts with a full planned complement of registered nurses show a concerning level of care left undone.

ETHICAL APPROVAL

Once in place, ethical approval was then obtained from [University of Sheffield and Royal College of Nursing] prior to the commencement of secondary data analysis.