2-Hour nursing shifts in critical care: A service evaluation

2-Hour nursing shifts in critical care: A service evaluation

! The Intensive Care Society 2018 Reprints and permissions: sagepub.co.uk/


DOI: 10.1177/1751143717748094 journals.sagepub.com/home/jics

Ceri Battle and Paul Temblett                                                                                                    


The aim of this single-centre study was to investigate the impact of the introduction of 12-h critical care nursing shifts on healthcare provider and patient care outcomes. A single-centre, prospective service evaluation was completed over a two-year period, comparing the 8-h and 12-h shifts. Outcomes included number of clinical incidents, levels of burn-out, sick rates, personal injuries and training. There were no significant differences between the clinical incidents, sickness rates, personal injuries and staff training between the two data collection periods. The results of the burn-out analysis demonstrate that emotional exhaustion and depersonalisation improved, from the 8-h to 12-h shifts (both p < 0.05). In conclusion, the results of this service evaluation have demonstrated that 12-h nursing shifts can be introduced safely into the critical care environment, without any detriment to patient or healthcare provider outcomes.


Critical care, extended shift patterns, nursing staff, patient outcomes, healthcare provider outcomes


Extended nursing shifts of 12-h or more, have become increasingly popular in the hospital setting. Limited research has been conducted investigating the impact of such extended shifts on either patient care or nursing staff healthcare outcomes.1 Although a number of studies have investigated various outcomes in 12-h shift patterns, these studies are generally of variable quality, as a result of the complex nature of the work

patterns involved.2-6 It is difficult to control extraneous variables, including shift sequence, overtime and break patterns. Age, grade and experience of the nurse may also influence study findings.

It has been reported that 12-h shifts lead to poor performance due to physiological strain, fatigue, burnout and job dissatisfaction, which consequently negatively impacts patient care and safety.6,7 A number of studies of US hospital nurses demonstrated that the risks of making an error are significantly increased when work shifts are longer than 12h, when nurses work overtime, or when they worked more than 40h per week.8,9 In a number of similar studies, nurses working shifts of 12-h or more and those working overtime, reported lower quality and safety, increased risk of errors and decreased nurses’ vigilance.5,10

In contrast, the 12-h shift has been reported to be favourable by nursing staff as they allow greater flexibility for personal responsibilities outside of the workplace.1-3 In a number of qualitative studies, 12-h shift patterns were seen as positive, contributed to staff satisfaction and also both a good recruitment and retention strategy.1-4,11 In a single-centre Australian study, results highlighted a positive impact on physical and psychological well-being and increased work satisfaction, for the nursing staff working the 12-h shift pattern. Increased continuity of patient care was identified as a positive outcome of

the 12-h shift.3

It is evident that there are conflicting conclusions reported by previous studies investigating differences between long and short day shift patterns. In the current climate of austerity and change within the National Health Service, it has been suggested that there is a need to investigate more efficient systems of work and shift patterns for NHS nursing staff. The aim of this single-centre study was to investigate and report any changes in healthcare provider and patient care outcomes over two data collection periods; one

Ed Major Critical Care Unit, Morriston Hospital, Swansea, UK

Corresponding author:

Ceri Battle, Physiotherapy Dept, Morriston Hospital, Swansea, SA6 6NL, UK.

Email: c..e@wales.nhs.uk

Battle and Temblett                                                                                                                                                                           215

year using the traditional 8-h shift patterns and the other year introducing the 12-h shift pattern.

Methods Setting

A single-centre, prospective service evaluation was completed over a two-year period in a large tertiary intensive care unit (ICU) in Wales. The ICU is a mixed medical and surgical unit, which admits over 1200 patients per year. The number of nursing staff working on the ICU during the two-year study period varied between 188 and 214, to cover a 28-bedded unit. This number does not include agency/locum nursing staff. In terms of skill mix, 6% were band 7 nurses, 11% were band 6 and 83% were band 5. (A newly qualified nurse would start at band 5 and their main role would be to provide patient care at the bedside. Band 6 nurses are more senior and would have additional responsibilities such as leading the team within the unit. Band 7 nurses are defined as advanced senior nurses with more managerial responsibilities away from the bedside.) Of these nursing staff, 65% worked full time and 35% part time. This ratio of full to part time nurses did not change over the two-year period.

During the first year of data collection (March 2015 to February 2016), nursing staff worked the traditional 8-h shift. In March 2016, the 12-h shift was introduced for those nurses who opted to work it and a second year of data collection was undertaken (long days). Agenda for Change pay bands 5 to 7 nurses were included in the study, regardless of age or level of experience. Locum nurses, practice development nurses and nursing management staff (any non-clinical role) were excluded from the study.

Definition of variables

Patient care outcomes were assessed using the reported number of clinical incidents and healthcare provider outcomes using levels of burn-out, sick rates, personal injuries and staff training. Clinical incidents were defined as an event that could have, or did result in, unnecessary damage, loss or harm such as physical or mental injury to a patient, staff, visitors or members of the public. This includes both clinical and nonclinical incidents. All clinical incidents in the Health Board are reported using a DATIX system, which remained constant throughout the two-year period.

Burn-out was assessed using the Maslach Burn-out Inventory (MBI; Human Services Survey)12 during the final two months of each data collection period. Recognised for over a decade as the leading measure of burn-out, this inventory incorporates the extensive research that has taken place in the 15 years since its initial publication.12 The survey addressed three separate components of burn-out; emotional exhaustion (nine items), depersonalisation (five items) and personal accomplishment (eight items). High scores of emotional exhaustion and depersonalisation, combined with low scores of personal accomplishment result in high scores of burn-out. Each item is scored from 0 (never) to 6 (every day). A number of studies support reliability, such as the three-factor structure and internal reliability.13,14 Iwanicki and Schwab13 reported Cronbach alpha ratings of 0.90 for emotional exhaustion, 0.76 depersonalization, and 0.76 for personal accomplishment. Furthermore, previous studies in the critical care setting have pointed out that the MBI was reliable for measuring

burnout in critical care staff.15,16 Surveys were distributed to nursing staff by hand, during the two-month data collection periods, in order to maximise the number of staff captured.

Sickness rates were reported as an in-month rate, that is a percentage of the workforce not present in work each month due to sickness. Personal injuries were defined as the number of physical injuries reported each month, that were sustained during the working hours and attributed to the working environment. All sickness data were taken from the official system used within the health board, which would include information taken from ‘return to work interviews, self-certification documents and GP certificates’. Training was defined as the number of hours of training delivered each month and included formal teaching, attendance at staff meetings or the weekly multi-disciplinary team meeting and in-house mandatory training. This excluded any mandatory training completed by the nursing staff outside of their working hours. Training data were provided by the practice development nurses responsible for training on the unit.

Statistical analysis

Data were recorded monthly for each variable throughout the two-year data collection periods. During the study period, all data were anonymised and stored on a hospital encrypted computer. Descriptive analysis was completed using numbers and percentages for categorical variables and median and interquartile range for continuous variables. Comparisons between the two data collection periods were completed using the Fisher’s exact test (categorical variables) and the Student’s t-test (continuous variables). Statistical analyses were performed using SPSS Version 22 (Chicago). Statistical significance was set at p<0.05.


The data collection for this study was considered to be a service evaluation as confirmed by the Health Board’s Joint Scientific Review Committee and therefore ethical approval was not required.

Journal of the Intensive Care Society 19(3)


Approximately 150 nurses participated in the study throughout the two-year period. As expected, this number varied slightly due to staff leaving and new starters. During the second year of data collection (12-h shift pattern), the percentage of nurses who worked the 12-h shift pattern varied between 75% and 81%. The remainder of the nursing staff opted to work the 8-h shift pattern, primarily due to childcare issues. Of the nurses working the 12-h shift pattern, 75% worked full time and 25% worked part time.

Table 1 demonstrates the results of the analysis comparing each outcome between 8-h and 12-h data collection periods.

The results of the analysis demonstrated no significant difference between the clinical incidents, sickness rates, personal injuries and staff training between the two data collection periods. The results demonstrated relate to the entire nursing workforce.

Response rates for the burn-out surveys were low for both data collection periods at 40% in the 8-h shift period and 42% in the 12-h shift period. The results of the burn-out analysis demonstrate that emotional exhaustion fell from a high to moderate level, from the 8-h to 12-h data collection periods (p<0.05). Depersonalisation fell from a moderate level in the 8-h data collection period to a low level in the 12-h data collection period (p<0.05). The normative data suggest that a low level of depersonalisation would score 46, a moderate score 7 to 12 and a high level would score 513. The low levels of depersonalisation reported in this study suggest that nursing staff did not display an unfeeling or impersonal response to their patients. Personal accomplishment remained at a moderate level between the two periods (p>0.05). Overall, using the three components of the inventory, nurses reported an average degree of burn-out across the two data collection periods, when compared to normative data.


The results of this service evaluation in a large ICU in Wales have demonstrated that there were no significant differences between 8 and 12-h shifts for nursing staff, other than an improvement in two components of burn-out (from 8 to 12-h data collection periods). A number of outcome measures were investigated over the two-year period, in an attempt to analyse both patient and healthcare provider outcomes. Table 1 demonstrates that number of clinical incidents, sickness, personal injuries and staff training did not differ between the two data collection periods in this study. Limited previous research exists, examining outcomes in the critical care field and most existing studies are primarily qualitative in design.1-6

The results of this study reported no difference in sickness rates between nursing staff working a 12-h shift pattern, in comparison to those working an 8-h shift pattern. Limited previous research has examined sickness rates between shift patterns as an outcome in critical care nursing. It could be suggested that sickness rates do not differ between the two groups, as although the 12-h shift has been previously demon-

strated to lead to higher levels of fatigue,5,10 it has also been shown to lead to improved family life and qual-

ity time off.1,2 Personal injuries have been investigated in a previous study in which 12-h shift patterns were not shown to lead to an increase in musculoskeletal complaints.17 Similarly, the results of this study found no difference in reported personal injuries. The number of reported clinical incidents did not differ between the two data collection periods in this study. Similar conclusions have been reported in previous research.18

The levels of burn-out reported by the nursing staff in this study remained at a moderate level throughout the two years, although an improvement in emotional exhaustion and depersonalisation was evident in the 12-h shift pattern period. Burn-out is a well-recognised problem in critical care staff, regardless of profession

Table 1. Results of analysis comparing all outcomes between 8-h and 12-h shifts.

8-h Shifts mean


12-h Shifts mean


p (95% CI)

Clinical incidents

27.1 (8.6/2.4)

27.3 (9.3/2.7)

0.946 (7.3 to 7.8)


N ¼ 60

N ¼ 63

Emotional exhaustion

28.4 (13.2/1.5)

17.3 (1.6/1.7)

0.001 (6.2 to16.0)


7.6 (6.9/0.8)

4.2 (4.3/0.7)

0.002 (1.3 to 5.5)

Personal accomplishment

34.1 (8.5/0.9)

37.3 (7.1/1.1)

0.541 (6.3 to 0.5)


9.0 (2.1/0.6)

9.8 (2.0/0.6)

0.290 (2.6 to 0.9)

Personal injuries

1.9 (1.4/0.4)

1.2 (1.4/0.4)

0.191 (0.4 to 1.9)

Staff training (h/month)

429.3 (149.6/43.2)

380.3 (175.8/50.8)

0.470 (89 to 187.2)

Note. Results gained from entire nursing workforce during both data collection periods.

   Battle and Temblett                                                                                                                                                                            217

and shift patterns. In a recent review, a number of factors were reported to be associated with burn-out, including age, sex, personality traits, work experience in an ICU, workload and shift work and end-of-life decision-making.19 In a similar study, additional organisational factors, such as ability to choose days off were associated with burn-out.20 It is possible that workingthe 12-hshiftpattern inthis study, allowed the nursing staff greater choice regarding days off work and more days away from work between shifts, thus avoiding end-of-life decision-making, which may have contributed to lowerburn-out levels. Researchremains inconclusive as to whether shift patterns contributes to improvement in levels of burn-out, and further studies are needed.

A number of previous studies recommend that the 12-h shift needs to be fundamentally well-managed, in order to ensure the safe running of the critical care department.1,2,4 Although there was no statistical difference in the hours of training delivered between 8 and 12-h data collection periods, actual organisation and delivery of training proved complicated. For example, shorter teaching sessions were introduced, but needed to be provided on a frequent basis. With the introduction of the 12-h shift pattern, alternative models for rostering need consideration, as they could facilitate staff development. For example, in order to make up full-time hours using the 12-h shift pattern, staff may have an additional 8-h shift per month that could be used for training. Evidently, new systems of delivery of training are needed, in order to facilitate optimal teaching and communication between senior staff and the nursing cohort.

There are a number of limitations to the design of this study, the most consequential being the inability to control the numerous confounders that could influence the study results. During the two data collection periods, it was not possible to control whether staff were working overtime, in particular, 12-h shifts during the 8-h data collection period. Similarly, a small percentage of nursing staff opted not to work the 12-h shift, during the 12-h shift data collection period. Due to the fact that both data collection periods were affected and data were collected over 12 months, it was assumed that the effect of this confounder would have been reduced. Similarly, a higher proportion of nursing staff working the 12-h shifts were full time, compared to the unit as a whole. Results should, however, be interpreted with caution.

There was also a low response rate for the burn-out questions which may have introduced reporting bias into the results. We are unable to comment as to whether the response rate to the surveys differed between full and part time staff, due to the anonymity maintained in the surveys. The lack of statistical difference may have been due to the small sample size; therefore, future adequately powered studies are recommended. Finally, the desire of some of the nursing staff to ensure results were positive and that the 12-h shift pattern would remain in place at the end of the study, could have led to bias in the completion of the burn-out surveys.

In conclusion, the results of this study have demonstrated no significant differences in any of the outcomes analysed (other than improvements in two components of burn-out) when comparing 8 and 12-h shifts in critical care nursing staff, working on a large ICU in Wales. In order for 12-h shifts to work in practice, there needs to be a willingness of all nursing staff involved to consider new methods of working, especially with regard to the delivery of training and communication methods.

Authors’ note

Ceri Battle has access to all the original data and takes responsibility for the integrity of the data.


We would like to thank Ceri Matthews, Miranda Williams, Tracy Owen, Ceri Lynch, Bethan James, Karen Devine, Diane Murphy, Dawn Davies, Bethan Lavercombe, Cheryl Jones and Charina Maldia for their assistance in completing this work. This work was completed at the Ed Major Critical Care Unit, Morriston Hospital, ABMU Health Board, Swansea, UK.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.


The author(s) received no financial support for the research, authorship, and/or publication of this article.


Try to use less than 50 words for each question.

1.     Why was the study done, and what research questions were the authors addressing?

2.     Does the data constitute a sample or an entire population? If a sample has been taken, identify both the sample and the population; otherwise, identify the population.


3.     In this article both qualitative and quantitative were observed. List at least one for each variable type, categorical and quantitative. For example: Weight in pounds (quantitative variable), Hair Color (qualitative variable).

4.     Was it an experimental or observational study? If a sample has been observed, is it likely to be representative of the population of interest? 

5.     Was it an inferential study? Explaining why do you think it is an inferential study?

6.     What statistical analysis methods(s) was(were) used?

7.     What is(are) the major conclusion(s) in this article?

8.     Was “p-value” used in the article? What was the conclusion drawn based on the p-value? (If many p-values were reported, please choose one p-value and explain the outcome based on this p-value.)

9.     Is there any misleading information presented in the paper? If there is, explain why it is misleading and what would you suggest to correct it.

10. Use your judgment to critically evaluate the articles. Name two possible biases that may occur in this research.

Calculate your order
Pages (275 words)
Standard price: $0.00
Open chat
Hello 👋
Thank you for choosing our assignment help service!
How can I help you?