The weekend effect: now you see it, now you don’tBMJ 2016; 353 doi: https://doi.org/10.1136/bmj.i2750 (Published 16 May 2016) Cite this as: BMJ 2016;353:i2750
All rapid responses
After listening to a recent podcast on the weekend effect (1) and reading the accompanying editorial (2), we would like to highlight a different perspective on this emotive problem. Adverse clinical outcomes (ACO) such as unanticipated critical care admission, cardiac arrest and death can occur for many reasons. We agree with the initial article response (3) that care of seriously ill patients is suboptimal throughout the week, as demonstrated by significant evidence from national audits, confidential enquiries and research (4-7). Development of quality improvement based solutions to the problem of ACO should therefore be a clinical priority. To do this, clinicians must fully appreciate the system in which care of deteriorating patients is provided, identify and understand sources of variation within the system and pursue a balanced iterative approach to developing, piloting and implementing change.
System: Suboptimal care is difficult to define. However, the final common pathway to ACO is failure to recognise and or rescue deteriorating patients. Rapid response systems (RRS) have essentially evolved from various attempts to optimise quality and safety of care for increasingly complex patients in high capacity and high turnover healthcare settings. Despite widespread implementation of RRS-based care to improve the recognition and resuscitation of seriously ill patients, many hospitalised adults still receive suboptimal care (4-7). Rapid response systems (8) are conceptually relatively simple, with linear afferent and efferent limbs for detection and recognition of deteriorating patients. In reality however they are highly complex and consist of many heterogeneous structural, electronic, paper and human elements. Paper and electronic components serve an array of functions from providing access to information such as laboratory, radiology and other clinical information to alerting healthcare professionals of established clinical abnormality and guiding clinical care. Human components of RRS are equally diverse. Critical and difficult to measure observations are often carried out by non-clinical members of the team and response to alerts by clinical professionals can lead to horizontal zero value consultations between junior clinicians of different specialties. The potential for random error to affect patient outcomes in this system is significant. An increased drive within resuscitation and critical care communities to move away from a focus on treatment of cardiac arrest to preventative strategies (9) highlights the importance of a RRS-based approach to the prevention of ACO from suboptimal care.
Variation: Problems with human performance in the RRS chain of prevention can be rationalised as random variation in unplanned or emergency clinical encounters. Identifying and understanding modifiable antecedents in this chain that could improve RRS performance in preventing of ACO requires detailed retrospective case record review (RCRR). However methodological weaknesses, logistic and economic burden limit use of detailed RCRR (10). Efforts to develop systematic and reproducible RCRR methods focus on identification of avoidable mortality (11, 12) which will potentially limit broader exploration of and learning from suboptimal care. Unpublished results from our on-going quality improvement work at several organisations in recent years suggest that unplanned clinical review is a critical safety link in the RRS chain of ACO prevention. In particular, analysis of case notes repeatedly shows variation in emergency assessments contributes to suboptimal recognition and response to evolving shock during prior to ACO. Assessments with insufficient content, evidence of cognitive bias or error and inadequate contingency planning were recurrent themes in opportunities to improve clinical care. Occasionally, RRS will fail to recognise or respond to a deteriorating patient because the underlying clinical picture was secondary to unusual in pathology or atypical progression. Such special variation is often an important learning opportunity for the system to deal with deviation from expectation without explanation.
Solutions: Improving the quality of care, in general, requires the realisation of three critical concepts: teams must summarise and simplify what to do, measure and provide feedback on performance and build an expectation of performance into the culture of those involved (13). Organisational learning efforts to improve awareness of systemic variation are often obstructed by significant cognitive and cultural barriers (14). Measurement of RRS performance by RCRR is hampered by methodological and logistical challenges. Culture change is something that requires significant input over near geological timescales. Improving acute care skills of frontline individuals and the team with organisational learning using RRS based RCRR is still an important intervention and further work is required on this front. Digital health technology will also have a role in both acute care provision and organisational learning. However, current products do not yet have an open and adaptable clinical information model to support clinical noting tools that provide sufficient ease-of-use and cognitive support at the patient-healthcare professional level. Work in Northern Ireland is currently exploring the use of an accelerated clinical translational informatics approach for the development and testing of open source rapid response based electronic structured exam tool (RESET). A pilot study on high fidelity simulation of medium acuity scenarios demonstrates utility and usability of a paper based structured noting solution as a platform for further development and testing of a digital prototype.
Conclusion: Mortality is a very poor proxy for care quality but there is significant evidence of problems with quality of acute care that should not be ignored. Acute care outcomes will improve if there is more focus on quality improvement approaches that bring demonstrable benefits to patients. Bringing about change that improves quality of acute care is an enormous challenge. To do this, we must fully understand the function of RRS in acute care and underlying contributors to variation within the system. As clinicians it is our responsibility, not just politicians and policymakers (3), to focus more energy and effort on finding ways to provide optimal care, as standard, for every acutely ill patient and less on epidemiological debate.
1. The weekend effect - what research can tell us (https://soundcloud.com/bmjpodcasts/the-weekend-effect-what-can-research-... las accessed 24/05/2016). Last accessed on 25th May 2016
2. Mckee M. The weekend effect: Now you see it now you don’t BMJ 2016;353:i2750
3. Stewart, K. Variations in care quality occur across the whole week, not just at weekends. (http://www.bmj.com/content/353/bmj.i2750/rr Last accessed 27/05/2016).
4. Time to Intervene? A review of patients who underwent cardiopulmonary resuscitation as a result of an in-hospital cardiorespiratory arrest. A report by the National Confidential Enquiry into Patient Outcome and Death (2012)
5. NELA project team. First patient report of the National Emergency London 2015
6. Petersen JA, Mackel R, Antonsen K, Rasmussen LS Serious adverse events in a hospital using early warning score - what went wrong? Resuscitation. 2014; 85:1699-703.
7. Garry DA, McKechnie SR, Culliford DJ, Ezra M, Garry PS, Loveland RC, et al. A prospective multicentre observational study of adverse iatrogenic events and substandard care preceding intensive care unit admission (PREVENT). Anaesthesia 2014; 69: 137-142.
8. Devita MA, Bellomo R, Hillman K, Kellum J, Rotondi A, Teres D, et al. Findings of the first consensus conference on medical emergency teams. Crit.Care Med. 2006; 34: 2463-2478.
9. Smith GB, Welch J, DeVita MA, Hillman KM, Jones D. Education for cardiac arrest--Treatment or prevention? Resuscitation. 2015;92:59-62.
10. Shah, A et al. Towards optimising local reviews of severe incidents in maternity care: messages from a comparison of local and external reviews. BMJ Qual Saf 2016;0:1-8.
11. Hutchinson A, Coster JE, Cooper KL, et al. A structured judgement method to enhance mortality case note review: development and evaluation. BMJ Qual Saf 2013;22:1032–1040.
12. Hogan H, Healey F, Neale G, et al. Preventable deaths due to problems in care in English acute hospitals: a retrospective case record review study. BMJ Qual Saf 2012;21:737–45
13. Bosk CL, Dixon-Woods M, Goeschel CA, Pronovost PJ. Reality check for checklists. Lancet 2009;374:444-445
14.Syed, M. Black Box thinking: the surprising truth about success. 2015 Hodder & Stoughton.
Competing interests: No competing interests
The “weekend effect” is an oversimplification. Attempting to address it is a distraction; it’s answering the wrong question. This is because hospital mortality is not a good measure of quality of care and weekends are not the only times when quality is compromised by the way in which services are organised.
The questions ought to be, firstly are there variations in quality of care depending on the time of day or the day of the week that emergency patients are admitted? Secondly, are these variations clinically important? Thirdly, if they are important, what are the underlying causes and, fourthly, does the NHS want to make the changes and investment that are necessary to address the causes?
We think that we can answer the first and second questions, at least for some groups of patients, using data from national clinical audits. Stroke patients receive high quality care if admitted on most weekday mornings, but quality deteriorates on weekday afternoons, overnight and at weekends. Aspects of care related to patient flow and bed capacity are also worse on Monday mornings (1). These are clinically important deficiencies. Hip fracture patients are less likely to receive evidence based care the later in the day that they are admitted on weekdays, or on Fridays and Saturdays (National Hip Fracture Database, unpublished data). This is also likely to be clinically important. Most patients dying in English hospitals will not be able to see a palliative care specialist after 5pm on weekdays or at any time on weekends (2). We predict that similar patterns of diurnal and weekly variation in quality will emerge for other conditions.
The times of the day and week when there are delays to time-critical treatments will not be a surprise to any medical registrar, general physician or bed manager; very similar complex and overlapping patterns of temporal variation exist in bed availability, patient flow, access to investigations and access to specialist advice when demand for care exceeds the capacity of the system to supply it.
Hospital activity follows largely predictable patterns. Hospitals are busiest on Mondays and Tuesdays when emergency admissions, elective admissions and outpatients are all highest, numbers of discharges are low and the effects of a large excess of admissions over discharges at the weekend has produced a high bed occupancy (3). Pressure eases as the week progresses, as discharges begin to exceed admissions and when elective and outpatient activity drops significantly on Fridays. The weekly variation is compounded by a diurnal variation in emergency admissions with numbers low overnight and early in the day, building up after lunch time, peaking a couple of hours later and remaining high until late evening.
By contrast, most hospitals provide “routine” services (radiology, pharmacy, laboratories, outpatients, physiotherapy, most clerical and ancillary staff etc) following a very different pattern, Monday to Friday 9am to 5pm. Outside these hours services are reduced, restricted to emergencies or absent. Most “routine” services are closing down just at the times when the demand from emergency admissions is highest. On weekdays, medical teams, especially consultants, may have responsibility for inpatients, outpatients and elective activity as well as that day’s emergency admissions, so adding to delays. With rising demand many hospitals now operate near to full capacity so even minor delays for a small number of patients will have serious knock on effects on many others.
So, variations in care quality occur across the whole week, not just at weekends, and relate to the working practices of a whole range of staff, not just doctors. The solutions will almost certainly involve changes to routine weekday working, not just weekends working, for a whole range of staff, not just doctors. For example, extending routine working hours in support services like radiology and pharmacy into early evenings, providing additional resources on Mondays, Tuesdays and weekday afternoons to support the predictable surge in emergency admissions and shifting low risk elective work and outpatients from Mondays to Fridays all might help.
The real challenge for politicians and policymakers, once they recognise the complexity of these variations, will be to then recognise that none of the potential solutions are likely to occur without significant investment.
1. Bray et al. The Lancet. 10 May 2016. http://dx.doi.org/10.1016/S0140-6736(16)30443-3
2. End of life care audit; dying in hospitals. National report for England 2016. http://bit.ly/1qiS8ep
3. The weekly pulse. CHKS. June 2012.
Competing interests: No competing interests