Between January 2005 and March 2009, hundreds of patients died due to poor care at Stafford Hospitala. The public enquiry into what had happened at the hospital was launched in November 2010. I was living in nearby Birmingham at the time, and stories of the horrors that had occurred at the hospital were frequently broadcast on the local news. Surgical equipment such as clips and clamps which should have been single-use were being used multiple times; patients on wards were left without water and some became so desperate they drank from vases containing flowers. Food was taken to vulnerable patients but not fed to them, and patients could be heard crying out from their beds due to lack of pain relief.
How could this happen?
As an NHS employee in a neighbouring healthcare trust, I found it hard to fathom how this could have happened. The Francis Report, the final report from the inquiry into Stafford Hospital, was published in February 2013. It suggested the causes were complex. There were significant staff shortages caused by the cutting of jobs in an attempt to meet a £10m budget deficit. When concerns were raised by staff or visitors through relevant forums, these were ignored. Importantly, staff became disengaged at all levels, and this in itself became a hard-to-measure but important cause of suffering.
In one tv news report a former patient, Nicola Monti, describes how she returned to hospital with a bowel condition following the birth of her second baby. She became sicker in the squalid hospital conditions and contracted MRSA. As she speaks about how the nurse communicated this information to her, by throwing the test results down on the bed, tears spring to her eyes. The results meant continued isolation from her two children and intense feelings of loneliness, but it is clear that simply the callous manner in which the nurse told her was in itself a significant cause of pain.
These patterns play out in more muted ways
Staff disengagement had a key role in contributing to the terrible events that occurred in Stafford Hospital. Thankfully the Staffordshire hospital failings are an outlier, but in my experience these patterns can play out in more muted ways. In a service I worked in where bullying was rife and staff were miserable, patient care did not receive the enthusiasm and energy it should have. On the other hand, in services I’ve worked in where staff felt they were part of a functioning and effective team, the mood was more positive and patients received the best that service had to offer.
Time for research
While I had seen these patterns in services I worked in, I wanted to find evidence to back up my anecdote. I arrived in my current job as a Lecturer at the University of Leeds and Bradford Institute for Health Research in October 2013 with this as my focus.
Staff: The greatest asset of the NHS
The first thing I discovered when I began my reading in this area is that the workforce is the greatest asset of the NHS. The NHS is the world’s 5th biggest employer, employing 1.7m staff altogether, including 141,000 doctors and 329,000 nurses, midwives and health visitors. The annual bill for employing clinical staff is £43 billion, which is around half of NHS Trusts’ budgets (National Audit Office, 2016).
The second thing I discovered is the extent to which turnover and sickness absence –clear outcomes of elevated burnout and disengagement – are causing major problems for the NHS. Between 2011 and 2015 the number of staff leaving their jobs each year due to poor work-life balance doubled. Each time a staff member leaves, costs are incurred due to 1) employing temporary staff to fill their role until a replacement is found, 2) advertising, interviewing and recruiting a replacement staff member and 3) training the new staff member. I was unable to find cost estimates for the NHS, but the price of replacing a nurse has been estimated to be $20, 561 in the USA, $26, 652 in Canada, $23, 711 in New Zealand and $48, 790 in Australia (Duffield et al., 2014).
NHS staff are off work due to sickness for twice the number of days as those working in the private sector. What is particularly concerning is that a significant proportion of these days are due to stress and anxiety. In doctors in acute trusts, 17% of all sickness absence days are in this category; in doctors in mental health trusts it is 26% (see my previous blog for more information on this). These absences can lead to delays and interruptions in care and divert NHS resources to paying for expensive agency cover. This problem is increasing: in 2010, £2.2 billion was spent on agency staff, but by 2015 the figure had risen to £3.3 billion. This leads to a vicious cycle: money that could be spent on enhancing services and supporting existing staff is diverted to agency fees. In turn, services suffer, staff are not as well supported and more leave and become unwell due to stress. This further increases the agency bill, and so on.
Impact on patient care
A large body of evidence shows that higher staff burnout is linked with poorer quality patient care. For example, in a study of emergency doctors published in 2015, those who were classed as suffering from burnout said they more frequently ordered extra tests, failed to treat pain in a timely manner, discharged patients to make the department manageable, did not discuss treatment options or answer patients’ questions, did not communicate important information in handovers and did not discuss treatment plans with appropriate staff (Lu et al., 2015).
This poorer care is reflected in lower patient satisfaction. In an analysis of the NHS staff surveys of 2009, 2010 and 2010 together with trust-level measures, patient satisfaction was higher when fewer staff worked extra hours, more staff felt valued by their colleagues, staff reported lower work pressure and higher levels of engagement (Powell et al., 2014).
The all-important question of patient safety
While there was a lot of research studies on the link between staff engagement, burnout and quality of care indicators, there was much less research on the link with patient safety outcomes. Exploring this was of key importance to me: in Stafford Hospital, patients weren’t just reporting low satisfaction, hundreds had died. Understanding and evidencing the link between staff engagement/burnout and patient safety was paramount to preventing this from happening again. When I told clinical colleagues of my plans, they told me not to bother: “It’s so obvious, of course staff wellbeing and patient safety are linked. You don’t need to research it”. I became slightly concerned I could be wasting my time, but decided to plough on anyway in the hope that like me, someone else might feel the need to evidence the obvious.
First step: A review of the literature
First we undertook the first systematic review of studies linking healthcare staff burnout and wellbeing with patient safety. Forty-six relevant studies were identified. The review found that 89% of the studies which measured wellbeing reported a significant association between wellbeing and patient safety, and 83% that measured burnout reported a significant association between burnout and patient safety. The review was led by Dr Louise Hall and published in PLOS One in 2016, and has since received over 100 citations according to Google Scholar metrics, making me glad that I didn’t heed the initial discouragement I received!
Which matters most – general mental wellbeing or burnout in particular?
The concept of more general mental wellbeing (e.g., depression) is distinct from burnout, which focuses specifically on negative work experiences and attitudes. I became interested in which of these areas may be most important to focus on in relation to patient safety. To address this we conducted a survey study in 323 nurses across 3 trusts. Participants completed measures of burnout, depression and perceptions of patient safety. Statistically speaking, we found that while both burnout and depression had direct individual associations with patient safety perceptions, the association between depression and patient safety was fully mediated by burnout. In other words, we found that depression IS linked with patient safety, but it is the portion of depression that overlaps with burnout that creates this link (See Figure 1). In short, burnout seems to be the more important concept to focus on. We concluded that interventions looking to improve patient safety may benefit from focusing on burnout in particular, for example using strategies to enhance staff engagement in work rather rolling out one-to-one therapy for depression. The study was published in Journal of Advanced Nursing.
Figure 1. It is the portion of depression that overlaps with burnout that is linked with poorer patient safety perceptions.
What are the mechanisms?
A final question we wanted to answer was into the mechanisms of this association: if a staff member is feeling burnt-out, how does this turn into poorer patient care? We addressed this in a qualitative study led by Dr Louise Hall, which was published in the Journal of Patient Safety. Five focus groups with 25 GPs were undertaken. GPs thought poor wellbeing and burnout affects care quality by reducing doctors’ ability to empathize, increasing their negative attitudes to patients and by increasing the likelihood that the GP will simply refer the patient on rather than manage them in clinic, even if the referral isn’t really appropriate. GPs thought that burnout impacts patient safety by reducing their mental functioning and decision making abilities and increasing their fatigue. As one GP said, when burnt-out, they may be less likely to ‘connect dots’ across time and realise that current symptoms may be indicative of a bigger picture indicating a more concerning problem; because of this they could risk missing an important diagnosis.
Interested in how healthcare staff burnout can be tackled? Please see this blog post.
This blog was written in conjunction with a talk given at the Practitioner Wellbeing Conference in Manchester on 14th June 2018. To download a copy of the slides, click here.
a The Francis report of February 2013 concluded that it would be unsafe to infer from these statistics that there was any particular number of avoidable or unnecessary deaths at the trust.