Negativity

In life we pay more attention to the negative things; the negative headlines in a newspaper, the negative reviews of an article we have written, the poor feedback about a talk we may have delivered.  We all remember the negative comments in an appraisal rather than the positive comments.  Even if there is a balance of positive and negative findings we put more weight on the negative ones and feel that the priority is to address the negative rather than the positive.  Even if we get one negative comment and nine positive comments, we feel we have to change ourselves to meet that one comment rather than accept that most people liked it so there is no need to change.  This in turn may mean that we change it so that only one person likes it and nine people don’t.  This attitude has ‘infected’ the world of safety.  Consequently negativity is considered more impactful that positivity.  We constantly feel we have to change based on the negative things that happen rather than the positive.

This is the same in life as well as it is in safety.  Steven Pinker (2016) says that many people face the morning news with trepidation and dread.  We are continually told that things are getting worse that modern life is much more negative than our past.  We pay attention to the stories of negativity and it leaves people longing for different times when it felt safer, kinder and more equitable.  If we translate this to healthcare, what people do is pay attention to the bad news, pay attention to the stories which point to the feeling that the health service is getting worse.  The media is particular effective at pointing out where healthcare fails rather than where healthcare succeeds.  And within healthcare safety we focus almost exclusively on the negative; what are the problems we need to address, what are the things we doing that are going wrong, data collection systems, the number of incidents, accidents, never events, serious incidents, deaths, complaints and claims.  The current methods used to study safety in healthcare are fixated on where we have failed in order to figure out how we can prevent those failures and improve the way we care for patients.

There is a desire to quantify the level of failure.  Asking if healthcare is getting worse or better. Or whether it is less safe or safer. The following describes what the researchers and safety experts over the years have tried to do so far to find out the scale of the problem with the ultimate aim of being able to tell if we are getting safer or not.

Studies of failure

Research in relation to failure can be tracked back to over a century and a half ago, since the time of Florence Nightingale and Ignaz Semmelweiss.  Florence Nightingale wanted to study why some of her patients died and some didn’t and what the causes of these were when she cared for the army in the Crimean War.  This was from April 1854 to March 1855 and she believed that most of the illness (or harm) which afflicted the army was caused by defects in the system.  She estimated that 1 in 7 of her patients died from preventable diseases rather than their battle wounds and that the things that would reduce the ‘harm’ or illnesses included good nutrition, warm clothing, good ventilation, cleanliness and hygiene (Huxley 1975).  The work of Ignaz Semmelweiss published in 1857 is often quoted as one of the first patient safety research studies into maternal morbidity and mortality and infection control.  His work concluded that increased hand washing resulted in a reduction in mortality from infections for mothers and babies in his care.  However, much of his findings were dismissed by his colleagues who refused to change their practice and Semmelweiss died well before his work would receive the recognition it deserved (Woodward 2017).

It is interesting to note that the contributory factors of both are similar to our knowledge today.  The factors Nightingale found in relation to nutrition, clothing, ventilation, cleanliness and hygiene and the finding by Semmelweiss that washing hands made a significant different to maternal morbidity and mortality have stood the test of time.  So there have been some clues and indicators for quite some time that we have yet to ‘fix’.

Also in the late 1800s another pioneer Ernest Codman took the radical step of publishing not only his patients’ outcomes but also his judgments on whether the results could have been improved and the probable causes of failure to achieve ‘perfection’.  Codman graduated from Harvard Medical School in 1895 and interned at Massachusetts General Hospital.  He joined the surgical staff of Massachusetts General and became a member of the Harvard faculty.  While there, he introduced the first morbidity and mortality conferences.  However, in a similar way to Semmelweiss staff were nervous about his work and in 1914 the hospital refused his plan for evaluating the competence of surgeons and he lost his staff privileges there.  Codman eventually established his own hospital (which he called the End Result Hospital) to pursue the performance measurement and improvement objectives he believed in so fervently.  To support his ‘end results theory’ Codman made public the results of his own hospital; in which for the 337 patients discharged between 1911 and 1916, Codman recorded and published 123 errors (Woodward 2017).

Retrospective case note reviews

Move forward over a hundred years and we find research studies which have tried to understand the scale and nature of the problem by auditing patient case notes.  There are very few documented early studies of the subject but an example published in 1974 conducted in California studied patient records and estimated that our of three million hospital admissions there were 140,000 injuries, 24,000 being considered due to ‘negligence’ (Woodward 2017).

Since the late 1990s the main method of study has therefore been to audit patient case notes which are usually described as ‘retrospective case note review’.  These case note review studies have been carried out across the world and have cited a range of incident rates ranging from 2.9% to 16.6% of all hospital admissions with preventable adverse event rates ranging from 1.0% to 8.6% (Woodward 2017).  The first of these was undertaken in New York State (Brennan et al 1991), the findings of which were extrapolated to suggest that as many as 98,000 patients in hospital settings in the US died each year as a result of problems related to their care.  One of the following studies was undertaken in the UK in two acute hospitals by a team under the leadership of Charles Vincent and colleagues (2001).  This has led to the now often used 10% statistic in the UK (and possibly worldwide) which is used in a number of different ways such as ‘on average there is a 10% error rate in healthcare’ or simplified to things like ‘10% of patients in healthcare are harmed’.

Other examples include a study from New Zealand that concluded that 3.4% of 118 deaths were related to preventable errors in healthcare (Briant et al 2006).  A large retrospective case record review study of 21 of the 101 hospitals in the Netherlands reported a figure of 4.1% adverse events contributing to death among deceased patients (Zegers et al 2009).

Helen Hogan and a team of researchers conducted a retrospective case record review study in the UK in 2009 (Hogan 2012).  In this study of 1000 adults who died in 2009 in 10 acute hospitals, reviewers judged 5.2% of deaths as having a 50% or greater chance of being preventable.  Extrapolating from these figures the authors suggest there would have been 11,859 adult preventable deaths in hospitals in England in 2009.  Hogan and her colleagues found that the problems associated with preventable deaths occurred in all phases of hospital care but were most likely in wards (44%) and involved poor clinical monitoring (31%), diagnostic errors (30%), or inadequate drug or fluid management (21%).  Most preventable deaths (60%) occurred in elderly, frail patients with multiple comorbidities judged to have had less than 1 year of life left to live.

In a further study of deaths in England by Hogan and her colleagues the reviewers identified a preventable death rate of 3.6%, lower than the results in 2009 (5.2%), and no significant variation in the proportion of preventable deaths between hospitals (Hogan et al 2015).  This study was to determine the proportion of avoidable deaths in acute hospital trusts in England and to determine the association with the tools used to assess and compare hospitals on their mortality data; the hospital standardised mortality ratio (HSMR) and the summary hospital level mortality indicator (SHMI).  The reviewers studied 34 English acute hospital trusts (10 in 2009 and 24 in 2012/13) randomly selected from across the spectrum of HSMR.  The difference between 2009 and 2015 is stated as:

• In the 2012/13 cohort the patients were sicker with a higher prevalence of several key comorbid conditions.  Whether or not this was a real difference or reflected greater propensity to record these comorbidities, the impact on reviewers is likely to mean they were less likely to judge a death as avoidable
• Reviewers’ awareness of the use of “do not attempt resuscitation” orders was probably greater as a result of the wider use of highly visible forms in the case records plus changes to the medical review form, which drew their attention to such orders
• There was a minor difference in the wording of the question about attribution of avoidability

There was a small but statistically non-significant association between HSMR and the proportion of avoidable deaths.  The authors concluded that the small proportion of deaths judged to be avoidable meant that any metric based on mortality is unlikely to reflect the quality of a hospital.  Therefore the authors recommended that measuring mortality should be focused on identifying ways of improving the quality of care and not used as an indicator of safety in a hospital.

Assessing mortality using either retrospective record review or a hospital wide mortality ratio is not a helpful or informative indicator of the safety of a hospital.  It is potentially misleading to the public, clinicians, managers, and commissioners to praise or condemn a trust on the basis of either measure.

Helen Hogan 2015