What we found
1. The process of investigating is not consistent, reliable or good enough.
We found that 40% of investigations were not adequate to find out what happened. Not only are trusts not identifying failings, they are also not finding out why the failings happened in the first place. For example, trusts did not find failings in 73% of cases in which we found them, and in over a third of cases where failings were found, trusts did not find out why something went wrong. This is in marked contrast to the perception of 91% of NHS complaints managers who were confident an investigation could find out what had gone wrong.
Serious incidents are not being reliably identified by trusts; we judged 28 of the cases we looked at to be serious enough to lead to a serious incident investigation, but only 8 had been treated as such by the NHS. Identification often relied on either clinicians to spot an incident or on a central risk team flagging incidents. It was clear from our visits to trusts that not all had reliable processes in place, contrary to the perception of complaints managers; 96% stated there was both a process and trigger to help identify a serious incident at their trusts.
We found wide variation between and within trusts in terms of how patient safety incidents are investigated. Perhaps more worrying, is a distinct absence of shared investigatory principles. How a case is investigated is subject to the individual investigator.
We are concerned that there is no national guidance for patient safety incident investigations which make clear:
- who should investigate and how independent of events they should be;
- the level of training an investigator should have for any particular type of investigation;
- broad requirements for the specific evidence needed. For example, statements, interviews or independent clinical reviews;
- how investigations should be independently quality assured;
- what general outcomes any good investigation should aim to achieve.
Worryingly, medical records, statements and interviews were missing from almost a fifth of investigations making it even harder for trusts to arrive at what went wrong and why. Organisations that provide care should not lose sight that it is patients, carers and families who are often at the heart of these investigations. They need to be involved in a meaningful way if investigations are to answer their questions. All of this has a huge impact on patients and families at the centre of any investigation. Our results show that in 41% of cases, complainants were given inadequate explanations for what went wrong and why. The two cases opposite highlight the tragic impact poor quality investigations can have on families and those raising complaints, and why it's important that lessons are learned.
Case study
A one-day-old baby received a blood transfusion to treat severe jaundice. Tragically, serious errors were made in delivering the transfusion resulting in Baby F's collapse, which led to permanent brain damage. Although a serious incident investigation was carried out, it was done so by a close colleague of the paediatrician in charge that day.
We considered that Baby F's collapse was avoidable and requested the trust carry out a review to find out why things went so seriously wrong. The trust acknowledged the investigation was a review of notes only, and clinical staff were not interviewed or asked to provide written statements.
It took three years for Baby F's parents to get a proper explanation for what happened to their baby, adding to their distress.
Case study
Mr M, a 36-year-old father, was taken to accident and emergency with sudden, severe chest pain. Medical staff suspected a heart attack however further tests revealed Mr M may have suffered a tear to the wall of his heart.
After being admitted to a medical ward, Mr M was later discharged with a possible blockage in the bowel with further investigation of his abdomen planned. The following day, Mr M collapsed and lost consciousness. Attempts at resuscitation failed and Mr M died.
Our investigation concluded had a CT scan taken place, Mr M would have been transferred for surgery giving him an 80% chance of survival. No serious incident investigation was conducted and two complaints meetings failed to give the family the answers they needed, despite a list of questions being submitted by the family in advance.
The hospital refused to provide an 'expert view' on whether the doctors' actions were appropriate, adding to the injustice and distress felt by the family.
2. Staff do not feel adequately supported in their investigatory role
There is no national, accredited training programme to support investigators and/or complaints staff in their role. Cultural issues can often be a barrier to getting to the heart of why something has happened.
Common reasons cited during our visits to trusts included a lack of respect; not being provided with protected time to investigate, and the lack of an open and honest culture despite the introduction of the duty of candour in November 2014.
Our visits suggest inequity in terms of who can lead different types of investigations. Our visits revealed that serious incident investigations would often be led by a named investigator with training; all other investigations which fell short of the serious incident criteria could be led by an 'appropriate person'.
Ultimately, staff need to be equipped and empowered to carry out investigations otherwise trusts risk adding to the distress felt by individuals and missing opportunities to make essential service improvements as the following case illustrates.
Case study
Ms G was concerned about changes to her breast and was referred by her GP to a breast clinic. An ultrasound scan led to a diagnosis of mastitis. At a follow-up appointment, a different breast specialist made the same diagnosis. When Ms G missed a follow-up appointment three months later, she was discharged from the breast clinic.
Fourteen months later, Ms G was diagnosed with incurable, advanced breast cancer that had spread to her bones, liver and brain. We found that the secondary cancers were allowed to develop because she had been misdiagnosed and that the two letters she had received confirming mastitis gave her false reassurance. We also found that the trust failed to fully investigate, and did not acknowledge the extent of the failings or the impact on Ms G.
The trust later acknowledged that it should have instigated a serious incident investigation when Ms G was diagnosed with cancer and had it done this, it could have considered learning and service improvements much sooner.
The trust identified a skills gap for staff responsible for investigating complaints, and developed and commissioned a complaint handling course with a local university; complaints management would now become part of their individual appraisals. The trust also established a quality approval process for complaints.
3. There are missed opportunities for learning.
Many complain because they do not wish the same thing to happen to somebody else. Therefore it was worrying to find that 25% of complaints managers were unsure that sufficient processes existed to prevent a recurrence of an incident, and a further 10% believed sufficient processes were not in place.
The impact of poor quality investigations that do not trigger a serious incident is felt most significantly by individuals and their families. However, it also results in missed opportunities to learn and make the relevant service improvements as the case opposite illustrates.
Action is needed in order for learning to take place and this requires people working together in a joined up way. NHS complaints managers, who are responsible for providing explanations to families and ensuring learning takes place, need to be joined up with clinical staff who are often tasked with leading patient safety incident investigations.
Our findings demonstrate that divisions within hospitals often work in isolation to each other; learning from investigations appears to be trapped in high level meetings; and learning across organisations often relies on goodwill and personalities rather than any established processes or mechanisms. Our advisory group reported that cross organisational learning tends to be led by the willing few rather than something that is a widespread practice across the NHS.
Case study
Mr D, a 77-year-old man, was admitted to A&E and seen by a junior doctor who suspected the cause of his symptoms was sepsis, a severe infection. He was not seen by a doctor for two-and-a-half hours, and antibiotics were then not given until two hours after they were prescribed.
Despite stepping up his treatment, Mr D died two days later. Concerns were raised by close family about the timeliness of Mr D's treatment and whether his death could have been avoided. In response to the complaint raised, the trust outlined chronological events using clinical records only.
Had a complaints investigation been done thoroughly, the trust would have found that clinical staff failed to recognise the severity of Mr D's illness, that he was not seen by a doctor for more than two hours, observations were not taken regularly and that a serious incident should have been triggered.
Our investigation concluded that the hospital missed an opportunity to give him the best chance of recovery by failing to give him more timely treatment.