St George's Hospital (Tooting)

On this page

• Overview
• Learning culture
• Safe systems, pathways and transitions
• Safeguarding
• Involving people to manage risks
• Safe environments
• Safe and effective staffing
• Infection prevention and control
• Medicines optimisation

We looked for evidence that safety was a priority for everyone, and leaders embedded a culture of openness and collaboration. We checked that people were safe and protected from bullying, harassment, avoidable harm, neglect, abuse and discrimination. We also checked people’s liberty was protected where this was in their best interests and in line with legislation.

At our last assessment we rated this key question as requires improvement. At this assessment the rating has changed to inadequate. This meant people were not safe and were at risk from avoidable harm.

We have not awarded this service a score for Safe.

Find out about when we will not publish a key question score and what we look at when we assess Safe.

The service did not always have a proactive and positive culture of safety based on openness and honesty. They did not always listen to concerns about safety and did not always investigate and report safety events. Lessons were not always learnt to continually identify and embed good practice.

Staff knew how to report incidents, but we were not always assured that they were reporting incidents appropriately. We saw evidence of a high number of incidents being reported between July to September 2024. Staff raised concerns through an electronic incident reporting system and discussions with managers. We reviewed documents of lessons learned from incidents such as clinical governance newsletters and learning from incidents forms which highlighted key learning outcomes for staff.

>

We reviewed trust responses to complaints and data showed 88 complaints were received between October 2023 and October 2024. Complaints were broken down between Green, Amber and Red categories to determine how many working days were allocated for a response. Data provided by the trust was not broken down by categories, therefore we were not assured that all complaints were aligned with trust policy or investigated in a timely manner. However further data provided by the trust showed complaint data broken down by categories. We were able to assess the raw data which showed that majority of complaints were closed in line with trust policy. Rationales were provided for some complaints closed outside of trust policy.

>

The duty of candour (DoC) is a regulatory duty that relates to openness and transparency and requires providers of health and social care services to notify patients (or other relevant persons) of certain `notifiable safety incidents' and provide reasonable support to that person. We reviewed two trust responses and were not assured that the trust upheld its Duty of Candour obligations.

The service did not manage or monitor people's safety. They did not make sure there was continuity of care, including when people moved between different services.

We were not assured that safety and continuity of care was consistent throughout patients care journey. There was not an effective system in place to manage patient acuity in the waiting room as we observed patients who arrived at the ED experiencing long waiting times before being assessed. This was a recurring theme during our conversations with patients. We observed the service lacked adequate clinical oversight, leading to long wait times for triage, sometimes exceeding 50 minutes, this was over the 15-minute national target for patients arriving at the ED. We observed patients with high-risk symptoms who should have been assessed urgently. We issued a section 29a warning notice as the service required significant improvements. The trust has now developed a standard operating procedure (SOP) to improve clinical oversight which was implemented 6 January 2025. This SOP outlines escalation procedures for wait times exceeding 15 minutes or when six or more service users are waiting.

Patients expressed confusion around the queueing process, as there was no clear guidance or signage directing patients in the waiting room. However, waiting times were visible in the waiting room. We reviewed data sent by the trust of multiple incidents relating to delays and overcrowding.

We reviewed data in August, September and October 2024 which showed a consistently high and increasing demand on the service which placed pressure on capacity, flow, and timely access to care. The service had a full capacity protocol which set out actions to take when the service was overcrowded including daily flow meetings and escalation to trust wide command. However, this was issued in early November 2024, during our assessment staff explained how the protocol was not yet fully embedded within the service.

>

Data provided from the service showed 74% of patients met the 4 hour waiting target between August and October 2024. This fell below the 76% national standard. Moreover, we noted that actual patient wait time from ED arrival was longer due to significant queues preceding triage. Staff reported significant queuing issues, with wait times sometimes exceeding an hour. This lack of clinical oversight and safety management systems resulted in delayed assessments, impacting patient care. Service users were not assessed by medical staff in a timely manner and were at risk of deterioration as a result until they were assessed by the streaming nurse, and their actual time of arrival was not recorded in this data. Therefore, we were not assured of the accuracy of these figures due to the issues identified regarding the triage and queuing processes.

>

Streaming to same day emergency care (SDEC) was considered during triage, using established criteria. However, this was restrictive having a very specific inclusion criteria for admission. Admission also required a discussion with a consultant before a patient could be sent to the SDEC. Staff told us the lack of flexibility in these protocols limited their ability to manage patient flow effectively. Additionally, there was no live quality or performance dashboard available for staff to view.

>

We reviewed 10 sets of electronic patient notes and found there were gaps in nursing documentation and a lack of detail within the notes, which meant there was an incomplete picture of patient care. We also noted a delay in doctors reviews. However, trust practice was for documentation to be maintained on the day of consultation or within 24 hours. For patients waiting in the ED for over 24 hours, there was a lack of detailed documentation, particularly regarding medication administration or patient updates. There was also a lack of detailed documentation relating to mental health patients who absconded. Notes we reviewed of a patient who had absconded did not reflect that they had left the ED before being assessed. The trust has developed written documentation standards for staff to adhere to and now undertake a daily audit in the service against basic documentation questions with outcomes shared with staff for improvement. A wider documentation audit was being developed to be available from April 2025 which will be monitored through ED quality and safety meetings. A documentation template is now being used across the service to ensure documentation standards are met.

>

Our review of ED morbidity and mortality meeting minutes demonstrated a proactive approach to risk management and escalation. During incident discussions, staff raised serious concerns around corridor care and formally escalated concerns to trust leaders and external regulators.

The service had a qualified streaming nurse and an additional consultant was sometimes available to support the streaming nurse by directly referring patients when needed. For patients who were offered a GP appointment, these were arranged by a dedicated navigator, allowing patients to leave the service without unnecessary delays.

>

The service had an SOP for paediatric referral services for children between 1 and 18 years old aiming to avoid hospital admission and ED attendances. However, this was only operational between 9am to 6pm and last referrals were at 4:30pm.

>

The service offered a hospital at home service, allowing patients to receive multidisciplinary support at home from nurses, therapists, and doctors. Clinicians worked closely with the hospital at home team to arrange follow-ups and personalised care at home.

The service did not always work well with people and healthcare partners to understand what being safe meant to them and how to achieve that. They did not always concentrate on improving people’s lives or protecting their right to live in safety, free from bullying, harassment, abuse, discrimination, avoidable harm and neglect. They did share concerns quickly and appropriately.

We reviewed the ED risk register and saw a risk of harm to children and young people due to delays in formal radiology reporting and reviewing the reports. Data showed that there were 3 incident reports of missed injury or illness to date. This means that we could not be assured that children were appropriately safeguarded.

Safeguarding policies were comprehensive and up to date. We requested data on safeguarding training numbers for all staff. Training compliance data showed that 90% of medical staff had completed safeguarding training for adults at Level 2, with compliance for children’s safeguarding at 94.7% for Level 2 and 89% for Level 3, exceeding the trusts target of 85%. Training compliance data for nursing and HCAs showed that 86.93% had completed safeguarding training for adults at Level 2, with compliance for children’s safeguarding at 89.29% for Level 2 and 88% for Level 3, exceeding the trusts target of 85%.

There was an understanding of safeguarding as staff described procedures and effective use of escalation processes. Staff generally identified and acted on safeguarding and domestic violence concerns, ensuring patients were supported appropriately. Staff told us they would raise safeguarding alerts or flag domestic violence concerns on the system for the assessment team to address in a private and secure area. During the assessment, we observed a streaming nurse identifying a potential domestic violence case, flagging it to the assessment team, and requesting further assessment. Patients benefitted from the proactive and collaborative approach taken by staff to address safeguarding concerns sensitively and appropriately.

Staff were able to articulate signs of abuse, neglect, or domestic violence and how they would respond, including referral to the safeguarding team. They demonstrated confidence in identifying safeguarding concerns and described clear processes for escalation. Staff reported receiving excellent responses and support from the safeguarding team, which they described as approachable and responsive. The trust demonstrated a commitment to taking action to keep people safe from abuse and neglect, including collaboratively working with partners. We reviewed data Between July 2023 and June 2024 that showed the service was making appropriate referrals to the local authority. We saw 569 referrals for adults and 1,807 for children. This demonstrated compliance with trust safeguarding policies.

The service is supported by a multidisciplinary safeguarding team, including a health visitor liaison and a named consultant lead for safeguarding in the paediatric emergency department.

The service did not work well with people to understand and manage risks. Staff did not provide care to meet people's needs that was safe, supportive and enabled people to do the things that mattered to them.

We reviewed the trust policy for NEWS 2, which stated that patients required hourly monitoring of NEWS scores between 0-8 hours. Nurses in the majors area monitored and entered NEWS scores into the electronic system. These scores appeared on the department tracking screen, but it didn't show when observations were last recorded. Post inspection the trust told us that observations were also recorded within the electronic patient record where the frequency of observations and when observations were due were confirmed. Staff escalated NEWS2 scores to the consultant in charge. However, for patients awaiting a speciality bed, staff told us that they discussed escalated NEWS2 scores with the speciality doctor. This caused delays as bleep numbers were unavailable, resulting in physical searches for doctors on the medical assessment unit. Post assessment the trust told us that was a staff directory for all staff bleeps to contact speciality doctors, which was easily accessible.

We reviewed the ED sepsis audit from April 2024 where 45 patients were audited. All patients should receive antibiotics within 3 hours, or within 1 hour of suspected neutropenic sepsis and for patients with refractory hypotension. Data showed that 1 patient that met the criteria for treatment via antibiotics waited 5 hours. Only 42.2 percent of patients received their antibiotics within 3 hours of arrival.

We requested PEWs audit which showed 100% compliance in June 2024, however there were no audits in July and August and in September only 80% of records showed completion of PEWS audit on admission. Therefore, we were not assured that PEWs audit were reviewed consistently due to the gaps in the data. In addition, we requested information regarding initial triage of paediatric patients being conducted within 15 minute of arrival. This information was not provided and therefore we were not assured triage was completed in a timely way.

We requested VTE data for ED, however the data provided was trust wide VTE audit results which indicated poor performance across the whole hospital. We were not assured the ED had oversight over their own performance and any actions taken to manage poor VTE performance. This was not documented on the ED risk register. Post inspection we were told that at the time of the inspection VTE audit reports did not include ED as a separate location. VTE audit data was captured with reference to risk assessment and prescribing within 14 hours of admission to a ward. This has since been extended to include VTE audit data collection within 14 hours of decision to admit as per national guidance. However, we were not provided with any additional supporting evidence to support this.

We reviewed electronic patient notes and found that risk assessments were not always completed, particularly for mental health risks. The service completed risk assessments for patients presenting with mental health needs, but these often lacked detail. While they outlined steps to minimise risks, it was unclear how decisions were made to categorise risks as low, medium, high, or very high. In cases where patients left the unit before completing treatment or referrals and presented a risk of harm to themselves or others, the service contacted the police. There had been 30 such incidents in the last six months categorised as low harm.

Adults and young people frequently waited 3 to 5 days, exceeding national standards, for a mental health bed once assessed as requiring admission under the Mental Health Act (MHA). Staff escalated these extended waits to senior management.

>

However, staff told us they were confident in escalating concerns related to long mental health waits or other risk factors. We observed detailed patient handovers, which included discussions of care plans, outstanding tasks, administered medications, and risks including falls. These handovers were supported using electronic patient records, to ensure staff taking over a patient's care were fully informed of a patient's condition. However, observed a set of patients notes which had missing National Early Warning Scores (NEWS), which are essential for identifying deteriorating patients.

>

The service held monthly multidisciplinary meetings, involving ED staff, the psychiatric liaison team, police, ambulance services, and social workers, to discuss concerns and improve processes. Staff described these meetings as collaborative and productive, highlighting strong relationships with partner organisations. Staff were regularly involved in discussions about safety, risk, and service pressures. In governance and multidisciplinary meetings, staff raised concerns about mental health patient waits, resource constraints, and patient behaviour in the waiting room. Notes from ED Morbidity and Mortality meetings demonstrated how staff contributed to incident review, action planning and scheduled follow-up reviews to support shared learning. Risk and performance updates were routinely shared with the team, and staff had opportunities to raise concerns through regular safety huddles and structured feedback routes.

>

We reviewed the trust's major incident plan, which was in date with the next review scheduled for December 2024.

>

The service did not always detect and control potential risks in the care environment. They did not make sure that equipment, facilities and technology supported the delivery of safe care.

Patients were not always cared for in safe environments as we observed overcrowding within the service including patients on trolleys in the corridor who had been waiting several hours to see a doctor. Staff described significant delays in securing mental health beds, sometimes lasting up to 7 days.

>

Staff expressed concerns about the overcrowding in the service, which they felt compromised safe care.

>

During our assessment, we found that two adult mental health rooms were intended to be ligature free, however, although these rooms had fixed fitted beds included, each room had an additional hospital bed and hospital trolley, which introduced new ligature anchor points, and significantly restricted space. There was no CCTV in these rooms, but it was provided in the main ED majors area. Following the assessment, the trust advised that they had reviewed the fixed beds in these rooms and would be providing alternative appropriate furniture.

>

Rooms for both paediatric and adult mental health patients included ligature anchor points, not all of which could be removed such as taps and hooks. Patients at risk of self-harm could not be left in these areas without supervision, which impacted upon their privacy, particularly during long stays.

>

Within the paediatric emergency department, the service did not have bed areas specifically designed to minimise the risks presented by patients with mental health needs. If the patient presented a heightened risk, the service would need to place them in the adult mental health room within majors. Patients could access toilet and shower facilities on both the adult and paediatric ED.

>

The mental health assessment room and other rooms used for mental health patients were located centrally within the adult and children EDs, presenting safety management concerns for patients who had behaviours that challenged other patients with significant physical health problems.

>

Staff had access to a quiet room which was used for distressed patients, but this was located off the main majors area, and was not ligature free, which posed a risk to staff and patients. There was a nursing station located near this room, but we observed it to be frequently unattended. However, since the hospital was not a mental health trust not all rooms were required to be ligature free.

We found examples of daily checks not completed in major B of resus trolleys along with out-of-date consumables. In the paediatric ED, we found trolleys had consumables that were found to be out of date including multiple non-adherent dressing.

>

We also found that the medicine cupboard in majors B and the control of substances hazardous to health (COSHH) cupboard in the dirty utility area were unlocked, despite signage indicating they should remain locked. Additionally, we found the dirty utility room door was left unlocked.

>

We reviewed meeting minutes where staff identified safety and dignity issues in shared spaces where privacy for procedures could not be maintained. Requests were made for additional equipment such as curtains and trolleys.

>

However, access to the service was secure, with a door buzzer system in place and staff carrying electronic passes to ensure controlled entry.

>

The service did not always make sure there were enough qualified, skilled and experienced staff. They did not always make sure staff received effective support, supervision and development. They did not always work together well to provide safe care that met people's individual needs.

Staff reported that safe care and treatment were sometimes compromised due to the high numbers of patients waiting for admission, extended stays of mental health patients, and resuscitation area capacity being exceeded. Paediatric staff expressed concerns about insufficient staffing levels in their area. This was not documented in the service's risk registers. However, we did observe that the risk register documented a lack of radiology reporting for paediatric patients.

>

Staff also shared that they were not always given protected time for training, but overall compliance rates met the trust's target. Junior doctors told us they felt well-supported, with excellent clinical supervision and high-quality education. All resident doctors we spoke to told us they would recommend the service to friends and family.

>

Staff explained how they could not always provide one-to-one, or two-to-one observations for mental health patients as recommended following assessment by the psychiatric liaison team, due to having insufficient staff. During our assessment, there were sufficient staff to meet patient needs. Staff told us that the trust directly employed registered mental health nurses and mental health support workers on each shift and more could be called in when needed (although there could be a delay in additional staff arriving).

>

Leaders acknowledged the pressures faced by the team, including the sustainability of the consultant rota. While there was 24/7 consultant presence, several consultants had come off the night rota due to stress, putting additional pressure on the team. The Royal College of Emergency Medicine (RCEM) recommends at least 34 whole-time equivalent (WTE) consultants to provide 24/7 care in the service. Although the service maintained strong 24/7 consultant coverage, the current number of WTE consultants fell short of this recommendation. We were not assured that the current staffing level was sustainable and presented long-term risks to staff well-being and service delivery.

>

Staff raised concerns about unsafe staffing levels at team meetings. Minutes recorded instances where only 3 medical assistants were available on shift, prompting escalation and review of rota fairness and staff wellbeing. During the inspection we spoke to an adult nurse working in paediatrics ED who had not had any additional training to look after children.

>

Registered general nurses within the EDs and agency RMNs did not carry out any restraint techniques on patients in line with trust policy. Only security staff were trained to do this. Staff said they would benefit from more de-escalation training, and more staff trained in physical restraints when needed. 69% of security staff had completed training in physical restraints of patients, with a further nine due to be trained in January 2025, bringing training compliance to approximately 94%.

>

Staff we spoke with did not report significant incidents of violence and aggression in the EDs over the last six months. Staff said that security staff were available when needed, and were needed more, since the police had introduced Right Care, Right Place, which meant police no longer stayed in the ED with mental health patients, unless they were detained under section 136 of the Mental Health Act.

Patients consistently described staff as competent, professional, and attentive to their needs. Patients described how staff followed established protocols and provided care that met their expectations. Despite pressures within the service, patients told us they felt the care they received was safe and appropriate.

>

The service was well-staffed at the resident doctor level, with 15 staff covering the early shift, 15 joining in the afternoon, and 9 on the night shift. Mandatory training compliance rates were high, with medical staff achieving 91.1% overall, including 96.2% for early warning scores, exceeding the trust target of 85%. Induction processes for doctors were thorough, and all doctors were required to undergo annual appraisals with 80.9% compliance at the time of the assessment. This was below the trust target of 90% and was not included on their risk register.

We reviewed meeting minutes where staffing risks were identified and discussed.

>

Staff raised concerns about limited support staffing and the need for clearer role responsibilities, particularly for healthcare assistants. Concerns around flexible working were also noted as contributing to rota pressures. The trust was progressing workforce development through policies to extend roles including physician associates and nurse-led analgesia in paediatrics. Staffing risks linked to overcrowding and corridor care were escalated through governance structures.

>

Volunteers within the service underwent comprehensive recruitment checks and a trust-wide induction, and they were supported and supervised by the nursing team. Evidence from the trust showed that bank staff had comprehensive training delivered as part of their bank training as well as a local departmental induction. Staff had access to trauma-informed care training and seminars, and 95% had completed conflict resolution training. Additionally, 89% of staff had completed Oliver McGowan training on supporting patients with learning disability and autism and 97% of staff had completed training in disability awareness.

>

The service did not always assess or manage the risk of infection. They did not always detect and control the risk of it spreading.

Mandatory infection prevention and control (IPC) training compliance for medical staff was 85.5%, which was above the trust target of 85%.

>

During our inspection, we observed corridor care for patients with Methicillin-resistant Staphylococcus aureus (MRSA). Corridor care is the practice of providing medical attention to patients in hallways or other non-designated clinical areas due to overcrowding or resource shortages. This posed infection, prevention and control issues due to limited access to gloves, bins, and sinks. Moreover, corridor care made it difficult to maintain appropriate isolation for infectious patients, which increased the risk of transmission to other patients and healthcare staff. Patient isolation and infection was listed as a risk on the open risk register, gaps in controls and assurance were documented however assurances and actions were left blank. This was recorded as an extreme risk on the open risk register.

It was reported on the open risk register that the ED would not meet IPC standards, which could lead to an increase in the prevalence of infection within the department and therefore impact patient safety and care, increasing Length of Stay (LOS), as well as staff sickness levels. Not all actions for this risk had been completed to mitigate this risk on the open risk register, this risk was recorded as a moderate risk.

Sewage leaks were regularly reported, most recently, week commencing 7 October 2024. This posed as a health and safety, and IPC risk to all patients and staff. A sewage leak would result in a short notice closure of the impacted areas of ED for maintenance. This would lead to increase space pressures within the department and a knock-on effect on admitted and non-admitted performance as well as considerable impact on staff morale and poor working conditions. This was documented on the risk register, with some actions competed but not all, this was recorded as a moderate risk.

However, in majors A, staff were observed practicing good hand hygiene by washing their hands after checking a patient's blood pressure, disposing of gloves, and wearing new personal protective equipment (PPE) before interacting with the next patient. During our assessment, we observed staff to be bare below the elbow.

>

In majors B, the treatment waiting area was visibly clean, with handwashing and sanitiser stations available. Curtains around beds were visibly well-maintained and had been changed regularly. Signage promoting good hand hygiene was displayed around hand washing basins. Facilities for mental health patients were also clean and well-maintained. We observed clinical areas had domestic and clinical waste bins and PPE was available.

>

Cleaning schedules and environmental audits for resus, paediatrics, majors A and majors B scored above the 98% compliance target. Hand hygiene audits from August to October 2024 demonstrated good compliance, with average scores above 98%. The latest PPE audit, conducted in May 2024, achieved 100% compliance. Isolation cleaning was also routinely completed prior to patient transfer. However, audits conducted in November 2024 identified some areas that failed but there were no action plans to address this. Areas that failed included hand wash dispensers and hand sanitisers.

In majors B, we saw that equipment was regularly checked, with portable appliance testing (PAT) in date for various equipment and plugs. All equipment observed was visibly clean. "Last cleaned" stickers were visible on trolleys and other appliances, and all were in date. All furnishings, such as chairs and flooring, were wipeable and easy to clean.

>

We reviewed policies provided by the trust including business continuity plans outlining protocols for managing service disruptions. This included emergency cleaning arrangements, the use of PPE, and the handling of isolation needs in the event of premises being compromised. The plan made specific reference to maintaining patient safety and continuity of care through access to essential clinical supplies, cleaning arrangements, and safely relocating patients.

>

The service did not make sure that medicines and treatments were safe and met people's needs, capacities and preferences.

We reviewed the records of 7 patients and found that medicines were not always given as prescribed, with some doses marked as not given retrospectively due to drug shortages. Medicines omitted included critical medicines such as inhalers, medicines for epilepsy and lithium. Staff told us that people did not always receive their medicine as the service often ran out of stock for medicines, requiring nurses to check other wards to obtain stock. Nurses told us that when they contacted the pharmacy for queries or support for obtaining stock or prescriptions, the response times were variable. Following our inspection, the trust has reviewed the medicines stock list to ensure it is up-to-date and reflect potential patient needs by conducting audits. The trust has updated and circulated the critical medicines list. The trust has also provided guidance to staff on how to access and obtain medicines that are not available in the service.

>

The service's pharmacy support was inadequate and did not match the recommended support outlined in the Royal College of Emergency Medicine (RCEM) guidelines. Pharmacists reported prioritising high-risk service users but were not able to review everyone in a timely manner. We saw that there was a pharmacist who was offering dedicated support to the whole emergency unit, including the acute medical unit (AMU). Following our inspection, the trust has conducted a risk assessment and has put mitigations in place including an additional interim pharmacist to increase support.

>

Staff highlighted how the resuscitation area of the service had regular discrepancies with controlled drugs (CDs). This was due to the acuity of the unit, and they did not always complete records as a result. Errors in CD record keeping books were crossed out which is not line with current legislation. Following our inspection, the trust has developed an improvement plan which is being monitored through weekly access meetings and governance meetings, including a spot check audit to review and manage concerns. The trust has increased the frequency of CD audits from quarterly to every other month to monitor progress with actions. Information on CD management has been displayed within the service and throughout the trust for staff.

>

Pharmacists told us they completed regular audits of antibiotic stewardship, venous thromboembolism (VTE) assessments and medicines management but improvement was inconsistent. This was in part due to constant rotation of resident doctors. Staff described seeing spikes in improvement but then a gradual decline. Therefore, we were not assured the audit process was effective.

>

Issues were identified with the electronic prescribing and administration record (EPMA) system, including weight-based prescribing errors, where incorrect weights were recorded in patient care records. Pharmacists identified prescribing errors linked to this issue. There was no system in place to flag overdue medicines or missed assessments, contributing to omitted doses and inconsistencies in medication administration. Following our inspection, the trust has provided assurances that mitigations are in place while this functionality is being developed, including drug check rounds twice a day to monitor missed doses.

>

The self-administration policy was not consistently applied across the service, leading to variation in practice. In majors A, patients self-administered medication without a formal assessment, while in majors B, medicines were stored in pod lockers, with administration prompted by nurses. Staff had different interpretations of how this was managed. Following our inspection, the policy for self-administration has been risk assessed and suspended. The service is now prescribing medication following an assessment by a clinician and this is being administered by nursing staff.

>

We issued a section 29a warning notice for medicine concerns and received an action plan from the trust describing mitigation steps for these risks.

We reviewed medicines optimisation group meeting minutes where staff raised governance risks around unlicensed medicines, highlighting delays and inappropriate reliance on junior doctors, reflecting staff awareness of prescribing safety.

>