SNAP3 (Frailty and Delirium)

SNAP3 was the third Sprint National Anaesthesia Project, managed by the University of Nottingham, the Royal College of Anaesthetists and the Health Services Research Centre (HSRC).

The number of older people undergoing surgery globally is increasing due to demographic changes and improvements in surgical and perioperative care. Frailty is an age-related syndrome that increases an individual's vulnerability to adverse outcomes in response to stressors such as illness, injury and surgery.

There is reasonable evidence that surgical outcomes are worse in the presence of frailty. Whilst the link between socioeconomic status, frailty and health outcomes is well-described in the community setting, less is known about these factors in the surgical setting. Furthermore, little is known about the importance of particular domains of frailty and the relationship with surgical outcome. Screening for frailty is increasingly advocated but there is a lack of consensus on which tool to use for screening and/or diagnosis in the perioperative setting. Many 'frailty' tools are instead counts of multimorbidity as opposed to multi-domain frailty assessment tools. Distinguishing between multimorbidity and frailty is important given the recognition that seven out of ten frail individuals also display multimorbidity, but only two of ten patients with multimorbidity are also frail.

Delirium is a distinct clinical syndrome associated with adverse outcomes following surgery. Less is known about the influence of severity, timing or form (hyperactive or hypoactive) of delirium on postoperative outcomes. To date, there is no effective pharmacological treatment to prevent delirium, with evidence instead supporting non-pharmacological multicomponent interventions aiming to reduce the incidence of delirium through targeting the triggers for the syndrome, in addition to reducing severity.

Whilst evidence supports commonality in the aetiology and pathogenesis of frailty and delirium, less is known about the interface of these distinct syndromes in the perioperative setting. The initial suggestion from the data is that these two conditions confer cumulative negative effect on postoperative outcomes, but this requires further exploration at scale.

A key question for researchers, clinicians, those responsible for planning perioperative services, and most importantly, patients and their families, is therefore, how to identify frailty and risk of delirium in routine clinical settings. If frailty and delirium can be identified then we can move on to answer questions like what should clinicians do with frailty, multimorbidity and delirium risk in the time before surgery and how should these conditions be managed during and after surgery. This study will generate a large, high-quality dataset on a cohort of older people undergoing a range of surgical procedures to help address these questions.

Study design
SNAP 3 involves three parallel studies:

• S1 Prospective observational patient cohort study
• S2 Organisational clinician survey of current perioperative care
• S3 Clinician survey of acute referrals and interventions to general and geriatric medicine

All patients who were 60 years or older, having surgery (day-case, elective and emergency) during a five-day period, were considered for inclusion. Patients were given a participant information sheet (PIS) whilst waiting in the preoperative areas.

• Consent Process 

1. Patients were identified from operating lists by clinical teams, given a PIS and referred to the research team if they were willing. Patients were then approached by the research team to discuss the study and consent.

2. Consent was taken either on an electronic device using electronic signatures with declarations and tick boxes or a traditional paper consent form.

3. Preoperative data collection
Preoperative data were collected primarily through a review of the medical notes, with participant confirmation if necessary. Medical data, admission information, demographic and socioeconomic data was sought.

• Frailty assessments

Four tools were used to assess presence and severity of frailty. Two tools required participant involvement and two were passive.

4a. The Clinical Frailty Scale (CFS) provides a word and pictorial representation of the frailty syndrome and is recommended in the UK as a national screening tool for frailty, with prior use in surgical populations. The use of the CFS requires observation of the patient and a brief discussion of their activities of daily living. This was completed by researchers before other frailty tools were seen to avoid confirmation bias. 

4b. The Reported Edmonton Frailty Scale is brief, feasible and has also been used in surgical populations. It involves answering 10 short questions and participating in drawing a clock face.

4c. The electronic Frailty Index (eFI) uses the deficit accumulation model of frailty. It is automatically calculated from primary care records. It isn't available in all areas of the UK, but was collected wherever it was recorded.

4d. The Hospital Frailty Risk Score can be calculated from Hospital Episode Statistics data at discharge. The central study team will report this, as it may be a useful automated method to highlight frailty to primary care colleagues.

• Process of care data

Information on process of care was collected primarily through a notes review with participant confirmation if necessary. This information was used to assess the process of preoperative assessment, modes of anaesthesia, use of a catheter and level of postoperative care.

• Delirium

The presence or absence of delirium was assessed on days one and three if the participant remained in hospital. The 4AT or CAM-ICU (delirium assessment tools) and a review of nursing and medical notes of delirium trigger words was used. Notes review was done manually by local researchers. These processes together optimise chances of detecting delirium.

• Postoperative morbidity

A Postoperative Morbidity Survey (POMS) (with appropriate speciality specific modifications for cardiac and hip fracture patients) was conducted on days three and seven if the participant remained in hospital. POMS is a tool used to assess postoperative morbidity. This is mainly a notes review but may require brief face to face interaction with the participant.

• Quality of life (QoL)

QoL was assessed using the EQ-5D-5L questionnaire and a patient/carer estimate of days alive at home (DAH) via telephone interview or electronic email questionnaire. The EQ-5D-5L is a six-question tool suitable for use over the telephone or electronic device. DAH is a patient preferred QoL outcome. We will cross check reported DAH with data linked by Hospital Episode Statistics/Office for National Statistics/Health and Social Care Wales/Electronic Data Research and Innovation Services/NHS Services Scotland (collected for DAOH). This will account for hospital length of stay and readmissions but not residence out of hospital but not at home (this specifically relies on patient/carer reports).

• Data linkage

Data linkage will be carried out with NHS Digital, Health and Social Care Wales, NHS National Services Scotland and individual trusts as appropriate. The following will be collected:

1. Length of stay: Acute hospital stay and days alive and out of hospital (DAOH) within 30 and 90 days

2. Mortality: in-hospital death, mortality at one year, two, five and ten years

3. Readmission: Readmission of participants within 30 days

4. Discharge destination

5. Socioeconomic status: postcode will be linked with indices of deprivation

Length of acute hospital stay (days) will be the primary outcome as it is expected to be affected by both medical complications and discharge planning issues. The other outcomes are important either as mechanistic explanations or as complementary patient-relevant metrics.

We emailed Principal Investigators a survey link which asked for details of existing services in perioperative care. For example, nurse-led and anaesthetist-led preoperative assessment clinics, geriatrician led perioperative medicine services and specialist nurse services.

A survey was conducted with on-call medical and geriatrician registrars (if applicable). The purpose of the survey was to gain an understanding of the workload that older surgical patients contribute to the on-call medical teams. This also offered a brief insight into the training medical doctors get in perioperative medicine.

Investigators approached the on-call doctors at the end of their on-call shifts (day and night) for seven consecutive days. No patient identifiable information was collected. 

Due to the study's population, it was expected that some patients would not have the capacity to give informed consent either before or during the study.

The three options for obtaining consent are summarised below.

A Consultee can be used in England, Wales and Northern Ireland. A Personal Legal Representative (PLR) is used in Scotland:

• If the patient is judged by the Clinician to have mental capacity, they were approached directly to seek informed consent for participation.
• If the patient lacks mental capacity, we sought advice from a relative or friend (Consultee or Personal Legal Representative). If the patient regained capacity, their consent to continue in the study was obtained. Separate information sheets and a consultee/Personal Legal Representative advice/consent form were produced for this situation.
• If it was not reasonably practicable to seek advice/consent from a Consultee/Personal Legal Representative within a suitable time frame, a Professional Consultee was approached for advice. As soon as possible after this, the patient or their representative was approached for consent.

The inclusion of participants who were unable to consent was essential to the scientific success of this study, to improve care for those with, and at risk of, frailty and delirium. To exclude them would be viewed as discriminatory and unethical. Their inclusion will add knowledge and improve care for surgical patients with frailty and delirium in the future.

• Patients who:
  • were 60 years and older
  • were undergoing surgery from 00:00 on the 21st March 2022 to 23:59 on the  25th March 2022
  • were a day case, or having emergency or elective surgery
  • were having general, neuraxial or regional anaesthesia
  • had capacity to consent or have an appropriate consultee/Personal Legal Representative available to agree to participation on their behalf.

This included patients who were undergoing surgery (which matches our other criteria) in an operating theatre, radiology suite, hybrid theatre or endoscopy suite.

Exclusion criteria were limited to those undergoing surgery that was superficial or minimally invasive performed solely under topical/infiltration local anaesthesia. For example, cataract surgery and endoscopy performed without general anaesthesia were excluded. Patients undergoing anaesthesia for a purely diagnostic procedure were excluded. 

• To characterise the epidemiology of frailty, multimorbidity and postoperative delirium in approximately 12 000 older people undergoing surgery in the UK
• Examine the relationship between frailty and perioperative outcomes separately by surgery types
• Examine the relationship between multimorbidity and perioperative outcomes separately by surgery types
• Examine the relationship between frailty and multimorbidity in the older person undergoing surgery
• Describe the variation in hospital-level and patient-level frailty-related interventions
• Identify associations between hospital-level and patient-level frailty-related interventions and outcome
• Develop and internally validate a risk-prediction tool for postoperative delirium.

Thank you SNAP3 Collaborators! We have confirmed the names of all SNAP3 Collaborators with local Principal Investigators. SNAP3 was a success because of each and every one of you! Click here for a list of all collaborators.

Please click here for SNAP3 Study Documents, including the protocol, ethics documents, and help sheets.