A Study to Evaluate the Introduction of New Staffing Models in Intensive Care: a Realist Evaluation (SEISMIC-R) (SEISMIC-R)

Background: Staffing in intensive care units (ICU) has been in the spotlight since the pandemic. Having enough nurses to deliver safe, quality care in ICU is important. However, what the skill mix should be (how many should be qualified nurses or have an ICU qualification) is unclear. Very little research has been done to look at which nursing staff combinations and mix of skills works best in ICU to support patients (described as 'staffing models').Research shows that there is a link between the quality of nurse staffing and poor patient outcomes, including deaths.

Aim: Our research plans to look at different staffing models across the UK. This study aims to examine new staffing models in ICU across six very different Trusts. This study will use a research technique called Realist Evaluation that examines what works best in different situations and help to understand why some things work for some people and not others. The design of this approach will help to better understand the use of different staff ratios across different ICU settings.

This study will examine what combinations of staff numbers and skills result in better patient care and improved survival rates. The aim is to produce a template that every ICU unit can use. To do this, this study will compare staffing levels with how well patients recover, and seek to understand the decisions behind staffing combinations.

Methods: This study will:

1. carry out a national survey to understand the different staff models being used, comparing this against the current national standard (n=294 ICUs in the UK including Scotland)
2. observe how people at work in 6 hospitals (called ethnography), watching how they make decisions around staffing and the effect on patients. The investigators will also conduct interviews (30 interviews plus 30 ethnographic observations) to understand staffing decisions.
3. look at ICU staffing patterns and models, and linked patient outcomes (such as whether people survive ICU) over 3 years (2019-2023) in those hospitals, including with a very different combination of staffing). The investigators will then carry out some mathematical calculations to understand the best possible staffing combinations, and how this varies.

Study Overview

• Status: Completed
• Conditions: Critical Illness; Intensive Care; Staff Attitude; Organisation
• Intervention / Treatment: Other: n/a (non-interventional)

Detailed Description

Background: Optimising deployment of the scarce nursing workforce in the intensive care unit (ICU) is paramount for patient safety, and staff wellbeing. ICU staffing models are determined by National Health Service (NHS) service specification, with 1:1 patient to registered nurse (RN) ratios for the highest acuity patients. A rapid expansion of ICU capacity during COVID19 led to adoption of alternative models, using more support staff, non-ICU qualified nurses and other professionals, reaching up to 70% at surge. The strengths, weaknesses, costs and effects of these models, and benefits of retaining them, remain uncertain. Lower nurse-staffing levels, and high workload, have been associated with adverse outcomes for patients, staff and organisations although ICU-specific evidence is limited. Studies focus on levels of RNs, contributing little to understanding consequences of changes retained post-COVID, or to guiding adoption of alternative staffing models. It is unclear how changes in staffing or specific models affect various outcomes.

Aim: To identify the key components of an optimal nurse staffing model for deployment in ICU.

Objectives/Methods: Guided by a realist framework, the investigators propose to interlink workstreams (WS) over 2 years to allow cross-fertilisation of ideas/hypotheses and inform emerging programme theories.

1. To identify and describe organisation of models, exploring intended mechanisms and outcomes for how different models work, the investigators will conduct:

   • a UK survey (WS 1) of all 294 ICUs in England/Wales/Northern Ireland (NI)/Scotland that will identify staffing models emerging/retained since COVID19, compared with United Kingdom (UK) service specifications.
   • a realist evaluation (WS 2, cross-cutting workstream) and detailed case studies involving six sites, and 30-40 interviews with: nurses/senior nurses; organisational leads; critical care network managers/commissioners; families/patients, to test emerging programme theories. Rapid ethnographies (n=30), will elucidate how staffing decisions are made.

2. To provide estimates of variability in demand for nursing staff and estimate associations between staffing patterns and patient outcomes, the investigators will:

   - use administrative e-roster (nurse staffing roster) data and patient data (WS 3) from the Intensive Care National Audit and Research Centre Case Mix Programme (2019-2023) to assess whether and how patient/staff outcomes vary with differing staff models between units/case study sites.

3. To develop simulation models to show the impact of models on capacity, cost and patient flow, the investigators will use simulation modelling (WS 4) to explore scenarios for different staffing policies given case mixes of case study units, swiftly and with no patient impact.

Analysis: Data integration occurs across all workstreams in WS 5. Theories developed from WS2 case studies will be further tested against WS 3 observational data and inform WS 4 mathematical simulation models of ICU capacity, patient outcomes and patient flow, to inform emerging propositions for the realist evaluation programme theories as context-mechanism-outcome configurations.

• Study Type: Observational
• Enrollment (Actual): 87

Contacts and Locations

Study Locations

• United Kingdom
  • Stevenage, United Kingdom
    • East and North Hertfordshire NHS Trust, Lister Hospital

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

Staff population: Workstream 1 and Workstream 2 sample is expected to be 294 (assuming 100% ICUs respond) plus 30 interviews and 30 ethnographic observation participants

Critical care patients population (de-identified data, no individual consent/sampling needed, using routine aggregated dataset): Workstream 3/4 the data set is likely to exceed 29300 Intensive care admissions

Description

Inclusion Criteria: Interviews with nursing staff

• Registered Nurses who have worked in ICU for at least six months. Inclusion criteria: Interviews with nurse managers
• Clinically-based nurse managers who have worked in ICU for at least one year. Inclusion criteria: Interviews with organisational leads
• Organisational leads who have been working in their role and in the ICU field for at least one year.

Inclusion criteria: Interviews with regional managers/commissioners • Regional managers/commissioners who have been working in their role and in the ICU field for at least one year.

Inclusion criteria: Interviews with patients/families

• Patient or family member over 18 years old.
• Patients who have been in General ICU for at least 48 hours in the last 6 months.
• Family members who have visited ICU for at least 20 mins on two days in the preceding 6 months.
• Patient discharged from hospital at least 2 weeks prior to the interview.
• Patient expected to be well enough, after hospital discharge, to attend the interview and to have capacity to consent.

Exclusion Criteria:

-

Study Plan

How is the study designed?

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Death from all causes within 30 days of ICU admission	Mortality	2019-2023

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Discounted Quality-adjusted Life Year (QALYs)	Quality-adjusted Life Years (QALYs)	2019-2023
Composite death/discharge to long term-care (LTC)	Death/discharge to LTC	2019-2023
ICU-acquired infection	ICU-acquired infection	2019-2023
Days of organ support in ICU	Days of organ support in ICU (per organ)	2019-2023
Cost of ICU stay	Cost of ICU and post-ICU stay (in hospital)	2019-2023
Staff absence	Sickness and absence (and associated costs)	2019-2023

Collaborators and Investigators

• Sponsor: University of Hertfordshire
• Collaborators: University of Southampton; University of Plymouth; Imperial College Healthcare NHS Trust; University of Exeter; Intensive Care National Audit & Research Centre; London South Bank University
• Investigators: Principal Investigator: Natalie A Pattison, University of Hertfordshire

Publications and helpful links

General Publications

• Hadley R, Dogan B, Wood N, Bohnacker N, Mouncey PR, Pattison N; SEISMIC-R investigator group. National survey evaluating the introduction of new and alternative staffing models in intensive care (SEISMIC-R) in the UK. BMJ Open. 2025 Apr 10;15(4):e088233. doi: 10.1136/bmjopen-2024-088233.

Study record dates

Study Major Dates

• Study Start (Actual): June 14, 2023
• Primary Completion (Actual): September 30, 2025
• Study Completion (Actual): September 30, 2025

Study Registration Dates

• First Submitted: May 4, 2023
• First Submitted That Met QC Criteria: June 14, 2023
• First Posted (Actual): June 26, 2023

Study Record Updates

• Last Update Posted (Actual): May 4, 2026
• Last Update Submitted That Met QC Criteria: April 28, 2026
• Last Verified: April 1, 2026

More Information

Terms related to this study

• Keywords: critical care; intensive care; nurse; Patient outcomes; realist evaluation; Staffing; healthcare organisation
• Additional Relevant MeSH Terms: Pathologic Processes; Disease Attributes; Pathological Conditions, Signs and Symptoms; Critical Illness
• Other Study ID Numbers: UH:02990

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: We have a data sharing management plan available on request

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No