Introduction + Context
Literature Review
Significance of the Research
Aim, Questions + Objectives
Methodology
Ethical Considerations
Impact
Contribution to Knowledge
Indicative Timeline
Bibliography
RECOVERING LOST CAPACITY IN ACUTE IMAGING AND INTERVENTIONAL RADIOLOGY:
RECOVERING LOST CAPACITY IN ACUTE IMAGING + IR:
A Phased Clinical Informatics Study On The Use Of Artificial Intelligence For Appointment Slot Optimisation In Clinical Imaging.
A Phased Clinical Informatics Study On The Use Of Artificial Intelligence For Appointment Slot Optimisation In Clinical Imaging.
PhD Project Proposal - [Digital Thesis]
Research / Data Analysis / System Design / Optimisation / Patient Experience Design
PhD Project Proposal - [Digital Thesis]
Research / Data Analysis / System Design / Optimisation / Patient Experience Design
[Last Updated - 01.05.26]
[Last Updated - 01.05.26]
1. Introduction + Context
1. Introduction + Context
Every day across NHS imaging and interventional radiology departments, scheduled slots go unused. A patient is cancelled at short notice because they were not adequately prepared, an infection risk is identified, a porter cannot be found in time, or a last-minute clinical decision changes the plan. The slot remains empty. Meanwhile, another inpatient or A&E attender might have been clinically suitable and operationally ready to fill it, but there is no system in place to identify them quickly enough.
This is not an occasional problem; it is a structural one, and it plays out in radiology departments across the country on every working day. The consequences are felt at every level. Patients wait longer for imaging or procedures that could inform urgent clinical decisions. A&E flow is slowed because attenders wait for scans before a bed decision can be made. Inpatient stays are extended unnecessarily. Radiographers, radiologists, imaging coordinators, and interventional teams absorb the administrative burden of managing cancellations manually. NHS Trusts lose capacity they have already paid for without any mechanism to recover it.
This research sets out to address that problem directly. The researcher works as an Interventional Radiology and Biopsy Coordinator at Medway Maritime Hospital, with oversight of imaging workflows across inpatient and outpatient pathways and direct responsibility for coordinating interventional procedures, mainly image-guided biopsies. That operational position provides daily, embedded access to the systems in which the problem manifests, and direct working relationships with the radiographers, radiologists, nursing staff, and administrators whose cooperation any solution would depend on. It also provides direct visibility of the interface between diagnostic imaging and interventional radiology; a junction that is consistently affected by cancellations and is rarely treated as a single operational unit in existing scheduling research.
The proposed solution is a co-designed, AI-assisted slot optimisation system that, when a cancellation or did-not-attend (DNA) occurs, surfaces clinically appropriate inpatients or A&E attenders who are already prepped or can be prepped in time, presents them to staff for confirmation, and supports rapid reallocation of the slot. A modest expansion of appointment time windows is built in to allow for transport, prep variability, and operational flexibility, without disrupting the underlying schedule. The intended outcomes are reduced unused imaging slots, improved patient flow, and measurable support for A&E throughput and bed capacity.
The immediate research focus is bounded to four service areas where the problem is most pronounced and the operational case for intervention is strongest: MRI, CT, ultrasound, and image-guided biopsies, for inpatients and A&E attenders.
This proposal describes the design, pilot evaluation, and framework derivation of that system, built through embedded, low-burden co-design with frontline staff and tested in a live NHS environment at Medway Maritime Hospital. It is positioned with strategic awareness of the broader operational context, in particular the formation of the Medway and Dartford & Gravesham NHS Trust group, which creates a credible pathway for tested, validated solutions to scale across sites once the pilot evidence is in place.
Every day across NHS imaging and interventional radiology departments, scheduled slots go unused. A patient is cancelled at short notice because they were not adequately prepared, an infection risk is identified, a porter cannot be found in time, or a last-minute clinical decision changes the plan. The slot remains empty. Meanwhile, another inpatient or A&E attender might have been clinically suitable and operationally ready to fill it, but there is no system in place to identify them quickly enough.
This is not an occasional problem; it is a structural one, and it plays out in radiology departments across the country on every working day. The consequences are felt at every level. Patients wait longer for imaging or procedures that could inform urgent clinical decisions. A&E flow is slowed because attenders wait for scans before a bed decision can be made. Inpatient stays are extended unnecessarily. Radiographers, radiologists, imaging coordinators, and interventional teams absorb the administrative burden of managing cancellations manually. NHS Trusts lose capacity they have already paid for without any mechanism to recover it.
This research sets out to address that problem directly. The researcher works as an Interventional Radiology and Biopsy Coordinator at Medway Maritime Hospital, with oversight of imaging workflows across inpatient and outpatient pathways and direct responsibility for coordinating interventional procedures, mainly image-guided biopsies. That operational position provides daily, embedded access to the systems in which the problem manifests, and direct working relationships with the radiographers, radiologists, nursing staff, and administrators whose cooperation any solution would depend on. It also provides direct visibility of the interface between diagnostic imaging and interventional radiology; a junction that is consistently affected by cancellations and is rarely treated as a single operational unit in existing scheduling research.
The proposed solution is a co-designed, AI-assisted slot optimisation system that, when a cancellation or did-not-attend (DNA) occurs, surfaces clinically appropriate inpatients or A&E attenders who are already prepped or can be prepped in time, presents them to staff for confirmation, and supports rapid reallocation of the slot. A modest expansion of appointment time windows is built in to allow for transport, prep variability, and operational flexibility, without disrupting the underlying schedule. The intended outcomes are reduced unused imaging slots, improved patient flow, and measurable support for A&E throughput and bed capacity.
The immediate research focus is bounded to four service areas where the problem is most pronounced and the operational case for intervention is strongest: MRI, CT, ultrasound, and image-guided biopsies, for inpatients and A&E attenders.
This proposal describes the design, pilot evaluation, and framework derivation of that system, built through embedded, low-burden co-design with frontline staff and tested in a live NHS environment at Medway Maritime Hospital. It is positioned with strategic awareness of the broader operational context, in particular the formation of the Medway and Dartford & Gravesham NHS Trust group, which creates a credible pathway for tested, validated solutions to scale across sites once the pilot evidence is in place.
2. Literature Review
2. Literature Review
Existing research on NHS scheduling has tended to focus on planning problems: how to build better appointment templates, how to predict did-not-attend rates, or how to manage outpatient waiting lists more efficiently (Cayirli and Veral, 2003; Ahmadi-Javid et al., 2017). These are valuable contributions, but they are not the question being asked here. This research is concerned with recovery: what happens after a cancellation occurs, and how quickly and accurately a department can respond.
Studies by NHS England and the Health Foundation have documented the scale of wasted capacity across NHS services, with significant numbers of appointment slots lost each year to late cancellations and DNAs (NHS England, 2023; Health Foundation, 2022). In radiology and interventional radiology specifically, published research on the root causes of cancellation and on the feasibility of real-time slot recovery for inpatients and A&E attenders is sparse. Work on AI-assisted scheduling in healthcare (Srinivas and Ravindran, 2018) has explored optimisation approaches in outpatient contexts, but those studies typically assume a planning horizon of days or weeks. They do not address the same-day or next-hour recovery problem that imaging coordinators face when an inpatient is cancelled an hour before their scan, or when a biopsy slot opens up unexpectedly.
Co-design and participatory health informatics literature (Greenhalgh et al., 2017) emphasises the importance of grounding digital tools in the realities of clinical workflow and the lived experience of frontline staff. However, much of this literature describes consultative engagement rather than embedded, practitioner-led research. A study conducted from within the operational role being studied is methodologically distinctive and addresses a recurring weakness in health IT evaluation: the gap between design assumptions and ward-level reality.
The integration of patient preparation status, clinical urgency, mobility data, infection status, and referral completeness into a real-time matching tool — covering MRI, CT, ultrasound, and image-guided biopsy pathways simultaneously — has not, to the researcher's knowledge, been studied in a live NHS Trust environment. That is the gap this research fills.
Existing research on NHS scheduling has tended to focus on planning problems: how to build better appointment templates, how to predict did-not-attend rates, or how to manage outpatient waiting lists more efficiently (Cayirli and Veral, 2003; Ahmadi-Javid et al., 2017). These are valuable contributions, but they are not the question being asked here. This research is concerned with recovery: what happens after a cancellation occurs, and how quickly and accurately a department can respond.
Studies by NHS England and the Health Foundation have documented the scale of wasted capacity across NHS services, with significant numbers of appointment slots lost each year to late cancellations and DNAs (NHS England, 2023; Health Foundation, 2022). In radiology and interventional radiology specifically, published research on the root causes of cancellation and on the feasibility of real-time slot recovery for inpatients and A&E attenders is sparse. Work on AI-assisted scheduling in healthcare (Srinivas and Ravindran, 2018) has explored optimisation approaches in outpatient contexts, but those studies typically assume a planning horizon of days or weeks. They do not address the same-day or next-hour recovery problem that imaging coordinators face when an inpatient is cancelled an hour before their scan, or when a biopsy slot opens up unexpectedly.
Co-design and participatory health informatics literature (Greenhalgh et al., 2017) emphasises the importance of grounding digital tools in the realities of clinical workflow and the lived experience of frontline staff. However, much of this literature describes consultative engagement rather than embedded, practitioner-led research. A study conducted from within the operational role being studied is methodologically distinctive and addresses a recurring weakness in health IT evaluation: the gap between design assumptions and ward-level reality.
The integration of patient preparation status, clinical urgency, mobility data, infection status, and referral completeness into a real-time matching tool — covering MRI, CT, ultrasound, and image-guided biopsy pathways simultaneously — has not, to the researcher's knowledge, been studied in a live NHS Trust environment. That is the gap this research fills.
3. Significance of the Research
3. Significance
The NHS is under sustained pressure to deliver more from existing resources. Waiting times for diagnostic imaging and interventional procedures have increased considerably in recent years, and there is growing recognition that recovery will depend not only on additional capacity but on using existing capacity more effectively. Wasted radiology and biopsy slots are a concrete, measurable example of capacity that already exists but is not being fully used. This research addresses that problem head-on.
What makes this project distinctive is not just the problem it tackles but how it tackles it. Most health IT research either develops technology without adequate grounding in clinical workflows, or studies clinical workflows without producing anything usable. This project does both, because the researcher occupies both positions simultaneously. Conducting the research from within the role of Interventional Radiology and Biopsy Coordinator means the system being designed is shaped by the same operational realities it will need to work within. Co-design — kept deliberately light-touch and asynchronous to respect staff time — reinforces this further, ensuring that radiographers, imaging administrators, and A&E clinicians are not consulted at the end of the process but are central to it from the start.
The contribution is also generalisable in a meaningful way. A validated implementation framework, derived from a real pilot in a real NHS Trust and tested against the governance, interoperability, and adoption constraints that make health IT projects fail, gives other Trusts something genuinely usable. It is not a proof of concept; it is a tested model, with documented conditions for success and an honest account of where the difficulties lie. The strategic context, the formation of the Medway and Dartford & Gravesham NHS Trust group, provides a credible scale-up pathway beyond the pilot site, should the evidence support it.
Beyond the immediate operational benefits, this research matters because the problem it addresses is only going to grow. As demand on NHS imaging and interventional services increases, the cost of each wasted slot rises. A system that can reliably recover even a proportion of those slots, replicated across multiple trusts, would have a meaningful cumulative effect on patient access, A&E flow, clinical decision-making, and the long-term sustainability of NHS imaging services.
The NHS is under sustained pressure to deliver more from existing resources. Waiting times for diagnostic imaging and interventional procedures have increased considerably in recent years, and there is growing recognition that recovery will depend not only on additional capacity but on using existing capacity more effectively. Wasted radiology and biopsy slots are a concrete, measurable example of capacity that already exists but is not being fully used. This research addresses that problem head-on.
What makes this project distinctive is not just the problem it tackles but how it tackles it. Most health IT research either develops technology without adequate grounding in clinical workflows, or studies clinical workflows without producing anything usable. This project does both, because the researcher occupies both positions simultaneously. Conducting the research from within the role of Interventional Radiology and Biopsy Coordinator means the system being designed is shaped by the same operational realities it will need to work within. Co-design — kept deliberately light-touch and asynchronous to respect staff time — reinforces this further, ensuring that radiographers, imaging administrators, and A&E clinicians are not consulted at the end of the process but are central to it from the start.
The contribution is also generalisable in a meaningful way. A validated implementation framework, derived from a real pilot in a real NHS Trust and tested against the governance, interoperability, and adoption constraints that make health IT projects fail, gives other Trusts something genuinely usable. It is not a proof of concept; it is a tested model, with documented conditions for success and an honest account of where the difficulties lie. The strategic context, the formation of the Medway and Dartford & Gravesham NHS Trust group, provides a credible scale-up pathway beyond the pilot site, should the evidence support it.
Beyond the immediate operational benefits, this research matters because the problem it addresses is only going to grow. As demand on NHS imaging and interventional services increases, the cost of each wasted slot rises. A system that can reliably recover even a proportion of those slots, replicated across multiple trusts, would have a meaningful cumulative effect on patient access, A&E flow, clinical decision-making, and the long-term sustainability of NHS imaging services.
4. Aim, Questions + Objectives
4. Aim, Questions + Objectives
Aim
To design, pilot, and evaluate a co-designed, AI-assisted slot optimisation system for acute imaging (MRI, CT, ultrasound) and interventional radiology biopsy services in an NHS Trust, and to derive a replicable implementation framework that supports its phased adoption across patient groups and trusts.
Primary Research Question
Can a co-designed, AI-assisted slot optimisation system reduce unused imaging capacity for inpatients and A&E attendees in an NHS Trust, and what implementation framework supports its phased adoption across patient groups and Trusts?
Sub-Questions
RQ1: What are the frequency, causes, and recoverability of unplanned cancellations and DNAs in MRI, CT, ultrasound, and image-guided biopsy services for inpatients and A&E attenders at Medway Maritime Hospital?
RQ2: What data collection methods make it possible to generate the quantitative and qualitative evidence this study requires while respecting the workload pressures on frontline staff at Medway, and what does an effective low-burden, embedded research design look like in this operational context?
RQ3: What patient-matching criteria, preparedness indicators, and user-experience design principles are necessary for a clinically safe and operationally usable slot optimisation system?
RQ4: What measurable effects does the system have on slot utilisation, DNA recovery, time-to-procedure, and A&E waiting time and patient flow indicators, and how do these effects compare against the pre-intervention baseline at Medway?
RQ4: Under what governance, technical, and organisational conditions can such a system scale beyond the pilot site to additional patient groups (GP referrals, outpatients, broader interventional radiology procedures)?
Objectives
Produce a validated taxonomy of cancellation and DNA causes across MRI, CT, ultrasound, and image-guided biopsy services, alongside a quantitative baseline of slot loss at Medway.
Co-design and specify a slot optimisation system that meets clinical safety, usability, and NHS interoperability requirements, using a low-burden engagement model that respects staff workload.
Evaluate the system through a live pilot, measuring slot utilisation, DNA recovery, time-to-procedure, and impacts on A&E waiting times and patient flow.
Derive a replicable implementation framework documenting the governance, technical, and organisational conditions under which the system can be adopted by other Trusts.
Scope future expansion of the framework to additional interventional radiology procedures and outpatient/GP referral pathways, without committing to deployment within the PhD timeline.
Aim
To design, pilot, and evaluate a co-designed, AI-assisted slot optimisation system for acute imaging (MRI, CT, ultrasound) and interventional radiology biopsy services in an NHS Trust, and to derive a replicable implementation framework that supports its phased adoption across patient groups and trusts.
Primary Research Question
Can a co-designed, AI-assisted slot optimisation system reduce unused imaging capacity for inpatients and A&E attendees in an NHS Trust, and what implementation framework supports its phased adoption across patient groups and Trusts?
Sub-Questions
RQ1: What are the frequency, causes, and recoverability of unplanned cancellations and DNAs in MRI, CT, ultrasound, and image-guided biopsy services for inpatients and A&E attenders at Medway Maritime Hospital?
RQ2: What data collection methods make it possible to generate the quantitative and qualitative evidence this study requires while respecting the workload pressures on frontline staff at Medway, and what does an effective low-burden, embedded research design look like in this operational context?
RQ3: What patient-matching criteria, preparedness indicators, and user-experience design principles are necessary for a clinically safe and operationally usable slot optimisation system?
RQ4: What measurable effects does the system have on slot utilisation, DNA recovery, time-to-procedure, and A&E waiting time and patient flow indicators, and how do these effects compare against the pre-intervention baseline at Medway?
RQ4: Under what governance, technical, and organisational conditions can such a system scale beyond the pilot site to additional patient groups (GP referrals, outpatients, broader interventional radiology procedures)?
Objectives
Produce a validated taxonomy of cancellation and DNA causes across MRI, CT, ultrasound, and image-guided biopsy services, alongside a quantitative baseline of slot loss at Medway.
Co-design and specify a slot optimisation system that meets clinical safety, usability, and NHS interoperability requirements, using a low-burden engagement model that respects staff workload.
Evaluate the system through a live pilot, measuring slot utilisation, DNA recovery, time-to-procedure, and impacts on A&E waiting times and patient flow.
Derive a replicable implementation framework documenting the governance, technical, and organisational conditions under which the system can be adopted by other Trusts.
Scope future expansion of the framework to additional interventional radiology procedures and outpatient/GP referral pathways, without committing to deployment within the PhD timeline.
5. Methodology
5. Methodology
This study uses a mixed-methods design combining quantitative data analysis, qualitative co-design research, prospective pilot evaluation, and external stakeholder engagement across four overlapping phases. The researcher's daily operational role at Medway, with oversight of imaging and biopsy workflows for inpatient and outpatient pathways, enables phases to run concurrently; which is what makes a four-year part-time timeline realistic.
A central methodological commitment of this study is that data collection should respect the workload of NHS staff. Frontline radiographers, radiologists, nurses, and coordinators are already operating under sustained pressure, and any research that requires substantial time commitments from them risks both poor participation and ethical strain. This proposal therefore prioritises asynchronous, low-burden data collection methods — short targeted email questionnaires, brief embedded feedback prompts, observational data drawn from routine workflow, and analysis of existing system logs — over time-intensive workshops or extended interviews. Workshops are limited to one or two short, focused sessions later in the project, used primarily for validation rather than generation. This is not a methodological compromise; it is a deliberate design choice intended to make the research feasible, ethical, and reflective of the operational environment it is studying.
Software systems referenced throughout this proposal such as the Trust's Radiology Information System (RIS), scheduling systems, and patient tracking tools will be confirmed and documented at registration stage. The methodology is intentionally platform-agnostic so that the framework derived in later phases is not dependent on any one vendor.
Phase 1: Baseline Audit (Quantitative)
A retrospective analysis of cancellation, DNA, and slot utilisation data drawn from the Trust's RIS and associated scheduling and patient tracking tools will establish a quantitative baseline across MRI, CT, ultrasound, and image-guided biopsy services. This phase produces a cancellation taxonomy that classifies reasons by type, frequency, modality, time of day, and patient pathway, and calculates the proportion of cancelled slots that were potentially recoverable — meaning cases where a clinically suitable patient existed on the waiting list at the time of cancellation but was not contacted in time.
Quantitative metrics in this phase include:
Slot utilisation rates by modality and patient pathway (inpatient, A&E, with outpatient included as comparator).
Unplanned cancellation and DNA rates, classified by cause.
Time from request to scan or procedure, by modality and urgency category.
Proportion of cancelled slots that were potentially recoverable.
Number of successfully reallocated cancelled slots under current manual practice — establishing the existing recovery rate against which the system will later be benchmarked.
Data sources include RIS exports, scheduling system records, departmental rotas, scanner downtime logs, A&E referral records, and patient tracking tools. Because this phase analyses retrospective, de-identified operational data rather than identifiable clinical records, it is anticipated that ethical evaluation and approval via the NHS Health Research Authority's service evaluation pathway will be appropriate, subject to confirmation with the Trust's Research and Development department.
Phase 2: Co-Design and System Specification (Qualitative)
Co-design in this study is deliberately structured to minimise demands on staff time. Engagement with key stakeholder groups — radiographers, imaging administrators, interventional radiology staff, A&E nurses and consultants, and radiology management — will primarily take the form of:
Short, targeted email questionnaires designed to elicit specific design inputs (e.g., matching criteria, preparation indicators, and workflow integration concerns).
Asynchronous feedback collection on draft scenarios, mock-ups, and decision rules, distributed via email with clear, time-bounded response windows.
Light-touch observational fieldwork drawn from the researcher’s day-to-day role, with reflective notes captured systematically.
Semi-structured interviews with key stakeholders (where availability permits), including radiology staff, coordinators, and relevant clinical teams. These will be focused, conducted opportunistically to minimise disruption and used to deepen insights gathered from questionnaires and observations.
One to two short validation workshops, scheduled later in the phase, to confirm that the synthesised requirements accurately reflect departmental needs before the specification is handed to a development team.
Thematic analysis will be applied to the combined qualitative data, with framework synthesis used to translate findings into a structured system requirements specification. The outputs of this phase are a validated system requirements specification and a co-designed interface prototype, which will be handed to an IT development team for build. The researcher's role within the PhD is design, specification, and evaluation; the technical build is carried out by others.
Phase 3: Pilot Evaluation (Mixed Methods)
Following information governance approvals and system build, the tool will be deployed in a live environment at Medway across the in-scope services (MRI, CT, ultrasound, and image-guided biopsy) for inpatients and A&E attenders. Prospective data will be collected and compared against the Phase 1 baseline.
Quantitative outcome metrics:
Slot utilisation rate, before vs after intervention, by modality and patient pathway.
DNA and cancellation recovery rate (proportion of lost slots successfully backfilled).
Time from request to scan or procedure, before vs after.
A&E waiting time impact for attenders requiring imaging or biopsy.
Bed occupancy and patient flow indicators, where access is available through Trust operational dashboards.
Number of successfully reallocated cancelled slots under the new system.
Operational metrics:
Time required to identify and prepare a replacement patient when a cancellation occurs.
Net workflow effect — whether the system reduces or increases coordinator and radiographer effort overall.
Staff coordination requirements introduced by the system (e.g., cross-team confirmations, porter requests).
Frequency and nature of overrides — where staff decline a system suggestion and why.
Qualitative data:
Brief, periodic email questionnaires capturing staff feedback on feasibility, usability, and perceived workload.
Asynchronous structured feedback at defined milestones during the pilot.
Patient experience data collected via existing departmental feedback channels, where this can be done without introducing additional patient burden.
Quantitative data will be analysed using descriptive statistics and appropriate inferential techniques (interrupted time-series analysis or pre/post comparison, depending on data volume). Qualitative data will be analysed thematically. Mixed-methods integration will follow a triangulation approach, ensuring that operational metrics, system data, and staff and patient experience are read together rather than in isolation.
Phase 4: External Stakeholder Engagement and Framework Derivation
The fourth phase moves beyond the Medway pilot to examine whether the findings and the system design are transferable to other NHS environments — including, in particular, Dartford & Gravesham as a strategic group partner — and other Trusts. This phase does not involve deploying the system at additional sites within the PhD scope; it involves structured engagement with stakeholders from other NHS radiology and interventional radiology departments to test the assumptions and conditions that the Medway pilot produces.
Engagement will use the same low-burden methods established in Phase 2: short questionnaires, asynchronous feedback on the draft framework, and a small number of focused validation conversations with imaging coordinators, radiology managers, and informatics leads from Trusts of varying sizes, demographics, RIS platforms, and digital maturity. The aim is to surface where the implementation conditions identified at Medway are generalisable and where they are context-specific; to understand the barriers other Trusts would face in adoption; and to identify what would need to be different about the system or the rollout in those environments.
The outputs of this phase feed directly into the final implementation framework. Rather than a framework derived solely from one pilot site, the finished model reflects a broader picture of NHS operational reality, stress-tested against the perspectives of departments that were not involved in building the system. This gives the framework greater credibility and practical utility as a resource for adoption across the Medway and Dartford & Gravesham group and beyond.
Framework synthesis will be applied across all four phases to produce a structured, evidence-based implementation guide covering governance requirements, technical integration considerations, staff adoption conditions, and a phased expansion pathway from inpatients and A&E attenders through to GP referrals, outpatients, and broader interventional radiology procedures.
This study uses a mixed-methods design combining quantitative data analysis, qualitative co-design research, prospective pilot evaluation, and external stakeholder engagement across four overlapping phases. The researcher's daily operational role at Medway, with oversight of imaging and biopsy workflows for inpatient and outpatient pathways, enables phases to run concurrently; which is what makes a four-year part-time timeline realistic.
A central methodological commitment of this study is that data collection should respect the workload of NHS staff. Frontline radiographers, radiologists, nurses, and coordinators are already operating under sustained pressure, and any research that requires substantial time commitments from them risks both poor participation and ethical strain. This proposal therefore prioritises asynchronous, low-burden data collection methods — short targeted email questionnaires, brief embedded feedback prompts, observational data drawn from routine workflow, and analysis of existing system logs — over time-intensive workshops or extended interviews. Workshops are limited to one or two short, focused sessions later in the project, used primarily for validation rather than generation. This is not a methodological compromise; it is a deliberate design choice intended to make the research feasible, ethical, and reflective of the operational environment it is studying.
Software systems referenced throughout this proposal such as the Trust's Radiology Information System (RIS), scheduling systems, and patient tracking tools will be confirmed and documented at registration stage. The methodology is intentionally platform-agnostic so that the framework derived in later phases is not dependent on any one vendor.
Phase 1: Baseline Audit (Quantitative)
A retrospective analysis of cancellation, DNA, and slot utilisation data drawn from the Trust's RIS and associated scheduling and patient tracking tools will establish a quantitative baseline across MRI, CT, ultrasound, and image-guided biopsy services. This phase produces a cancellation taxonomy that classifies reasons by type, frequency, modality, time of day, and patient pathway, and calculates the proportion of cancelled slots that were potentially recoverable — meaning cases where a clinically suitable patient existed on the waiting list at the time of cancellation but was not contacted in time.
Quantitative metrics in this phase include:
Slot utilisation rates by modality and patient pathway (inpatient, A&E, with outpatient included as comparator).
Unplanned cancellation and DNA rates, classified by cause.
Time from request to scan or procedure, by modality and urgency category.
Proportion of cancelled slots that were potentially recoverable.
Number of successfully reallocated cancelled slots under current manual practice — establishing the existing recovery rate against which the system will later be benchmarked.
Data sources include RIS exports, scheduling system records, departmental rotas, scanner downtime logs, A&E referral records, and patient tracking tools. Because this phase analyses retrospective, de-identified operational data rather than identifiable clinical records, it is anticipated that ethical evaluation and approval via the NHS Health Research Authority's service evaluation pathway will be appropriate, subject to confirmation with the Trust's Research and Development department.
Phase 2: Co-Design and System Specification (Qualitative)
Co-design in this study is deliberately structured to minimise demands on staff time. Engagement with key stakeholder groups — radiographers, imaging administrators, interventional radiology staff, A&E nurses and consultants, and radiology management — will primarily take the form of:
Short, targeted email questionnaires designed to elicit specific design inputs (e.g., matching criteria, preparation indicators, and workflow integration concerns).
Asynchronous feedback collection on draft scenarios, mock-ups, and decision rules, distributed via email with clear, time-bounded response windows.
Light-touch observational fieldwork drawn from the researcher’s day-to-day role, with reflective notes captured systematically.
Semi-structured interviews with key stakeholders (where availability permits), including radiology staff, coordinators, and relevant clinical teams. These will be focused, conducted opportunistically to minimise disruption and used to deepen insights gathered from questionnaires and observations.
One to two short validation workshops, scheduled later in the phase, to confirm that the synthesised requirements accurately reflect departmental needs before the specification is handed to a development team.
Thematic analysis will be applied to the combined qualitative data, with framework synthesis used to translate findings into a structured system requirements specification. The outputs of this phase are a validated system requirements specification and a co-designed interface prototype, which will be handed to an IT development team for build. The researcher's role within the PhD is design, specification, and evaluation; the technical build is carried out by others.
Phase 3: Pilot Evaluation (Mixed Methods)
Following information governance approvals and system build, the tool will be deployed in a live environment at Medway across the in-scope services (MRI, CT, ultrasound, and image-guided biopsy) for inpatients and A&E attenders. Prospective data will be collected and compared against the Phase 1 baseline.
Quantitative outcome metrics:
Slot utilisation rate, before vs after intervention, by modality and patient pathway.
DNA and cancellation recovery rate (proportion of lost slots successfully backfilled).
Time from request to scan or procedure, before vs after.
A&E waiting time impact for attenders requiring imaging or biopsy.
Bed occupancy and patient flow indicators, where access is available through Trust operational dashboards.
Number of successfully reallocated cancelled slots under the new system.
Operational metrics:
Time required to identify and prepare a replacement patient when a cancellation occurs.
Net workflow effect — whether the system reduces or increases coordinator and radiographer effort overall.
Staff coordination requirements introduced by the system (e.g., cross-team confirmations, porter requests).
Frequency and nature of overrides — where staff decline a system suggestion and why.
Qualitative data:
Brief, periodic email questionnaires capturing staff feedback on feasibility, usability, and perceived workload.
Asynchronous structured feedback at defined milestones during the pilot.
Patient experience data collected via existing departmental feedback channels, where this can be done without introducing additional patient burden.
Quantitative data will be analysed using descriptive statistics and appropriate inferential techniques (interrupted time-series analysis or pre/post comparison, depending on data volume). Qualitative data will be analysed thematically. Mixed-methods integration will follow a triangulation approach, ensuring that operational metrics, system data, and staff and patient experience are read together rather than in isolation.
Phase 4: External Stakeholder Engagement and Framework Derivation
The fourth phase moves beyond the Medway pilot to examine whether the findings and the system design are transferable to other NHS environments — including, in particular, Dartford & Gravesham as a strategic group partner — and other Trusts. This phase does not involve deploying the system at additional sites within the PhD scope; it involves structured engagement with stakeholders from other NHS radiology and interventional radiology departments to test the assumptions and conditions that the Medway pilot produces.
Engagement will use the same low-burden methods established in Phase 2: short questionnaires, asynchronous feedback on the draft framework, and a small number of focused validation conversations with imaging coordinators, radiology managers, and informatics leads from Trusts of varying sizes, demographics, RIS platforms, and digital maturity. The aim is to surface where the implementation conditions identified at Medway are generalisable and where they are context-specific; to understand the barriers other Trusts would face in adoption; and to identify what would need to be different about the system or the rollout in those environments.
The outputs of this phase feed directly into the final implementation framework. Rather than a framework derived solely from one pilot site, the finished model reflects a broader picture of NHS operational reality, stress-tested against the perspectives of departments that were not involved in building the system. This gives the framework greater credibility and practical utility as a resource for adoption across the Medway and Dartford & Gravesham group and beyond.
Framework synthesis will be applied across all four phases to produce a structured, evidence-based implementation guide covering governance requirements, technical integration considerations, staff adoption conditions, and a phased expansion pathway from inpatients and A&E attenders through to GP referrals, outpatients, and broader interventional radiology procedures.
6. Ethical Considerations
6. Ethical Considerations
Several challenges have been identified, with explicit mitigation strategies for each.
Clinical Safety
The most important consideration in any system that influences patient scheduling is clinical safety. Patients must be appropriately prepared before any slot reassignment occurs; clinical appropriateness for the modality, contrast use, and procedure type must be confirmed; and no element of planned patient care may be compromised by the operational gain from filling a slot. These safeguards are built into the system design through a strict human-in-the-loop model: every backfill suggestion produced by the system must be reviewed and confirmed by a qualified member of staff before any booking is made or any patient is contacted. The system supports clinical decision-making; it does not replace it. Clinical safety boundaries — including which patients may never be auto-suggested for a slot, and which prep states must be verified directly rather than inferred — will be specified explicitly during co-design and documented within the system requirements.
Information Governance and Data Access
Access to RIS and scheduling data requires Trust information governance approval. The researcher's existing employment at Medway and familiarity with the Trust's R&D processes provide a credible pathway. Phase 1 is anticipated to qualify as service evaluation; Phases 2 and 3 will require NHS Research Ethics Committee review.
Staff Adoption and Workload Sensitivity
Automated tools in clinical environments frequently encounter resistance, particularly where staff feel their professional judgement is being bypassed or their time imposed upon. The light-touch co-design methodology, the human-in-the-loop design, and the explicit decision to use asynchronous and email-based engagement methods are all intended to address this. The research process itself models the principle that staff time is a constrained, valuable resource.
Part-Time Timeline
Conducting a four-year part-time PhD alongside full-time employment carries real workload pressures, particularly in Year 4 where analysis and write-up converge. This is mitigated by writing in parallel from Year 2 onwards and by the researcher's embedded position at Medway, which removes the access delays that typically extend this kind of research for external candidates.
Scope Management
The phased expansion model is designed to keep the research focused. The PhD covers design through to framework derivation, with the live pilot scoped to MRI, CT, ultrasound, and image-guided biopsies for inpatients and A&E attenders. Broader interventional radiology procedures, GP referrals, and outpatient pathways are acknowledged as future expansion targets within the framework, but are not committed deployment work within the PhD timeline.
Several challenges have been identified, with explicit mitigation strategies for each.
Clinical Safety
The most important consideration in any system that influences patient scheduling is clinical safety. Patients must be appropriately prepared before any slot reassignment occurs; clinical appropriateness for the modality, contrast use, and procedure type must be confirmed; and no element of planned patient care may be compromised by the operational gain from filling a slot. These safeguards are built into the system design through a strict human-in-the-loop model: every backfill suggestion produced by the system must be reviewed and confirmed by a qualified member of staff before any booking is made or any patient is contacted. The system supports clinical decision-making; it does not replace it. Clinical safety boundaries — including which patients may never be auto-suggested for a slot, and which prep states must be verified directly rather than inferred — will be specified explicitly during co-design and documented within the system requirements.
Information Governance and Data Access
Access to RIS and scheduling data requires Trust information governance approval. The researcher's existing employment at Medway and familiarity with the Trust's R&D processes provide a credible pathway. Phase 1 is anticipated to qualify as service evaluation; Phases 2 and 3 will require NHS Research Ethics Committee review.
Staff Adoption and Workload Sensitivity
Automated tools in clinical environments frequently encounter resistance, particularly where staff feel their professional judgement is being bypassed or their time imposed upon. The light-touch co-design methodology, the human-in-the-loop design, and the explicit decision to use asynchronous and email-based engagement methods are all intended to address this. The research process itself models the principle that staff time is a constrained, valuable resource.
Part-Time Timeline
Conducting a four-year part-time PhD alongside full-time employment carries real workload pressures, particularly in Year 4 where analysis and write-up converge. This is mitigated by writing in parallel from Year 2 onwards and by the researcher's embedded position at Medway, which removes the access delays that typically extend this kind of research for external candidates.
Scope Management
The phased expansion model is designed to keep the research focused. The PhD covers design through to framework derivation, with the live pilot scoped to MRI, CT, ultrasound, and image-guided biopsies for inpatients and A&E attenders. Broader interventional radiology procedures, GP referrals, and outpatient pathways are acknowledged as future expansion targets within the framework, but are not committed deployment work within the PhD timeline.
7. Impact
7. Impact
The most immediate impact of this research is straightforward: patients at Medway are imaged or biopsied sooner. When a slot becomes available at short notice and the system identifies a suitable patient to fill it, that is a patient who receives their imaging or procedure earlier than they otherwise would. For inpatients awaiting scans that inform treatment decisions, for A&E attenders who need urgent imaging to determine their care pathway, and for patients waiting for image-guided biopsies, that difference in timing can be clinically significant. Improved A&E imaging turnaround also has direct implications for emergency department flow and bed capacity — pressures that radiate well beyond radiology.
Beyond individual patient outcomes, this research addresses a problem that is common across NHS imaging and interventional radiology departments but currently has no standard solution. Coordinators across the country are managing cancellations manually, relying on experience, informal knowledge of waiting lists, and phone calls to identify replacement patients. That approach is slow, inconsistent, and dependent on individual staff knowledge that leaves departments when people move on. A validated, evidence-based alternative gives departments a better tool and gives Trusts a model they can implement, with documented evidence of where it works and what it requires.
The strategic context strengthens this further. With Medway and Dartford & Gravesham forming a Trust group, a successful pilot at Medway has a credible pathway to wider adoption across sites that share operational challenges. The implementation framework produced by this study is designed from the start to support that kind of phased scale-up, with documented conditions for success, an honest account of the governance and interoperability requirements involved, and a structured pathway for incremental expansion across patient groups and procedures.
The longer-term significance is about the sustainability of NHS imaging and interventional services. Demand is growing, and the gap between capacity and demand will not close through new scanners and additional staff alone. Making better use of existing capacity, by recovering slots that would otherwise go to waste, is part of how these services remain viable as that demand increases. This research contributes a practical, tested approach to that goal — grounded in the realities of how NHS departments actually work.
The most immediate impact of this research is straightforward: patients at Medway are imaged or biopsied sooner. When a slot becomes available at short notice and the system identifies a suitable patient to fill it, that is a patient who receives their imaging or procedure earlier than they otherwise would. For inpatients awaiting scans that inform treatment decisions, for A&E attenders who need urgent imaging to determine their care pathway, and for patients waiting for image-guided biopsies, that difference in timing can be clinically significant. Improved A&E imaging turnaround also has direct implications for emergency department flow and bed capacity — pressures that radiate well beyond radiology.
Beyond individual patient outcomes, this research addresses a problem that is common across NHS imaging and interventional radiology departments but currently has no standard solution. Coordinators across the country are managing cancellations manually, relying on experience, informal knowledge of waiting lists, and phone calls to identify replacement patients. That approach is slow, inconsistent, and dependent on individual staff knowledge that leaves departments when people move on. A validated, evidence-based alternative gives departments a better tool and gives Trusts a model they can implement, with documented evidence of where it works and what it requires.
The strategic context strengthens this further. With Medway and Dartford & Gravesham forming a Trust group, a successful pilot at Medway has a credible pathway to wider adoption across sites that share operational challenges. The implementation framework produced by this study is designed from the start to support that kind of phased scale-up, with documented conditions for success, an honest account of the governance and interoperability requirements involved, and a structured pathway for incremental expansion across patient groups and procedures.
The longer-term significance is about the sustainability of NHS imaging and interventional services. Demand is growing, and the gap between capacity and demand will not close through new scanners and additional staff alone. Making better use of existing capacity, by recovering slots that would otherwise go to waste, is part of how these services remain viable as that demand increases. This research contributes a practical, tested approach to that goal — grounded in the realities of how NHS departments actually work.
8. Contribution to Knowledge
8. Contribution to Knowledge
This research makes three principal contributions to the field of clinical informatics and NHS operational research.
First, it produces an empirically grounded taxonomy of cancellation and DNA causes across acute imaging and image-guided biopsy services in an NHS Trust — a baseline that does not currently exist in the published literature in this combined form. By treating diagnostic imaging and image-guided interventional radiology as a single operational unit rather than separate scheduling silos, the taxonomy reflects how these services actually function in practice and provides a foundation for future research on slot loss in acute settings.
Second, it produces a co-designed system specification and validated pilot evidence for a clinically safe, operationally usable approach to real-time slot optimisation for inpatients and A&E attenders. This addresses a gap that existing scheduling research, focused on planning rather than recovery, has not filled. The contribution is methodological as well as practical: the use of embedded, low-burden co-design conducted from within the operational role being studied offers a replicable model for health informatics research that prioritises ecological validity and respects the workload constraints of NHS staff.
Third, it produces a stress-tested implementation framework that documents the governance, technical, and organisational conditions under which such a system can be adopted by NHS Trusts, with explicit pathways for phased expansion across patient groups and procedures. By incorporating external stakeholder engagement from Trusts beyond the pilot site, the framework moves beyond single-site findings to offer a tool with broader transferability — particularly relevant in the context of NHS Trust group formations such as Medway and Dartford & Gravesham.
Together, these contributions translate operational knowledge currently held informally by clinical coordinators into a structured, transferable evidence base, and offer a worked example of how practitioner-led clinical informatics research can produce outputs that are usable as well as publishable.
This research makes three principal contributions to the field of clinical informatics and NHS operational research.
First, it produces an empirically grounded taxonomy of cancellation and DNA causes across acute imaging and image-guided biopsy services in an NHS Trust — a baseline that does not currently exist in the published literature in this combined form. By treating diagnostic imaging and image-guided interventional radiology as a single operational unit rather than separate scheduling silos, the taxonomy reflects how these services actually function in practice and provides a foundation for future research on slot loss in acute settings.
Second, it produces a co-designed system specification and validated pilot evidence for a clinically safe, operationally usable approach to real-time slot optimisation for inpatients and A&E attenders. This addresses a gap that existing scheduling research, focused on planning rather than recovery, has not filled. The contribution is methodological as well as practical: the use of embedded, low-burden co-design conducted from within the operational role being studied offers a replicable model for health informatics research that prioritises ecological validity and respects the workload constraints of NHS staff.
Third, it produces a stress-tested implementation framework that documents the governance, technical, and organisational conditions under which such a system can be adopted by NHS Trusts, with explicit pathways for phased expansion across patient groups and procedures. By incorporating external stakeholder engagement from Trusts beyond the pilot site, the framework moves beyond single-site findings to offer a tool with broader transferability — particularly relevant in the context of NHS Trust group formations such as Medway and Dartford & Gravesham.
Together, these contributions translate operational knowledge currently held informally by clinical coordinators into a structured, transferable evidence base, and offer a worked example of how practitioner-led clinical informatics research can produce outputs that are usable as well as publishable.
9. Indicative Timeline
9. Indicative Timeline
The timeline below is proposed for a part-time PhD over four years. The central compression strategy is running the literature review, ethics applications, and baseline audit concurrently in Year 1, which is made possible by the researcher's daily access to RIS and scheduling data at Medway. Writing begins in Year 2 and continues alongside each subsequent phase, so that Year 4 is focused on analysis and completion rather than starting the thesis from scratch.
The timeline below is proposed for a part-time PhD over four years. The central compression strategy is running the literature review, ethics applications, and baseline audit concurrently in Year 1, which is made possible by the researcher's daily access to RIS and scheduling data at Medway. Writing begins in Year 2 and continues alongside each subsequent phase, so that Year 4 is focused on analysis and completion rather than starting the thesis from scratch.
Year
Year
Year 1
Year 1
Year 2
Year 2
Year 3
Year 3
Year 4
Year 4
Key Activities
Key Activities
Literature review and ethics/IG applications running concurrently with the live RIS and scheduling-system audit. Daily operational access enables formal documentation and analysis to begin immediately, without the access delays that typically slow this phase for external researchers.
Literature review and ethics/IG applications running concurrently with the live RIS and scheduling-system audit. Daily operational access enables formal documentation and analysis to begin immediately, without the access delays that typically slow this phase for external researchers.
Cancellation taxonomy finalised. Light-touch co-design with radiographers, IR staff, imaging administrators, and A&E nursing staff conducted via short email questionnaires and asynchronous feedback. System requirements specification completed and prototype handed to the IT development team by year end. Thesis chapters 1 and 2 drafted in parallel.
Cancellation taxonomy finalised. Light-touch co-design with radiographers, IR staff, imaging administrators, and A&E nursing staff conducted via short email questionnaires and asynchronous feedback. System requirements specification completed and prototype handed to the IT development team by year end. Thesis chapters 1 and 2 drafted in parallel.
Usability validation (one to two short workshops). IG approvals for live pilot. Pilot deployment at Medway across MRI, CT, ultrasound, and image-guided biopsy services. Prospective data collection on slot utilisation, DNA recovery, A&E waiting time impact, and patient flow. Periodic email-based staff feedback throughout. Thesis chapters 3 and 4 drafted in parallel.
Usability validation (one to two short workshops). IG approvals for live pilot. Pilot deployment at Medway across MRI, CT, ultrasound, and image-guided biopsy services. Prospective data collection on slot utilisation, DNA recovery, A&E waiting time impact, and patient flow. Periodic email-based staff feedback throughout. Thesis chapters 3 and 4 drafted in parallel.
Quantitative and qualitative data analysis. External stakeholder engagement (including Dartford & Gravesham and other comparator Trusts) for framework validation. Implementation framework finalised. Scoping of future expansion to broader IR procedures, GP referrals, and outpatients. Full thesis write-up and submission, drawing on chapters drafted across Years 2 and 3.
Quantitative and qualitative data analysis. External stakeholder engagement (including Dartford & Gravesham and other comparator Trusts) for framework validation. Implementation framework finalised. Scoping of future expansion to broader IR procedures, GP referrals, and outpatients. Full thesis write-up and submission, drawing on chapters drafted across Years 2 and 3.
The four-year timeline is ambitious but grounded in a realistic assessment of what the researcher's position at Medway makes possible. Most of the access, relationship-building, and operational familiarisation that takes external PhD candidates one to two years to achieve is already in place. That is the foundation this timeline is built on.
The four-year timeline is ambitious but grounded in a realistic assessment of what the researcher's position at Medway makes possible. Most of the access, relationship-building, and operational familiarisation that takes external PhD candidates one to two years to achieve is already in place. That is the foundation this timeline is built on.
10. Bibliography
10. Bibliography
Ahmadi-Javid, A., Jalali, Z. and Klassen, K.J. (2017). Outpatient appointment systems in healthcare:
A systematic review of optimisation and simulation-optimisation methods.
European Journal of Operational Research, 258(1), pp.3–34.
Cayirli, T. and Veral, E. (2003). Outpatient scheduling in health care: A review of literature.
Production and Operations Management, 12(4), pp.519–549.
Greenhalgh, T., Wherton, J., Papoutsi, C., Lynch, J., Hughes, G., A'Court, C., Hinder, S., Fahy, N., Procter, R. and Shaw, S. (2017). Beyond adoption: A new framework for theorizing and evaluating nonadoption, abandonment, and challenges to the scale-up, spread, and sustainability of health and care technologies. Journal of Medical Internet Research, 19(11), e367.
Health Foundation (2022). Reducing waiting times in the NHS: What can we learn from the evidence?
The Health Foundation.
NHS England (2023). Diagnostics: Recovery and transformation.
NHS England.
Srinivas, S. and Ravindran, A.R. (2018). Optimizing outpatient appointment systems using machine learning algorithms and scheduling rules: A prescriptive analytics framework. Expert Systems with Applications, 102, pp.245–261.
Ahmadi-Javid, A., Jalali, Z. and Klassen, K.J. (2017). Outpatient appointment systems in healthcare:
A systematic review of optimisation and simulation-optimisation methods.
European Journal of Operational Research, 258(1), pp.3–34.
Cayirli, T. and Veral, E. (2003). Outpatient scheduling in health care: A review of literature.
Production and Operations Management, 12(4), pp.519–549.
Greenhalgh, T., Wherton, J., Papoutsi, C., Lynch, J., Hughes, G., A'Court, C., Hinder, S., Fahy, N., Procter, R. and Shaw, S. (2017). Beyond adoption: A new framework for theorizing and evaluating nonadoption, abandonment, and challenges to the scale-up, spread, and sustainability of health and care technologies. Journal of Medical Internet Research, 19(11), e367.
Health Foundation (2022). Reducing waiting times in the NHS: What can we learn from the evidence?
The Health Foundation.
NHS England (2023). Diagnostics: Recovery and transformation.
NHS England.
Srinivas, S. and Ravindran, A.R. (2018). Optimizing outpatient appointment systems using machine learning algorithms and scheduling rules: A prescriptive analytics framework. Expert Systems with Applications, 102, pp.245–261.
Note: References listed above are foundational sources identified at proposal stage. A full and updated literature review will be completed during Year 1 of the study.
Note: References listed above are foundational sources identified at proposal stage. A full and updated literature review will be completed during Year 1 of the study.
View My Devlog
for the latest updates on upcoming projects
View My Devlog
for the latest updates on upcoming projects