Considerable planning with weekly meetings was needed for the RCR-MedCurrent CDS pilot project at the London Northwest and Hillingdon Hospitals Trusts which serve a population of 1.2 million.
(Fig 1.)
Fig. 8
Figure 1a
Fig. 1: Figure 1 a&b. The location of the pilot project is in a diverse sector of greater London where there is a wide range in income, ethnicity and education. Provision of healthcare is by 4 primary care Clinical Commissioning Groups (Brent, Ealing, Harrow and Hillingdon), working with 4 major hospitals (Central Middlesex, Ealing, Hillingdon, Northwick Park), providing the majority of imaging with further contribution from independent providers.
Figure 1b.
Figure 1 a&b. The location of the pilot project is in a diverse sector of greater London where there is a wide range in income,
ethnicity and education.
Provision of healthcare is by 4 primary care Clinical Commissioning Groups (Brent,
Ealing,
Harrow and Hillingdon),
working with 4 major hospitals (Central Middlesex, Ealing,
Hillingdon,
Northwick Park),
providing the majority of imaging with further contribution from independent providers.
Identification of key stakeholders was an early step (Figure 2).
Driven by a multidisciplinary innovation programme and led by an experienced project manager,
inclusion of core,
steering and extended groups enabled collaboration of technical,
supply,
clinical and patient representatives.
Eight work streams were set up to provide leadership,
objectives and deliverables on time (Figures 3&4):
1. Clinical Decision Support;
2. Technical Interfaces;
3. Technical Implementation at Hospital 1;
4. Technical Implementation at Hospital 2;
5. Information Governance;
6. Communications and Engagement;
7. Change Management; and
8. Value Audit.
Fig. 2: Figure 2. Stakeholders were identified early in the process to facilitate involvement and buy-in. The core group meets weekly feeding back from their meetings and raising issues requiring input from other work streams.
Figure 2. Stakeholders were identified early in the process to facilitate involvement and buy-in.
The core group meets weekly feeding back from their meetings and raising issues requiring input from other work streams.
Fig. 3: Figure 3. Work streams and activities discussed at weekly teleconferences.
Figure 3. Work streams and activities discussed at weekly teleconferences.
Fig. 4: Figure 4. Gantt chart for phased implementation of Clinical Decision Support (CDS) for imaging requests. Early work was needed for the technical interface between CDS, electronic requesting and radiology information systems. A phased introduction was planned for each of two imaging providers, with feasibility studies in secondary and primary care (phase 1 and 2) followed by systems roll-out to primary care (phase 3) and finally secondary care (phase 4).
Figure 4. Gantt chart for phased implementation of Clinical Decision Support (CDS) for imaging requests.
Early work was needed for the technical interface between CDS,
electronic requesting and radiology information systems.
A phased introduction was planned for each of two imaging providers,
with feasibility studies in secondary and primary care (phase 1 and 2) followed by systems roll-out to primary care (phase 3) and finally secondary care (phase 4).
The CDS tool (Figure 5) incorporates the intellectual property of the RCR iRefer Guidelines: Making the best use of clinical radiology [11,12] and MedCurrent’s Orderwise CDS platform for decision making logic and rules.
Data mining,
machine learning and analytic data are all features of the system providing value to the referrer and justifying practitioner.
Generic ICRP Level 2 justification becomes available to the practitioner to facilitate vetting: authorisation,
level 3 justification and protocolling of the exam.
Fig. 7
Figure 5a.
Fig. 5: Figure 5. Clinical decision support (CDS) for imaging requests. The presenting complaint of headache (a) needs further clarification through an additional question (b), which may have previously required practitioner-referrer clarification, delaying justification. The recommendation (c) is backed by grading, radiation dose and cost information. Data mining enables the patient’s renal function and imaging history to be displayed, avoiding unnecessary repeat investigations or contrast induced nephropathy. The smart system also provides analytic data to reinforce the use of appropriate imaging by both referrer (client) and practitioner (provider).
Figure 5b.
Fig. 6
Figure 5c.
Figure 5. Clinical decision support (CDS) for imaging requests.
The presenting complaint of headache (a) needs further clarification through an additional question (b),
which may have previously required practitioner-referrer clarification,
delaying justification.
The recommendation (c) is backed by grading,
radiation dose and cost information.
Data mining enables the patient’s renal function and imaging history to be displayed,
avoiding unnecessary repeat investigations or contrast induced nephropathy.
The smart system also provides analytic data to reinforce the use of appropriate imaging by both referrer (client) and practitioner (provider).
Metrics for personalised value (to patients,
clinicians,
radiologists and radiographers); technical value (for imaging appropriateness,
technological interface and efficient process); and allocative value (for resource,
human and financial factors) are all pertinent.
Audit of personalised value is assessed by a questionnaire with a balanced 5-point Likert scale,
taking into consideration satisfaction,
decision regret [13-16],
time-commitment and user friendliness.
Measures for technical value of appropriateness of imaging are CDS-generated and backed by pre- and post- implementation retrospective audits using methodology from the RCR National Audit of Appropriate Imaging [17].
Measures for allocative value take into account metrics for all imaging,
targeted areas of over-utilisation and measures for turnaround time.
In selecting metrics,
the experience of previous CDS implementations have been taken into account [18-20].
