10.

For people admitted to hospital with a medical emergency, consider providing the following, accompanied by local evaluation which takes into account current staffing models, case mix and severity of illness:

• Consultant assessment within 14 hours of admission to determine the person’s care pathway
• Daily consultant review, including weekends and bank holidays
• More frequent (for example, twice daily) consultant review based on clinical need.

Mortality, quality of life, avoidable adverse events and patient and/or carer satisfaction were considered by the committee to be critical outcomes.

Early diagnosis, hospital admission, number of diagnostic tests, length of stay, GP visits, referrals from admission, unplanned readmission, discharge and staff satisfaction were considered to be important outcomes.

The committee considered that avoiding readmission was likely to be particularly important for people who have a chronic condition as this has an impact on mortality and also could have an impact upon psychological wellbeing and the ability to maintain independence.

Emergency Department

A single RCT was identified. The committee decided that the Senior Work up Assessment and Treatment (SWAT) intervention had most similarities to current systems in the NHS (Rapid Assessment and Treatment [RAT]) compared to the non-SWAT intervention because for consultants to work effectively, they need the support of a team and therefore seeing patients alone would not be productive. Indeed, in the UK, consultants do not normally see patients in isolation.

The comparison of SWAT versus control data suggested that SWAT may provide a benefit in increased proportion of patients achieving the National Emergency Access Target (NEAT), which is to be seen and discharged from the ED within 240 minutes of triage; proportion of admitted patients who met NEAT; and proportion of discharged patients who met NEAT. However, there were more patients admitted and fewer discharged with early consultant review. The committee surmised that early consultant review might, in some circumstances, be disadvantageous if it took place before definitive investigations were available which might have permitted safe discharge on later review. Therefore, review prior to all the relevant information being present may result in a greater number of patients admitted. However, the fact that more patients were admitted, although increasing demand, may be a positive step as it may ensure that certain patients receive the inpatient care their condition requires. The presence of a senior decision maker may identify these patients.

The committee discussed their experience of the Rapid Assessment and Treatment system (the UK system of immediate consultant triage at presentation to ED). Perceived benefits included more rapid diagnosis, earlier administration of antibiotics and analgesics, and more appropriate triage. However, such outcomes are not normally measured in trials whereas admission, discharge and length of stay are affected by a wide variety of factors, and therefore may not accurately capture the whole effects of early consultant triage.

Six observational studies suggested that early consultant review may provide a benefit in reduced length of ED stay, 30 day unscheduled re-admissions, admissions, patients achieving NEAT, discharged patients achieving NEAT, admitted patients achieving NEAT, patients seen within the recommended time and patients who did not wait to be seen. There was a possible increase in mortality but this was discounted by the committee as there was only a difference of 1 case between the 2 groups.

No evidence was identified for early diagnosis, quality of life, GP visits, avoidable adverse events, diagnostic test number, patient and/or carer satisfaction, referral from admissions and staff or trainee satisfaction.

Acute Medical Unit

A single observational study was identified suggesting that early consultant review may provide a benefit in reduced length of stay, and the proportion of patients discharged on the day of admission. There was no effect on mortality during admission; there was a possible increase in the proportion of patients discharged within 24 hours and readmitted within 1 week for the same clinical problem.

No evidence was identified for hospital admission, readmission, early diagnosis, quality of life, GP visits, avoidable adverse events, diagnostic test number, patient and/or carer satisfaction, referral from admissions and staff or trainee satisfaction.

Stroke patients:

No evidence was identified in a stroke care setting. The committee felt that the results from ED and AMU could be extrapolated to stroke patients.

Intensive (or critical) care unit:

No evidence was identified in an intensive care unit (critical care unit) setting. Studies of resident versus non-resident intensive care specialists were considered too indirect to be employed as substitutes for early consultant review. Given this lack of evidence, the committee considered that studies in ED and AMU patients might be used to inform recommendations relating to the ICU.

Overall

The committee noted that the effect of early consultant involvement is dependent upon the staffing model, the presenting case mix and the disease process. For example, conditions with a well-defined treatment pathway may benefit more from early consultant involvement if this results in earlier diagnosis and entry to the pathway. In settings where patients are presenting with often unclear disease processes (for example, in an emergency department), the benefit of early consultant involvement might be realised if consultants’ greater knowledge results in earlier diagnosis, or diminished if the diagnostic process is complex. The committee noted that a range of models for early consultant involvement were used in the studies examined, and that the model used within a UK context may differ from those included in the studies. For example, the Rapid Assessment and Treatment model implemented within some emergency department settings in the UK was a model containing a range of interventions, including early consultant involvement. It was felt to be similar but not identical to the SWAT model in the RCT for EDs.

Overall, the evidence was mixed but suggested some benefit in outcomes over usual care for the ED and AMU. No evidence was identified to suggest harm in early consultant involvement and the committee were not aware of any negative outcomes that might occur. They therefore chose to make a consensus recommendation to consider early consultant involvement in care of a patient with an acute medical emergency. However, there was insufficient evidence to recommend specific models such at RAT.

No relevant economic evaluations were identified. Unit costs of staff time, emergency department visits and relevant hospital admissions and stays were presented to the committee.

One RCT, described above, set in the emergency department showed that the SWAT arm of the trial was associated with a trend for more patients meeting the 4-hour target; however, there was also a trend for more admissions and less discharges compared to the control arm. The committee felt that without information on the appropriateness of the decisions to admit or discharge, it would be difficult to fully assess the impact of the SWAT model. Anecdotally, the committee felt that the equivalent model in the UK (Rapid Assessment and Treatment or RAT) had shown some clinical benefit in terms of timely diagnosis and treatment. These benefits might be expected to result in saving in downstream costs.

For the AMU, the observational study included in the clinical review suggested that there was a reduction in length of stay, which would translate into possible cost saving.

The committee noted that the economic impact of early consultant assessment would be dependent on how it could be achieved or implemented in practice. Possible scenarios discussed included increasing the number of consultants, increasing their contracted hours (which might include working out-of-hours or being on-call) or accommodating the required changes in the consultants’ current rotas by prioritising early patient assessments over other duties, which can be undertaken by other staff members.

The committee commented that the most likely scenario in large hospitals is that consultant rotas could be tailored to accommodate prioritising assessing patients given current capacity levels and the limited number of NHS consultants, which precludes the possibility of recruiting more consultants. However, this may not be feasible in smaller hospitals.

New cost-effectiveness analyses were conducted for 2 areas of early consultant assessment with the results presented to the committee. A cohort model and a simulation model were built to assess the cost-effectiveness of early consultant assessment. Both models used inputs from bespoke data analysis, national data and treatment effects (primarily length of stay reduction and modest reductions in adverse events) that were informed by the above review but elicited from the committee members. The full model write up can be found in Chapter 41.

Rapid Assessment and Treatment in the Emergency Department (RAT)

The models compared RAT in the ED with no RAT. RAT involves an immediate assessment by a consultant in the ED, using additional resources in terms of consultant time at an incremental cost to normal care.

Both models found that RAT was cost increasing with assumed no impact on quality of life, hence no gain in quality-adjusted life-years. The committee noted that RAT is a costly intervention a, with additional consultant time for all ED major patients. An optimistic sensitivity analysis found RAT to cost £98,000 per QALY gained – far from being cost effective. The main impact of RAT is likely to be on hospital flow, not taken into account by the cohort model. The simulation model saw a reduction in 4-hour breeches from 10% to 8%.

The committee concluded that RAT is a costly intervention that is probably not cost effective in general, although it might still have a positive impact on hospital flow in hospitals operating at sub-optimal levels of efficiency within the emergency department.

Extended hours for consultants in Acute Medical Units (AMU)

The model compared consultant assessment available in the AMU 08:00-18:00 with consultant assessment available in the AMU 08:00-22:00. Therefore, the intervention involves the presence of a consultant to assess and treat on the AMU for an additional 4 hours in the evenings, 7 days a week. This uses additional resources in terms of consultant time at an incremental cost to normal care.

The results of the cohort model found that extended hours on the AMU was cost increasing with a small impact on quality-adjusted life-years. However, the QALYs gained were not large enough in the base case or optimistic sensitivity analysis to allow an incremental cost-effectiveness ratio under the £20,000 threshold, £45,500 per QALY gained in the base case and £25,500 in the optimistic treatment effects sensitivity analysis. The committee noted the results of the cohort model with an ICER close to the £20,000 threshold in the sensitivity analysis. However, they also noted that extended hours in the AMU was likely to have an impact on hospital flow, not taken into account by the cohort model. However, the AMU could not be properly assessed by the simulation model because too many runs would be required.

The committee noted that the intervention allows earlier decision making, potentially avoiding an overnight admission or facilitating earlier discharge. They also noted that extended hours in the AMU could have a positive impact on the hospital flow and patient outcomes, and therefore may be cost-effective at local level. However, extended hours to the AMU should only be implemented alongside local evaluation.

Conclusion

The committee felt that early consultant assessment could be cost effective in some settings. It is associated with some clinical benefit and, in some settings, the cost might be completely offset by savings from increased efficiencies in the hospital pathway. However, it was agreed that this would not be the case nationwide and any intervention should only be implemented at the local level alongside evaluation.

For some Trusts, the resource impact of this recommendation will be more hours of consultant time in the AMU and other high care units. This should be partially offset by reduced length of stay and fewer complications. Some Trusts might want to disinvest in RAT, which would mean savings in terms of ED consultant staff time. There are benefits of early consultant assessment that were not captured in the model and are difficult to quantify, including impact on quality of life from quicker diagnosis and more appropriate location of/better quality of death.

Overall, the evidence was not very strong and therefore the committee felt that neither immediate consultant assessment, such as RAT, nor extended hours could be recommended. However, there is still a need for consultant assessment at the earliest practical opportunity.

Current pragmatic recommendations from professional organisations recommend initial consultant review within 14 hours for patients admitted to acute medical units [Society for Acute Medicine{ ACT2015}, and within 12 hours for patients admitted to intensive care units [UK Faculty of Intensive Care Medicine{FICM2016}]. The committee concluded that in the absence of definitive evidence, these professional recommendations were reasonable, but should be subject to local audit and evaluation.

Emergency department:

One RCT was identified which was based in Australia and was graded low to moderate quality due to risk of bias and imprecision. The committee considered whether the study was applicable to a UK setting as in a non-UK setting, patients may present more frequently to secondary care as a first contact. However, the committee chose not to downgrade this study for indirectness as the model was applicable. The observational evidence was all graded as very low quality due to lack of randomisation and the presence of additional confounders, such as the intervention group also receiving point of care testing in addition to early consultant review.

Acute medical unit:

One observational study was identified and the outcomes were graded as very low quality due to risk of bias, imprecision and indirectness. There were some baseline differences in the conditions for which patients in both groups were being assessed and multivariate analysis had not been carried out.

No evidence was identified for stroke care, intensive care or critical care units.

Original health economic modelling was assessed to be directly applicable but still had potentially serious limitations due to the treatment effects being based on expert opinion, albeit conservative and informed by the guideline’s systematic review.

Due to the quality of the evidence the committee decided to make a cautious recommendation for providers to consider consultant review within 14 hours.

The committee noted that, in practice, many of the competencies required to implement a model of early consultant review may be delivered by other members of healthcare staff. However, it is the knowledge or expertise that the consultant brings to the assessment that is crucial. Consultants do not work in isolation and need support of other staff; therefore to implement, this will require reconfiguration of rotas and changes in the availability of healthcare professionals.

The committee were aware of observational evidence across a range of healthcare settings which was not included in the review because of either the availability of higher quality evidence or because it did not meet the inclusion criteria for the review. The committee noted that this observational evidence supported their recommendations for early consultant involvement in these settings.

Although no evidence was found on patient and/or carer satisfaction, the committee noted that it was probably the preference of patients to be seen quickly, spend minimal time in ED and AMU and receive an accurate assessment of their condition with appropriate admission and discharge decisions.

The committee was interested in how early the consultant review should be to demonstrate an improvement in clinical outcome. The definitions for an early consultant review as presented in the evidence was highly variable, most of which were unclear and vague. For example, one study defined an early consultant review as a review within 24 hours, whereas another study defined an early consultant review as when a consultant was present 4 days out of 5 during the working week from 9am-5pm.

The committee referred to the RCP’s Acute care toolkit 4 and the Society for Acute Medicine clinical quality standards: Delivering a 12-hour, 7-day consultant presence on the acute medical unit which includes the following 2 key recommendations:

1. During the period of consultant presence on AMU, all newly admitted patients should be seen within 6 to 8 hours, with the provision for immediate review as required according to illness severity.
2. A newly admitted patient must be seen by a consultant within 14 hours after arrival on AMU.

It was felt by the committee that, although there was no evidence from other acute care units such as the CCU, HASU or ICU, this way of working could be extrapolated to those centres. Indeed, in some of these units it is already occurring, that is, PCI in ST elevation MI which is often performed by a consultant cardiologist, or the delivery of thrombolysis in patients with stroke being covered by a consultant stroke thrombolysis rota.

The Academy of Royal Colleges provided a report called the benefits of consultant delivered care². In this report they highlighted the benefits of consultant delivered care:

• Rapid and appropriate decision making
• Improved outcomes
• More efficient use of resources
• GPs access to the opinion of a fully trained doctor
• Patient expectation of access to appropriate and skilled clinicians and information
• Benefits for the training of junior doctors.

As part of the implementation of 7 day services, hospital trusts are expected to meet 10 clinical standards produced by NHS England. The standards were drawn up by the national medical director, Sir Bruce Keogh, and his colleagues at NHS England in 2013, informed by an Academy of Medical Royal Colleges report published in 2012. Trusts are expected to meet 4 priority standards by the end of this financial year. The standards are:

• Time to first consultant review—patients should be seen as soon as possible but within at least 14 hours
• Inpatients should have 7 day access to a range of diagnostics
• Inpatients should have access to a range of key interventions
• All acute patients must be seen and reviewed by a consultant twice daily.