Inpatient illness severity surveys provide essential data for planning capacity and managing patient flow in the acute hospital setting

Introduction

The United Kingdom NHS is facing unprecedented pressure to deliver ever more efficient care. Significant focus has been placed on improved patient safety, superior outcomes and lower mortality rates. In the acute sector, workload is increasing at a rate where many hospitals struggle to provide high-quality care within budget. ¹ Demands on hospital systems are greatest in unscheduled emergency care, which, in turn, has a significant adverse impact on scheduled elective care, particularly in the winter. ²

There is variation in systems dealing with emergency care across the United Kingdom. ³ NHS acute hospitals strive to maintain best practice standards for all patients. On a hospital-wide scale, this primarily requires optimal management of patient flow through the hospital, and on ensuring that all patients receive timely intervention in an appropriate clinical setting from the specialist teams relevant to their condition. The systems of bed management and flow control underpinning this approach often vary between hospitals ⁴ depending upon historical service development and culture. On an individual patient-level scale, there has been recent emphasis on the importance of recognition of the severity of individual patients’ illness and its impact on patient physiology and physiological reserve. These factors are coupled to plan the extent of treatment escalation or palliation in the event of deterioration. ⁵ Patient acuity has become a term often used to describe the severity of illness and the associated physiological impact.^6,7

Tools exist to enable acute hospitals to identify acuity in their inpatient populations. The Modified Early Warning Score (MEWS) developed by intensive care physicians was recently renamed as the National Early Warning Score (NEWS) by NHS England as part of a drive to improve acuity recognition in acute NHS hospitals and trigger early referral for specialist input.^8,9 The Association of United Kingdom University Hospitals (AUKUH) acuity tool is widely used to characterise individual patient care needs by assigning a level of care to each individual; this is generally a tool used within High Dependency and intensive care areas.^7,10 We believe that it is possible to model the inpatient hospital population using these tools in a manner that allows tailoring of service provision to the acuity of the local population. In particular, the relative simplicity of the MEWS system makes it appealing, if it can be correlated to the more comprehensive AUKUH scale.

Methods

Details of this study were forwarded to the Research and Development Department at the Royal Cornwall Hospital NHS Trust. It was deemed that no ethical approval was required to conduct this audit as no direct patient contact was involved.

We used the MEWS (Table 1) and AUKUH (Table 2) tools to assess every inpatient in a rural, acute general hospital serving a population of 480,000 in the UK. The hospital admits 100,000 patients through 565 acute beds per year. We initially assessed all inpatients over a five-day period in November 2011 using both tools (Survey Period 1).

OPEN IN VIEWER

Table 1.

Modified early warning scoring system. ¹¹

Score	3	2	1	0	1	2	3
Respiratory rate (min−1)		≤8		9–14	15–20	21–29	>29
Heart rate (min−1)		≤40	41–50	51–100	101–110	111–129	>129
Systolic BP (mmHg)	≤70	71–80	81–100	101–199		>200
Urine output (mls/kg/h)	Nil	<0.5
Temperature (°c)		≤35	35.1–36	36.1–38	38.1–38.5
Neurological				Alert	Reacting to voice	Reacting to pain	Unresponsive

OPEN IN VIEWER

Table 2.

AUKUH dependency tool. ⁷

Levels of care	Inclusion criteria	Guidance on care required
Level 0	Elective medical or surgical admission. Routine post diagnostic/surgical procedure care. May have underlying medical condition requiring on-going treatment. Patient awaiting discharge.	Routine post-op/post procedure care (including ½ hourly obs until stable). Regular observations for 2–4 h, ECG monitoring to establish stability, fluid management, PCA, oxygen therapy 24%–40% (specialist surgical areas only – single chest drain). Requires routine nursing assistance.
Patient requires hospitalisation. Needs met through normal ward care.
Level 1	Observation and therapeutic intervention. “Step down” from Level 2 care. Post op care following emergency admission requiring immediate therapeutic intervention. Deteriorating condition or fluctuating vital signs.	Instability requiring continual observation/invasive monitoring. Support of outreach team but NOT higher levels of care. Oxygen therapy greater than 40% ± chest physio 2–6 h. Arterial blood gas analysis – intermittent 24–48 h following tracheostomy, insertion of central lines/epidurals/chest drains.
Appropriately managed on in-patient ward but requires more than baseline.
Level 1a
Acutely ill patient requiring intervention or those who are unstable with a greater potential to deteriorate.
Level 1b	Severe infection. Sepsis. Complex wound management. Compromised immune system. Psychological support/preparation. Requires continual supervision. Spinal instability/Mobility difficulties.	Complex drug regimes. Patient and/or carers require continued support owing to poor disease prognosis or clinical outcome. Completely dependent on nursing assistance for all activities of daily living. Constant observation due to risk of harm.
Patients who are in a stable condition but have an increased dependence on nursing support.
Level 2	Deteriorating/compromised single organ system. Post-operative management following major surgery. Post-operative optimisation/extended post op care. “Step down” from Level 3 care. Uncorrected major physiological abnormality.	Patients requiring non-invasive ventilation/respiratory support. Routine short-term post-operative ventilation. First 24 h following tracheostomy insertion. Requires a range of therapeutic interventions including: greater than 60% oxygen, continuous ECG and invasive pressure monitoring. Vasoactive drug infusions (amiodarone, potassium, inotropes, GTN, magnesium). Haemodynamic instability. Pain management: IV analgesic infusions, CNS depression of airway and protective reflexes. Neuro monitoring.
Patients who are unstable and at risk of deteriorating and should not be cared for in areas currently resourced as general wards (May be managed within clearly identified, designated beds resourced with the required expertise and staffing level OR may require transfer to a dedicated Level 2 facility/unit.)
Level 3	Monitoring and Supportive therapy for compromised or collapse of two or more organ systems.	Respiratory or CNS depression/compromise requires mechanical/invasive ventilation. Invasive monitoring, vasoactive drugs, treatment of hypovolaemia/haemorrhage/sepsis or neuro protection.
Patients needing advanced respiratory support and therapeutic support of multiple organs.

Information was sought via discussion with ward staff and review of nursing observation charts and medical notes. For each patient, we calculated MEWS, AUKUH acuity care level and recorded whether the patient was receiving oxygen therapy, had an intravenous cannula or urinary catheter in situ and whether the patient had a “Do Not Attempt Cardiopulmonary Resuscitation” (DNACPR) order in place.

The results and subsequent analysis of this exercise directly informed a decision to expand the Critical Care facility to provide more high dependency (Level 2) capacity. Following this expansion, the exercise was repeated over a second 5-day period in February 2012 (Survey Period 2).

Individual patient data were recorded on an encrypted laptop on a predesigned Microsoft Excel spreadsheet. ¹² The data were analysed by MEWS and AUKUH care level separately. In addition, we undertook a comparative analysis to determine if there was any correlation between MEWS and AUKUH that may inform changes to care escalation triggers in the MEWS. We performed a priori sub-group analysis of the Level 2 patients, stratifying by care location (General ward vs. Critical Care). Statistical comparisons were made between study periods using the Mann–Whitney U test, stratified by AUKUH care level. In addition, Chi-squared analysis was undertaken to compare proportions of Level 2 patients managed in and out of Critical Care across the two study periods. Significance was presumed at a confidence level of p < 0.05.

Results

In total, 3850 patients were reviewed: 1854 patients during Survey Period 1 and 1996 patients during Survey Period 2. The data for each survey by AUKUH care level are presented in Table 3. The differences at each care level between Survey Periods did not reach statistical significance (p = 0.575).

OPEN IN VIEWER

Table 3.

Showing patient acuity by care level during Survey Period 1 and Survey Period 2 (n = 3850).

AUKUH care level	Level 0	Level 1	Level 1a	Level 1b	Level 2	Level 3	Total
Survey Period 1	1336 (72.1)	12 (0.6)	162 (8.7)	267 (14.4)	51 (2.8)	26 (1.4)	1854 (100.0)
Survey Period 2	1777 (89.0)	0 (0.0)	28 (1.4)	128 (6.4)	38 (1.9)	25 (1.3)	1996 (100.0)

The location of Level 2 patients is further delineated in Table 4 according to patient location at the point of survey. The proportion of Level 2 patients managed outside Critical Care fell significantly (62.7%–31.6%; p = 0.0036).

OPEN IN VIEWER

Table 4.

Location of Level 2 patients.

	General ward	Critical care	Total
Survey Period 1	32 (62.7)	19 (37.3)	51 (100.0)
Survey Period 2	12 (31.6)	26 (68.4)	38 (100.0)

AUKUH and MEWS analysis

Figure 1 illustrates the range of MEWS within each of the AUKUH groups. Figure 2 demonstrates the sub-group analysis of the relationship between MEWS and AUKUH for Level 2 stratified by care location. Those managed on the ward have a median MEWS of 5 compared with a MEWS of 1 for those managed in Critical Care. OPEN IN VIEWER

Figure 1.

Distribution of MEWS amongst patients stratified by AUKUH care level (n = 3850).

OPEN IN VIEWER

Figure 2.

Distribution of MEWS for Level 2 patients stratified by care location (n = 89).

Discussion

We believe that this survey is the first in the United Kingdom to examine the acuity of all inpatients in an acute hospital. The data generated by both surveys confirm that most patients in an NHS acute hospital have low acuity. Of the 3850 patients surveyed, 3335 (86.6%) required Level 0 care. This demonstrates that only one in six inpatients have care requirements that are not met by a standard ward environment. Potentially the MEWS tool may therefore be used to identify those patients who require increased levels of care in an environment appropriate to their illness severity.

In total, 587 patients required Level 1 care (1, 1a or 1b). This increased level of care requirement may suggest a benefit from cohorting patients to maximise benefits of both economy of scale and safety. By concentrating relatively small numbers of appropriate specialist medical and nursing staff in one area, higher levels of support can be delivered with the least economic and logistical burden. These areas should be specialty specific, including both medical and surgical units.

Within the Level 2 population, there is an apparent skew towards a MEWS of 4–5, but equally, a similar proportion of equally acute patients had a MEWS of 0 (Figure 1). Superficially, this tends to suggest that MEWS has a poor sensitivity for high acuity patients although it may have a greater specificity at higher MEWS. However, when ward-based and Critical Care–based Level 2 patients are analysed separately (Figure 2), a much clearer trend emerges towards a MEWS cut-point which could be used as a surrogate marker of those requiring Level 2 care. Those on Critical Care meanwhile may still fulfil the definition of Level 2 care but have normalised physiology and a correspondingly low MEWS, e.g. the non-invasive ventilation patient who has optimised oxygen saturations and respiratory rate albeit with organ support. This post-intervention bias could undermine the potential use of MEWS for triggering care escalation in ward patients; in our opinion, these patients are a different Level 2 population, requiring consideration separate from those within a Critical Care environment when determining acuity assessment methods. Our stratified analysis accounts for this.

We recommend that acuity surveillance undertaken by bed management teams, at least daily, will allow high acuity patients to be identified and transferred to Critical Care immediately. In addition, regular reporting of patient acuity in Critical Care in the same forum, will lead to timely discharge from Critical Care as patients recover, accepting that both level of care and MEWS must be considered jointly to avoid post-intervention bias. This will reduce the likelihood of delayed admission to Critical Care of Level 2 patients and reduce the associated potential harm.

Conclusion

We have demonstrated a relationship in the distribution of patient acuity, using the AUKUH tool and the MEWS using data from 3850 patients. Current care pathways for patients in the MEWS systems recommend escalation, if the MEWS rises above 5. ⁸ Our data confirm that the patient acuity increases towards Level 1b and 2 when the MEWS increases above this point. We have also demonstrated that those with Level 2 dependencies being managed outside of Critical Care have an increased median MEWS compared with those managed within Critical Care due to the effects of organ support.

We successfully used these data to evidence the requirement for Critical Care expansion in our hospital with a significant reduction in higher dependency patients being managed in a general ward environment. We believe that this methodology can be used both for effective service planning and development as well as to deliver daily improvements in flow of patients to the right care areas and plan to apply this for the benefit of our Level 1 patient population.