5.1 Introduction

Last year, an analysis of emergency admissions was included as a chapter in Key Health Data (Key Health Data 2004).  It has been retained as a subject this year owing to concerns in the commissioning community that trends in this kind of service use are more fully understood.  Also, under Payment by Results, the programme of reform of Health Service financing, (Department of Health 2002) there is increasing transparency in the resource flows which accompany patients who are admitted to hospital.

These health needs are difficult to plan for, as by their very nature they are unpredictable.  They are both the product of the epidemiology of the populations involved and also of health services, which may be able to intervene to meet patients’ needs outside of the context of admission to hospital.

Last year, overall trends in emergency admissions were presented.  The use of a case mix analysis to look at admissions of patients with less intensive health needs, was used to highlight concerns over an increase in short admissions for investigation and observation.  This year, some of last year’s headline statistics have been generated for financial year 2004 / 2005.  Additionally analyses comparing population admission rates in a range of case mix groups has been calculated.  These groups have been chosen as they represent common reasons for unplanned admission.  They are admissions of children, admissions for respiratory diagnoses, admissions for injury, admissions for ischaemic heart disease and admissions for stroke.

Anecdotal feedback from readers suggested that the trends in financial year 2004 / 2005, which could not then be determined, were likely to show considerable acceleration in admissions, particularly for less intensive admissions.

In this chapter, both simple count data and rates have been included.  Whilst it is important to consider rates to understand needs in populations, it is also important to consider changes in absolute numbers, as since the NHS financial reforms, the aggregate of these counts have measurable financial flows attached to them.

5.2 Results

Emergency admissions to hospital have seen a modest year on year rise for the last five years, although the age sex standardised rate has been very stable (Figure 5.01) over this period. 

The pattern within individual populations has been considerably more variable, reflecting local needs, demography and service factors (Table 5.01 and Figure 5.02).

Table 5.01 Counts of all Emergency Admissions, by PCT Resident Population, for the West Midlands Region, Financial Years 2000/2001 - 2004/2005

Figure 5.01 Emergency Admission Rate, Directly Standardised for Age and Sex, per 100,000 population, for Residents of the West Midlands Region 2000/2001 - 2004/2005

Figure 05.01 Emergency Admission Rate, Directly Standardised for Age and Sex, per 100,000 population

Figure 5.02 Directly Standardised Rates of Emergency Admission, per 100,000 Population, All Diagnoses, by PCT Resident Populations, Financial Year 2004/2005

Figure 05.02: Directly Standardised Rates of Emergency Admission, per 100,000 Population.

As can be seen, PCT populations in deprived areas tend to be the populations with the highest rates in 2004/2005.  The variation recorded in previous years can be seen in similar figures in a previous KHD report (Key Health Data 2004).  Some populations have seen a growth in excess of 10% in total admissions compared to last year which will have a significant impact upon budgets, particularly if unforeseen.

These data have been disaggregated to examine individual case mix groups and the directly standardised rates compared across the populations for the financial year 2004/2005.

Firstly, in line with last year’s chapter, a less intensive admission group was analysed.  This comprised a wide range of short non-surgical admissions where a lack of significant diagnosis was recorded with either signs and symptoms only, or superficial injuries were noted as the primary diagnosis. 

Last year it was hypothesised that overall this group would increase markedly.  Over the five year period the population admission rate in the region has been rising faster than that for admissions generally (Figure 5.03), but it is questionable whether the increase in 2004 / 2005 was as much as had been anticipated by some commissioners.

Figure 5.03 Directly Standardised Rates of Emergency Admission, per 100,000 Population, Lower Intensity Admissions by PCT Resident Populations, Financial Year 2004 / 2005

Figure 05.03 Directly Standardised Rates of Emergency Admission, per 100,000 Populations, Lower Intensity Admissions

As can be seen there is also considerable variation across the populations (figure 5.04).  One interesting feature is that whilst in general we see broadly the more deprived populations at the higher end of the range and the affluent populations at the lower end, there are some affluent populations amongst the higher rates and vice versa. 

In each of the case mix groups (Figures 5.4 to 5.9) we notice rates that are arguably higher or lower than expected levels of need given the known characteristics of the population. In each case there is potential to investigate what local epidemiological or service factors may be driving this variation.

Figure 5.04 Directly Standardised Rates of Emergency Admission, per 100,000 Population, Lower Intensity Admissions by PCT Resident Populations, Financial Year 2004 / 2005

Figure 05.04 Directly Standardised Rates of Emergency Admission, per 100,000 Population, Lower Intensity Admissions

Figure 5.05 Directly Standardised Rates of Emergency Admission, per 100,000 Population, Patients Under 16 Years Old, All Diagnoses, by PCT Resident Populations, Financial Year 2004 / 2005

Figure 05.05 Directly Standardised Rates of Admission, per 100,000 Population, Patients Under 16 Years Old

Figure 5.06 Directly Standardised Rates of Emergency Admission, per 100,000 Population, Main Primary Diagnosis of Respiratory Disease, by PCT Resident Populations, Financial Year 2004 / 2005

Figure 05.06 Directly Standardised Rates of Emergency Admission, per 100,000 Population, Main Primary Diagnosis of Respiratory Disease.

Figure 5.07: Directly Standardised Rates of Emergency Admission, per 100,000 Population, Main Primary Diagnosis of Injury, by PCT Resident Populations, Financial Year 2004 / 2005

Figure 05.07 Directly Standardised Rates of Emergency Admission, per 100,000 Population, Main Primary Diagnosis of Injury

Figure 5.8: Directly Standardised Rates of Emergency Admission, per 100,000 Population where Main Primary Diagnosis is Ischaemic Heart Disease, by PCT Resident Populations, Financial Year 2004 / 2005

Figure 05.08 Directly Standardised Rates of Emergency Admission, per 100,000 Population where Main Primary Diagnosis is Ischaemic Heart Disease

Figure 5.09: Directly Standardised Rates of Emergency Admission, per 100,000 Population where Main Primary Diagnosis is Stroke, by PCT Resident Populations, Financial Year 2004 / 2005

Figure 05.09 Directly Standardised Rates of Emergency Admission, per 100,000 Population where Main Primary Diagnosis is Stroke

5.3 Discussion

The response to emergency care needs in a PCT population is largely determined by the dominant provider of emergency services.  For many there will be just one hospital which provides the majority of care.  Although due to geographical reasons some populations will be more heterogeneous in their choice of provider.  Individual policies of these providers can have an effect on the rates observed here.

For example some populations have access to services where the balance between community and hospital interventions is different as a matter of policy.  For example some areas have a service to manage less serious strokes in the community without automatically admitting e.g. intermediate care services.

Also the kind of unit that can be most easily accessed may have an impact. For example we see in the children’s analysis (figure 5.05), Heart of Birmingham having a rate not very far from the regional rate, however, in many other groups this is one of the populations with the highest need for emergency care.  This may be due to having ready access to a specialist Children’s hospital where more interventions and investigations can be undertaken in the A&E setting before admission is required.

Hence it is important not to use these data as relative indicators of the state of health of these populations.  Whilst undoubtedly the overall incidence of serious illness is the most important component of demand for emergency admission, other factors may be sufficiently significant to distort the population admission rate, so the link between it and epidemiology is not constant across populations.  Where variation is a cause for concern considerable efforts should be made to elucidate the underlying causes before the assumption is made that there is an unacceptably high variation in incidence of disease.

References:

1. Key Health Data for the West Midlands 2001, University of Birmingham DPHE 2002
   
   
2. Reforming NHS financial flows, Introducing payment by results, Department of Health, October 2002