3 Are there any new signs audible in the chest to suggest pneumonia?

4 Are any features of severity present?
- (raised respiratory rate, low blood pressure, confusion of recent onset: see Box 1)

This guideline distinguishes three different populations within LRTI:

• Patients with no new signs in the chest, and who have underlying chronic respiratory disease and have an increase in dyspnoea and sputum purulence - e.g. acute exacerbation of COPD
  with purulent sputum (see section 4).
• Patients with new focal signs in the chest (crackles or altered breath sounds) - suspected
  community acquired pneumonia (see section 5).
• Patients with no signs in the chest, and who have been previously well and do not have other features of severity - non-pneumonic LRTI (see section 6).

The presence of significant comorbidity in the first two groups may influence disease management decisions. Similarly, patients without chest signs who have co-morbidity (e.g. chronic cardiac, renal, hepatic or neurological disease or have diabetes mellitus) or are elderly, require more careful management when they acquire an LRTI. In the absence of direct evidence, the consensus is that such patients should be considered for a sputum culture. Appropriate antibiotic prescription can then be made. However, optimal management of comorbidity is outside the scope of this guideline.

2.2 Incidence of lrti

Approximately 1.7 million people aged between 16 and 79 years were treated for LRTI in England and Wales in 1991, representing an incidence of 46/1,000 population.¹³ UK studies documenting the frequency of LRTI in the adult population in general practice suggest an overall incidence of 44-84 cases per 1,000 population per year.^13,14 Extrapolation of these figures suggests that in Scotland there are 150,000 -350,000 cases of LRTI in adults each year.

LRTI is a particularly common and serious illness in the elderly. In a population-based retrospective cohort study in Rochester, Minnesota, USA, approximately one in 18 residents older than 65 years experienced one or more episodes of bronchitis or pneumonia over one year, with an overall 30-day mortality of 10.7%.¹⁵ A UK prevalence-based burden of illness analysis estimated the direct health care cost to be £1,364 million annually at 1992/93 prices.¹⁶ Over 60% of these costs occur in hospital, although inpatient care accounted for only 1.4% of all LRTI episodes. A total of 13% of these admissions were discharged within 24 hours and further calculation suggests that managing 80% of these admissions in the community could reduce NHS costs by between £10 million and £49 million annually.

Figure 1 represents the estimated incidence of LRTI in both hospital and community in Scotland. These figures were derived from studies in general practice in Nottinghamshire and extrapolated to a Scottish population of 4.2 million adults. The diagram emphasises the fact that most individuals with LRTI do not consult their GP and very few require admission to hospital. These figures are likely to be underestimates of the actual incidence.

Figure 1: The LRTI pyramid

Incidence of LRTI and pneumonia in adults in the community and in hospital in Scotland (adapted with permission from Macfarlane¹⁷).

2.2.1 THE SPECTRUM OF LRTI IN GENERAL PRACTICE IN SCOTLAND

In a UK general practice study of LRTI, 100% of 206 patients presented with cough and/or sputum production, 71% with breathlessness or wheeze, 68% with chest pains or aches, 66% with sweats, and 64% with sore throat. On examination, 49% had an inflamed throat, 25% focal chest signs, and 7% cold sores.¹³

Continuous Morbidity Recording (CMR) data for LRTI in a subset of 51 practices in Scotland in 1999 show that the incidences are similar between the sexes (Figure 2). Incidence rates are increased in pre-school children (110-140/1,000 population), lowest for 5-64 year olds (<70/1,000 population), increased in those aged 65-74 (110-130/1,000 population) and highest in those aged 75 years and over (140-180/1,000 population). There is a strong seasonal variation with a peak incidence in December and January (Figure 3). Deprivation, based on postcode, also influences the incidence of LRTI, with females living in the most deprived areas having the highest incidence (Figure 4). The CMR data also confirm the prevalence of asthma and COPD in patients consulting with LRTI (Figure 5).

Figure 2

Figure 3

Figure 4

Figure 5