9.1. Introduction

It is estimated that 40% of the population consult their general practitioner on at least one occasion during the year because of infection and nearly one third of all consultations in general practice are concerned with infection (Fleming et al. 2002). Eighty percent of antibiotics prescribed are used in primary care.

Clinical diagnoses made by General Practitioners (and symptoms reported to NHS Direct) offer the ability to monitor systematically a variety of syndromes/symptoms which could give early warning of a health protection issue (whether this be microbiological or chemical in origin).

The UK primary care structure (with registered lists, geographically defined populations and expert staff) offers important potential for health protection. The recent strategy for combating infectious diseases (Department of Health 2002) stresses the need for co-ordinated surveillance at the primary care level including the need to establish reliable denominators and links between databases (e.g. antibiotic prescribing and resistance).

The surveillance task in primary care is two-fold – to give early warning of new public health hazards and to monitor the incidence of known hazards. Staff at the Regional Surveillance Unit (Health Protection Agency West Midlands) take a national lead within the Health Protection Agency for health protection surveillance in primary care. The following sections summarise recent collaborative work, with particular emphasis on work in the West Midlands.

9.2. NHS Direct

A commitment by the Government to provide people with quick and easy access to health care (Department of Health 1997) led to the implementation of NHS Direct (NHS Direct Online). NHS Direct is a national nurse-led telephone helpline which aims to provide the public with rapid access to professional health advice and information about health, illness and the NHS. NHS Direct nurses use clinical decision support software, the NHS Clinical Assessment System (NHS CAS), to respond to calls. The NHS CAS contains over 200 algorithms consisting of a series of questions relating to the symptoms of the person about whom the call is made. When symptoms are reported (‘triage calls’), nurses use clinical judgement to select the most appropriate algorithm to handle the call. Most calls result in either advice for self care; a routine GP referral; an urgent GP referral; an Accident and Emergency Department referral; or a ‘999’ call-out. These outcomes are termed ‘dispositions’. It is important to remember that callers are not diagnosed by NHS Direct but classified (‘triaged’) on the basis of symptoms described to determine priority of need and appropriate place of treatment. NHS Direct also handle large numbers of health information queries.

NHS Direct is organised into 22 sites in England and a single site covering all of Wales (see Figure 9.1). The West Midlands region is covered by two NHS Direct sites, West Midlands NHS Direct (population 2.4 million) covering Birmingham, the Black Country and Solihull, and Midland Shires NHS Direct (population 3 million) covering Staffordshire, Shropshire, Worcestershire, Herefordshire, Warwickshire and Coventry. These two NHS Direct sites have call centres based in Dudley and Stafford respectively. The data presented here are from these two call centres. Figures 9.2, 9.3, 9.4 and 9.5 show typical call data broken down by hour, month, age and outcome.

	Figure 9.1. Map of the 23 sites in England and Wales with the 2 local sites
	Figure 9.2. Hourly call distribution
	Figure 9.3. Monthly call distribution
	Figure 9.4. Age distribution
	Figure 9.5. Call outcomes

9.3. NHS Direct Symptom Surveillance

Recently it has become possible to use NHS Direct call data to add to existing methods of community surveillance. Since 1999 a scheme for ‘influenza like illness’ surveillance has been developed, based on monitoring NHS Direct data (Harcourt et al. 2001; Cooper et al. 2002). Recent terrorist activity has highlighted the need to improve surveillance systems for the early detection of chemical or biological attacks. Therefore, in 2001 the NHS Direct influenza like illness surveillance scheme was expanded to encompass a wider range of symptoms (Baker et al., in press). The aim was to identify an increase in symptoms indicative of the early stages of illness caused by the deliberate release of a biological or chemical agent, or more common infections.

Data relating to ten key symptoms/syndromes (‘cold/flu’, fever, cough, difficulty breathing, diarrhoea, vomiting, rash, lumps, double vision and ‘eye problems’) are received electronically from the 23 NHS Direct call centres and analysed on a daily basis by the Regional Surveillance Unit (HPA West Midlands). These ten key symptoms are intended to be indicative of the early stages of a range of illnesses caused by biological or chemical weapons. Significant statistical excesses (‘exceedances’) in calls for any of these symptoms are automatically highlighted and assessed by a multi-disciplinary team. If an exceedance is thought to represent a threat to the public, information is passed to local health protection teams. Regular bulletins are also disseminated to relevant public health and emergency service teams. Examples of the daily proportions of some of the ten key symptoms are presented below.

	Figure 9.6. Examples of 10 key symptoms (Cold/flu, Cough, diarrhoea, Difficulty Breathing, Fever
	Figure 9.7. Examples of 10 key symptoms (Double vision, Eye problems, Lumps, Rash, Vomiting)

9.4. Recent Trends in the diagnosis of Upper Respiratory Tract Infection in Primary Care.

The Royal College of General Practitioners Weekly Returns Service (WRS) collects morbidity data from a network of 78 sentinel practices across England and Wales. These practices provide weekly summaries of patient consultations and currently cover a population in excess of 650,000.

Infections of the respiratory tract account for 14% of all consultations (Fleming et al. 2002) and are the main reason for the prescribing of antibiotics in general practice in England and Wales. A recent decline in antimicrobial usage in primary care has been noted (Wrigley et al 2002; Majeed et al. 2002) and attributed to a decreased incidence of respiratory infections between 1993 and 2000 (Fleming et al 2003).

To examine recent trends in Upper Respiratory Tract Infection (URTI) presenting to general practice the number of first and new episodes were extracted for the period 1994 to 2002 for the following conditions using the International Classification of Infectious Diseases (9th revision) :- common cold (ICD 460, 465); sinusitis (acute) (ICD 461); otitis media (ICD 381.0, 382.0, 382.9) and acute tonsillitis/sore throat (ICD 462, 463). Data were available by gender for separate age groups. The results are presented for all URTI combined.

In 1995 the mean weekly incidence rate of all URTI peaked at 502 per 100,000 and then decreased by 46% to 269 per 100,000 in 2002 as shown in Figure 9.8.

Figure 9.8. All reported Upper Respiratory Tract Infection, England and Wales 1994 to 2002 (mean weekly incidence per 100,000)

Figure 9.9 shows that the incidence of URTI was highest in infants and decreased progressively with age. Most URTI occurred in the winter months.

Figure 9.9. All reported URTI – mean weekly incidence by age and quarter 1994 to 2002 (males and females combined)

The WRS data has shown that the mean weekly incidence of URTI conditions diagnosed in general practice in England and Wales has decreased by nearly 50% over the period 1994 to 2002. The reasons for the recent decrease are not clear.

Upper respiratory tract infections are all diagnosed clinically and only rarely would microbiological investigation be needed. Recent validation exercises have shown a high degree of internal consistency of WRS data (Ross 1994; Fleming et al 2000; Fleming et al 1999).

In recent years the WRS has moved to a fully automated electronic data recording system making data capture and analysis much simpler. Clinical data are recorded directly in patient electronic medical records and data extraction from these is fully automated. There is no longer the need for duplicate data entry for clinical management and for surveillance purposes. As computer use increased in the early 1990s it was suspected that at least some of the increase seen in 1994 and 1995 could be explained by simplification of data entry procedures. However, for many of the practices the benefits of these improvements were not realised until the latter half of the 1990s when incidence was falling.

Recent campaigns urging people not to consult their GPs with mild viruses may have had an impact. The pressure on doctors (and patients) not to prescribe antibiotics for self-limiting viral illnesses may have resulted in a diagnostic shift towards lower respiratory tract infections in order to rationalise the decision to prescribe. However, WRS data for the incidence of acute bronchitis, of asthma and of influenza-like illness in those age groups not normally vaccinated (Fleming, Sunderland et al. 2000; Goddard et al. in press) have fallen simultaneously and by similar margins as URTI.

Recent trends in the incidence of URTI presenting to general practitioners suggest a substantial reduction which may, in part, explain the recently reported reduction in antibiotic prescribing.

9.5. General Practice Research database

The General Practice Research Database (GPRD) is a database of anonymised patient data owned by the Medicines & Healthcare products Regulatory Agency (MHRA). Data are provided by over 500 practices, which record all significant consultations. The database contains data on 2.7 million patients (around 4.7% of the UK population). It has been widely used to study morbidity and pharmaceutical adverse events. Information collected includes demographic data, medical diagnoses, prescriptions, referrals, hospital admissions and their outcomes and preventative care. One of the strengths of the GPRD is the ability to link diagnoses to prescribing data, which allows the study of conditions common in general practice and their treatment. Several validation studies have shown that there is a high degree of completeness and reliability of data (Jick et al. 1991; Nazareth et al. 1993; Hansell et al. 1999).

We have investigated the possibility of using the GPRD as a surveillance tool by using data from the West Midlands to show trends in diagnosis and treatment of head lice. Diagnoses of head lice increased from a level of 7.5 per 1000 patient years at risk in 1994 reaching a peak of 28.2 per 1000 patient years at risk in 1997. Rates of head lice diagnosed in females were on average twice those for males (Figure 9.10).

Figure 9.10. Annual diagnoses of head lice episodes by sex

The highest diagnosis rates were seen in the 5 to 11 year age group which reached a peak of 139.8 per 1000 patient years at risk in 1997 (Figure 9.11).

Figure 9.11. Annual diagnoses of head lice episodes by age

These findings are comparable with data from other sources but an advantage of the GPRD is the ability to link diagnoses to prescribing data, as illustrated in >Figure 9.12. Total prescriptions for head lice increased from 7.3 per 1000 patient years at risk in 1994 to a peak of 27.1 per 1000 patient years at risk in 1997. Malathion and permethrin were prescribed most often. Carbaryl was prescribed less often and usage fell after 1995. Rates of head lice prescriptions closely followed independent diagnosis rates.

Figure 9.12. Annual insecticide use for head lice

We also looked at further prescriptions for parasiticidal preparations (defined as a second prescription for a parasiticidal preparation issued to the same patient for the same condition within 30 days of the initial prescription). Figure 9.13 summarises our findings. The proportion of further parasiticidal prescriptions issued within 30 days of the initial prescription increased to a peak of 11.5% of prescriptions in 1997. The insecticides prescribed for further prescriptions were compared to those of the original prescription to see whether the same or a different insecticide had been prescribed. In all years the proportion of further prescriptions for the same insecticide was greater than that for a different insecticide. The ratio of the same:different further prescriptions changed during the study period, reaching a high of 5:1 in 2000.

Figure 9.13 Further prescriptions for parasiticidal preparations

The methods developed for this study are being applied to the surveillance of other communicable diseases.

Originally published in Communicable Disease and Public Health (Smith et al. 2003). Reproduced with the permission of the Health Protection Agency.

References

Baker M, Smith GE, Cooper DL, Verlander NQ, Chinemana F, Cotterill S, Hollyoak V, Griffiths R. Early Warning and NHS Direct: a role in community surveillance? J Pub H Med 2003:24(4); 362-368.

Cooper DL, Smith GE, Hollyoak VA, Joseph CA, Johnson L, Chaloner R. Use of NHS Direct calls for the surveillance of influenza – a second year’s experience. Commun Dis Public Health 2002; 5(2): 132-136.

Department of Health, The New NHS Modern and Dependable, 1997.

Fleming DM, Zambon M, Bartelds AIM, de Jong JC. The duration and magnitude of influenza epidemics: a study of surveillance data from sentinel general practices in England, Wales and the Netherlands. European Journal of Epidemiology 1999;15:467-473.

Fleming DM, Cross KW, Sunderland R, Ross AM. Comparison of the seasonal patterns of asthma identified in general practitioner episodes, hospital admissions, and deaths. Thorax 2000;55(8):662-665.

Fleming DM, Charlton J, Charlton JRH, Nicoll A, Smith GE. Impact of infections on primary care – greater than expected. Commun Dis Public Health 2002;5(1):7-12.

Fleming DM, Ross AM, Cross KW, Kendall H. Reducing incidence of respiratory infection and its relation to antibiotic prescribing. British Journal of General Practice 2003;53: 778-783.

Goddard NL Kyncl J, Watson JM. 2003. Appropriateness of thresholds currently used to describe influenza activity in England. Commun Dis Pub Health (in press).

Hansell A, Hollowell J, Nichols T, McNiece R, Strachan D. Use of the General Practice Research Database (GPRD) for respiratory epidemiology: a comparison with the 4th Morbidity Survey in General Practice (MSGP4). Thorax 1999;54:413-419.

Harcourt S, Smith GE, Hollyoak V, Joseph CA, Chaloner R, Rehman Y, Warburton F, Ejidokun OO, Watson JM, Griffiths RK. Can calls to NHS Direct be used for Syndromic Surveillance? A pilot at three sites using influenza as an example. Commun Dis Public Health. 2001: 4(3); 178-82.

Jick H, Jick SS, Derby LE. Validation of information recorded on a general practitioner based computerised resource in the United Kingdom. Br Med J 1991;302:766-768.

Majeed A, Wrigley T. Antibiotic rates in England are falling. Br Med J 2002;325: 340.

Nazareth I, King M, Haines A, Rengel I, Myer S. Accuracy of diagnosis of psychosis on general practice computer system. Br Med J 1993;307:32-34.

NHS Direct Online. http://www.nhsdirect.nhs.uk.

Ross AM, Fleming DM. Incidence of allergic rhinitis in general practice 1982-92. Brit Med J 1994; 308:897-900.

Smith S, Smith GE, Heatlie H, Bashford J, Ashcroft D, Millson D. Head lice diagnosed in general practice in the West Midlands between 1993 and 2000: a survey using the General Practice Research Database. Commun Dis Public Health 2003; 6(2): 139-43.

The Chief Medical Officer, Getting Ahead of the Curve. A strategy for combating infectious diseases (including other aspects of health protection). Department of Health, 2002.

Wrigley T, Tinto A, Majeed A. Age- and sex- specific antibiotic prescribing patterns in General Practice in England and Wales, 1994 to 1998. Health Stat Q 2002;14: 14-20.

[ Chapter 8 | Top | Chapter 10 ]