Public Health and Epidemiology at a Glance

Margaret Somerville, K. Kumaran, Rob Anderson

• Publisher: Wiley-Blackwell
• Year: 2012
• ISBN:
  9781118308646
• Edition: 1
• Topic: Medicine
• Subtopic: Biostatistics

This brand new title in the market-leading at a Glance series provides a highly-illustrated, visual introduction to the key concepts of population-level disease prevention. Accessible, double-page spreads help you understand and appreciate the determinants of health which impact on healthcare services and their effectiveness.

Public Health and Epidemiology at a Glance features high-yield information on all the topics covered at medical school, including:

• Measurement of population health
• Epidemiological concepts of cause and association, surveillance, and risk
• Use of epidemiology in clinical decision-making
• Epidemiology of major health problems such as cardiovascular disease and cancer
• Health promotion
• Basics of health economics
• Basics of health policy, needs assessment, and evaluation

Also featuring self-assessment questions to help test learning, this new title will provide an invaluable resource for medical and healthcare students, junior doctors, and those preparing for a career in public health.

1

Introduction to public health

What is public health? Why do I need to study it? We hear this question a lot from medical students just starting out on their medical careers. There is, of course, the standard definition:

Public health is the science and art of preventing disease, promoting health and well-being and prolonging life through the organised efforts of society (Faculty of Public Health)

… but what does this really mean?

The difference between the clinical and public health roles of doctors (and health services) is often illustrated by the image of people pulling others out of a river (Figure 1a). So busy are these people with saving those who are drowning that nobody has thought to go back upstream to find out why people are falling in to begin with. Public health aims to go upstream to find out why people are drowning. As well as understanding the problem, public health also tries to prevent it or reduce the harm resulting from it. Such action may involve persuading decision-makers to put up effective barriers to stop people falling into the river, repairing damaged river banks or controlling flooding, as well as providing information in the right way to prevent risky behaviour near the river. It may also be appropriate to make sure that the people saving those who are drowning are well trained and at the right place on the river bank to save as many lives as possible effectively and efficiently.

Doctors and other health care professionals spend their time dealing with people with health problems – those drowning people – and in treating individuals as effectively as possible. But many individuals’ ability to obtain and follow medical advice is limited by circumstances outside their control. They may not be able to get to a clinic or hospital or afford the tests, drugs or other treatment once there; they may not understand the advice or treatment because of educational, language or cultural barriers, or may find it impossible to follow because of their domestic or social circumstances. Understanding these ‘upstream’ determinants of health is vital to providing health services that are sensitive to people’s needs and effective in improving health. Methods of addressing them include legislation (e.g. wearing seat belts or motorcycle helmets), fiscal policy (e.g. taxing alcohol and tobacco), local and national social initiatives (e.g. literacy programmes, housing improvements and cycle paths) as well as more specific disease prevention programmes (e.g. immunisation). Taking such action requires a very different approach to that of the traditional healer, one that recognises that doctors and health care professionals may not be able to act directly themselves, but can work with and influence others to take action to improve health. It involves working with many different people, professionals, organisations and communities both within and outside the health sector.

There can be tensions between the traditional clinical approach to individuals’ health problems and this population approach: what leads to improvement in the health of a population as a whole may not mean health improvement for every individual within it. Conversely, doing what is clinically best for the individual patient may mean others are excluded from getting appropriate, or even any, health care. Getting this balance as right as possible is a public health concern.

So public health is not just about acquiring a detailed knowledge base or a specific set of skills; it is also about an approach to health and health problems that is population-based, rational, transparent and fair. The public health approach seeks to identify and quantify health problems at a population or community level and then develop, introduce and evaluate interventions to improve health, monitoring progress to see whether the actions have made a difference. Epidemiology, the study of disease patterns, is the key discipline that helps us to understand population health, but in order to fulfil the role set out in the previous sentence, public health needs to draw on a wide range of other disciplines and knowledge. Statistics, sociology, psychology, health economics, health promotion, management and leadership, health systems and policy all contribute to the public health approach. This book attempts to give you an introduction to this complex and fascinating subject, which is fundamental to the good practice of medicine.

Domains of public health

The scope of public health, as described above, is very wide-ranging, but is generally recognised as falling into three domains (Figure 1b). All three domains draw on the academic disciplines listed above and all collect or make use of information relevant to health, such as population data from the census, data on health service use (e.g. prescribed drugs, hospital admissions or consultations with health professionals), registrations of births and deaths and disease and risk factor prevalence levels (e.g. alcohol consumption or diabetes).

• Health protection covers communicable diseases and environmental hazards, such as exposure to toxic chemicals and poisons. Exposure to hazardous substances at work is covered by the separate discipline of occupational medicine.
• Health improvement includes understanding the wider determinants of health, such as housing, education, poverty and lifestyle risk factors and seeks to improve health through health promotion and disease prevention.
• Improving services is concerned with how the quality of health services can be improved through evidence-based planning, the provision of effective and cost-effective treatment and ensuring that services are available to everyone who can benefit from them.

In the first section of this book (Chapters 2–11), we cover the main epidemiological concepts and methods that underpin evidence-based practice, whether public health or clinically focussed. The second section (Chapters 12–21) covers the types and sources of information used to assess population health status and need for healthcare. The third section (Chapters 22–30) covers health improvement and the final two sections health economics (Chapters 31–34) and health services (Chapters 35–37).

Epidemiology and Evidence-Based Practice

2

Incidence and prevalence

Definitions

Epidemiology: the study of the occurrence and distribution of health-related states or events in specified populations, including the study of the determinants influencing such states, and the application of this knowledge to control the health problems

Incidence (I): the number of instances of illness commencing, or of persons falling ill, during a given period in a specified population; more generally, the number of new health-related events in a defined population within a specified period of time. It may be measured as a frequency count, a rate or a proportion.

Prevalence (P): the total number of individuals who have an attribute or disease divided by the population at risk of having that attribute or disease either (a) at a specified time (point prevalence), or (b) over a specified period (annual, lifetime, one year) (period prevalence).

(All the above definitions come from M. Porta and J.M. Last, Dictionary of Epidemiology, 5th edition, OUP, 2008).

The relationship between incidence and prevalence is modified by the duration of the disease:

High prevalence may result from a high incidence or a long disease duration or both. For example, suppose that in a population of 100 people, 5 new cases of disease occur during the first year of observation (Figure 2a). Assuming that no one in the population had the disease at the start of our observation period, both the prevalence of the disease in the first year and the annual incidence is 5 per 100 people or 5%. If the disease lasts 10 years or more, but no more new cases arise during our second year of observation, then the incidence in the second year is zero, but the prevalence at the end of that time is still 5%.

In Figure 2b, each horizontal bar represents one individual followed over time. The blue section indicates when they have had a particular illness; the green section indicates when they have been well. We can see that two individuals have been well throughout the whole 5-year period that this group has been under observation, while another has also been well for the 3.5 years that he/she has been under observation. One person has been ill for the whole period and another has been ill and then lost to follow-up, or has died, at 2.5 years. Four people developed the illness over the 5-year period, giving us a cumulative incidence over the 5 years of 40%. The seven individuals who were ill during the whole 5-year observation period had a total of 18.5 years of illness, giving us an average duration of disease of 2.6 years.

Rates

Incidence and point prevalence, as used in Figure 2b, are expressed as proportions (percentages), although they were originally expressed as numbers of new and existing cases. In order to express incidence and prevalence as rates, however, the denominator needs to take into account both the number of individuals and the length of time each has been under observation. In Figure 2b, incidence at 3 years is given by the number of new cases arising in the 3-year period (3) divided by the time the individuals have been under observation (29.5 years), giving us an incidence of 10.2 cases per 100 person-years of observation. The period prevalence for years 2–4 is given by the number of prevalent cases (7) divided by the length of time that all individuals were under observation during that time (18 years) giving us a period prevalence of 38.9 cases per 100 person-years of observation. Point prevalence is always expressed as a proportion, as there is no duration of observation to take into account.

The terms incidence and prevalence are frequently used loosely to refer to proportions, rates or numbers, as indicated in the definitions above.

Descriptive Epidemiology

Conventionally, incidence and prevalence are described by time, place and person. Figure 2c shows that mesothelioma death rates in men are much higher and increasing in older age groups (men aged 65–74 years and over 75 years), but have been declining at younger ages since around 1990. The map in Figure 2d shows the geographical distribution of mesothelioma deaths in men from 1981–2005 as the standardised mortality ratios (SMRs, see Chapter 6) for men by local and unitary authorities in Great Britain. The areas with the highest SMRs are shown in shades of pink to red.

Describing prevalence and mortality in this way can suggest possible explanations for the observed disease patterns in time, place and person. Hypotheses generated in this way can then be investigated further with specific studies.

3

Risks and odds

In this chapter, we will look at the main statistical measures used to quantify risk in epidemiological studies.

Risk, Relative Risk and Absolute Risk Difference

A risk is defined as the number of events divided by the total population at risk over a given time period (Box 3.1).

For example, if there are 150 cases of seasonal flu occurring in 1500 staff and students at the medical school in a year, the risk of a staff or student developing seasonal flu during the academic year 2009–10 is:

In epidemiology, useful information is obtained by comparing two groups. Extending the above example, let us assume that there are 1000 students in the school of whom 50 develop seasonal flu. Therefore the risk of students developing flu during the year is 50 ÷ 1000, i.e. 5% (or 0.05 as a proportion). If 100 of 500 staff develop flu, the risk of staff developing flu is 100 ÷ 500, i.e. 20% or 0.2. How can we compare the risks in these two groups? Box 3.2 illustrates two methods of comparing risks in these two groups.

Although there are technical differences, and purists may not agree, in practice relative risk is commonly used as an umbrella term to describe risk ratio and rate ratio. While both relative risk and absolute risk difference are useful to compare two groups, the absolute difference takes into account how common the underlying condition is (i.e. its prevalence) in the population. It has important implications for practice as illustrated by the following hypothetical example.

Example

In a hypothetical study examining the risk of developing leukaemia amo...