Health Statistics Research Paper

The term health statistics system can be used to describe any organized or structured mechanism that involves health statistics. It can therefore include systems that are used for collecting, analyzing, interpreting, and making available quantitative data in relation to health status or health services activity, as well as referring to the actual collections of quantitative health related data. Increasingly these statistics systems are computerized, and the increased availability and ease of access of information technology means that this trend towards computerization of health statistics systems is likely to continue in the future. However in many countries, paper based systems are still used for the collection of health statistics and, in some cases, the aggregation and dissemination of these health statistics are also not computerized. Therefore focusing solely on computerized health statistics systems, when looking for health information can mean that much valuable information is missed.

Decisions about health and health care, whether made by policy makers, by health professionals or by the public, are influenced by the information that is available. Much of this information is derived from official and other health statistics. Effective health care is therefore dependent on the process of gathering, analyzing, and disseminating timely and appropriate information. The quality and accessibility of the health statistics systems, from which this information is derived, can have a major influence on how health and health care systems are run and used.

When discussing statistical systems the terms data and information are often used interchangeably, but it is important to recognize the differences. Data are specific facts recorded in some way for possible future use, for example, the annual expenditure for a specific department in a hospital. Data becomes information when it is has meaning, usually by putting it in some relationship with other data. For example, relating expenditure for a hospital department to the services provided by that department and comparing this with the data for similar departments in other hospitals, provides information on how expensive it is to deliver the service in the hospital concerned compared to other hospitals. Health statistics are derived from data, and should provide information.

Another term that is increasingly used in relation to health information is knowledge. By putting information in the context of other information, including both quantitative data and textual information, deductions can be made that increase knowledge. Knowledge provides the basis for making decisions in a rational way. Knowledge, based on information, allows appropriate actions to be selected in order to achieve specific objectives. This is often referred to as evidence based decision-making. This is one of the main purposes of gathering data and producing health statistics.

1. Scope Of Health Statistics

In some respects, the terms health information system and health statistics system can be used inter- changeably. Where the focus is on the collection of the data, or on the statistical calculations that need to be carried out on the data to produce useful information, then the term health statistics may be used more often. However, if the focus is on the results of this data collection and analysis, and on making this information available, then the term health information is used more often. The use of telemedicine is expanding the scope of health care information, and so the definition of health information is growing to include not just health statistics but also multimedia audio, video, and high-resolution still images (Balch and Tichenor 1997).

The scope of what health statistics systems cover can be defined in several ways:

(a) by the type of data and information included; (b) by where the data are produced;

(c) by where and how the information is held and made available; and

(d) by the purposes for which the information is used.

A narrow definition of health statistics could conclude that the purpose for collecting statistical information should be primarily to satisfy the needs of government. This would mean that health statistics systems would just cover those systems designed to produce official health statistics for government use. However, the current view in many countries, and of many of the producers of official statistics, is that the statistics are produced to serve not only government but also the wider community. The members of this wider community of health statistics users come from different organizational backgrounds, including legislative bodies, government departments, health service organizations, researchers, voluntary organizations, and patients, as well as the public in general. The health service organizations that need to use health statistics can also be wide ranging, depending on the health system in operation in the country concerned. They are likely to include both the public and private sector, and both providers of health care services and purchasing organizations.

This broader view of health statistics users is both driven by, and enabled by, the rapid spread of the use of the Internet and the World Wide Web. Familiarity with accessing information via the web has raised the expectations of both health professionals and the public in relation to the availability of health information. Increasingly producers of health statistics have started to respond to this by making the information they produce available in different formats and through different mechanisms to suit the differing needs of the user community.

The widespread use of the Internet has meant that this is now used as a common method of disseminating information about statistics, as well as providing access to the actual data in some cases. Access to statistics held on data archives, many of which include health statistics, is also possible via the web. Web-based approaches can also allow information from different systems to be brought together, either by making these available via a web site, or by using web based technologies to access information held on different systems.

The design and coverage of health statistics systems, described later in this research paper, therefore needs to be seen as being currently very much in a transitional stage. For many systems, developments are ongoing which will take advantage of the opportunities, and also meet the challenges, of the technology that is now available for bringing together information from different sources and for providing wider access to this information (Sandiford et al. 1992).

2. Use of Health Statistics

There are many different individuals and organizations involved in planning, managing, running, investigating, and using the health care system. Health statistics systems should ideally be designed to meet the information requirements of this health care community. Some of the main uses of health statistics are summarized in Sect. 2.1 to 2.9.

2.1 Information For Policy Makers And For The Management Of Health Care Delivery

Policy makers at both national and local level need information to support decisions on planning and funding of services. This applies both to policy makers responsible for publicly funded services, and to organizations such as health insurance companies involved in the funding and provision of private health care services (Roos 1999, Roos et al. 1999). Health information systems can provide decision-makers with the capability to:

(a) Make critical comparisons across regions and subregions of residents’ health status, socioeconomic risk characteristics, use of hospitals, nursing homes, and physicians.

(b) Carry out analyses, in relation to the population served, of demographic changes, expenditure patterns, and hospital performance.

(c) Carry out outcome research across hospitals and countries to support recommendations on effective treatments,

(d) Make comparisons of expenditure relative to outcomes to support purchasing decisions

(e) Carry out longitudinal research on the impact of health policy changes.

This information helps to answer questions such as what health care services do particular population groups need and, by comparing this need to the current service provision, which populations need more services and which need less? (See Slater 1999, Azubuike and Ehiri 1999, De Kadt 1989.)

In addition, health information systems are important support tools in the management and improvement of the delivery of health care services. Experience in several countries suggests that comparison of hospital-specific data influence hospital performance. If statistics on provider performance are made available this ensures not only prudent and costeffective health care purchasing, but also it gives health care providers comparable information to enable them to improve the care they deliver.

2.2 Health Surveillance Systems

In health terms, surveillance is the close monitoring of the occurrence of selected health events in the population. This implies that the events to be monitored can be specified in advance. This may include setting up a notification system for the health event, to supplement data available from routine health data collection systems, for example in the case of infectious diseases such as tuberculosis. However, the events monitored can also include certain signs and symptoms which can be seen as ‘indicator events’, to give warning of unforeseen problems.

Surveillance methods have been developed as part of efforts to prevent or control specific health events. Surveillance systems are therefore dynamic, in contrast to archival health information, which can be used for epidemiological analysis. The events that can be controlled in surveillance systems include infectious diseases, maternal and child health, chronic noninfectious diseases, injuries, and the effects of exposure to occupational or environmental risk factors.

2.3 Types Of Health Information

The ultimate aim of any statistical work in the health sector should be to assist in the maintenance and improvement of the health of the population and in the delivery of health care services. Health information can therefore be defined as any information that is needed to support these activities.

It is important to be able to access both health service and non-health service data when looking at health needs, since the determinants of health often lie outside the health service. Work in relation to prevention of illness and the promotion of good health therefore require access to data from a wide range of sources. In response to this requirement health statistics systems need to cover, or provide access to, information not only on all aspects of the health care systems, but also on factors that influence health and the availability and use of health care.

There are different models of health systems used in different countries, and these have an impact on the health statistics systems, both in terms of the data collected, and hence the information potentially available, and in terms of where the information is held and who has responsibility for it. There may also be variations in the organization of health care, hospitals and hospital information systems within a given national environment (Saltman and Figueras 1997). However, the differences in health system models usually relate mainly to funding mechanisms, rather than to the basic activities that need to be carried within the health sector. This means that the types of information held are usually similar, even if the organizations involved are different.

The types of information covered usually include:

(a) Health status

—Health of the population

—Vital statistics (births, deaths)

—Health surveillance—for both infectious and non infectious diseases

(b) Registers

—Patient registers e.g., persons insured, persons registered with a family doctor

—Disease registers e.g., cancer

(c) Health service resources

—Health care facilities

—Health care staff

—Other resources e.g., medical equipment, drugs

(d) Health service activity

—Hospital and other secondary care activity

—Primary care activity

—Financial information

(e) Explanatory factors

—Economic information

—Sociodemographic information

—Environmental information

(f) Other related data (often government or public sector statistics)

—Social care sector activity

—Police—crime figures, road traffic accidents

—Education

Some of the main types of information listed are discussed in more detail in later sections. In addition to the quantitative data listed in Sect. 2.3, the interpretation of information from health statistics systems often requires reference to textual information in relation to current policies and guidance, and also to research findings.

Some of the information is obtained using administrative data, which are routinely collected as part of the process of running and monitoring the health system. In addition information is available from censuses, either using data from national population censuses or from censuses of specific groups, e.g., census of patients in hospital at a particular time. Health statistics are also available from a wide range of survey data. These can be international, national, regional, or local surveys, and can include surveys that are part of a regular survey program as well as ‘one off’ surveys carried out for a specific research or policy analysis study.

The increased use of computerization in health statistics systems has already been mentioned. Computers can store large volumes of data and can process them quickly. This means that if data are entered on to a computer at source it may not be necessary to summarize data before onward transmission. For example, every patient’s record can be stored and sent rather than just counts of patients with particular diseases. Consequently converting data into information can be postponed until a specific issue raises the need for a particular piece of information. Much more flexibility is therefore available in how the data are aggregated to create information.

Also data held in electronic form is no longer confined to being in one place at a time. Several people can simultaneously be reading the same record in quite different locations. This creates possibilities for sharing information, which were never feasible with paper systems. However it also increases the risk of unauthorized access to data, and data confidentiality issues in relation to health statistics systems are discussed in a later section.

2.4 Health Of The Population

Health monitoring requires measures of incidence, prevalence, and severity of illnesses, and the consequences of illness, classified by age, sex, socioeconomic, and geographical characteristics of the population. There are no ideal measures, and several sources need to be used in combination, for example using both health surveys and information about patients in hospital. The varied sources of population health information provide different pieces of the overall health picture.

Ultimately the objective of health care expenditure is to improve the health of the population not for its own sake but so that people can live happier more prosperous lives. Gains in health status might be regarded as the ultimate measure of a health program’s success. Composite measures of health status such as the DALY (Disability Adjusted Life Year) and the QALY (Quality Adjusted Life Year), although developed from different disciplinary backgrounds, both attempt to summarize the potential gains from medical interventions (Williams 1999, Jenkinson 1994). Whereas health economists developed QALYs, DALYs have their roots in epidemiology. As a consequence, the perspective of the first is based on patient assessment of pain and suffering while the second is disease driven with a medical assessment of the amount of suffering.

2.5 Vital Statistics—Mortality Data

The most absolute measure of ill health is death, and mortality statistics are one of the principal sources of information about the changing patterns of disease and the differences in these patterns between countries. Since mortality statistics are widely available death rates have traditionally been used as crude, but often effective, surrogates for more comprehensive measures of disease. This is often justified by the fact that death is unambiguous and the data collection is complete. However use of mortality data alone can provide misleading impressions of the burdens of disease in society. For example, in certain parts of the world, musculoskeletal disorders (including rheumatism and arthritis) are the most important current causes of limiting long-standing illness but are one of the least prevalent causes of death.

Death rates vary by age and by sex and so, if comparisons are to be made between populations with different age and sex composition then some adjustments are needed. A common approach is to use age standardized rates and ratios (often referred to as Standardized Mortality Ratios or SMRs). These are calculated by applying the rates found for a particular population to an agreed standard population. Standardization is also used when comparing other measures, such as hospitalization rates, which are likely to be affected by the age and sex structure of the population. Another measure commonly used when considering premature mortality or avoidable deaths is Potential Years of Life Lost (PYLL). This highlights the loss to society as a result of early deaths (for example, below age 65 or 75). The figure is the sum, over all persons dying from that cause, of the years that those persons would have lived had they survived to the stated age (e.g., 65).

2.6 Vital Statistics—Births Data

Although systems vary between countries, the systems from which birth and maternity data can be derived are likely to include civil registration of births, the notification of births, and hospital-based data collection systems. Birth registration is the oldest source of data about births. In most countries the primary reason for setting up systems for registering births, marriages, and deaths was to provide civil documents to record the events for legal purposes. To obtain a full picture of birth and maternity care, data also needs to be collected about facilities and staffing of the maternity services and about the socio-economic context in which childbirth occurs.

2.7 Patient Registers

Where health insurance systems are in operation, there will be registers for the population covered by the schemes. These registers are likely to include information such the individual’s insurance status, name, date of birth, address, and identification number, plus possibly other socio-economic or medical status data, e.g., whether the person is a diabetic. Other data related to individuals would be the health insurance contributions made on their behalf from employers or through other mechanisms. Information on dependents covered by an employed person’s contributions may also be included.

In many health systems patients will be registered with a family doctor or primary care center. These registers are likely to include some of the same basic patient information as for the health insurance registers, as well as some health status information. Records of preventive services such as immunizations carried out by family doctors may also be included.

2.8 Disease Registers

Separate registers are often kept, usually at local or regional level, for patients with specific diseases, and cancer registration is an important example of this. Since, in many parts of the world, cancer-related services are consuming ever-increasing health resources; the information from cancer registries is important in helping health care planners, researchers, and policy-makers formulate strategies to meet this challenge.

Cancer registration is the collection and classification of data on all cases of cancer occurring within a defined territory. The primary source of data on which cancer registration is based is the patient’s clinical case notes. This holds true whether the information is derived from hospital departments, primary or community care, or the private sector.

Cancer registration is a worldwide process. The longest established cancer registries are in Europe and North America, but increasingly they are being developed in the countries of Asia, Africa, and South America. Attempts to record comprehensive information on the incidence and distribution of cancer have been made as part of individual censuses and studies since the last century, but it was not until the late 1940s that cancer registration really got underway. In 1959 the World Health Organization (WHO) produced recommendations for the establishment of cancer registries, and, in 1965 the International Agency for Research on Cancer (IARC) was established as a specialist research center of WHO.

2.9 Hospital Data

A wide variety of data is available from hospitals for the purpose of health service analysis and performance monitoring. Where computerized systems are not available, much of this is obtained as aggregate counts of patients, and of resources used, falling into predefined categories. However, where computerized health information systems are in operation, individual records of care can be accessed for each patient which contain wide ranging details of the diagnoses, the treatments delivered, and the resources used.

2.10 Non-Health Service Data

As has already been mentioned, it is important to be able to access non-health service data when looking at health needs, and when comparing statistics on health service activity. In many countries there is also an increasing emphasis within the health sector on partnership working with other agencies, and the need to share information on common and related issues. Health and health care activities can be affected by activities in other sectors, and monitoring and planning of health services activity need to take this into account.

Non-health service data sources include population estimates, vital statistics, and census data. Vital statistics covering data on births and deaths have already been mentioned. Population estimates and census data are discussed elsewhere in this publication. Various deprivation indices, some of which are derived from Census data, are also used when analyzing and interpreting statistics relating to the health sector.

The links between housing and health have been well established. Some data on housing can usually be obtained from population censuses, and information about housing stock may also be available from local government statistical systems. Poor air quality is also considered to be detrimental to health. Once again some information may be available from government statistical systems supplemented by special studies. Other statistics relevant to health that may be available from government sources include information on education, income, unemployment, transport, and crime.

3. Standard Coding Classification Systems

If health statistics are to be compared across different organizations, regions, and countries, it is important that common coding and classification systems are used throughout the health sector and that they are used consistently. The WHO International Classification of Diseases is a classification of specific medical conditions and groups of conditions which is determined by an internationally representative group of experts who advise the World Health Organization (WHO), which publishes periodic revisions. It was first introduced a hundred years ago and the version in current use worldwide is the tenth revision (ICD 10). In addition to the use of international classifications such as ICD10, many countries and international communities have undertaken initiatives to standardize the health data they collect in order to facilitate comparisons across countries. An example of this is the European Union’s program ‘Community Action on Health Monitoring’ that was adopted for a five-year period with effect from 1 January 1997 (for EU health information projects, see CORDIS website). Its stated objective was to contribute to the establishment of a Community health monitoring system that makes it possible to:

(a) Measure health status, trends and determinants throughout the Community.

 (b) Facilitate the planning, monitoring and evaluation of Community programs and actions.

(c) Provide member states with appropriate health information to make comparisons and to support their national health policies.

(d) Develop methods and tools necessary for analysis and reporting on health status, the effect of policies on health and on trends and determinants in the field of health.

The program is structured in three pillars as follows:

(a) Pillar A deals with the establishment of Community health indicators and with conceptual and methodological work related to the process of making the data comparable and for identifying and developing suitable indicators;

(b) Pillar B deals with the development of a Community wide network for the sharing and transferring of health data between member states, the European Commission and international organizations;

(c) Pillar C deals with the development of methods and tools necessary for analysis and reporting and the support of analyses and reporting on health status, trends and determinants and on the effect of policies on health.

These activities are being carried out in close cooperation with the member states and also with institutions and organizations which are active in the field of health monitoring, in particular: the World Health Organization, the Organization for Economic Cooperation and Development, the International Labor Organization (OECD Health Data, WHO Health for All Database, WHO World Health Report 2000, WHO Health in Transition profiles).

4. Managing Confidentiality

One important aspect of health information, and therefore of the systems that relate to health information, is confidentiality. This is because the information held may relate to identifiable individuals. Many countries, as well as international communities such as the European Union, have strict laws requiring organizations, including government agencies, to register the types of personal information they hold, the purposes they use it for and who they share it with. The law gives people the rights to inspect the data which organizations hold about them and to challenge inaccuracy and mis-use. Information systems and data flows have to be designed to respect privacy with personal identifier data removed when it is not needed and security precautions in place when it is needed.

Some principles that commonly apply to collecting, holding, and using health data are that:

(a) Health authorities should justify the collection of personally identifiable information,

(b) Patients should be given basic information about data practices,

(c) Data should be held and used in accordance with fair information practices,

(d) Legally binding privacy and security assurances should attach to identifiable health information with significant penalties for breach of these assurances,

(e) Disclosure of data should be made only for purposes consistent with the original collection,

(f) Secondary uses beyond those originally intended by the data collector should be permitted only with informed consent.

5. Recent Developments

Some of the trends in technology affecting health information systems have already been described. Another feature of planned developments in systems is a move towards integrated, patient-centered computing. Currently information is often collected according to the activity carried out e.g., by hospital department, which can make it difficult to obtain a complete set of information for an individual patient, or group of patients. The patient centered approach means that the computer systems and files are designed around capturing information along the whole continuum of care for a patient (Nold 1997).

It is also expected that health statistics systems will move towards higher-speed data transmission, and greater network capacity, following national and international computing standards. This will allow more rapid access to health statistics, as well as facilitating the bringing together of information from different sources—leading to improved information and greater knowledge about population health and health care (Mendelson and Salinsky 1997).

Bibliography:

1. Azubuike M C, Ehiri J E 1999 Health information systems in developing countries: Benefits, problems, and prospects. Journal of the Royal Society for the Promotion of Health 119(3): 180–4
2. Balch D C, Tichenor J M 1997 Telemedicine expanding the scope of health care information. Journal of the American Medical Informatics Association 4(1): 1–5
3. CORDIS—website http:/dbs.cordis.lu/
4. De Kadt E 1989 Making health policy management intersectoral: Issues of information analysis and use in less developed countries. Social Science & Medicine 29(4): 503–14
5. ICD10 http://www.who.int/whosis/icd10
6. Jenkinson C (ed.) 1994 Measuring Health and Medical Outcomes. UCL Press, London
7. Mendelson D N, Salinsky E M 1997 Health information systems and the role of state government. Health Affairs 16(3): 106–19
8. Nold E G 1997 Trends in health information systems technology. American Journal of Health-System Pharmacy 54(3): 269–74
9. OECD Health Data—website http://www.oecd.org/els/health/
10. Roos N P 1999 Establishing a population data-based policy unit. Medical Care 37(6 Suppl): JS15–26
11. Roos N P, Black C, Roos L L, Frohlich N, De Coster C, Mustard C, Brownell M D, Shanahan M, Fergusson P, Toll F, Carriere K C, Burchill C, Fransoo R, MacWilliam L, Bogdanovic B, Friesen D 1999 Managing health services: How the Population Health Information System (POPULIS) works for policy-makers. Medical Care 37(6 Suppl): JS27–41
12. Saltman R B, Figueras, J 1997 European Health Care Reform—Analysis of Current Strategies. WHO Regional Publications, European Series, No 72
13. Sandiford P, Annett H, Cibulskis R 1992 What can information systems do for primary health care? An international perspective. Social Science & Medicine 34(10): 1077–87
14. Slater C H 1999 Outcomes research and community health information systems. Journal of Medical Systems 23(4): 335–647
15. Williams A 1999 Calculating the global burden of disease: Time for a strategic reappraisal? Health Economics 8(1): 1–8
16. WHO Health for All Database—website for Health for All in the 21st Century, http://www.who.int/archives/hfa/
17. WHO Regional Office for Europe, European Health Observatory Health in Transition profiles, http://www.observatory.dk/frame/index/main.htm
18. WHO World Health Report 2000 http://www.who.int/
19. World Bank Data, http://www.worldbank.org/