Intended for healthcare professionals

Education And Debate Health needs assessment

Assessing health needs in developing countries

BMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7147.1819 (Published 13 June 1998) Cite this as: BMJ 1998;316:1819
  1. John Wright (wrightj{at}brihosp.mhs.compuserve.com), consultant in epidemiology and public health medicineb,
  2. John Walley, senior lecturer in international public healthb
  1. a, bNuffield Institute for Health, University of Leeds, Leeds LS2 9PL
  1. Correspondence to: Dr Wright

    This is the last in a series of six articles describing approaches to and topics for health needs assessment, and how the results can be used effectively

    In most developing countries, the evolution of health services has been dominated by Western models of health care. These have rarely taken into account how local people explain illness, seek advice, or use traditional healing methods. The emphasis has been on hospitals and curative care rather than on trying to address local health needs equitably and effectively. Since the Alma Ata declaration on primary health care, more attention has been given to increasing coverage of basic services and preventing common diseases. However, the bias in resource allocation towards secondary care and urban areas remains.

    Summary points

    Timely and accurate information is essential if health services in developing countries are to meet the needs of their populations

    Routine health information can provide an epidemiologically based assessment of ill health and identify what health services are needed

    Community appraisals can provide valuable insight into patients' needs as well as empowering communities

    Emergency health needs are similar whatever the disaster. Community involvement, good surveillance, and foresight are important

    The global burden of disease can be represented by disability adjusted life years; these can help to identify international health needs

    Health needs are changing and new challenges from chronic diseases and HIV infection must be faced. Better coverage of preventive and essential healthcare services has led to greater emphasis on improving the quality of health care and ensuring that the most efficient use is made of scarce resources. For example, infant mortality has fallen dramatically in the past two decades through interventions such as oral rehydration for diarrhoea and immunisation programmes. With fewer children dying there has been greater emphasis on the need to tackle the causes of infant and child morbidity. Families can be smaller, and this has highlighted the need improve the availability of family planning.

    If health services are to respond to the changing health needs of their local populations, then planners and managers need useful and timely information about the health status of these populations. Some of this information can come from routine data sources or may be collected from large, one-off population studies. Some information can be obtained from community surveys.

    Figure1

    Top 10 causes of admission over five years, all hospitals, Mashonaland Central Province, Zimbabwe. Figures exclude normal deliveries, which comprise 27% of admissions in 1995 and 25% in 1994

    Routine information

    Information about diseases or use of health services can help to build up a picture of the health needs of a local population.1 Such epidemiological information can come from national, regional, or local sources.

    • National census data can provide information on the age and sex distribution of a population. This information can be used to calculate crude birth rates and fertility rates

    • Death certification and registers can provide information on the cause and place of death. Infant mortality rates can be calculated from the number of liveborn infants who die in the first 12 months of life

    • Hospital inpatient records can be used to obtain numbers of admissions, cause of admission, and length of stay, and outpatient consultations can be used for numbers of patients and diagnoses (figure)

    • Disease notification systems can provide information on important infectious diseases

    • Maternity unit statistics can describe births rates, maternal ages and parity, numbers of low birthweight (<2500 g) babies, and maternal mortality

    • Pharmacy information provides information on the use of essential and non-essential drugs

    • Laboratories can provide information on the appropriate use of tests and numbers of positive tests (for example, sputum samples for pulmonary tuberculosis, malaria blood slides)

    • Workplaces can provide data on absences due to sickness, occupational injuries, and regular employment health checks.

    This information provides a snapshot of a population's health—but without comparative information this will be of limited use in planning health services. Comparison can be with other populations (national or regional) or with the same population over time.

    The disadvantage of routine information is that it is often inaccurate, incomplete, and out of date. For example, outpatient records may give only the main complaint of patients attending and may not distinguish new patients' visits from repeat visits. Notifiable diseases may be missed, and when they are picked up they are often not reported. It is also difficult to make generalisations about a local population from routine data. For example, people who attend a hospital are more likely to reflect a more affluent and urban population. One-off studies can provide more detailed, relevant, and accurate information on a specific topic (box) but are time consuming and costly.

    Combining different methods of needs assessment2

    Bacterial and tuberculous meningitis is an important cause of morbidity and mortality in developing countries despite the availability of effective treatment.

    Epidemiological assessment—A national study was undertaken in Swaziland to describe the epidemiology, clinical features, and outcomes in each case of meningitis admitted to hospital. The overall case fatality was found to be 42% in all ages and 63% in adults. Significant association with a period of drought was found, and the increasing contribution of HIV infection was highlighted. The results also identified the age distribution and aetiology of meningitis in the country and allowed an assessment of the potential impact of immunisation programmes.

    Community appraisal—Semistructured interviews were carried out on a random sample of mothers attending a health centre. These were used as the basis of a focus group discussion with a purportedly selected group of health workers. The need for education about the awareness of symptoms and the importance of prompt referral and treatment was identified.

    Action—To reduce the high mortality from meningitis by reducing delays in treatment, a coordinated education campaign for the public and health workers, using posters and outreach teaching sessions, was undertaken.

    Community appraisals

    Community appraisals describe approaches to needs assessments that emphasise involvement of local people. A confusing number of terms describe similar methods: rapid evaluation methods, rapid appraisal methods, rapid community surveys, rapid rural appraisal, relaxed rural appraisal, participatory rural appraisal.3 4 5 6 7 The development of rapid appraisal methods during the 1980s came in recognition of the time consuming and rigid nature of traditional epidemiological and questionnaire surveys. Experience with these appraisal methods showed that when they were done well they provided valuable, reliable, and timely information on health status, knowledge, attitudes, and behaviours. More recently, emphasis has been placed on encouraging people to participate in their own appraisal (for example, participatory rural appraisal).3 4 5Many of the principles behind these techniques stem from the formative work of Paulo Freire in enabling oppressed people to understand and address their own educational needs.8

    In community appraisals the assessors support and facilitate community understanding and action rather than just record information (see box above for example of one programme). Local communities can be empowered by the opportunity to participate in health planning, and health workers have the opportunity to appreciate the perceived strengths and weaknesses of services.

    Community appraisal: an example

    Factors affecting participation in nutrition, health, and development in commercial farms in Zimbabwe9

    The workers and their families on commercial farms are one of the most disadvantaged groups in Zimbabwe. A farm health programme has been operating for 15 years in Mashonaland Central Province, including child health and preschool and nutrition activities. As malnutrition in children under 5 remains more common on the communal farms than elsewhere, a better understanding of the factors influencing nutrition, health, and development is needed.

    Eight farms, ranging from well developed to underdeveloped, were selected. Permission of each commercial farmer was requested by telephone and followed up by an explanatory letter delivered by hand. On each farm the commercial farmer or representative was interviewed.

    Participants for group discussions were recruited randomly among workers with preschool children, aiming for 6-8 female workers, 6-8 permanent male workers, and 6-8 seasonal workers. Anyone who seemed to hold some kind of authority was tactfully removed from the group discussion by asking them to assist in drawing the social map, which was drawn on the ground and then copied.

    The research investigated:

    • Knowledge, attitudes, and practices relating to health

    • Felt needs, priority problems, opportunities, and solutions

    • Factors affecting communication

    • Factors affecting participation in health activities

    • Factors likely to assist or hinder an intervention programme.

    Results

    Children's nutrition was not viewed as a priority problem by farm workers or farm owners. Farm workers gave poor working conditions, working hours, low salaries, and lack of family food as priorities; health care for children came much lower on the priority rankings.

    The workers are a fragmented community with no sense of belonging to a group. There is tension between permanent workers, who have better conditions, and seasonal workers.

    An unhealthy child is described as dirty, sick, thin, eats cold food and has a pot belly, and is miserable. Contributory factors include parental fighting, inadequate food, sickness, and lack of child care at home or at preschool.

    Issues likely to influence negatively participation included zvondo jealousy and mistrust among women: for example, not organising a cooking roster for the preschool, as they don't want the woman whose turn it is to cook to benefit from the food. Another example is poor response by the commercial farmer to efforts to improve workers' health once the toilet pits were dug, the farmer failed to provide cement and a builder to finish the job.

    The information collected in community appraisals is used to develop acceptable and sustainable programmes in partnership with the community. These may be programmes of health care, nutrition, or family planning that improve services for the community. The same methods can be used to monitor and evaluate the developments.

    Whatever method is used for appraisals, the emphasis is on qualitative techniques of interviewing and listening to people.9 Methods of community appraisal include the following:

    • Summarising existing information from routine sources or previous surveys (for example, causes of morbidity and mortality)

    • Exit interviews after a clinic visit to obtain the patient's perspective on the quality of care and understanding of the health messages received (for example, checking that the mothers of children with diarrhoea understand how to make up oral rehydration solution)

    • Interviews with health workers (for example, to assess people's perception of local needs, interviews can be structured with a standard list of questions, or semistructured, with just a list of topics that need to be covered)

    • Ranking of priorities or preferences (for example, asking local people to produce a “league table” of needs)

    • Case note review and audit (for example, examining the recording of tasks and health education given to patients)

    • Household survey to assess family health needs (for example, seasonal variation in food intake and accessibility to clean water)

    • Focus group discussion to obtain the opinions of a specific population group (for example, a facilitator guides the group of purposefully selected informants through a framework of questions that aim to stimulate discussion and communication of opinions; an assistant takes notes of the discussion for later analysis)

    • Direct observation of chosen indicators or behaviours (for example, the performance of health workers in communication or clinical skills).

    The assessors need to have good listening skills, a recognition that communities know their own needs, and common sense in analysing the results. Some training is necessary to provide the assessors with the skills needed to undertake appraisal techniques and generate good quality, reliable findings. They must beware of generating false hopes in the community for what can be achieved.

    Steps in community appraisal

    • Define aims of appraisal

    • Identify community for assessment

    • Identify study team and train in qualitative techniques

    • Examine available information

    • Define key questions and issues

    • Pilot questions in interviews or questionnaires

    • Identify key informants

    • Choose and use appropriate methods

    • Analyse information after each interview

    • Write report and develop action plan

    The choice of subjects for questionnaires or interviews will determine whether the results can be generalised. This sampling can be done randomly, systematically (every fifth house in a village, say), or by purposefully selecting key informants (people with expert knowledge: patients, mothers, sex workers, chiefs, elders, church leaders, shopkeepers, health workers, government officials). Care should be taken when selecting key informants that they reflect the range of different interest groups.

    Ideally a combination of methods should be used when assessing health needs—for example, analysis of routine health data plus a questionnaire or focus group. This allows cross checking and validation of results, and it increases their relevance or generalisability to the study population. Routine population data can be superficial and inaccurate; however, they do allow a quantitative comparison with other population data. A small number of interviews may not provide opinions representative of the whole community but can show people's true priorities.

    Language and literacy barriers may arise in discussion of complex health issues. Techniques to overcome these barriers in non-literate populations include community mapping, seasonal calendars, Venn (chappati) diagrams, and dramatisation techniques. 4 5 These visually based methods provide opportunities for local people to explore and analyse their needs in their own terms and enhance their involvement in the assessment.11

    People oriented planning

    The United Nations High Commission for Refugees has developed a simple needs assessment tool called people oriented planning to help guide decisions about refugee needs:

    • Which foods should be supplied, and to whom?

    • How should they be distributed?

    • Who should live where?

    • What are the critical medical needs?

    • What are the cultural patterns of health care?

    • How are target groups best reached?

    This is approached through an analysis of the refugee population profile, activities, and use of resources. Specific questions about the refugees help to clarify what activities people did (farming, teaching, social, political, house building) before their displacement, who did what and when.

    Emergency needs assessment

    Quick decisions and actions are imperative in the aftermath of a disaster. The immediate, life supporting needs after any major disaster are similar whether the cause is of gradual onset, such as drought, famine, or war, or sudden onset, such as floods or earthquakes. These include clean and adequate water and sanitation; adequate food rations; shelter—including clothing and blankets; and essential medical care.12 Information must be obtained not only from government or other agencies (including, increasingly, the international media) but from the affected community. This community will have the capacity to help itself, and any disaster response should build on this.

    Involving the community is essential in assessing the effects of the disaster and targeting vulnerable groups (young children, elderly people, pregnant women). It is also vital to avoid cultural problems. Some problems (such as sending pork products to Islamic countries) can be avoided with intelligence. Others require more insight: a famine relief programme ran into problems because the affected population, which was used to a staple of white maize, had strong traditional beliefs that the yellow maize being distributed was inedible and poisonous.13

    In addition to considering immediate needs, it is important to plan for the future. A community dependent entirely on donor food supplies will be vulnerable when these are withdrawn, especially if normal food production is still disrupted. Good surveillance systems to monitor health and malnutrition are also vital. For example, anthropometric surveys of children in refugee camps or outreach clinics, measuring weight for height or upper arm circumference, can provide valuable nutritional assessments.12 Monitoring of infectious diseases such as measles can prompt timely immunisations.

    Global needs

    National and international health needs are also important in planning health services. Most assessments of the relative importance of different diseases are based on how many deaths they cause. This convention has certain merits: death is an unambiguous event, and the statistical systems of many countries routinely produce the data required. There are, however, many diseases or conditions that are not fatal but that are responsible for great loss of healthy life: examples are chronic depression and paralysis caused by polio. These conditions are common, can last a long time, and often lead to considerable demands on health systems.

    Global needs are represented by the global burden of disease. This burden of disease includes both morbidity and mortality. Morbidity can be assessed according to the amount of disability—for example, from blindness—and mortality can be expressed in terms of life years lost. The needs can then be expressed through a combined measure of such as the disability adjusted life year (DALY).14

    The proportion of disability and loss of life varies from disease to disease so there will be more disability due to leprosy but more years of life lost from tuberculosis. Overall the global burden of disease, when calculated as disability adjusted life years, is made up of about two thirds from years of life lost (mortality) and one third from disability (morbidity).

    Disability adjusted life years can be used to rank diseases in order of magnitude of burden of disease in developing countries. The existing rankings can be compared with a prediction of the future.15 The table shows the scale of the demographic and epidemiological transition anticipated by 2020, with depression and traffic accidents predicted to be the biggest burdens of disease.

    Projected burden of ill health in the developing world*)

    View this table:

    Disability adjusted life years should be interpreted with caution because of the assumptions that are made.16 For example, the combination of discounting and age weighting means that an infant's death equates with the death of a young adult. Disability adjusted life years are based on incomplete, internationally available data that may contain inaccuracies, and they are calculated on the basis of specific diseases or disease groups. Many diseases have multiple outcomes, and interventions may reduce the burden for more than one disease. For example, treatment of diabetes will reduce the risk of stroke, coronary heart disease, and renal failure.

    To date, disability adjusted life years have been calculated globally and by WHO region. Attempts are being made to estimate the national disease burden, as in Ghana, but limitations in data make this a difficult task.

    Medical information needs

    As in developed countries, evidence of effectiveness is an essential component of needs assessment. Attempts are currently being made to improve access to research information and effectiveness information using the internet, including:

    EpiInfo

    EpiInfo is a software package developed by the Centres for Disease Control and Prevention in the United States. It allows easy questionnaire design (EPED), data processing, and analysis. The analysis module provides a user friendly statistical package. It is considered public domain and may be freely copied. Its simplicity and free availability make it ideal for researchers in developing countries. (Contact: Division of Surveillance and Epidemiologic Studies, Epidemiology Program Office, Centers for Disease Control and Prevention, Atlanta, Georgia 30333, USA.)

    Despite these limitations, disability adjusted life years are the only data available that combine morbidity and mortality into a simple indicator of burden of disease. This can be used to identify current and future international health needs and to plan essential national health services.17

    Acting on the assessment

    The hardest part of any needs assessment is translating the results into policies and practices that will provide beneficial change. The involvement of health workers in techniques such as rapid or rural appraisal will encourage changes at an individual level. Local workshops can provide an opportunity to review the lessons learnt with other health workers. If this change is going to be sustainable and adaptable then the appraisal should be a continuous process with ongoing feedback. Implementation of strategic changes can be facilitated if the policy makers themselves are active in the process.

    These articles have been adapted from Health Needs Assessment in Practice, edited by John Wright, which will be published in July.

    Acknowledgments

    We are grateful to Anthony Zwi for comments and advice, to Dr Aad van Geldermalsen for the figure, and to Margaret Haigh for secretarial support.

    Funding: None.

    Conflict of interest: None.

    Footnotes

    • Series editor: John Wright

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.