|Medicine - Epidemiology (ECHO - NOHA - Network on Humanitarian Assistance) (European Commission Humanitarian Office, 1994, 120 p.)|
Network on humanitarian assistance
A - Types of data
Raw data of an investigation consist of observations made on individuals. In many situations the individuals are people, but it needs not to be. For instance, they might be red blood cells or hospitals. The number of individuals is called the sample size. Any aspect of an individual which is measured, like age, weight or sex, is called a variable.
It is often useful to distinguish between three types of variables: qualitative, discrete, and continuous. Discrete and continuous variables are often called quantitative. Qualitative (or categorical) data arise when individuals fall into separate classes. These classes may have no numerical relationship with one another, e.g. sex : male, female; eye colour: brown, grey, blue, green.
Discrete data are numerical, arising from counts. Their values are integers (whole numbers), like the number of people in a household or the number of cases of pertussis in a week. If the values of the measurements can take any number in a range, such as height or weight, the data are said to be continuous.
In practice there is an overlap between these categories. Most continuous data are limited by the precision with which measurements can be made. Human height, for example, is difficult to measure more precisely than to the nearest millimeter and is usually measured to the nearest centimeter. So, only a limited set of possible measurements is actually available, although the quantity “height” can take an infinite number of possible values. The measured height is really discrete. However, the methods described below for continuous data are as appropriate for discrete variables.
B - Frequency distributions
If data are purely qualitative, the simplest way to deal with them is to count the number of cases in each category.
The count of individuals in each category is called frequency of that category, for example, the frequency of death through earthquakes is 389,700. The proportion of individuals in this category, related to all deaths, is called the relative frequency. The relative frequency of deaths through earthquakes is 389,700/1,011,200 = 0.385 or 38,5 per cent.
If the categories are ordered, we can use another set of summary statistics, the cumulative frequencies.
The 2,115 victims of accidents were classified according to rough impressions based on their injury. Such a classification could be useful for planning the resources for medical help. The cumulative frequency of a value of a variable is the number of individuals with values less than or equal to that value. Thus, if we order grade of injury from slight to critical, the cumulative frequencies are 81; 521(= 81 + 440); 1,567 (= 81 + 440 + 1,046) etc.. The cumulative relative frequency of value is the proportion of individuals in the sample with values less than or equal to that value. For example, they are 0.038 (= 81/2115), 0.246 (= 521/2115) etc.. Thus, we can see that the proportion of victims with at the most serious injuries is 0.741 or 74.1 per cent.
This frequency distribution is not a very informative summary of the data, most of the values occurring only once. The cumulative frequencies are quite satisfactory.
To get a useful frequency distribution we need to divide the height scale into class intervals, e.g. from 155 to 160, from 160 to 165 and so on, and count the number of individuals in each class interval. The class intervals should not overlap, so we must decide which interval contains the boundary point to avoid it being counted twice. It is usual practice to put the lower boundary of an interval into that interval and the higher boundary into the next interval. Thus, the interval starting at 155 and ending at 160 contains 155 but not 160.
C - Histograms and other frequency graphs
The frequency distribution can be calculated easily and accurately by using a computer. Without using a computer data should be ordered from lowest to highest value before making the interval boundaries and counting. This is rather like starting from Table 3.
Graphical methods are very useful for examining frequency distributions. Figure 1 shows a graph of the cumulative frequency distribution for the height data. This plot is very useful for calculating some of the summary statistics presented later. The most common way of depicting a frequency distribution is by a histogram. This is a diagram where the class intervals are on an axis and rectangles with heights or areas proportional to the frequencies erected on them. The vertical scale shows the relative frequency of observations in each interval.
We often want to summarize a frequency distribution in a few numbers, for facilitating reporting or comparison. The most direct method is to use quantiles. The quantiles are sets of values which divide the distribution into a number of parts so that there are equal numbers of observations in each part. For example, the median is a quantile. The median is the central value of the distribution, so that half the points are less than or equal to it, and half are greater than or equal to it. If we have an even number of points, we choose a value mean between the two central values. For the height example, we have 22 values, so we have to take the middle bet-ween the two central values (11th and 12th of theorderd values) (176 + 178)/ 2 = 177 cm, which we easily get from the cumulative frequencies in . We can get any quantiles easily from the cumulative frequency distribution.
We often want to summarize a frequency distribution in a few numbers, for facilitating reporting or comparison. The most direct method is to use quantiles. The quantiles are sets of values which divide the distribution into a number of parts so that there are equal numbers of observations in each part. For example, the median is a quantile. The median is the central value of the distribution, so that half the points are less than or equal to it, and half are greater than or equal to it. If we have an even number of points, we choose a value mean between the two central values. For the height example, we have 22 values, so we have to take the middle bet-ween the two central values (11th and 12th of theorderd values) (176 + 178)/ 2 = 177 cm, which we easily get from the cumulative frequencies . We can get any quantiles easily from the cumulative frequency distribution.
In general, we estimate the q quantile, the value so that a proportion q will be below it, as follows: We have n ordered observations which divide the scale into n + 1 parts: below the lowest observation, above the highest and between each adjacent pair. The proportion of the distribution which lies below the i-th observation is estimated by i / (n + 1). We set this equal to q and get i = q (n + 1). If i is an integer, the i-th observation is the required quantile. If not, let j be the integer part of i, the part before the decimal point. Then we take the (j + 1)th observation as the q quantile. Other quantiles which are particularly useful are the quartiles of the distribution. The quartiles divide the distribution into four equal parts. For the height data the first quartile is 174 cm: i = 0.25 x 22 = 5.5. Therefore, the 1st quartile is the 6th observation which we get again from the frequency distribution. We often divide the distribution into centiles. For the 10th centile of height i = 0.1 x 22 = 2.2, so the 10th centile is the 3rd observation, 168 cm. We can estimate them from Figure 1 by finding the position of the quantile on the vertical axis, e.g. 0.1 for the 10th centile or 0.9 for the 90th centile, drawing a horizontal line to intersect the cumulative frequency graph, and reading the quantile off the horizontal axis.
A convenient figure summary of a distribution is the box and whisker plot, which uses the median, quartiles, maximum and minimum of the observation.
D - The mean
The median is not the only measure of central value for a distribution. Another is the arithmetic mean or average, usually referred to simply as the mean. It is found by taking the sum of the observations and dividing it by their number. For the height example the sum of all values is 3906 , so the mean is 3,906/22 = 177.5.
At this point we need to introduce some algebraic notation, widely used in epidemiology. We denote the observations by :
x1, x2,..., xi, ...xn
There are n observations and the i-th of these is xi.
The summation sign is an upper-case Greek letter, sigma, the Greek S. When it is obvious that we are adding the values of xi for all values of i, which runs from 1 to n,
The mean of the xi is denoted by x, pronounced ‘x bar', and x = 1/n - xi.
In this example the mean is very close to the median, 177. If the distribution is symmetrical the mean and median will be about the same, but in a skewed distribution they will not.
E - Variance and standard deviation
The mean and median are measures of the central tendency or position of the middle of the distribution. We shall also need a measure of the spread, dispersion or variability of the distribution.
One obvious measure is the range, the difference between the highest and lowest value. This is a useful descriptive measure, but is has two disadvantages. First, it depends only on the extreme values and so it can vary a lot from sample to sample. Secondly, it depends on the sample size. The larger the sample size, the further apart the extremes are likely to be.
The most commonly used measures of dispersion are the variance and standard deviation, which we shall describe now. We start by seeing how each observation differs from its mean. Table 6 shows the deviations from the mean of the 22 observations of height. If the data are widely scattered, many of the observations will be far from the mean, and so many deviations will be large. If the data are narrowly scattered, very few observations will be far from the mean and so few deviations will be large. We square the deviations and then add them, as shown in Table 6. This gives us :
In the example equal to 1,269.5. For an average squared deviation, we divide the sum of squares by (n - 1), not n.
The estimate of variability is called the variance, defined as follows:
The variance is calculated from the squares of the observations. This means that it is not in the same unit as the observations, which limits its use as descriptive statistic. The obvious solution is to take the square root, which will then have the same unit as the observations and the mean. The square root of the variance is called the standard deviation, denoted by s:
(s =_ variance )
A - The denominator
The notations introduced in this chapter are used in epidemiological studies for the description of diseases. Usually some exposure (like smoking) is regarded with respect to certain diseases (e.g. lung cancer). Measures of effect describe the association between exposure und disease. Although these notations sometimes may sound strange, in the context of disasters the measures however give sense: exposures may be living locations (see Table 7), and “disease” usually are death or injury. This chapter is orientated on a paper on Cesar G. Victora (1993).
This tornado was placed among the severest 3 percent of all tornadoes in the United States. In the 2 weeks following the tornado Glass and his coworkers had interviewed families of the deceased Wichita Falls residents and persons who were seriously injured. Based on this study the authors were able to estimate the number of people at risk in the different locations. A statement like “equal number of fatal and serious injuries took place in ‘mobile homes' and ‘apartments' ” being based on the frequency count of 4 may be misleading. Obviously, much more people have been in apartments and the risk of being injured is low (1.1 per 1,000) compared to mobile homes (13.3 per 1,000). This pitfall “floating numerators” can be solved by using the appropriate denominator.
Choice of the appropriate denominator is one of the most important tasks of an epidemiologist. The most commonly used denominators are
- total number: the total number of persons under study at a given
- non diseased: the number of persons who do not have the disease of interest at a given time
- person time: the number of persons at risk multiplied by the time for which each remains at risk.
These denominators are essential for measuring disease frequency. Before describing these measures, however, it is useful to recall some basic definitions. A ratio is the quotient of any two numbers. For example, the female to male ratio is greater than one in most communities. Ratios used in epidemiology range from 0 to + -. A proportion is a special type of ratio in which the denominator contains the numerator. For instance, 0.53 % of people living in the area of the Wichita Falls tornado have been injured. A proportion must range from 0 to 1, or 0 % to 100 %. Odds are the number of events divided by the number of non-events. Odds, although common in betting, are harder to interpret than proportions. They vary from 0 to + _, often being expressed as 1:2 (that is, 1 case per 2 non-cases). In the example of the Wichita Falls tornado the odds of serious to fatal injuries is 52:35 or about 15:10, that is 15 seriously injured people on 10 fatally injured people. Different types of epidemiological studies allow calculation of different measures of disease frequency.
The figure represents the group under study. At time t0, a0 individuals already have the disease of interest and c0 do not. Of c0, b1 will acquire the disease by t1, while c1 remain healthy. At the end of the study (t2), c2 will still be unaffected.
This scheme assumes that the disease occurs only once in each individual and that there are no losses to follow-up.
B - Prevalence
In cross-sectional studies, subjects are examined once. The number of cases may then be divided by the total number of persons studied (denominator ‘total number'). This is usually called prevalence. In the figure at t0, the prevalence is equal to a0 / (a0 + c0), while at t2 it equals (a0 + b2)/(a0 + b2 + c2). The prevalence is a proportion. In the example the prevalence of people staying in single family houses during the tornado was 59.1 % (9,705/16,420).
C - Incidence
The situation is more complex in so called cohort or incidence studies in which subjects are followed over time. Incidence studies usually exclude individuals who are already affected at the beginning (a0). A first choice of denominator is therefore the number of non diseased persons, the initial population at risk (c0). If they are followed up until t2, the number of new or incident cases (b2) divided by c0, gives the so called incidence risk, also known as cumulative incidence. Cumulative incidence is a proportion. An example of a (short) follow-up-study is given in Table 8. The data have been collected after the 1980 earthquake at Compania, Italy. The earthquake trapped c0 = 548 people. Until the first time point t1 = 12 h, b1 = 134 people have been extracted, giving (cumulative) incidence for extraction 134 / 548 = 0.24. At t3 = 2 days the cumulative number of extracted people is 436, giving a cumulative incidence of 436/548 = 0.80.
The cumulative incidence has 2 disadvantages: firstly, subjects who develop the disease, who die from other causes or who are lost to follow-up can no longer be detected as incident cases. For our example of the earthquake this is not a serious problem because the follow-up time is very short. Secondly, the cumulative incidence may cover very different forces of being extracted on different days. This leads on to the development of an incident rate. Its denominator is expressed as person-time units. The incidence rate may be calculated for various time intervals. If for the earthquake example we take one day as a time-unit, with exception of the first day the incidence rate is equivalent to the proportion extracted in the table: 0.26, 0.34, 0.22, 0.11 for day 1 to day 4. For the first day the time interval is only half a day. So, the incidence rate (per day) is 0.48 for the first 12 hours and 0.90 for the second 12 hours.
An incidence rate is a ratio, ranging from 0 to + -, because the numerator (events) is not contained in the denominator (person-time). It reflects the velocity of change in some characteristics of the population. For recurrent diseases, incidence rates may be greater than 1 per person-time unit. For example, in many developing countries there are around 3 diarrhoea episodes per child-year.
D - Proportionate and case-fatality rates
Even without population data a denominator can still be obtained. For instance, in the given example of Wichita Falls Tornado the number of fatal injuries in vehicles may be related to the overall number of fatal injuries. This proportion is often called proportionate mortality rate.
Proportionate rates are not as useful as those based on the population at risk. For instance, the number of deaths due to a particular cause may be related to the overall number of deaths in the same period. This denominator allows calculations of a (proportionate) mortality rate, which is actually a proportion.
The proportion of deaths among new cases of disease in a given period is the case-fatality rate. This is often erroneously referred to as mortality rate: “rabies is a disease with high mortality” is not true in most places, although its case-fatality is high everywhere. A case-fatality rate is a proportion, ranging from 0 to 1.
E - Measures of effect
An effect relates to the association between an exposure and a disease. Effects may be expressed in relative or absolute terms.
Relative effects are expressed as ratios, that is, quotients of two frequency measures. They are often referred to as relative risks. Their general form is:
frequency among exposed
frequency among unexposed
Because both frequencies must be expressed in the same units, such a ratio is dimensionless, ranging from 0 to + -. For example, people in vehicles have had about a seven times higher risk of being injured than those staying in single-family-houses.
The relative risk does not take the absolute number of injuries into account. This is done by a so called (population) attributable risk which measures the percentage of injuries that could have been avoided if all people had stayed in a location with minimal risk. So 68 % of the injuries could have been avoided if these persons, who stayed in their vehicles, would have been in apartments. As it is shown in Table 3 a high percentage (57 %) of injured people would have been avoided if single family houses were more resistant against earthquakes. Although the relative risk is rather small, a high number of people staying in that location would yield a high attributable risk.
A - Objectives of study
The starting point of any investigation must be to define clearly its objectives, since these will determine the appropriate study design and the type of data needed. Objectives may be categorized into one of three main types as listed below. An investigation usually has several objectives, which can of course be of different types.
- Estimation of certain features of a population. For example, what is the average number of diarrhoeal episodes per year experienced by under-5-year-olds in Bangladesh?
- Investigation of the association between a factor of interest and a particular outcome, such as disease or death.
- Evaluation of a drug or therapy or of an intervention aimed at reducing the incidence (or severity) of disease. For example, does the use of sleeping nets reduce the risk of malaria, and if so does spraying the nets with insecticide afford additional protection ?
B - Observational studies
In general, it will be necessary to carry out a special study to collect the relevant data to answer the specific objectives. Estimation and association objectives lead to studies which are observational in nature; the natural history of disease is observed with no attempt made in the study to alter it. Evaluation objectives may be answered by either observational or experimental studies, depending on the type of measure being evaluated and whether it is already in use.
Observational study designs may be divided into three major groups, cross-sectional, longitudinal (including cohort) and case-control studies. People may be selected individually or in clusters. The available resources and logistic difficulties often mean, however, that is is not possible to examine a sufficient number of clusters to gain a representative picture, and results are not uncommonly based on a survey in just one community. Judgement of representativeness is then subjective rather than statistical.
C - Cross-sectional study
A cross-sectional study is carried out at just one point in time or over a short period of time. Cross-sectional studies are relatively quick, cheap and easy to carry out, and straightfortward to analyse. Since they provide estimates of the features of a community at just one point in time, however, they are suitable for measuring prevalence but not incidence of disease, and associations found may be difficult to interpret. For example, a survey on onchocerciasis showed that blind persons were of lower nutritional status than non-blind. There are two possible explanations for this association. The first is that those of poor nutritional status have lower resistance and are therefore more likely to become blind from onchocerciasis. The second is that poor nutritional status is a consequence rather than a cause of the blindness, since blind persons are not as able to provide for themselves. Longitudinal data are necessary to decide which is the better explanation.
D - Longitudinal study
In a longitudinal study individuals are followed over time, which makes it possible to measure incidence of disease and changes over time and easier to study the natural history of disease. Occasionally the acquisition of data may be restrospective, being carried out from past records. More commonly it is prospective and, for this reason, longitudinal studies have often been alternatively termed prospective studies.
In the majority of cases, the simplest way to carry out a longitudinal study is to conduct repeated cross-sectional surveys at fixed intervals and to enquire about, or measure, changes that have taken place between surveys, such as births, deaths or the occurrence of new episodes of disease. The interval chosen will depend on the factors being studied. For example, to measure the incidence of diarrhoea, which is characterized by repeated short episodes, data may need to be collected weekly to ensure reliable recall. To monitor child growth, on the other hand, would require only monthly or 3-monthly measurements.
The population under study may be either dynamic or fixed. In a dynamic situation, individuals leave the study when they no longer conform to the population definition, while new individuals satisfying the conditions may join. An example would be the study of incidence of diarrhoea in under-5-year-olds, in which monitoring of children would cease when they attained their fifth birthday, while newborns would be recruited into the population as they were born. In a fixed situation, on the other hand, the population is defined at the onset and, apart from deaths, migrations, and other losses to follow-up, its composition remains unchanged throughout the study. A fixed population is often called a cohort, such as the birth cohort of 1990 that is all people born in 1990.
Longitudinal studies tend to be more costly and to pose many logistic problems in their execution.
E - Case-control-study
A case-control study is used to investigate the association between a certain factor and a particular disease. The design is very different to other types of studies because the sampling is carried out according to disease status rather than exposure status. A group of individuals identified as having the disease, the cases, is compared with a group not having the disease, the controls.
F - Questionnaire design
In most studies it will be necessary to prepare a specially designed record form or questionnaire for collecting the data. This should be kept as brief as possible, and the temptation to ask every conceivable question should be resisted. Overlong questionnaires are tiring for interviewer and interviewee alike, and may lead to unreliable responses. Questions should be clear and unambiguous and written exactly as they are to be read out. Technical jargon or long words should be avoided, as should negative questions, leading questions, and hypothetical questions.
Careful thought should be given to the order in which information is collected and to whether the questionnaire should be self-administered or completed by an interviewer. There should be a logical progression through the form which is easily followed. Questions are best arranged in sections. It should be clearly indicated whenever the completion of a section is dependent on the response to a previous question, and the starting point of the next section should be easily identifiable. It is best to minimize the number of skips or jumps which can occur, since too many can be confusing, and sections can be accidentally missed. The questionnaire should be clearly labelled with the titel of the study and with the respondent's name and study identification number. These should be repeated on the top of each page. The next section usually consists of general identifiying information such as age, sex, and address. In general it is a good idea to arrange subsequent sections in order of importance of the information to the study, so that the most important information is collected when the interviewer and interviewee are freshest and least bored. Any sensitive questions are, however, probably best left to the end.
G - Open and closed questions
Questions may be in one of two forms, open or closed. Open questions are used to search for information, and the interviewer records the replies in a freely written form. There are no preconceived ideas about what the possible responses might be. In a closed question, on the other hand, the response is restricted to one of a specified list of possible answers. This list should include a category for ‘Other' with space to write in the details and a category for ‘Don't know'. The interviewer may either lead the respondent through the list category by category, or ask the question in an open form and then tick the category which most closely corresponds to the answer given. Open and closed forms each have advantages and disadvantages, and the choice very much depends on the particular context. Responses to closed questions are considerably more straightforward to process, but open questions can yield more detailed and in-depth information. One possibility is to use an open form during the pilot phase of the study and to use the results of this to draw up the list of answers for a closed question form for the main study.
H - Coding
Numerical data should be recorded in as much detail as possible and as individual values rather than precoded on the questionnaire into groups. Consider the example of age. The preferred option is to record date of birth and later to calculate age from this and the date of interview. The next best is to record the respondent's age in, for example, years for adults, months for young children, weeks for infants, and days for newborns. The least satisfactory approach is simply recording in which age-group, such as 0-4, 5-9, 10-14, 15-24, 25-44, 45-64 or 65 + years, the respondent belongs. The units of any measurements should also be clearly specified, for example whether weight is to be recorded in kilograms or pounds, and the number of digists of accuracy required should also be indicated. With closed questions, the corresponding numerical codes should be printed alongside the listed choice of responses. With open questions, it will be necessary to code the replies after the questionnaire has been completed and space should be allowed for this.
If data are to be entered onto a computer then this should be borne in mind when designing the questionnaire. For example, it may speed up the data entry procedure if all the information to be entered is coded into boxes arranged down the right-hand side of the form. In most cases it will be necessary to assign numerical codes to the responses to non-numerical variables, such as code 1 for male and 2 for female. One box is needed for each digit (or letter), the total number of boxes required for a variable being determined by the number of digits in the maximum response likely to be recorded for the variable. Where possible the use of code zero should be avoided, since on many computer systems and in many statistical software packages it is impossible to distinguish zero from a blank response, meaning no data.
I - Multiple response questions
Multiple response questions require special consideration. They can be dealt with in two different ways. For example, in rural West Africa a family may use one or more of eight possible sources (rainwater, borehole, well, spring, river, lake, pond, stream) for their drinking water. The first method is to assign a separate coding box to each possible response, in this case eight. Each box would contain either code ‘1' for source used or ‘2' for source not used. Thus if a family used rainwater and also collected water from a river, the rainwater box (number 1) and the river box (number 5) would contain code 1, while the other six boxes would contain code 2. The second method is to instead assign codes 1 to 8 to the eight different sources and to decide on the maximum number of responses any family is likely to give. Suppose we decided that three responses was the limit. We would then allocate three separate coding boxes to the question and enter in these the code numbers of the sources named. The family using rainwater and river water would be coded as 1 (rainwater) in box 1 and as 5 (river) in box 2. Box 3 would be left blank. The codes should be entered either in numerical order, as done here, or in some other logical order, such as amount of usage of source.
J - Data checking
Each questionnaire should be carefully checked after completion, and again once the data have been entered onto the computer. The importance of this should not be overlooked. Checking should take place as soon after data collection as possible in order to allow the maximum chance of any resulting queries being resolved. Checks are basically of two sorts, range checks and consistency checks. Range checks exclude, for example, the erroneous occurrence of code 3 for sex, which should only be code 1 (male) or 2 (female). Consistency checks detect impossible combinations of data such as a pregnant man, or a 3-year-old weighing 70 kg. Scatter diagrams showing the relationship between two variables are particularly helpful in doing this, as they allow odd combinations of, for example, height, and weight to be easily spotted.
Three basic precautions are recommended to minimize errors occurring during the handling of data. The first is to avoid any unnecessary copying of data from one form to another. The second is to use a verification procedure during data entry. Data are entered twice, preferably by two different persons, as this gives an independent assessment of any poorly written figures. The two data sets are then compared and any discrepancies resolved. The third is to check all calculations carefully, either by repeating them or, for example, by checking that subtotals add to the correct overall total. When using a computer, all procedures should be tried out initially on a small subset of the data and the results validated by hand.
Planning involves determining, with respect to a particular system, precise objectives, and implementing the necessary resources to achieve them within the time allowed.It provides a rational basis for decision-making.
The notion of PRIORITIES and OBJECTIVES (once again) raises the question of whether health planning is really justified. Why plan?
A - Economic rationale
Resources are (relatively) scarce and the needs (relatively) great > necessity for choices. Hence the concern for (economic) efficiency in the use of resources.
Possible objection: the market could take care of these choices. "Orthodox" (i.e. liberal) economic theory seeks to demonstrate that the allocation of scarce resources in order to satisfy numerous needs is best achieved by giving free rein to the forces of supply and demand. "Market equilibrium" will thus be attained through pricing, providing that is:
- free competition is allowed to operate (in this particular instance, providing there is no monopoly over the supply of health care and no State intervention in order to regulate the price of health care services)
- the participants in the exchange (supply and demand) are fully informed about the state of the market.
B - Political (or social) rationale
This implies the reference to value judgements such as:
- "access to health care (or even just "health") is the
fundamental right of every individual"
- "health is one of the essential goals of social development"
- "health is a means of progress" (by increasing the productivity of labour).
These value judgements are fairly widely accepted, irrespective of political systems, civilisations and cultures.
The mechanisms of the market however (where supply partially creates its own demand, where demand is only ever solvent demand, where perfect information does not exist, etc.) are not capable of fulfilling the requirements contained in these value judgements. Hence the (social/ political) concern for the equitable distribution of health care services and opportunities to enjoy good health.
C - Scientific/technical (or public health) rationale
In the health care sphere, there is always a slight mismatch between supply/demand/needs (Cf. diagram below). This is true of many other goods or services (e.g. education, housing, food, clothing, insurance) but the health care sphere is special in that the information required to form some idea of
- the significance of the distinction between (stated) demand and
- the match between demand and the available supply (= health care)
Normally eludes the individual or community concerned and / or is beyond their powers of judgement. This is because in the health care sphere
- there is a very powerful "emotional charge":the focus is on the body, suffering, vitality, appetite for life, in other words "goods" with regard to which, the general public and individuals alike, can be both very demanding and highly irrational
- the "demander" (i.e. the patient or anyone seeking some form of medical care) is invariably in an inferior position in relation to the "seller" (the health care system):handling information is hard enough for someone in good health, and illness further diminishes their ability to determine whether supply, demand and needs are suitably matched
- needs can be defined in a very precise manner by a particular authority (medical science) generally recognised as being impartial and unconnected with the "health care market".
In the health sphere more so than in the case of other goods and services, we are thus a very long way from the theoretical model of political economy, which posits a "supplier" and a "demander" who are perfectly informed about the state of the market and behave in a rational manner. Hence the (public-health) concern for voluntarist improvements to the existing health care system:greater efficiency (by matching supply to needs), high-quality care, etc.
These three types of rationale remain equally valid if you believe that health care policy should be much more comprehensive than simply organising the medical profession and care.Thus,
- as regards the availability and quality of food,
- dietary habits,
- environmental health,
- drinking water supplies,
- safety (road, domestic),
- general hygiene,
- working conditions, etc.
there are no "markets" as such, where trade in goods or services could ensure a proper match between supply, demand and needs.
Diagram:Supply, demand and needs in the health sector
Important - In public health, in order to avoid lengthy debates about the concept of "need", the accepted definition is an operational (practical) definition, rather than a conceptual (theoretical) one:needs are anything defined as such by members of the health care profession in the light of the available knowledge.
Example - Health care needs = "the estimated manpower and quantity of services required, in the opinion of professionals and according to the state of the art in medical science, in order to provide an optimal level of health care" (T. Hall).
However hard the various authors of public health manuals may try to get away from this operational concept of needs, in everyday practice it is invariably the latter which prevails.
Faced with these arguments, the question naturally arises of why we plan so little, particularly in Western countries.
A - Technical factor - Lack of data
In some cases, the most basic statistics, health cards, levels of coverage per service, etc. are quite simply lacking.
B - Economic factor - The illusions fostered by the affluent society
In those countries which have experienced "the affluent society", the relative abundance of resources obscured the need for planning:what was the point of planning when, to all appearances, the full range of needs could (or would eventually) be met?
Why bother about priorities when it seemed, all of the existing needs could be satisfied? Hence, with the present budgetary constraints weighing heavily on welfare expenditure, the need for a sort of "retrograde (or retrospective) planning", is perceived by those involved in the health care sector as a policy of belt-tightening and rationing.
C - Institutional factor - The interests of institutions
Because of their organised structure and lobbying power, the interests of institutions (hospitals, peripatetic or stationary preventive medicine services, etc.) and professional associations (doctors, nurses, physiotherapists, etc.) carry more weight than any analysis of the needs and level of demand of the community concerned.In the health care sphere, the latter is still largely unorganised:the "patients" or "potential beneficiaries of preventive measures" do not constitute a force capable of deploying itself on the socialfront.Those who claim to speak on their behalf (politicians, doctors, trade unions, mutual associations) are themselves suspect because governed by their own institutional mentality.
D - Cultural factor - The "culture" (or lack thereof) of the medical world
The failure to train and educate doctors and paramedics in the aspects of evaluation, self-evaluation, resource management etc. of public health is a legacy of the liberal practice of medicine in most Western countries (postulates of liberal medicine = complete freedom for the patient to choose his or her doctor, complete freedom for doctors to choose the form of treatment, and where they wish to work, complete freedom as regards setting their fees, on-the-spot payment for treatment).
In the former colonies, by contrast, the modern health service, after the pattern of the civil service and armed forces, was instituted from the outset in the form of a public service, the move towards privatisation being a recent phenomenon. Since resources were much scarcer than in the industrialised countries, receptiveness to the notion of health care planning tends to be more immediate (although not necessarily effective!). The Implications for planning :
1) Emphasis on inputs
The deep-seated reluctance to think of health care systems, medical professions or medical services in terms of results (outputs) (= effectiveness) and yield (output / cost) (= efficiency) breeds a tendency to focus purely on the inputs, whether financial, human, material or technical:"Since you force me to plan, tell me what my budget allocation is and leave me to get on with my job".
2) Tendency to operate in isolation
Each component of the health care system tends to operate in isolation, with precious little regard for the quality of the relations between the various components: GPs / specialists / hospitals / school medicine / industrial medicine / multiple preventive services / administration.
3) Too much emphasis is placed on the two-way doctor-patient relationship
Treating illnesses in the patient's own home takes precedence over promoting health in the community at large.
4) Pathological fear of any "political meddling"
The liberal practice tradition inclines one to see any voluntarist and concerted attempt to reorganise the health care system as unwarranted political interference in the "art of healing".
All of this results in various impediments to planning:
> the proliferation of fragmented views of problems
> contradictory or redundant "plans" and policies
> tendency to latch onto false problems.
Under such circumstances:
a) health planning can never be anything more than
- a sporadic activity, lurching from one crisis or fashion to the next (AIDS, home-based health care)
- a partial activity, confined to a handful of problems leading to a more balanced form of lobbying rather than the fulfilment of needs (e.g. the former "hospital planning" scheme in Belgium)
- an activity which has no impact on the way the system operates.
b) there is no sense of direction, in other words
- what are the priorities?
- how much is the health care system going to cost and what will it deliver in return?
- which areas need to be evaluated and improved?
A - The preconditions
* political will or impetus
* legal framework / health care policy
* consult and inform the socio-professional groups concerned, with a view to:
> at the very least, harmonising their opinions (over and above
the various sectoral viewpoints)
> if possible, motivating them
> at best, getting them to work together
* administrative capacity
* initial census of "problem sectors"
In many countries (including our own), any form of general planning is doomed from the outset... owing to lack of political will or impetus.
When the political will is there, getting the relevant socio-professional groups involved can sometimes pose an insurmountable obstacle, inasmuch as institutional interests tend to prevail over considerations of public health.
In some cases (as in certain Third World countries) it is the administrative capacity which is lacking.
B - Identifying needs/problems
Remark - Different authors do not always use the same terms in the same way.Pineault and Daveluy for instance, interpret the term "health problem" as meaning a "problem such as it is perceived by the population", as opposed to need.
Identifying needs and problems hinges on the COLLECTION and ANALYSIS of data.
a) approach based on indicators (Cf. "Community diagnostics")
- socio-demographic indicators:mortality rate, birth rate, age
- health indicators:prevalence, incidence, attributable risk, etc.
- takeup of health care services
Often the data is too sketchy to justify this indicator-based approach, in which case one must use:
b) a survey-based approach
c) a dead-reckoning approach (scientific literature)
d) a comparative approach (i.e. compare two neighbouring areas one of which is better known)
e) a consensus-seeking approach
- key informers
- Delphi method
- nominal group
- community forum
- community's impressions
f) inventory of health care resources and their use
Surprisingly often, one ends up with a surfeit of data, in which case the big mistake can be to think they all have to be used. No planner can process everything.He must be able to pick out the relevant data.
This raises, among other things, the question of the quality of the data available.
- are sometimes designed, presented and processed in such a bureaucratic manner that the health officer in charge of recording them will tend to enter any old figure rather than leave a particular column blank (he knows that no-one will bother to check falsified figures, whereas he will get into trouble if all the columns are not completed!)
- even when they are "reliable", still only represent the diagnosed sickness ratio (section 1 of the figure below).
The Diagnosable sickness ratio represents a broader measure (section 2) : diagnosis errors and forgetting to record all of the cases account for the difference. Sometimes, however, Diagnosed sickness can tend to overstate a particular problem (too many fevers in tropical regions for example, are labelled "malaria").
Actual sickness is even broader: it requires in effect that those who are ill realise they are ill, think to contact the health service for treatment and are actually able to do so (notion of geographical, cultural and financial accessibility of health services).
Finally too, one cannot overlook the fact that the current trend towards training peripheral, auxiliary staff in Third World health services on the basis of "specific problems to be solved" (often using decision trees) is not conducive to data collection based on "cases of clearly defined pathologies". One needs to be quite clear however, what it is one wants: to place data collection in the service of basic practitioners or vice versa... There is no reason why we should not have a data collection systembased on "specific problems to be solved", although it may mean entrusting the in-depth analysis of the prevalence and incidence of particular pathologies to more selective epidemiological studies.
C - Establishing priorities
1) To begin with, one requires a precise definition of the health problem concerned, which must be clearly delimited. Thus for example, rather than "combatting malaria" , the problem will be defined as "combatting infant mortality due to malaria" or as "combatting sickness in pregnant women due to malaria".This makes it possible to plan and organise the different stages more effectively.
2) The golden rule is as follows:
Any given health problem constitutes a "priority" from the point of view of planning health care measures, if it is important and if it is vulnerable.
The importance of any given problem depends on the weighting assigned to the following three components:
- its importance / urgency for the community in question
This urgency can be due to a correct appraisal of the situation by the public at large or can be the "distorted" result of inadequate treatment of a particular issue by the media.In this latter instance, health care planners can make this very area of better public information, one of their goals.
- its intrinsic seriousness (as measured by the specific sickness
ratio and death rate)
- its frequency (as measured by prevalence and incidence)
To disregard the public perception of the problem is to lay oneself open to a technocratic concept of planning, of the sort practised by the promoters of so-called "selective" primary health care (cf. details of this debate elsewhere). On the other hand, focusing too exclusively on people's requirements - as in the case of Western liberal medicine -leads to a medical profession which merely responds to demand without considering the needs, and without reorganising its offerings in the light of those needs; in other words, without establishing priorities.
The vulnerability of any given problem can be broken down into a number of elements :
- its technical vulnerability:is there an effective method (preventive or curative) of combatting this problem?
- The degree of certainty, in medicine, concerning the validity of any given method of prevention or treatment, hinges, in descending order of reliability, on:
- the randomised controlled trial
- case-control study
- the studies prior to/after or with/without the given programme
- experts' opinion
- personal experience, etc.
- its operational vulnerability (= practical vulnerability, in the field), sometimes referred to as feasibility, which in turn depends :
- on the operational possibilities (infrastructures, personnel, climatic conditions, means of transport, etc.)
- on the costs, in terms of : costs for the individuals / for a third party who foots the bill immediate / recurrent costs absolute / relative costs (cost-effectiveness, cost-benefit)
- what result am I getting and at what cost ?
- what am I saving in another sector of the health service if I plan such and such an activity (e.g. curative health care costs saved by a prevention programme) ?
- what will I have to forgo if I devote my resources to such and such an activity (notion of opportunity cost) ?
The relative costs study (using the notions of cost-effectiveness and in some cases, cost-benefit) throws up Several choices and thus paves the way for options in planning.
- on the acceptability of the activities envisaged (legal, political, cultural, ethical acceptability).
3) How do we select priority problems ?
One could start by subjecting all of the problems listed in point 4.2. to the following test :
Please note - that what we are dealing with here are priority measures and not exclusive measures. That is to say, a problem which is deemed to be "unimportant" and "non-vulnerable" should not necessarily be excluded from any health care measure.It is just that since it is not a priority, it can only warrant a small share of the available resources.If one were to exclude the vast majority of "minor complaints" from health care measures however, one would end up with a technocratic and indeed inhuman health service, which would rapidly alienate those it purported to serve.
Even if he follows the steps outlined above, the planner will still find himself faced with a daunting list of priorities. How can we grade these priorities further?
D - Establishing objectives
1) Defining possible alternative forms of action
No planning process is complete without examining alternative solutions for each chosen objective
This is done by :
- basing oneself on the data contained in scientific literature
(compared with the practical feasibility)
- and/or using the same techniques as those used to select problems (4.2.)
2) Defining the target group, at whom the project is aimed
The entire population? A particular age group? A particular high-risk group? Etc.
3) Defining the general objectives
Example : to reduce the incidence of tetanus by x% within such and such a time-limit the number of deaths due to tetanus by z% within such and such a time-limit or, more specifically, umbilical tetanus, or tetanus in children, or adults, etc.
4) Defining the operational objectives
vaccinate --% of the target population with 3 injections
inform --% of the overall population
train --% of midwives
retrain --% of auxiliary staff
ensure that --% of sufferers contact the health service.
E - Determining the activities
The idea at this stage is to determine the type and number of specific activities which need to be undertaken: e.g. number of vaccines, method of conveyance, the freezer chain, information medium, etc.
This stage is more than just an administrative one. Not just because success on the ground depends on the practical organisation, but mainly because this stage requires us to answer two important questions:
1 ) Preventive or curative measures ?
- primary prevention: any measure aimed at reducing the incidence
of illness in the community
- secondary prevention: any measure aimed at reducing the prevalence of illness in the community
- tertiary prevention: any measure aimed at mitigating the consequences of illness in the community (in terms of permanent disability, loss of independence, etc.)
Early detection of illnesses (via screening or case research) generally comes under secondary prevention (the illness is already present but at a preclinical stage: detecting it with a view to treatment thus reduces the prevalence).Detection comes under primary prevention when it concerns a particular risk factor or a precursory stage of the illness (the illness has not yet developed: detection has the effect of reducing its incidence).
In short, a preventive measure is only justified if one can answer "yes" to the following five questions :
1_ Is the illness sufficiently frequent and serious to warrant early intervention (= is it a priority)?
2_ Does the illness have a sufficiently long preclinical phase to allow early intervention?
3_ Is there a valid, reliable test for identifying the illness at the preclinical phase?
4_ Is there some form of intervention (treatment) which is more useful or more effective when it is applied following detection than when it is applied following a diagnosis formed under the usual circumstances?
5_ Is the community at large well-disposed towards screening and does it accept the relevant modes of application?
2) Vertical or horizontal strategy?
The choice of intervention strategy is of cardinal importance: it determines the type of medicine that will eventually be offered to the population concerned.
The first type of strategy involves organising a vertical action programme : from the Department of Health right down to grassroots level (be it a New York district, a station in the Tanzanian bush or a village in the Bolivian Altiplano), the planners organise everything (staff, vehicles, equipment, pay, managerial structure, information flows, etc.) in the form of an independent, self-sufficient channel, separate from the rest of the health service. The thinking behind this vertical planning is as follows: given a particular health problem, what service needs to be organised in order to deploy the methods/ activities provided to combat it?
The second type of strategy involves incorporating health care and programmes into existing services (or services which may be set up or revised on this occasion).Rather than introducing some new activity (i.e. the programme) from the top down, one endeavours to reorganise the existing services in order that they may take on the activities envisaged by the planners. The thinking behind this horizontal format is as follows: given a particular health service, how can we organise it so as to incorporate the solution of priority problems into the existing range of activities, which require a comprehensive, integrated and consistent approach to health care?
Each of the two strategies has its pros and cons.
Opting for one or other has fundamental repercussions on the overall functioning of a country's health service. The vertical approach tends to overlook this latter aspect: the "priorities" are taken care of by the vertical programmes and there is a danger that those in charge will regard the actual functioning of the permanent health services as being outside their province. Yet it is to these permanent services that the community turns on a daily basis for all its problems, "priority" or otherwise.
In the Third World, the debate is even keener. Organising primary health care involves integrating care at a peripheral level; the attitudes of the major aid and intervention agencies (who often advocate selective primary health care) encourages countries to adopt a series of vertical programmes.
"Selective" primary health care classes as "priorities" those problems where the sickness ratio/ death rate is high and where effective prevention or treatment can be readily implemented. The opinion of the communities themselves is unimportant. Non-priority problems are discounted. In practice, this leads to the selection, depending on the location, of fewer than ten or so priority measures, most of which are geared towards pregnant mothers and young children.
At the end of the day, it is a question of opting for the type of health care one prefers. For proponents of vertical planning, the key question is: "Which illnesses ought to be combatted first?" whereas the horizontal structure hinges on the question of "What type of medicine should we offer the population?" (Cf. D. Grodos et X. de Bune, Les Soins de santrimaires sctifs: un pi pour les politiques de santu Tiers Monde, Soc Sc Med, Vol. 26, N_9, pp 879-889, 1988).
F - Deploying and co-ordinating resources
The focus shifts here to the management of the actual programme.
This often neglected stage is nonetheless essential. It alone can highlight the sort of questions which will help correct any measures taken in future:
- were the objectives aptly chosen?
- were the results what we expected?
- are the methods used the most suitable?
- were the resources deployed the most appropriate?
A - Nature of economic evaluation
These notes summarise the introductory chapters of the work : MichaelF. Drummond, Greg L. Stoddart and George W. Torrance, Methods for the Economic Evaluation of Health Care Programmes, Oxford Medical Publications, Oxford University Press, 1987 (181 p).
The underlying concept of economic analysis is that of opportunity cost. The decisions taken by economic players (the employer who hires or invests, the consumer who saves or buys) hang on choices which reflect these players' preferences when faced with multiple needs and limited resources. Choosing one thing means sacrificing something else: allocating 1000 francs to the purchase of compact discs means forgoing a 1000 franc meal at a restaurant, for example. Similarly, in the health care sphere, the real cost of a particular health care programme is not the number of francs shown in the programme budget, but rather the "value" of the results in health terms of any other programme passed over in favour of the first programme. It is this "opportunity cost" which economic evaluation endeavours to measure and compare with the results of the programme assessed.
- Is a particular service, programme or activity worth setting up compared with what we could obtain otherwise, with the same resources?
- Is the way in which a particular service, programme or activity operates satisfactory compared with some other mode of operation?
- Should all hospitals be equipped with tomodensitometers or should preference be given to the geriatric departments?
This latter type of question highlights the fact that the answer is seldom "yes" or "no" but more often "to what extent"?
Economic analysis will endeavour to:
- relate the costs (inputs) to the effects or consequences (outputs)
- evaluate the possible choices in relation to each other or to make implicit choices explicit (many medical activities are performed out of habit and have never given rise to an economic evaluation).
Economic evaluation is an analysis of the alternative forms of action in terms of costs (inputs) and consequences (outputs), which must be identified, measured and where appropriate, optimised.
B - Types of economic evaluation
Remarque - For the purposes of this report, we will use the term "programme" to refer not just to health care programmes in the proper sense, but also services, activities, schemes, institutions, procedures, methods, etc.)
These apparently clear distinctions are not always evident in scientific literature.
The widespread confusion concerning the terms used and the description of the type of approach employed has prompted the suggestion that we classify economic evaluations as follows:
A cost study is purely concerned with costs. If it compares several programmes with different costs, the preferred term is cost-minimization analysis.
A cost-effectiveness study relates the effects of a given programme, measured in physical units (life-years gained, number of accurate diagnoses, number of cases detected) to their cost. Ordinarily, such a study does not consider the fact that these effects may not be worth pursuing; they are assumed to be desirable.
A cost-utility study relates the effects of a given programme, measured in quality-adjusted life-years (QALYs) to their cost. It is therefore a type of cost-effectiveness analysis, which is particularly useful for programmes where it is the result in terms of sickness ratio which matters, or in terms of life-years gained. These extra life-years are "modulated" or "adjusted" according to the quality of life afforded: sound health is rated 1, death 0, a minor handicap (side-effects of chronic treatment, etc.) 0.2 (for example) and a major handicap (impotence, incontinence, etc.) 0.7 (for example).The "utility" thus weights the life-years gained by the quality of the remaining life.
A cost-advantage study or cost-benefit study (synonymous) relates the effects of a given programme, measured in monetary terms (francs, dollars, etc.) to their cost. In theory, this ought to be the broadest and most sophisticated form of economic evaluation, in so far as one compares all of the costs with all of the advantages, translated into monetary value. In practice however, translating all of the effects into monetary terms is difficult and cost-benefit studies often confine themselves to that which can be easily assessed in this manner. In many cases therefore, it constitutes a more limited approach than a cost-effectiveness analysis.
Broadly speaking, economic analysis establishes the connection between costs and effects: it is an evaluation of yield or efficiency (synonymous).
2) Classification Criteria:
- Is a particular situation compared (i.e. is some attempt made to examine one or more alternatives) or simply described ?
- Are the costs (inputs) AND the consequences (outputs) of the different alternatives examined ?
"Partial" does not mean pointless!It does mean, however, that there is no attempt to address the issue of efficiency, making it possible to choose between several options.
In cases 1A, 1B and 2, there is no comparison between various alternative solutions: what we have is a straightforward description rather than an economic analysis:
- description of effects in 1A
- description of costs in 1B
- description of cost-effectiveness in 2.
Please note - This latter type of study is sometimes mistakenly referred to in literature as "cost-benefit analysis". E.g. Reynell PC and Reynell MC, The cost-benefit analysis of a coronary care unit, Br. Heart J. 34, 897-900, 1972. The authors present data on the costs of a coronary care unit and the number of lives saved. They do not compare the costs and effects of a coronary care unit with the costs and benefits of some alternative solution however. What is more, there is no attempt to convert the effects into monetary value.
Cases 3A and 3B offer comparative analyses, but never of both costs and effects at the same time.
Case 3A is obviously that of randomised controlled clinical trials relating to a new treatment (or screening procedure)
A cost analysis (case 3B) will compare, for instance, the respective cost of two vaccination strategies, but without considering their effectiveness.
Full economic evaluation
This is represented by case 4.There are basically four types of analysis.
a) Cost-minimization analysis
Cost-minimization analysis compares the costs of two programmes whose effects are identical.
Cost-minimization analysis differs from straightforward cost analysis (case 3B) in that it forms part of a study (often a randomised controlled trial) which proves that the two programmes are equally effective (e.g. conventional surgery and ambulatory surgery for one and the same pathology).Cost analysis often postulates or does not even seek to discover whether the treatments differ in terms of their results.
b) Cost-effectiveness analysis
The cost-effectiveness analysis compares the effects and costs of two or more programmes. E.g.: life extension following a kidney transplant or dialysis. One calculates:
- a cost per life-year gained
- or the number of life-years gained per franc spent.
A cost-effectiveness analysis does not necessarily compare two programmes applied to the same health problem: one might very well compare the number of life-years gained as a result of heart surgery, kidney transplants and wearing a safety belt. What matters is that the effects of the activities compared can be measured using the same criterion (life-years gained, number of days unfit for work avoided, etc.).
Furthermore, the effect does not have to be a treatment result, but may very well be a diagnosis or detection result: cost per case detected according to a particular procedure; cost per case diagnosed according to a particular diagnostic strategy. E.g.: Hull R et al., Cost-effectiveness of clinical diagnosis, venography and noninvasive testing in patients with symptomatic deep-vein thrombosis, N. Engl. J. Med., 304, 1561-7, 1981.
c) Cost-benefit analysis
Often it is not possible to measure one particular effect of several alternative solutions. Either the effects are of a different nature, or they are multiple. How in that case, can we reducemultiple or different effects to a common denominator?
Examples: compare home-based dialysis, hospital dialysis and kidney transplants not only in terms of life-years gained but also in terms of frequency of medical complications (multiple effects); compare high blood pressure screening in terms of life-years gained and anti-flu vaccination in terms of days unfit for work avoided (different types of effects).
In these instances, one has to go beyond the actual effect itself and assign a monetary value to the various effects being compared. This is basically a cost-benefit analysis.
The result can be presented in the form of a ratio (cost in $ / benefits in $) or a difference (net benefit of a particular programme - net benefit of some other programme... or of having no programme at all).
Translating life-years gained, days of unfit for work avoided and medical complications avoided or entailed into francs is no easy task. Which is why we sometimes quantify just some of the costs, namely those which can be easily measured.
Sometimes too, we measure costs and benefits from a particular point of view: from the point of view of the Nation as a community, or the social security system, the State, households, companies, etc. The differences can be considerable (e.g. cost-benefit of vaccinating workers in the catering industry against hepatitis A: the costs and benefits will vary markedly depending on your point of view).
d) Cost-utility analysis
Cost-utility analysis also attaches a value to the effects obtained but rather than being expressed in monetary terms, this value is measured in terms of the degree of utility of the effects obtained. That one and the same state of health (or illness) can be of varying utility depending on the individuals concerned can be seen from the following simple example: twins, one of whom is a bank clerk and the other a violinist, break their second finger on their left hand; although both are in the same state, if we ask each to rate on a scale of 0 to 10 the inconvenience of not having the use of their second finger (hence the utility of the treatment), we will receive very different replies.
This notion of utility allows us to take account of the quality of life afforded by different effects and at the same time provides a common denominator for comparing the costs and effects of different programmes.
This common denominator is normally expressed in "days of good health" or in "quality-adjusted life years" (QALYs).The number of life-years gained is weighted by an index expressing the degree of utility attached to the state of survival. Such an index must form the subject of opinion polls (conducted among patients or the general public) if it is to be valid.
C - Choosing a type of analysis
1) There is no fixed order of preference for these four types of economic evaluation.
2) Often, the researcher does not know in advance exactly what type of analysis he can apply; it may depend on the results of a clinical study associated with the evaluation: two treatments which ultimately prove to be equivalent will reduce a cost-minimization analysis to a cost-effectiveness description. Or a cost-benefit analysis may be combined with a cost-utility analysis, for particularly tricky problems (e.g. neonatal care).
3) The most important thing is to ascertain whether the complexity of the analysis is really commensurate with the question posed:
a) cost-benefit or cost-utility analyses primarily seek to determine whether a particular programme is "worth the trouble" compared with some other programme.
b) cost-minimization or cost-effectiveness analyses tacitly assume that the effect of the programme concerned is "worth the trouble".
4) Economic evaluation does not exempt us from having to think. It simply highlights the options or renders explicit certain options which have been ill-discerned or accepted without discussion. Whether or not economic considerations should prevail in the final decision is a matter for the decision-maker.
D - Types of cost and effects
- to the patient (and those close to him):cost of the treatment,
of travelling, etc.
- to the health service (in terms of staff, premises, equipment, running costs):
- variable costs: these vary in proportion to the volume of activity
- fixed costs: these do not vary (or do so only by leaps and bounds, by thresholds) according to the volume of activity.
Indirect costs (to the patient):
- loss of productivity (time wasted) in so far as it results from
participation in the programme
- psychological costs (loss of quality of life)
Externalities (external costs)
These are the costs occasioned by the programme beyond the scope of the health care system and the patient (or his family) - and tend to be fairly imponderable.
1_ Therapeutic effects (effectiveness, measured objectively, without any value judgements):reduction in the sickness ratio, reduction in the death rate, etc.
2_ Direct and indirect benefits (benefits, advantages)
Direct benefits for the health service: resources saved (although
this benefit can sometimes be a "negative" one: helping people to live longer
can increase the amount of use made of the health service!) Direct benefits for
the patient: resources saved (in terms of money or leisure time)
Indirect benefits for the patient: production gains (much debated area).
3_ Quality of life (utility):needs to be assessed separately.
E - Three important concepts
1) Marginal cost, marginal benefit, marginal effectiveness
This notion is defined here with regard to costs; it may equally be applied to benefits and effectiveness.
Average cost: total costs divided by total units of operation (patients, persons screened, vaccinated, etc.).
Marginal cost: additional cost entailed by one extra unit of operation.
Famous example: cost per detected case of cancer of the colon by looking for blood in the stools. The cost per case detected (average cost) obviously increases with the number of tests, but the cost per case detected thanks to the 6th test (marginal cost) was estimated at 47 million dollars in 1975!A prime example of how economic evaluation can clarify the cost of a particular option and make decision-makers aware of it.
2) Timescale adjustments to costs and benefits: "Present-worthing"
Quite apart from the effects of inflation, 100 francs available today is "worth" more than 100 francs available in one year's time: it is always preferable to obtain a particular benefit immediately and to defer the costs until later. The concept of "time preference" as used in economics, is a subjective one, which can vary from one society to another.
Estimating the present value of future costs and future benefits is done by means of a "present-worthing" (or discounting) procedure.
Discounting (in economics) is an operation whereby someone (normally a bank) advances the amount of a negotiable instrument ahead of its maturity date, less a deduction (discount rate), in return for ownership of this instrument.
The interest rate on loans operates according to the same mechanism; it "compensates" for the "loss of value" due to the "time preference". Thus, today's franc (F) will be worth more in n years (F') according to the equation:
F' = F (1 + r)n
where r is the "present-worthing" rate, the discount rate or the interest rate (the same formula applies for compound interest).
Conversely, the present value (F) of a franc in n years (F') is equal to:
F = F' / (1 + r)n
Since the time preference is not an exclusively financial concept, "present-worthing" also needs to be applied in cost -effectiveness and cost-utility studies (even though the effects here are not measured in monetary terms).The techniques for doing this can be examined elsewhere.
Some countries, in order to calculate the cost of the various programmes envisaged (health care, public works, etc.) officially recommend using a specific discount rate.
Reminder: the discounting procedure must be strictly distinguished from adjusting costs for inflation.
3) Sensitivity test
Economic evaluations can vary according to the degree of uncertainty, inaccuracy or controversy over the methods employed.
- the number of attacks due to a flu epidemic can vary from one year to another and cause the effects of a vaccination programme to vary
- the costs of hospitalisation can change rapidly
- various discount rates may be applied
- the indirect costs and benefits may be included or excluded from the analysis
- certain parameters are only roughly known and assumptions must be made as to their true value.
A sensitivity analysis will test the variation in the results (effects and costs) by simulating different variations in the parameters considered.
If a major variation in the parameters has little effect on the outcome of the analysis, the latter is said to be sound. Otherwise, some attempt will have to be made to discover the true value of the parameters by means of a validation procedure (test the working hypothesis under actual conditions, in the field).
A - Disasters and public health - Introduction
Over the last decade, public health has become a serious concern in response to natural disasters. There has been an increasing realization that the effects of natural disasters on the health of populations are amenable to study by epidemiological methods. Death rates can be computed for different types of disaster and attack rates can be calculated for the various types of disorder that occur in survivors, and indices of this sort can be used in planning appropriate measures for rescue and relief. There is also considerable interest at present in prevention, especially in the field of earthquake engineering, and such indices will provide essential tools for evaluating the effectiveness of various structural designs and building regulations in reducing death and injuries. The effectiveness of various types of assistance and the long term effect of aid on the restoration of the pre-disaster situation could be assessed as well, if adequate pre-disaster information were available. This article discusses the various epidemiological indices relevant to disaster situations and their value in planning preventive and rescuemeasures.
Death rates in disasters are highly variable, depending on a number of factors such as the type of disaster, the density and distribution of the population, conditions of the environment, degree of preparedness, and opportunity for warning. While the Yellow River floods in China in 1931 are said to have caused several million deaths, the earthquake in 1964 at Anchorage, Alaska, one of the severest ever recorded, claimedonly 115 victims.
While death rates are of little or no use as an epidemiological indicator for the planning of relief and rescue, they are of considerable interest for evaluating the effectiveness of preventive measures aimed at mitigating the effects of disasters, especially in the case of earthquakes.
The number of deaths per hundred houses destroyed in Turkey, a well-known area of tectonic instability, strongly points to some inadequacy in building techniques, making houses particularly lethal in this part of the world. The geographical distribution of damage and loss of life has not been uniform throughout the country, but has been higher in the east, despite a smaller population at risk in that part of the country. This may be due to the type of building material available, since, as one travels east, adobe (unburnt, sundried brick) is the main material used. In Iran, high death rates are associated with houses that are built of insufficiently reinforced adobe in which large concrete slabs are not properly supported.
High case-fatality rates from earthquakes are thus associated with the building techniques use in rural areas in certain countries. This highlights the need for control measures, such as appropriate building legislation and education, to reduce the number of deaths.
In other types of disaster, such as floods, the number of deaths may depend on early recognition of the impending disaster and on an appropriate warning system which allows the population enough time to leave the area or seek refuge. The timing of the warming is then all - important - whether to be on the safe side and run the risk of giving a false alarm, or to wait for definite signs of a disaster and risk giving the warning too late. The problem is similar to the dilemma of sensitivity versus specificity of methods of case-detection that is well known to epidemiologists and public health specialists involved in preventive medicine. The number of deaths may thus be used to check the adequacy of the criteria adopted for early warning of disasters.
Age-specific mortality rates after a disaster may also provide interesting insights into the reaction of the people at the time of impact. In Guatemala, following the 1976 earthquake, age-specific mortality rates showed a bimodal profile.In the town of Patzicia (377 deaths) death rates above the overall rate of 3.5 % occurred among the 5-9 years age group (5.6 %) and among persons aged 60 years and over (5.5 %).
A similar profile was obtained in the town of Sumpango (244 deaths). In both towns, the observed mortality was lower among infants and children 1-4 years old than in the 5-9 years age group, suggesting that parents took special care of their younger and more defenceless children. Following the Bangladesh cyclone in 1970, age-specific mortality also showed a bimodal, but different, distribution : 29 % and 20 % respectively, in the very young (O-4 years) and over 60 years age groups. Data of this kind may thus identify disaster-vulnerable groups, and suggest ways in which appropriate education in communities exposed to natural disasters could reduce the number of deaths.
Effects on the prevalence and incidence of communicable diseases form another important aspect of disaster epidemiology. There may be increased transmission of diseases following the collapse of control systems (for example, interruption of insecticide spraying) and this aspect requires the implementation of a disaster-adapted epidemiological surveillance system with appropriate indices and trends. The consequences of the disruption of routine health services in disaster situations as regards mortality and morbidity are often overlooked. Such effects are more relevant in urban areas and in developed countries. Recently, there have been cases where mortality from ischaemic heart disease, renal failure, and possibly from obstetric causes, have apparently been affected by the unavailability of normal health services. Mental health is another recently identified and perhaps underestimated problem related to disasters, especially in urban areas. Studies are under way to assess the long-term impact on mental health of the Nicaragua earthquake. However, indices of mental health are hard to design, because of the large variety of cultural patterns found in the disaster-struck community and our ignorance about what should indeed beconsidered an adjusted reaction to disasters in exposed individuals.
Another major disaster-related problem is nutrition. It has been widely studied over the last 10 years, following the political uprising in Bangladesh, the civil war in Nigeria, and the droughts in countries south of the Sahara and in Ethiopia. Methods of nutritional assessment have been developed but the indirect effects of famine on the community extend much beyond the epidemiological field. The nutritional indices developed so far, such as weight-for-height, consider only one facet of the problem. Malnutrition, and its corollary food aid, affect the population through a variety of mechanisms, which may range from effects on fertility and abortion, to price and market structures, and the distribution of wealth. Much more study is needed to develop appropriate measurements of the effects of malnutrition and to evaluate the appropriateness of various forms of food aid.
B - Disaster Preparedness: Overview of issues
Natural disasters which occur relatively frequently and have a significant impact may be classified into four main categories: floods, earthquakes, cyclones and droughts. Other catastrophic events, such as landslides, avalanches, snow and fires occur less often and threaten smaller proportions of the populated world. The destructive agents in four main classes mentioned above are wind, water (a lack or excess thereof) and tectonic forces. While all these generally cause structural damage, their mortality and morbidity effects are variable.
The disaster cycle can be differentiated into five main phases, extending from one disaster to the next. The phases are : the warning phase indicating the possible occurrence of a catastrophe and the threat period during which the disaster is impending; the impact phase when the disaster strikes; the emergency phase when rescue, treatment and salvage activities commence; the rehabilitation phase when essential services are provided on a temporary basis; the reconstruction phase when a permanent return to normalcy is achieved.
The disaster-induced mortality and morbidity differ between these phasesand are also a function of the prevailing health and socio-economic conditions of the affected community. As a result of this, global statistics on disasters seem to indicate a significantly higher frequency of natural disasters in the Third World than the industrialized countries. Assuming nature not to be conscious of economic differences, a disaster may be defined by the vulnerability of the population to a natural event and not by the mere fact of its occurrence (de Ville de Goyet and Lechat 1976).
Special characteristics of disasters
It is useful to start by locating the four main disaster types on relative scales of lethality, predictability, onset time and impact scope. This ranking provides some guidance towards understanding the variation in mortality impact noted among disaster events across time and space. Figure 1 displays the four scales with the relative positions of the disaster types.
Although drought-related famine is a very special class of disaster, it nevertheless falls within the general paradigm which characterizes natural disasters. Famines are disasters of high predictability. With the exception of the Great Bengal Famine of 1941-43, almost all important famines since then, and certainly the ones of Sahelian Africa and Ethiopia, were more or less foreseen. Famines, in fact, provide an excellent illustration of the fact that the knowledge of impending disaster does not imply that a community can or will take responsive action.
At the other end of the scale, earthquakes tend to be least predictable disasters, striking with little warning. Japan is one of the few high risk countries that have an effective warning and evacuation system, as well as excellent community education programmes (Nakano et al 1974). The earthquake of Niigata (16th June 1964) registered 7.7 on the Richter scale. Although 20,000 houses were destroyed, only 13 people were killed and 315 injured. Due to the quality of its preparedness programmes, Japan suffers very limited mortality despite the high number of seismic shocks it registers (Akimoto 1972).
In terms of lethality, earthquakes present the greatest risk of death to those affected (Table 1). Onset delay is also the shortest in earthquakes, which is related, to a certain extent, to its low predictability. Famine, on the other hand, has a slow build up period before it reaches acute emergency proportions. Floods can be somewhat ambiguous in their onset characteristics. They can be slow-developing and fairly predictable such as the annual floods in the plains of the Ganges in India or in the Itajai River basin in Brazil, but nevertheless regularly cause a number of deaths and a certain amount of damage (Civil Defense of Santa Catarina, personal communication, 1983). Acute and catastrophic floods are usually generated by cyclones or tidal waves; examples are the ones in Philippines (1984) and Bangladesh (1985). Floods cause somewhat lower mortality than other disasters, but the scope of damage is generally wider and more pervasive.
Trends and differentials in disaster related mortality
On a global level, the mortality generated by natural disasters showssome interesting tendencies, creating the beginnings of an analytical framework within which specific impacts may be systematically analysed for robust indicators, efficient needs assessment or preparedness and rehabilitation planning. The mortality form disasters is a function of risk, development and coping or adjustment capacity (preparedness). Table 2 displays comparative data on these from a number of countries.
The official disaster data reveal two important variations in disaster mortality : a temporal increase and a geographical correlation.
Time trends in disaster mortality
Mortality per event - Between the two ten-year periods, 1960-69 and 1970-79, a significant increase in average mortality per event is noted in all categories except perhaps in floods where direct mortality is generally low (Table 3).
The greatest increase is noted in earthquakes, which takes a quantum leap from one period to the next. The mortality in 1960-69 was 750 deaths per earthquake whereas in the following ten-year period the death toll per event went up to 4 871 deaths per earthquake.(It is interesting to note here that the total number of earthquakes requiring international assistance did not increase significantly from one period to the next). The huge increase in earthquake mortality is partially explained by the Tangshan strike of 1976 in China which contributed more than half of the entire ten-year period death toll. The official estimate of 224 000 dead accounts for exactly 47 per cent of the total number dead due to earthquakes during this time. But even allowing for the Tangshan quake, the death mortality per strike remains as high as 1,780 per earthquake compared to 750 in the previous decade. Population density (Lechat 1984), structural quality (Glass et al 1977), time of strike (De Bruycker et al 1983) and intensity of seismic activity (Alexander 1985) seem to be the main risk factors, but they fail to explain adequately the high mortality in earthquakes.
Local conditions, evidently, play a bigger role than expected in determining disaster mortality.
Mortality per 1 000 exposed - The mortality rates per 1,000 exposed to disasters increase significantly over the two decades for all types of disaster except floods, although the increase is relatively slight in earthquakes. This stability in the mortality rates of earthquakes is mainly due to its being a high risk disaster with comparatively localized effects. The greatest increase is observed in drought-related famines where the population get progressively weaker from previous famines and succumb in each successive crisis in greater numbers. Floods show a slight improvement, as it were. However, the mortality impact of floods may be hypothesized as being typically spread over the period following the flood rather than being a direct and immediate effect of the event. This increase in the mortality rate possibly reflects the inability of current disaster management policies to reduce the vulnerability of a community. Despite significant disaster assistance, and aid of nearly one billion dollars in the 1970-78 period, the increase in mortality, controlling for the number of events, indicates a steady decline in the resistance of the populations to disasters (Stephens et al 1982).
Regional differentials in disaster mortality
Geographically, the mortality generated by disasters is consistently and positively correlated to the level of the economy. Table 4 presents some figures of mortality classified into three national income categories.
Mortality rates, controlling for the number of disaster events, are substantially higher in poor countries than in the richer ones. The classification is, of course, gross and the data demand closer analyses for better definition of risk factors and vulnerability patterns amongst the severely affected populations. Such analyses can have direct impact on programme planning and policy-orientation. Table 4, however, does serve to indicate the important influence of the prevailing socio-economic conditions on the eventual disaster impact (Cuny 1983, Shah 1985). For predictive and needs assessment purposes then, the prevalent socio-economic and health conditions prevalent in the affected community could be a better determinant of the epidemiological impact than the physical characteristics of the event.
As seen in Table 4, disaster-generated mortality increases dramatically as economies descend the income scale. Barring a deliberate selectivity of nature in her allocation of high intensity disasters to low income countries, a less "natural" explanation is the differential power of communities to resist and recuperate from shock. Table 5 presents some data on the 1971-72 earthquakes of Managua(Nicaragua) and San Fernando Valley (USA) (Seaman 1984).
The comparison reveals some interesting points. Speaking "naturally" of the two earthquakes, the seismic activity level of the California earthquake was significantly higher, registering 6.6 on the Richter scale compared to 5.6 in Managua. (One unit increase is an important proportion due to the logarithmic scale of Richter readings). On the Mercalli scale (measuring the extent of physical damage over surface area) the California quake caused major damage (IX-XI level damage) over 100 km2, whereas Managua registered a lower level of damage to a smaller area of land. The population directly affected by the earthquake in California was 13 times that of the earthquake in Managua. Despite all physical conditions indicating to the contrary, the mortality in Managua was somewhere around 5,000 deaths compared to 60 deaths in California. Similarly, in 1974, Hurricane Fifi left an estimated 8,000 dead in Honduras, crashing through at a windspeed of about 250 km/h and causing80 per cent disruption of impact area. In the same year, Cyclone Tracy killed 49 in Darwin, Australia, with a similar windspeed and proportion of impact zone disruption. The selectivity of impact can also be observed on a more localized scale. The Guatemala earthquake in 1976killed about 1,200 people and left 90,000 without homes in the cityalone, but almost exclusively from the slum populations. Shanty towns and slum areas in the burgeoning metropolises of several high risk Third World countries are especially fragile in all kinds of disasters. The unprecedented increase in these slum populations has contributed to the inflation of the disaster victims in the recent years. The Jakarta shanty towns, for example, where floods frequently cause the canals-cum-latrines to overflow into the living quarters of the slum-dwellers, have created epidemics of typhoid and skin and gastro-intestinal diseases and raised infant and child mortality. Flooding in low-lying areas of Bangladesh has exacerbated endemic cholera and other diarrhoeal diseases.
Disaster related morbidity
The data on morbidity (namely injuries and disease) after a disaster are remarkable by their absence or incomparability. The definition of injury, when registered, is largely unstated and reporting of diseases largely incomplete. This has resulted in a series of observations, some anecdotal, some systematic, but nearly all fragmentary. There is clearly an urgent need for standardized reporting of injuries and cause of death (preferably using a standard format such as the International Classification of Disease). Without such standardization, disaster planning and management remains an ad hoc activity and analyses of thekind attempted here can only be undertaken somewhat superficially.
Injury profiles of natural disasters
There are some recorded figures available on injuries sustained in earthquakes where authorities registered and published morbidity data. It is uncertain what qualified as injury and, more importantly, the bias introduced by those who were not treated in hospital. There are even fewer data on non-traumatic morbidity. Classification bias and general incomparability pose important problems for those making analyses for programme or policy purposes.
In the case of earthquakes, the disaster type most prone to causing traumatic injury, fractures constitute the major portion of the impact. Fractures of the extremities are significantly more frequent than any other sort. In the Tashkent (1966) and Ashkabad (1948) earthquakes, two of the few instances where injuries were classified according to type, about 69 per cent were fractures of the limbs (Beinin 1981). Fractures of the extremities as a proportion of injuries sustained in Managua (1971) and Iran were 77 per cent and 58 per cent respectively (Whittaker 1974, Saidi 1963). In the recent volcanic eruption of Colombia, gaseous gangrene was the second most serious morbid condition next to suffocation. The Mexico earthquake seemed to have relatively few injuries compared to deaths, although the official figures are not yet available. Most injuries, be they lacerations in cyclones or fractures in earthquakes, tend to occur during the catastrophe itself or in the very immediate post-impact phase.
Clearly, in both earthquakes and cyclones, structural quality of housingis a major determining factor of the extent and type of injury, which is, in effect, a proxy variable for the socio-economic level of the community or the houselhold (Haas et al 1977). The relationship between injury and death and other variables not directly related to the catastrophe is discussed later in the paper.
Disease profiles of natural disasters
Despite popular belief, major epidemics are fairly rare events after natural disasters (de Ville de Goyet and Lechat 1976, Seaman 1984), especially in industrialized countries. Some risk exists in developing countries where sanitation is poor and endemicity of many communicable diseases is normally high. A severe malaria epidemic occurred after Hurricane Flora in Haiti in 1964, ostensibly caused by the multiplication of breeding places for mosquitoes in the damaged areas. However, the interrelationships between the disaster and the hurricane are more complex and epidemiologically interesting than this explanation indicates (Lechat, personal communication, 1985). Overcrowding and breakdown of fragile sanitation systems could, conceivably, provoke epidemics in developing countries. An epidemic ofleptospirosis was reported in Recife, Brazil, after floods in 1975 (Correa et al 1972). These are, generally, fairly unimportant in scope. Of a more serious nature are those brought on by famine conditions, such as the cholera epidemic in Somalia in 1985 and meningitis in Ethiopia earlier. Camp conditions, reduced resistance and breakdown of the social systems are possibly the provoking factors for these epidemics, but to what extent and how these factors are associated with the diseases cannot be conclusively established without specific study. Usually, however, disasters do not generate "new" diseases unless they are brought in by migrating populations, as has been observed in the recent African famines. Floods tend to exacerbate endemic communicable diseases in populations, especially if sanitary and sewage systems are primitive or, as is more often the case, non-existent. A regional variation in diseases similiar to that seen between developed and developing countries is noted within a developing country. The incidence rates during a cholera epidemic in Bangladesh were found to becorrelated to education and income. The poorer sections of the affected region used canal water for drinking and washing purposes and presented, furthermore, a generally lower resistance to infections. The incidencerate of the disease per 1,000 families with no schooling was 16.3whereas it was 8.2 among families with at least one high school graduate (Levine et al 1976). In famines, the synergy between malnutrition and infectious diseases gives it an altogether different dimension as compared to others. Communicable and nutritional deficiency diseases in famine disaster are, in fact, the principal manifestation of the event.
Long-term impact of disasters
The long-term impact of disasters, possibly the most pervasive and destructive phase, expresses itself variously. Disaster-induced death and disability of an earning member of a family implies a lifetime's loss of revenue and possible destitution.
A study by Karakos et al (1983) after the earthquake of 1980 in Thessaloniki, found that 50 per cent of all the families with at least one death lost their only working member and thus experienced a direct decrease in income. In developing countries, where the informal sector is an important source of revenue for a large proportion of the population and social security is less developed, such a loss can be fatal to the surviving members of the family.
In flooding disasters, salt-water contamination of subsistence and marginal land indicates not one, but several, harvests lost. For nutritionally and economically fragile populations, this means a rise in mortality as a secondary effect of the disaster.
Similiarly, death of breeding stock of herdsmen and loss of capital or tools of trade due to water damage, cyclones or earthquakes effectively destroy the means of livelihood of these families (PAHO Disaster Reports 1981). Finally, the death of a mother has a devastating effect on small children, raising the morbidity rates among them (Patil and Koshy 1984). A great deal of empirical work, clearly, needs to be done in order to evaluateaccurately the complete health impact of a disaster. So far, impact evaluation stops at counting the dead.
Policy implications for the health sector
The increasing interest shown in the impact of natural disasters by researchers in disciplines other than engineering, geology and meteorology has had a salutary effect of raising questions on current international and national disaster policy and relief action. The growing body of literature emphasizing the importance of discriminating between the geophysical event and its human consequences is provoking organizations and governments to take another look at disaster relief. This is an encouraging turn of events given the large calamities witnessed in the last three years and the worrisome increase in the number of victims, dead or destitute (Wijkman and Timberlake 1984, OFDA 1982.
Futhermore, with non-emergency developmental aid and co-operation between the first and third worlds grinding to a slow halt, the substantial resources generated by public appeals for disaster assistance demands an efficient and cost-effective use, oriented towards a long-term resolution of the problem instead of an emergency, stop-gap measure.
It is not difficult, even with the patchy and incomplete data available, to demonstrate that the physical characteristics of a natural event do not satisfactorily explain its impact. This paper has cited several comparative instances illustrating this point. Most natural disasters and the damage associated with them are characteristic rather than accidental features of the afflicted community (Hewitt 1983). Although this is well proven (but not widely accepted) in the case of famines, it is less well documented in the more acute disasters. Two of the largest famines since World War II have been in countries with a normal or more than normal food production during the famine year. Rivers (quoted in Wijkman and Timberlake 1984) observed that Ethiopia was a net exporter of food in 1973, and both Bangladesh and Bengal produced more grain in 1974 and 1941 respectively than in the preceding years (Wijkman andTimberlake 1984, Sen 1983). Drought sometimes serves as a trigger mechanism for a famine, but the disaster remains a largely poverty-related catastrophe with a very weak causal relationship to food supply. Similarly, the impact of other disasters is a function of the physical and economic resistance of the population.
This distinction between the event and the impact brings into a so far simple situation a host of messy, complex and difficult socio-political and economic considerations. The essence of the vulnerability issuelies in the fact that the communities can "cope with earthquakes but not with their fellows" (Brecht 1965). Japan, as cited earlier, has minimized earthquake-generated damage despite frequent shocks of high magnitude, where Nicaragua and Guatemala are unable to withstand quakes of much lower intensity.
Disaster relief has traditionally been based on policy formulated from charitable motives drawing on critical and emergency care approaches. This has made it a primarily medical activity, involving surgical units flown out with specialist teams and field hospitals equipped with sophisticated life-saving apparatus. On the other hand, research and development in natural catastrophes has concentrated almost exclusively on climate monitoring, radar tracking, flood barriers and other middle to high technology devices. Finally, the charitable nature of disaster relief permits policy to bedictated by the principle that any aid is good aid, thereby generating anecdotes in the literature on the superfluous, inappropriate and frequently, absurd relief packages.
Lechat (1981) has pointed out that the even volunteers, who tend to descend upon the scene, can be a serious liability in a crisis situation, due to their inexperience andredundancy.
As far as policy is concerned, two observations are in order before discussing further the implications. First, although medical care (in so far as surgical and critical care is implied) appears largely unnecessary, health and public health, having played thus far a small but significant role, have a responsibility and potential in this area. Second, the central issue in management and relief in natural disasters has to be recognized as the reduction of vulnerability to disasters of the community and, within it, the population at risk. The progressive diminution of mortality and morbidity impact of a catastrophe should be the main health policy objective of disaster management.
The extraction of disaster relief from its comfortable niche of charity and emergency care would force planners to address structural issues of the availability, equity and appropriateness of health care in the community. This widening of scope permits and, indeed, demands the inclusion of health planning for disasters in the on-going health plans of the region, thus providing an existing working base and infrastructure. A disaster response could be successfully incorporated into the training of the health workers and their activities. The fundamental tenets of primary health care may be applied to disaster preparedness and prevention programmes, in so far as both involve community participation, multi-sectoral objectives, the use of local resources and building human resource capability.
These aspects of primary health care are especially relevant in disaster management for the several reasons. External emergency disaster assistance is rarely, if ever, either on time or particularly appropriate. This is not due to any sluggishness on the part of the distant agencies, but rather the result of inadequacies in communication, assessment of need, accessibility and other difficulties. It has been observed that families, friends and neighbours search, evacuate and extricate their own in the immediate aftermath of a disaster and that by the time external relief teams are functional on site, a very large majority of the total dead have already died (DeBruycker 1983). In other words, those that die do so within the very first hours of the event, and immediate emergency rescue and care are provided by the local inhabitants. Those that survive the first twenty four hours generally survive without major bodily harm. In addition, the survivors, contrary to expectations, are rarely in a panic and disorganized. In fact, they act with calm and common sense to manage to the best of their abilities their own affairs (Wijkman and Timberlake 1984). External emergency relief, therefore, is largely expensive, wasteful and not particularly effective.
This, however does not imply that disaster relief should be abandoned, because in a catastrophe people need help and the resources exist.
It simply indicates that external disaster relief should focus on reducing population vulnerability and invest in structural changes in health care organization and accessibility of the population. It should also provide training and education at the local level for emergency activities such as evacuation, first aid and so forth. This rationalization of policy would provide the communities with the tools and knowledge by which they could defend themselves against future hazards. Disaster assistance resources could be deployed to expand the primary health care structure and train their personnel in emergency shelter management, rapid epidemiological surveillance and control, food distribution and needs assessment and recording and registration.
A - Rapid needs assessment: national response to disaster
Following a natural disaster, there is usually a lack of specific reliable information on the extent of damage and of medical need. Often the problem lies rather with an excess of contradictory ill-founded or exaggerated information. The rapid acquisition accurate information or of estimates of known accuracy are prerequisites both for planning of anational relief programme and for guiding international assistance. Data collection need not impede initial relief work since this can begin on the basis of the information obtained. When adequate preparedness fordisaster exists, this information will become available and will include accurate and up-to-date information relevant to the disaster affected area including an inventory of the medical facilities, available personnel, material, transport and communications, both within the disaster area and in adjacent parts of the country. This knowledge will allow a reasoned first estimate to be made of the extent and nature of the likely need created by a given event, and of the resources availableto meet that need. The assessment of needs and priorities involves several steps. Data must be collected, interpreted, and disseminated to potential users, and finally utilized in defining priorities and making decisions. Only the first two aspects are considered here.
Types of data required
Two broad categories of information are required by health administrators :
1) General statements about the extent of damage, the area and population affected, functional damage to public services, telecommunications, highways and roads, power and other utilities. The collection of this information is independent of the health administration.
2) Specific medical, epidemiological and administrative information on health problems and available resources. No universal answers can be given as to the health information which will be needed, nor on the most effective and least costly methods of collection.
Here, as in most other areas which are under discussion, the solution most suitable to local conditions must be selected and measures for implementation included in the predisaster plan of operations. The following types of data have been shown to have practical value following recent disasters:
Number and proportions (or rates) of injuries : several categories of injuries need to be defined. The most pragmatic classifications are based on the site (e.g. fractures of arms and legs) or the severity of trauma. The latter classification is the basis for effective triage in the field. Gross data on the total number of persons injured are to little value to relief officials unless some indication is provided on the type of treatment required, for instance, ambulatory treatment or major surgery.
Epidemiological data on sex and age are also required and can be obtained on a sample basis. Spotsurveys in Guatemala following the earthquake of February 1976 suggested that children under five and adults over 50 have a considerably higher rate of trauma. Such information has direct implications for the type of treatment facilities and material required.
Incidence of communicable disease
The risk of outbreaks of communicable diseases is usually of major concern both to the general public and to the administration. A quick survey of the site of impact will provide baseline data and should lead to the setting-up of an epidemiological surveillance system based on reporting of suspect cases of selected diseases. This can be made using simple symptom reporting from out-patient and hospital departments. The time necessary for disease transmission and therefore for the building-up of an epidemic, precludes outbreaks in the first few days of the emergency.
Inventory of remaining health facilities
A detailed inventory of functional health facilities will include much information not included in a general survey of damage. For relief purposes, the emphasis is placed on describing the damage to existing facilities. Field surveys by experienced professionals combined with low-flying reconnaissance missions by helicopter have been effective in providing a quick assessment, but a more detailed inventory must be obtained by ground survey.
Inventory of medical supplies
A quick survey of supplies available at the site of the disaster may indicate that urgently needed drugs or material could be easily salvaged and used. In Nicaragua, a survey of the medical warehouses in Managua a few days following the devastating earthquake which destroyed the city in 1972, showed that over one million dollars worth of medical and surgical supplies were immediately salvageable although the warehouses had been reportedly totally destroyed. The assessment of available supplies should not be limited to the site of the disaster but, if necessary, should be carried out at national level.
Inventory of essential services available within or outside the stricken area
The assessment of available resources should not be limited to the area of impact. For instance, a daily monitoring of beds available in all hospitals within and outside the disaster area following the Nicaragua and Guatemala earthquakes dramatically improved the use of resources and reduced the need to evacuate patients to hospitals in other countries.
Water supply and sanitation facilities
Health problems are not limited to the management of mass casualties : the provision and repair of damaged water supplies and sanitation systems are important. A survey of the water supply systems in urban or rural areas should be included in the early assessment of health needs. In the assessment of water supply and sanitation it is important that assessment teams be constituted of experts familiar with the sanitary conditions existing before disaster. Experience has shown that engineers unfamiliar with poor rural conditions - and especially foreign experts - tend to overestimate the emergency requirements and falsely attribute to the effect of the disaster deficiencies in sanitation and water supplies which actually reflect long-term development problems and which existed before.
Techniques and methods
The techniques and methods for the quick collection of data exist with a sufficient level of accuracy for emergency purposes. However, the personnel with the necessary experience for this task may be under strong emotional or political pressure : exaggeration is often the rule and may reflect the interest of the administrators in attracting maximum attention for the population for whom they are responsible. Long delays may occur before information is transmitted to the central level unless telephone or radio is available. Reports received during the first twelve hours generally speak only of casualties and damage in the most accessible areas. For example, following the Guatemala earthquake, it was not until forty-eight hours after the impact that any specific information on the extent of the damage outside Guatemala City was available, although there had been earlier rumours of widespread damage in rural areas. The major advantage of this source of information lies in the fact that it is part of a normal and therefore organized reporting process and can be improved to meet the special requirements of an emergency situation. It is particularly valuable when a continuous flow of quantified data is required to monitor major changes in the situation.
A systematic rapid sample survey of the areas affected by the disaster is necessary to provide an immediate overview of the situation. One cannot overemphasise that the survey should cover the entire area and not be limited to sectors reported damaged. Failure to comply with this conditions may result in isolated localities being left unattended for aconsiderable period of time.
Advantages of a survey being promptly carried out are many : it quickly provides standardized information which would otherwise not be available until normal communications are re-established. Following the earthquakes in the rural parts of Peru(1971) and Guatemala, field surveys provided the national emergency committees with a comprehensive overview of the extent of damage within three to five days. The participation of a trained statistician is recommended in the planning of a survey. The results of the survey following a disaster will only be of value if it has been properly planned and data adequately processed. If this is not done the data may be invalid. With regard to the large variety of data to be collected, survey teams should be multidisciplinary and call on the best expertise available in the country. In the health sector, epidemiologists are best qualified to assume this responsability and provide a diagnosis on the public health situation.
Aerial and satellite photography
Recently, greater use of sophisticated technology such as photography by high-flying aircraft and satellites has been made in quantifying the extent of damage. The accuracy is known to be useful for the assessment of macro-damage, but its potential in health remains to be determined. The technology is not readily available to most disaster-affected countries and its cost may limit future application.
Implementational aspects of a needs assessment exercise
The objective of the assessment is to determine which human or material resources will be required to cope with the emergency. These resources must be readily and promptly available : if too much time elapses between the identification of a need and the arrival of supplies at the side, they may no longer be useful. For example, after an earthquake first-aid and surgical supplies must be delivered within the first few days to outlying health services. The international community, by-and-large, is not often in a position to respond so quickly. Following the Nicaragua and Guatemala earthquakes, the number of aircraft arriving with medical supplies was reaching a peak two weeks after the impact, by which time the priority had already shifted to the provision of sanitation and other forms of reconstruction.
The responsibility and authority for conducting an assessment of needs lies with the government of the disaster-affected country. Should the national resources be found insufficient for relief, external assistance from bi-lateral or international sources will have to be sought. However, when a disaster is of such magnitude as to make extensive international aid necessary, it often happens that nationals with the necessary expertise are too scarce to conduct on their own an adequate assessment of the extent of the damage and of the needs. Moreover, potential donors need objective and detailed information on material requirements in order to plan their action and make estimates of the expenditure involved.
It must not be forgotten that they have to justify their action to their governing bodies. Consequently, if often happens that teams or individuals are sent to the affected area by each organization to carry out an independent survey and provide first-hand data on the requirements. The participation of the international community in a joint assessment of disaster situations, under the authority of the disaster-affected country would contribute greatly to a better and more efficient response to the challenge of natural disasters.
B - Medical supply management
During the emergency, most medical supply needs can be met with the use of local supplies. The primary problem in crisis management is distributing the supplies that are localy available. Other significant areas are inventory preparation and control and the use of donated international supplies are as follows : centralized reporting system ; preparation of a list of essential medical supplies ; inventory preparation and control ; compiling the inventory of supplies available in the Nation ; medical supply distribution, field receipt (sorting and storing), and requisition of supplies ; cheking distribution networks ; arrival of incoming supplies.
C - Disease surveillance after disaster
Field investigation or rumors and reports of communicable disease
Rumors and unconfirmed reports frequently circulate after a major disaster, but until recently epidemiologists were not asked to take part in relief efforts except when there was need to investigate the more serious of these. In recent major disasters, the appropriate evaluation of rumors has been made possible through this increasingly earlier involvement of epidemiologists in the relief response. This can be attributed to two factors. The most obvious is that prompt investigation can take place before a situation gets out of hand. Perhaps more important, however, has been the existence of the opportunity to educate members of disaster agencies, the media and the national health authorities about appropriate ways to interprate and respond to rumors.
Gaining access to laboratories to obtain definitive diagnosis and support for epidemiologic investigations
When the epidemiologist, investigating a rumor, encounters patients with symptoms compatible with the disease in question, it is imperativ to collect specimens appropriate for diagnosis, and to properly handle and transport them to a competent laboratory, where they should receive priority attention. Selected laboratory investigation of symptoms or symptom complexes (such as fever-diarrhoea) reported to be increasing may also be required for undertaking appropriate public health measures and developing guidelines for proper management of patients.
There are reasons that it may be necessary to obtain laboratory confirmation of selected notifiable diseases from a sample of patients. The first of these is that not all notifiable communicable disease can be diagnosed with confidence on the basis of clinical criteria alone. Second, the public health laboratory is essential to the promotion of efficient communicable diseases control. The medical officers are primarily concerned with communicable diseases in general population, rather than in individual patients. For such persons, the diagnosis of typhoid fever or measles in a hospitalized patient only represents the tip of an iceberg. Because epidemiology units do not have the authority or resources to adequatly carry out control measures, it is critical to, as effectively as possible, present information from surveillance and the field investigation to key decisions makers. Epidemiologic information, implications, and an outline of alternatives of action must be summarised in the presentation in nontechnical terms understandable to laymen.
D - Emergency assessment of nutritional status
Suitable methods must be adopted for the rapid objective measurment of the nutritional status:
- of individuals eligible for special food relief
- of communities in order to detect changes with time and decide priorities in food distribution (nutritional surveillance).
Weight-for-height is the best indicator for the diagnosis of nutritional status, nutritional surveillance and individual screening. Weight-for-age and arm circumference are less reliable for assessment and screening but can be used to measure changes with time. Oedema rates are a valuable indicator when kwashiorkor is the prevalent form of PEM in the area. Results of surveys and surveillance must be interprated with caution. That can be misleading unless the individuals measured are representativ of the whole population and the technique is standardized and properly used.
Why measure malnutrition in emergencies ?
During a nutritional emergency the relief foods may be scarce and should be given to the people in greatest need. Since much of a population may be able to supply part of all of its own food, it is very useful to have an objective and quantifiable measure of nutritional status. Measurment of nutritional status in emergencies relies mainly upon talking body measurment (anthropometry), particularly height, weight, and arm circumference. Valuable information may also be obtained from simpler methods, for example monitoring clinic records or measuring the prevalence of oedema. The commonest reasons for measuring malnutrition in a relief program are:
- Initial assessment : a rapid survey of population should be done before initiating a relief program, in order to identify the areas or groups that are most affected. Surveys of this type need to be carrefully designed and conducted by an experienced team. They will not be considered further.
- Individual screening : body measurment may be used to select the malnourished individuals eligible for food relief for themselves or their whole family.
- Nutritional surveillance of the population: the repeated measuring of entire communities gives an idea of differences among the various population groups and changes in nutritional status with time. It may be used to decide priorities in the contribution of relief and will also provide some information about the effectivness of the relief program. In nutritional surveillance one is not interested in monitoring the progress of a child, but in knowing whether the overall nutritional condition of a village (or camp) A, is good or bad, is better or worse than that of village B and C (and so requires more supplies and personal), and whether it is improving or deteriorating with time. Nutritional surveillance should not be confused with the "surveillance" or follow up of an individual child in nutritional centers or health services.
Indicators of malnutrition
Clinical signs of PEM or specific deficiencies - Clinical signs are signs that can be rapidly assessed by touching or examining the child concerned rather than by instruments or tests.
Body measurements - They are used to detect malnutrition, but not food shortage, since malnutrition can also be caused by ignorance of faulty feedings habits in the presence of sufficient food. The results of body measurements can be misleading if considered in isolation. Chronic under nutrition leads to a slowing in a child's rate of growth. A chronically malnourished child will be short for his age ("stunted") although he may be of otherwise normal proportion. An acute episod of severe undernutrition results in a loss of muscle and fat which are used up to provide energy and the individual becomes thinner without significant effect upon height ("wasting").
In an emergency what is important is the measurement of acute malnutrition, the effects of chronic malnutrition being of less concern. Because both stunting and wasting result in low weight-for-age, relating body measurements to age is not recommanded. Two measurements are commonly used to access acute malnutrition ("wasting"):
- arm circumference (AC).
Presence of diseases associated with PEM - These, include measles, diarrhoea (defined for instance as three or more loose stools per day), whooping cough, etc.
Mortality data - PEM is associated with increased mortality among young children (e. g. from measles, etc.). The data collected should expressed as rates ; for example, the rate per thousand of marasmus among infants (aged 0-1) in a refugee camp is :
Number of infants with marasmus in the camps
--- x 1000
Total number of infants in the camps
Very great efforts should be made to measure children accurately. Small errors (e. g. 2 - 3 cm in height) in the measurement of a younger child may lead to significant errors in the classification of child's nutritional status. Only one indicator should be selected. Weight-for-height, the recommended body measurement in times of emergency, is a sensitive indicator of acute malnutrition. It is fairly independent of sex, race, and age (up to about 10 years of age). It requires a sufficient number of robust scales and adequate training of personnel. Neither condition is easy to meet in an acute emergency situation. If ages are not known, arm-circumference-for-height is the best alternative. Measuring arm circumference instead of weight results in only a marginal saving of time compared to that required for traveling and assembling people. Several techniques such as the QUAC stick (Annex 5) have been divided to simplify field work and are useful for the screening of large numbers of children. As a second alternative, measurement of arm circumference alone (without measurement of height) is acceptable in situation where resources are extremely limited. Considerable time is saved by not measuring height. The sensitivity of the measurement as an indicator is poor but sufficient when PEM is severe and whitespread.
Other indicators of the evaluation of relief programs
The following indicators can be useful in evaluating a relief program:
- age distribution of children attending relief centers compared with the age ditribution from census data,
- monthly attendance rate of children registered (this is obtained by deviding the monthly average number of those attending by the total number of children registered),
- malnutrition rates in people attending relief centers compared with similar rates obtained by an occasional survey of random sample and house-to-house visits in the same area.
E - Organisation of nutritional relief : general food distribution, mass and supplementary feeding
There are four ways in which food relief may be organised :
- general food distribution : dry food is distributed to people who are able to prepare their own meals,
- mass feeding : prepared meals from a central kitchen are served to the population,
- supplementary feeding : in addition to the ration (dry foods or meals) for the whole family, vulnerable groups receive an extra meal or ration to meet their particular needs,
- intensive or therapeutic feeding of PEM cases.
Food must be nutritionaly valuable as well as acceptable to the local population. Remember that foods that are not consumed have no nutritional value ! Average rations must be cumulated to provide at least 6.3 MJ (1500 K cal.) / person / day for a few weeks and 7.5 MJ (1800 K cal.) / person / day for longer periods. Organisation and planning (ration cards, distribution schedule) are the keys to the sucess or failure of a relief program.
The type of food distribution employed will depend entirely upon local circumstances. A refugee camp, where individuals have cooking facilities may be adequately served by the distribution of dry rations alone. Where a large rural population is affected, a special arrengement may be needed e. g. some people with full rations, some with partial rations and selected groups with supplementary rations. Wherever possible, assist people at their homes and avoid setting up refugee camps. Distributing food to nomadic groups is difficult, and no easy way of doing so has been found. Points at which people congregate (e. g., water sources) may be selected as the best places where distributing food.
Basic considerations of selecting foods : the food must:
- correspond to the nutritional needs and food habits of the
- fulfill special logistic requirements, i. e. be easy to transport, store, and distribute, and
- be available in sufficient quantities.
Calculating dry rations, to be done on a family rather than an individual basis, since in this way the number of people attending distributions will be reduced and administration simplified.
Organising a distribution : the key to run a successful food distribution program is to be well organised. The participation of the community in the relief program and in decision-making will help towards an orderly distribution. However, responsible post (store keeping, administration) must be given to reliable individuals outside the community to rule out personal bias, preferences, or vulnerability to pressure.
Mass feeding (cooked meals) is usually limited to institutions and refugee camps. Choice of food, calculating food rations and organisation of cooking facilities (kitchens, personnel and equipment, fires and fuel, hygien and food storage).
Major deficiency diseases in emergencies
Protein-energy malnutrition (PEM) is the most important health problem during a nutritional emergency. Severe PEM can present several forms : nutritional marasmus is characterized by a severe wasting away of fat and muscle ("skin and bone"). It is the commonest form in most nutritional emergencies. Kwashiorkor is characterized by oedema, usually starting at the lower extremities. Marasmic kwashiorkor is a combination of wasting and oedema. Mineral and vitamin deficiencies may also be important. Severe anemia is common and requires a daily intake of iron for an extended period of time. Vitamin A deficiency, the most important vitamin deficiency, is characterized by night blindness and / or eye lesions which may lead to permanent total blindness. The severe forms are usually associated with PEM. Other deficiency conditions are less common ; beri-beri, pellagra, scurvy, rickets.
Mineral and vitamin deficiencies must be identified and the individuals affected or at risk treated by administration of the missing nutrient.
Protein-energy malnutrition (PEM)
It is a problem in main developing countries even in normal times. In times of nutritional emergency it is primarily the most acute forms of PEM that have to be dealt with. These are characterized by a rapid loss of weight and may be evident in a much wider range of age groups than usual. For example, significant numbers of older children, adolescents, and adults may also be affected. Past experience has shown that many emergencies affect the supply of food to only a proportion of the population concerned. It is often the case that only a small proportion of the total population presents clinical signs of severe PEM. For each case of severe clinical PEM there may well be ten moderate cases and hundred children of "near normal" nutritional status. Progession from moderate to clinically severe forms is rapid.
Nutritional results from prolonged starvation : the main sign is a severe wasting away of fat and muscle. It is the most frequent form of PEM in cases of severe food shortage.
Kwashiorkor : the main sign is oedema, usually starting at the lower extremities and extending in more advanced cases, to the arms and face.
Vitamin A deficiency and xerophtalmia : vitamin A deficiency is the leading cause of permanent blindness in pre-school children. It is almost always associated with some degree of PEM. Xerophtalmia is the term used to describe the eye signs caused by vitamin A deficiency which is most likely to be a problem in areas where the diet of the very poor, even in normal times, do not meet requirements. Since vitamin A is stored in the leaver, a sudden deterioration in the diet does not necessarily produce an immediate sharp rise in the incidence of cases, and there may well be a delay of several months until vitamin A deficiency occurs.
Opportunity cost is the sacrifice in real terms suffered by an
economic subject who makes a choice between several possible actions: when this
subject engages in a particular activity, e.g. a production activity (let's say,
a health care programme), the cost borne by him consists in the value of the
opportunities which he passed over, i.e. in the value of the goods and services
which cannot be created elsewhere because the resources employed are no longer
available for other uses (...).The principle of opportunity costs highlights the
profound significance of a firm's production costs [or of a service, such as a
health care programme].They represent the value of the production factors
[labour, capital, etc.] in the best possible other forms of use to which these
factors could be assigned". Raymond Barre, Economie politique, volume 1, P.U.F.,
1978, pp 15 and