10. Surveillance systems

The preceding chapters have been concerned with the assessment team's retrospective survey of a flood's impact on health. It is quite likely, however, that the overall effect may take some time to become completely evident. Therefore a quick and simple monitoring mechanism should be in place which gives health officials an early warning of negative changes in the health status of the community. The changes detected should then be investigated and verified to design corresponding control measures.

By keeping a close watch on the community, the monitoring or surveillance system simultaneously allows health officers to deploy personnel and supplies according to objective need and to formulate logical relief requests.

Epidemiologic surveillance systems for communicable diseases are the most widely known and used in normal times to monitor the health status of a community. Yet in disaster conditions, noncommunicable diseases may also have to be monitored. If the initial survey indicates that risk factors have changed for the worse or are about to, the usually protracted duration of floods calls for setting up additional monitoring systems as a special precaution. These systems follow.

Nutritional surveillance In floods perhaps more than in any other disaster with the exception of droughts, crops are likely to be damaged and long-term nutritional deficiencies may result.

Water-quality monitoring: Experience has shown that gastrointestinal disorders are a concern related to contaminated water supplies.

Entomologic surveillance: Floods tend to alter ecological habitats in a way that may be favorable to vector breeding and increased human exposure.

One of the responsibilities of the assessment team is to determine whether a surveillance system exists and, if so, whether it adequately monitors the flood-created conditions. If not, the team should identify those areas in need of surveillance and recommend guidelines for establishing simple and rapid monitoring and reporting systems.

These systems should target the population at greatest risk, as defined in Chapter 1. They will be most effective if the team tries to integrate rather than bypass existing reporting systems, no matter how rudimentary these may be.

The basic guidelines for each of these monitoring systems are outlined below.

EPIDEMIOLOGIC SURVEILLANCE

Although no serious outbreaks have occurred after recent disasters, the potential (as mentioned in Chapter 3) is there. To avoid an out-of-control situation, the usual monthly reporting system discussed in Chapter 3 is not adequate-the reporting is too infrequent and takes too long to reach the decision-making levels; precious time is then lost in instituting control measures.

A simple reporting system should be established which is transmitted on a daily or, at most, on a weekly basis. It should not depend on physician diagnoses but symptoms or signs that can be detected by health staff with the most basic training (or even by nonhealth personnel).

A questionnaire for this purpose should be designed or adopted (Figure 10.1) (Western, 1982) based on symptoms of the minimum number of conditions likely to occur. The assessment team should make copies available at health centers in high risk areas so that the reporting system is enhanced and the staff is alerted if an outbreak is probable.

The conditions to consider are those endemic to the region, those seen by the assessment team to be on the rise, and those that experience has shown to be common during and after floods.

This system will unquestionably overestimate the prevalence of certain conditions (this was demonstrated in Bolivia during the 1982-83 floods, when a simplified, symptom-based system was compared with the "institutionalized" diagnosis-based reporting system. However, it should be kept in mind that such a mechanism is for early warning, and the information is to be confirmed before control measures are taken. In a disaster situation, it is better to be overly cautious than unprepared.

The team may want to include noncommunicable conditions that are not normally reported in routine epidemiologic surveillance programs. During the "El Niño" floods of 1982-83, for example, Bolivia and Ecuador decided to include animal bites and skin lesions in their system.

Sources of Information

Obviously, the health centers should report data for the simplified surveillance system. The main value of the system, however, is to collect information from sources closer to the community, such as the health promoters or community health workers, and in certain circumstances school teachers and staff at the evacuation camps.

Frequently the Red Cross and other voluntary, religious, and social agencies organize health dispensaries that may be manned by paramedics and "first-aiders" but are often not supervised by medical personnel. These may be the only places for miles around in which some sort of health care is provided and thus become an important source of information.

FIGURE 10.1. Daily report of disease surveillance.

Easy-to-recognize symptoms or diseases

Health installation or aid group:...

Date:...

	Cases		Deaths
	Younger than 15 years	Older than 15 years	Younger than 15 years	Older than 15 years
Fever (without diarrhea/cough)a
Fever with diarrheab
Fever with coughc
Measles
Meningitis
Dogbite
Snakebite
Burns
Trauma
Energy-protein malnutrition
Other
Daily total

Comments:_________

a Indicative of malaria, dengue.
b This can be subdivided according to blood, mucus, vomiting.
c Indicative of respiratory infection.

Source: PAHO, Emergency Health Management after Natural Disaster, Sci Pub No. 407, Pan American Health Organization, Washington, D.C., 1981, p. 26.

Data Processing and Reporting

Each reporting unit processes its own data; by calculating simple percentages of cases over attendances, it compares the proportion of certain conditions on different days. This information should be passed on to the nearest health unit on a daily basis, if possible, or at least weekly.

The health unit-be it a dispensary, health center, or hospital-should compile the data from the different reporting sources within its geographic area of influence. It then processes, analyzes, and interprets the information and relays the results to the next level health unit. The "primary" or "basic" health unit normally will be the first to investigate an increase in the number of cases of a given symptom or symptom complex and, once they are verified, will attempt to diagnose the clinical entity causing the symptom or symptoms. If the outbreak is confirmed, this unit will take the first steps to control it, and will inform the next administrative level of the outcome. The primary unit may request assistance in the diagnosis or control phase.

The interpretation of the processed data and the control measures taken at each level are the core of the simplified epidemiologic surveillance system in disasters. Without this process of investigation, verification, diagnosis, and control, the system is worthless.

Each health unit should be prepared in advance to carry out the process for two reasons: (a) the unit may be isolated and unable to request assistance, and (b) it may be critical to take control measures as soon as possible. Transmitting the information may be difficult: normal means of communications may be broken down as a result of the disaster and not all communities have a police post with a radio or "ham" operator. At times it may only be possible to send data with a returning rescue or relief party.

Sources of Error

As has been mentioned earlier, a system based on symptoms tends to overestimate the potential for outbreaks of certain diseases. The symptoms may be vague and ill-defined for a given condition. In general, this system is sensitive but not very specific.

A common source of error is a reporting system that mixes diagnoses and symptoms' or that has nonmedical personnel making diagnoses. All too often numerous cases of "malaria" or "typhoid fever" are reported by first-aiders with tremendous good will but very little training.

NUTRITIONAL SURVEILLANCE

Monitoring the nutritional status of the community allows the authorities to decide whether to establish, continue, or discontinue a food aid or supplementary feeding program for vulnerable groups. It also pinpoints those communities most in need of nutritional assistance.

Young children are particularly sensitive to the effects of food shortages and are the first to show signs of malnutrition. For that reason, the surveillance system should concentrate on them.

Sources of Information

A. Health centers and dispensaries:

The weight and sometimes the height of children attending health centers are usually measured and recorded. Unfortunately, in too many instances, the data are never analyzed or compared with a standard of reference or used for an intervention program. Such information is key to establishing a simplified system of nutritional surveillance.

Most countries have adopted a standard of reference for weight-for-age or weight-for-height and established criteria for degrees of nutritional status. These are based on the Gomez classification that is common throughout Latin America, on the WHO classification, or on some other local standard. Whatever standard and classification is used in the area should be the bands for the surveillance system. A simple form of recording data is shown in Figure 10.2.

FIGURE 10.2 Nutritional Surveillance.

Health center________

Date________

Parameter (weight-for-height, weight-for-age, etc.)

Classification (Gomez, WHO, national, etc.)

Age group	Normal	Deficient	Total
Under 1 yr.	/ / / / / / / / / /	/ / / / / / (6=37.5%)	16
1 to 4 yrs.	/ / / / / / / / / / / / / / /	/ / / / / (5-25X)	20
Total	25	11= 30.5%	36

The form illustrated in Figure 10.2 can be filled out every day, every clinic session, or every month, depending on the circumstances.

B. Evacuation centers, schools, other shelters:

Evacuation centers and other shelters can use a system similar to that suggested for health centers. In schools or evacuation centers, for instance, children can be weighed every week or every month and the community's status can be monitored according to the percentage of cases below the normal level ("deficients"). It may be more practical, however, to use arm circumference measurements, as will be seen below.

Information can also come from those attending food distribution centers or emergency kitchens.

Indicator To Be Utilized

Whenever possible, the surveillance system should be based on a functioning infrastructure. If health centers measure weight, the system should be based on weight-for-age; if both weight and height are measured, the parameter should be weight-for-height.

A weight-for-height figure is better than one of weight-for-age because the former differentiates acute undernutrition, which may be due to the flood's impact, from chronic undernutrition, which has nothing to do with the disaster (although aggravated by it). Furthermore, weight-for-height is independent of age. However, measuring height properly requires calibrated boards, and these are not as widely available as balance scales.

When properly done, measuring arm circumference has an advantage in that only a tape measure is needed; it is cheap and can be carried in a pocket. For a quick survey or when the system is based on data-collecting posts other than health units, this is the simplest of the three methods to use but it also is the least sensitive. The measurement obtained can be compared with the reference values suggested in Table 7.1.

Reporting the Information

The assessment team should proceed in the same manner as for epidemiologic surveillance. However, daily or weekly reporting is not necessary since no "epidemics" of undernutrition occur over short intervals, as may happen with communicable diseases; monthly reports are adequate.

Unless the increase in undernutrition is from an outbreak of gastroenteritis (which is usually accompanied by loss of weight), there are not many corrective measures that the primary health worker can take. Undernutrition is frequently a socioeconomic problem; in disasters it often stems from the logistics of food supplies and distribution, in which many different agencies may be involved and over which the health worker has no control. The main responsibility of the health worker is to inform responsible authorities (forcefully, if necessary) of any deleterious changes in the nutritional status of his or her community.

Sources of Error

The most common source of error is that of technique. Frequently the scale is bad or the measuring technique is poor. Length/height may be particularly difficult to measure. When there is no length board, length is often measured from crown-to-heel and, if not stated as such, will be a source of gross mistakes in the interpretation of the results.

WATER-QUALITY MONITORING SYSTEM

After a disaster, one of the most critical needs is for an adequate supply of water (PAHO, 1982b). A system to monitor and control water quality should be established as soon as possible.

In floods, critical aspects in water safety are the amounts of residual chlorine and the detection of Escherichia coli in water samples sent to the laboratory. In a disaster situation, it may be more practical to analyze samples locally than to rely on laboratories. This can be done by using simple diagnostic kits based on colorimetry (in the case of chlorine) and membrane filters for bacteria (Assar, 1971; Vargas de Mayo, 1984) (Fig. 10.3).

FIGURE 10.3 Field kit for bacteriological analysis of water supply.

Sources of Information

It is essential to establish the points from which samples are to be taken and the frequency with which this should be done. Usually the samples are taken from the water source and at different points along the supply network. If water is distributed in water trucks or similar container vehicles, samples should be taken at the point where the truck gets its supply and toward the end of the run.

It may be difficult to set up a system based on sampling different points if transport and communications are difficult or impossible to arrange. The team may have to limit itself to the area that can be covered on foot, which may or may not include the source of water supply.

Reporting the Information

The local health authorities should identify the source of contamination, and in the event that residual chlorine is absent or below recommended standards (Assar, 1971), they should investigate whether the amount used is insufficient or absent. The findings should be reported to the health authorities in charge of emergency rehabilitation measures.

Although the corrective measures are usually beyond the health sector's responsibility, it should nonetheless report its findings to the agencies in charge of the water supply and pressure them to repair and treat the system accordingly.

ENTOMOLOGICAL SURVEILLANCE

Chapter 6 reviews at some length the assessment of the flood's impact on vector populations, particularly mosquitoes, and mentions the need for an entomological surveillance system. The full effect of the floods on vector-borne disease transmission will not become evident until some months after the flood's onset or in fact months after the waters begin to subside. Waiting for an alert from epidemiologic surveillance may take too long - monitoring breeding sites and mosquito density is essential.

Sources of Information

There must be permanent points for adult mosquito captures and monitoring of larva breeding sites. There should also be surveillance of meteorological activity in the area, particularly rainfall, to determine its influence on the vector population. Migration of people and animals should also be checked and mapped.

The role of the assessment team is to verify whether an entomologic surveillance system exists and how is it functioning, the type of reporting, its frequency, and the response by the relevant agencies. In the absence of a system, the team should set up the methodology for one.

CHECKLIST

Basic Questions

· What diseases may become epidemic as a result of the disaster?

· To whom and where should food aid be directed?

· When should food aid be started or discontinued?

· Is there a vicious cycle of "gastroenteritis-malnutrition"?

· Is drinking water contaminated?

- Where and why?

· Is there an increase in the vector population?

- Which ones?
- Where?

Epidemiological Surveillance

· List of signs and symptoms

- Criteria for selection

* endemicity
* result of initial evaluation
* previous experience

· Source of information

- Health centers and posts
- Health promoters/community health workers Schools
- Evacuation camps
- Health posts of voluntary agencies

· Data processing

- Simple percentages
- Processing at each level
- Analysis and comparison of the results, daily or weekly

· Flow of the information

- Sources of the data
- Flow to nearest health unit

* collection and compilation
* analysis
* interpretation
* investigation/verification/diagnosis
* corrective measures

- Flow to health unit or higher level for:

* information
* request for assistance

· Communication system

- Regular

* telephone
* radio
* road, railroad, etc.

- Irregular

* rescue groups
* boat, mules, etc.

· Frequency

- Daily
- Weekly

Nutritional Surveillance

· Sources of data

- Health centers and posts
- Health promoters/community health workers
- Schools
- Evacuation camps
- Health posts of voluntary organizations
- Food distribution centers
- Emergency kitchens

· Questionnaire to utilize

· Parameter

- Weight
- Height

* length boards

- Arm circumference
- Sources of error

* inadequate equipment
* inadequate technique

· Frequency of reporting

- Monthly

Water-Quality Monitoring

· Test for residual chlorine
· Amount of Escherichia coli
· Laboratory
· Portable kits
· Specify

- Sampling points
- Frequency of sampling
- Responsible person
- Flow of information

· Investigation of contamination source

Entomological Surveillance

· Permanent capture points: frequency of observations
· Permanent larval breeding sites: frequency of observations
· Meteorological data
· Migration of people and/or animals
· Frequency of observations
· Data flow

- System
- Frequency