Document(s) 15 of 19

A survey of participants was taken during the workshop to better determine their interests and needs regarding GIS applications and functions (Table 1). This paper serves to bring forward the salient points from that survey. It is hoped that the results will facilitate the development of better solutions to the problems encountered by the growing community of GIS users in the health field.

Survey responses are presented in Table 2. Responses range from I (very useful) to V (not useful) for questions 1, 2, and 3; and Yes or No for question 4. For each variable, the highest frequency has been printed in bold. Question 5 was open-ended, and will be commented on in the conclusion.

Questions 1 through 4

Question 1 dealt with the usefulness of GIS in different health-related fields. GIS are seen as most useful for monitoring and control, which was given a score of I by 12 respondents, and a score of II by 7 respondents. GIS were also believed to be very useful for research, with 85% of the respondents rating their usefulness as I or II. Responses varied more across health policy: 12 respondents gave it a score of I or II, while 7 gave it a score of III. Finally, the usefulness of GIS for health education was not seen to be as great by many of the participants: only eight persons gave them a score of I or II.

Question 2 looked at the geographic scale of GIS applications, ranging from the local to the national level. The level at which GIS were seen to be the most useful was the national. However, if scores of I and II are considered together, all levels are considered relatively equally: 14 respondants chose the national level as that at which GIS would be most useful, 14 also chose the district / large urban settings level, and 15 chose the regional/provincial level.

Question 3 was more technical and dealt with GIS operations. This question was characterised by a great homogeneity of responses. Not a single GIS operation, among the 28 listed in the question, was rated poorly. At the same time, however, few GIS operations were rated as "very important". Rating III was the most common answer, a result which is examined below. It is interesting to note that satellite images input received the highest number of very important scores (6).

Question 4 considered the human resources required to operate a GIS, a complex technology which requires skilled personnel to realize its full potential. Most participants indicated that they have skilled personnel for all the tasks listed, except spatial modeling, but there were also a significant number of persons who do not have the required personnel.

Project planning and management is the least problematic, with only 4 persons not having skilled personnel. Database management is more problematic: 6 respondents face a lack of personnel in attribute database management, while 7 respondents have personnel problems in spatial database management. Mapping brings more difficulties: almost half of the respondents do not have personnel skilled in this area. The most important problem is with spatial modeling: nearly all respondents (85%) lack skilled personnel in this area.

Question 4.2 was prepared for the respondents who answered "No" in the preceding question. However, the data shows that not only these people responded: the number of responses in question 4.2 exceeds the number of persons who answered "No" in question 4.1.

This makes the data rather ambiguous and difficult to interpret, but it nevertheless permits an important observation: many persons who do not have skilled personnel for some GIS tasks in their immediate environment can find them elsewhere within their institution. This is only part of a solution, however, as there remain clear problems at the institutional level with attribute data management (one case), spatial data management (four cases), mapping (six cases), and spatial modeling (15 cases). Most respondants indicated that if they could not find skilled GIS personnel within their own institution, they could them elsewhere in the country, except for persons skilled in spatial modeling, which still causes problems in four cases.

Discussion

There have been attempts to identify a few key patterns in the survey data presented in this text, using methods such as nonlinear principal components analysis and cluster analysis. The hypothesis was that different kind of GIS users (for instance users from universities versus users from ministries, or research-oriented users versus application-oriented users) would have different profiles. These attempts to pattern the responses were unsuccessful, however, because the variability of the data is very small. In other words, the population of GIS users investigated seems to be very homogeneous; in general, they share the same vision, interests, and difficulties. A few points, such as the use of GIS for health education, or their application at the community or rural level, show a wider variation of opinions. This general homogeneity, as well as the divergence on these specific points, are worth of consideration.

A second issue which should be noted is that the "application questions" (questions 1 and 2) clearly received better scores than the more technical items of question 3. This probably means that current users of GIS in the health field are more concerned with the application side of the technology and less concerned by its more technical aspects.

These are not seen as not important (columns IV and V have frequencies of 0 for all GIS functions!), but they clearly have fewer high ratings than application questions. The exceptions should be noted: map algebra ( reasoning , i.e., following a logical or mathematical path with maps), digital elevation modeling (the importance of the physical environment could be stressed here), and map design are seen as specially important among the GIS functions.

The responses concerning the availability of skilled personnel to assist in the operation of a health-oriented GIS indicate that many participants have problems at this level: the answers indicate that 42% of personnel needs remain unmet. It should be stressed that the GIS operations for which the unmet needs are the greatest (spatial modeling, mapping, and spatial database management) are precisely at the heart of a GIS, and are probably its most characteristic functions.

The survey data presented indicate that to exploit to the full potential of GIS, health specialists need a system that offers a wide variety of "user friendly" operations. Specifically, a system is needed that offers the required tools but that allows the user to concentrate on application problems. More could probably be done to devise such a system, perhaps by adapting existing packages. The data concerning the availability of skilled personnel leads to the same conclusion: an effort should be made to better adapt the spatial components of the GIS to the health field.

Conclusions

Increasingly more people believe that enormous progress in public health can be made by better understanding the links between health and the human and physical environment. GIS are certainly among the most useful tools that make possible the analysis and representation of these complex relations. Users of GIS in the health community still appear as pioneers; their experience shows that the use of GIS is not without difficulties and that the tool could probably be better adapted to their situation.

The survey presented here indicates that if the tool is to be used to its full potential, efforts should focus on several issues.

• Developing a software system that offers a range of GIS functions in a more user-friendly way to people who are not "hard GIS specialists." The ideal Health GIS program should let the user concentrate on the substance of the problem rather than on the tool.

• Finding a solution to the scarcity of skilled personnel. Development of a system as described in the preceding paragraph should help alleviate this problem. The preparation of a "health-oriented tutorial package" to accompany such a system, presenting examples of health/environment data integration, analysis, and presentation could also help solve this problem.

• Increasing accessibility of the spatial data. There is a wealth of remotely sensed environmental data (climate, relief, vegetation, land use, transport networks, human habitat, to name a few), but where is that material? How is it obtained? How much does it cost? This seems to be far from clear for health specialists. Some kind of information centre could be imagined to facilitate the access of health specialists to this data.

Table 1. GIS for health — quick workshop survey.

To assist identification of directions for follow-up of this workshop, would you kindly answer the following questions, based on your former GIS working experience and what has emerged from the workshop.

Please answer questions 1 and 2 using a number between I (very useful) and V (not useful).

1. What is your opinion about the usefulness (I-V) of applying GIS

a. in health research b. in health monitoring and control c. in health policy d. in health education e. other (specify)

___ ___ ___ ___ ___

2. What degree of usefulness (I-V) would a health GIS have when applied at the following levels

a. Small community/rural setting b. District/large urban setting c. Region/province d. Nation

___ ___ ___ ___

3. Taking into account your work setting and conditions, and your specific needs and problems, please indicate on page 2 the relative importance of the listed GIS operations. If you think that some operations should be added, please do so on a blank line.

4. The main tasks for operating a GIS are listed below (for details of these tasks please refer to table on page 2). Adequately trained personnel are necessary to realize these tasks.

4.1 Do you have personnel (Y or N) with appropriate skills for:

a. Spatial database management? b. Attribute database management? c. Spatial modeling? d. Mapping? e. Project planning and management?

___ ___ ___ ___ ___

4.2 If no, do you have access to people with these skills elsewhere in your institution or in your country?

	Country	Institution
a. Spatial database management b. Attribute database management c. Spatial modeling d. Mapping e. Project planning and management	___ ___ ___ ___ ___	___ ___ ___ ___ ___

5. It has been suggested that a network of people interested in the health field be created following this workshop. Could you express your idea about how this network could be of mutual benefit? Do you have suggestions about how it should function, what it should offer?

Please indicate your

Document(s) 15 of 19