4.2. Analysis of the Health Information System

Uwe Wahser: Construction of an Adapted Health Information System
Previous File	Overview	Next File
	Detailed Table of Contents

4.2. Analysis of the Health Information System

The analysis of the health information system in the Kabarole district was carried out during the first eight weeks of the second phase of the systems engineering project. The activities which were carried out during the systems analysis had to be arranged according to the respective time scheduling of the involved parties. The results were summarized in an intermediate report (compare annex R "Report: 'The Kabarole District Health System - Description, Opportunities and Suggestions for Improvement'", chapter 1 "Description of the Kabarole District Health Information System"). The following sections list the single activities which were carried out in the course of the analysis. Experiences with and alternations of the methodology from section 3.2. "Analysis of the Health Information System" are described along with the respective activities. The following sections correspond to the elements of the proposed methodology in section 3.2.4. "Assessment of the Objects of Interest".

4.2.1. Document Analysis

The analysis of documents during the first phase concentrated on a project progress report of the basic health project of GTZ ([OKWERO et al. (1992)]), copies of some publications on the district health system and a collection of forms, which are used in the district. These documents were posted to the author of this thesis by the district health team and gave already a first impression of the district health system and the main reporting lines. But these documents were also the only source of documented information on the district health system which was available prior to the deployment. Unfortunately the assumption from section 3.2.4. "Assessment of the Objects of Interest", that the German headquarters of GTZ could make available information on the district health system, turned out to be wrong. It was not possible to receive any documents on the basic health project. During the second phase of the systems engineering project, which took place in the district, the collection of forms could be supplemented and updated. A list of all collected forms is given in table 6 of annex R "Report: 'The Kabarole District Health System - Description, Opportunities and Suggestions for Improvement'". In the second phase it was also possible to inspect forms, which were already used for reporting. A random selection of the central summary report form from the health units was carried out (compare MF77 in table 6 of annex R). In the same way a random collection of supervision reports was inspected (compare Sup01 and Sup02 in table 6 of annex R). Documentations on the forms and how these have to be filled out were not existent in the district. For the analysis of the business level, the so called district operational manual was helpful. It contained an organigraph, the description of aims and objectives of health programs and services, a collection of forms and miscellaneous pieces of information on the district health system. However, the manual proved to be not well sorted. The information of the different services ranged from multi-page documents to single-line, handwritten notes per service. Especially the respectively responsible staff members were not specified. The collection of forms was partially outdated, not documented and far from being complete. A comprehensive report on the district health system was in preparation during the time of this analysis. From this report a collection of maps, which depicted the providing health units for each service, could already be used for this analysis. Population statistics were available from the governmental data center in Kampala ([STATISTICS DEPARTMENT (1981)] and [STATISTICS DEPARTMENT (1992)]). Country information on Uganda from the German Statistics Department ([STATISTISCHES BUNDESAMT (1991)]) provided supplementary information for the considerations on the target population in section 1.2.2. "Target Population" of annex R. The ethnological background information could be acquired from [NZITA and MBAGA-NIWAMPA (1993)].

4.2.2. Unstructured Interviews with Representatives of the District Health Team

As a lucky coincidence, a seminar on quality assurance was carried out for the members of the district health team during the first week of the deployment in Fort Portal. By attending the seminar, the author of this thesis was introduced to the district health team. It was possible to see most of the district health team members at one time. This facilitated the identification of key informants for the unstructured interviews. Table 4.2-1 lists the unstructured interviews, which were conducted with staff members in exposed positions. Main topics were the general setup of the district health system and the priorities of the deployment in Fort Portal.

Tab. 4.2-1: Unstructured interviews were conducted with members of the district health team in higher positions in order to gain an overview on the district health system.

Date	Duration	Interviewee	Function
10.6.	1 h	Mr. Geoffrey Kabagambe	District medical officer
10.6.	2 h	Mr. Peter Okwero	Coordinator of quality assurance activities
16.6.	˝ h	Mr. Tom Rubaale	Project Manager
27.6.	1 h	Mrs. Beatrice Ssempebwa Mrs. Hamet Kimara Mr. Silver Kasoro-Atwooki Mr. Fred Ntegyereize	Health unit supervisors with additional functions in the district health team
29.6.	1˝ h	Mr. Walter Kipp	Project Consultant

Besides these official interviews, which had to be done with fixed appointments, it was also possible to have many informal conversations with other staff members during the daily activities in the district health team offices. These are mentioned in section 4.2.6. "Participatory Observation of the District Health Team".

4.2.3. Unstructured Interviews with Representatives of External Parties

The identification of possible informants from external parties was usually realized in a completely open procedure. It was done by entering offices of institutions which might be of relevance and inquiring for persons which could provide the needed information. If the institution turned out to be not competent in the matter, possible alternatives were inquired. This method was easily feasible in Fort Portal, but proved to be very strenuous in Kampala. Table 4.2-2 lists the interviews with external parties which were conducted in Fort Portal:

Tab. 4.2-2: The interviews which were conducted in Fort Portal concerned population dynamics.

Date	Duration	Site/Interviewee	Topics
5.7.	˝ h	Fort Portal Municipality - Mr. Richard Monday, town clerk	Migrations into and from the district capital.
5.7.	˝ h	Kabarole District Administration, Fort Portal - Mr. Elias Byamungo	Migrations into and from the district.

The interviews in Fort Portal were conducted to gain statistical data amd information on migrations. The impact of refugee movements seemed to be of importance in the district, also. The interviews did not provide the wanted information, though. One reason was the absence of up to date summary data at the district administration. Another reason, especially concerning refugee movement, was the feeling of the interviewees that this information might be confidential and should be provided from the national level. References to institutions on the national level were given. The opportunity for a visit at national authorities arose in the seventh week of deployment. It was possible to join district health team members on a five day trip to Kampala. While two days accounted for the journey, the interviews from tables 4.2-3 to 4.2-4 could be conducted in the remaining three days. The first set of interviews concentrated around the national health information system. These were done at the health planning unit of the ministry of health in Entebbe. The need for an additional interview at the UNICEF offices arose because UNICEF had asked the district health team to implement a management information system. This system was designed by UNICEF as a self assessment tool for health units with a software application product for an assessment at the district level. Table 4.2-3 lists the conducted interviews which concerned information systems:

Tab. 4.2-3: Interviews concerning information systems were conducted at the ministry of health and the UNICEF offices.

Date	Duration	Site/Interviewee	Topics
13.7.	7 h	Health Planning Unit of the Ministry of Health, Entebbe - Mr. Dirk van Damme, WHO consultant - Mr. Vincent Ndazima, health planner - Mr. Sumuju, statistician - Data entry clerks	- Existing national health information system. - Further processing of district data. - Newly planned national health information system. - Software support of both.
14.7.	1 h	UNICEF offices, Kampala - Mrs. Jessica Kafako	Offered management information system of UNICEF.

The interviews at the ministry of health turned out to be very fruitful. Information on the existing health information system and a supply of missing forms could be obtained. A newly planned health information system was demonstrated and requirements of a software application product at district level were discussed. Also the further processing of district data could be inspected. The interview at UNICEF was not satisfying. Neither a competent representative for information on the management information system nor a full demonstration installation of the software application product were available. The last set of interviews concerned population statistics of Kabarole district. Table 4.2-4 lists the interviews which could be conducted:

Tab. 4.2-4: The last set of unstructured interviews concerned population statistics.

Date	Duration	Site/Interviewee	Topics
14.7.	˝ h	Data Center, Ruth Tower, Kampala - Mr. Mukulu	National household and population censi of 1980 and 1991.
15.7.	˝ h	Department of Community Development/Refugee Department, Udyam House, Kampala - Mr. Twesigomwe	Refugee movement in Kabarole District.
15.7.	˝ h	Population Secretariat, Impala House, Kampala - Mr. Kabera	Migrations into and from Kabarole District.

Especially for the last two interviews a letter of recommendation from the district medical officer was very helpful. The given information was up to date background information which was not documented elsewhere.

4.2.4. Semi Structured Interviews with Coordinators of Health Services

After the services and programs of the district health team and their coordinators were identified, semi-structured interviews were conducted. The selection of coordinators was done mainly because of the representation of their program in the selection of forms. Especially the family planning program and the mother and child health care services accounted for many forms. The selection of the other services and programs was mainly limited by the possibility to fix appointments with the involved co-ordinators. Table 4.2-5 lists the semi-structured interviews which were conducted:

Tab. 4.2-5: The variety of conducted programs and provided services resulted in a variety of interviews with co-ordinators of these activities.

Date	Duration	Interviewee	Coordinated Service/Program
22.6.	2˝ h	Mrs. Margaret Okwero	Family Planning Program
27.6.	1˝ h	Mr. Samuel Bamuhiiga	Onchocerciasis Control Program
4.7.	1 h	Mr. Steven Kaliba	Local Drama Groups
4.7.	1 h	Mr. Hosea Mpuga	Health Education Program
5.7.	ž h	Mrs. Beatrice Ssempebwa	Mother and Child Health Care
6.7.	1˝ h	Mrs. Gudrun Sahlmüller	AIDS Control Program/Control of Sexually Transmitted Diseases
8.7.	1 h	Mr. Albert Kilian	Malaria Control Program

4.2.5. Assessment at the Health Unit Level

The assessment of the implementation of the health information system at the health unit level was constrained by the availability of transport. The time scheduling of visits at health units was therefore strictly dependent from the logistics of the car pool. Because of the unknown geography a driver had to be available as well. Therefore the methodology from section 3.2.4. "Assessment of the Objects of Interest" had to be modified. The selection of health units was not done to be representative. It was done according to the availability of transport. Despite the immense car pool of 14 landcruisers, the planned observations had to be restricted to the fifth week of the deployment, when there were supposed to be no supervisory activities. On the other hand there were some good opportunities to participate in inspections of health units by experienced members of the district health team. The following description of activities during health unit assessment is therefore structured according to the degree of self determination of the author of this thesis. Tables 4.2-6 to 4.2-8 list the health units which were visited. Since the time for travelling also had a great impact on time scheduling, the respective specifications of the duration of each visit were supplemented with the total time spent on visiting health units on that day, including transport. During the first three weeks there were some opportunities to accompany members of the district health team on inspections of health units. These visits gave the opportunity for unstructured observation at the health units and discussions of background information with the respective health team members. Table 4.2-6 lists these visits:

Tab. 4.2-6: Unstructured observations were conducted when health units were visited as part of an inspection by district health team members. The total duration of the activities including the time for travelling is given in brackets.

Date	Duration	Health Unit	Remarks
2.6.	˝, ˝, (7) h	Hapuyo, Kasule	Inspection of newly constructed buildings by Mr. Walter Kipp and Mr. Fred Ntegyereize
7.6.	1, 1, (4) h	Two selected health units in Bunyangabu County	Field visit with ten health team members as part of the quality assurance workshop. Special emphasis on health information.
13.6.	1, 1, (6) h	Kakuka, Nyahuka	Bundibugyo District:
14.6.	(8) h, ~ ž h each	Butuma, Kasulenga, Mirambi, Kyengi, Bubukwanga, Ntandi	Part of the initial assessment of the district health services by Mrs. Andrea Knigge, Mr. Henning Mohr and the respective supervisors from the Bundibugyo district health team.
15.6.	(7) h, ~ 1 h each	Rwebisengo, Ntoroko, Kanigutu

The unstructured observations during inspections were helpful for getting a general idea of how health services were provided at the health unit level. Especially a series of visits at health units during the third week of the deployment was very informative. Although the visited health units were not in Kabarole district, but in the neighboring Bundibugyo district, the general setup of the health units was reported to be the same as in Kabarole. The inspections at the respective health units were conducted as a general initial assessment for further activities of the GTZ project in Bundibugyo district. The broad based nature of the inspection gave a good insight into the work of the health units. A first impression of the main components of the health information system could be obtained as well. The next opportunity for launching a series of health unit visits arose during the fourth week of the deployment. Because of the limited travelling opportunities, the selection of health units was reduced to those with an almost complete set of provided services. Health units with a weak reporting were left out since the problems of these health units were reported to be part of a general weakness due to miscellaneous reasons. The visits during that week were attached to routine supervisions. Therefore it was possible to conduct only semi-structured interviews on these visits. Table 4.2-7 lists the visits with only semi-structured interviews:

Tab. 4.2-7: Semi-structured interviews were conducted when the visit was attached to a regular supervision of the health unit. The total duration of the activities including the time for travelling is given in brackets.

Date	Duration	Health Unit	Accompanied Supervisor
20.6.	1, ˝, (4) h	Butiiti, Mbale	Mrs. Hamet Kimara
24.6.	1˝, (3) h	Rutete	Mr. Silver Kasoro-Atwooki

The semi-structured interviews without observation of the data collection were informative only to a certain degree. For one reason the time needed for discussing all available forms was underestimated, so that the interviewees were conducted too hasty. For another reason the presence of the supervisors as official representatives from the district level put the interviewees in the position of being tested on their knowledge rather than giving information on the health information system. Also the official introduction of the author of this thesis by the supervisors was, although being polite, a little hindrance in setting up a casual atmosphere. The fifth week brought about the opportunity for two complete observations. By this week the regular health unit supervisions of the month were almost completed, leaving transport capacities for the author of this thesis. Transport for two days could be obtained. Table 4.2-8 lists the visits in that series:

Tab. 4.2-8: Two observations on data collection at health units were conducted. The total duration of the activities including the time for travelling is given in brackets.

Date	Duration	Health Unit	Observed Services
30.6.	3, (7) h	Nyabanni	Outpatient department, drug distribution, ante natal care
1.7.	4, (6˝) h	Kyarisozi	Outpatient department, drug distribution, family planning consultation

Resulting from the experience of the semi-structured interviews from table 4.2-7, the author of this thesis introduced himself to the incharge of the respective health unit without showing the letter of recommendation to avoid unnecessary formalism. The fact was emphasized, that the author was a student without clinical knowledge, who wanted to learn about the health information system. The wish was expressed to watch data collection as it is done practically, to see where disadvantages of the forms are. By making clear that the author wanted to learn and not to teach or criticize, a friendly atmosphere could be created, in which data collection was explained freely. It was possible to watch services in the outpatient department and other areas. Doubts of the author, whether this procedure could invade the privacy of the clients were put aside by the health unit staff. Spare times between consultations were used to conduct parts of the semi-structured interview. The limitations of this procedure are obvious: calculating an average consultation time of about 15 to 20 minutes per client, only three sessions can be observed per hour. Therefore the number of observed consultations can only be very limited. Almost each observed consultation brought up a surprising situation with regard to the recording procedure. It is therefore obvious, that a significant greater number of consultations has to be observed in order to obtain a somewhat comprehensive picture of the limitations of the health information system. Since some of the data collection methods were handled differently at each health unit, a greater number of observed health units would have been useful as well.

4.2.6. Participatory Observation of the District Health Team

As expected, the participation in activities of the district health team gave opportunity for many informal discussions with staff involved in the procession of data on the one hand, but also on topics concerning organizational aspects of the district health system. The main activities were centered around the statistics office, but also on all kinds of problems concerning computers in different departments. Already on the first day of deployment computer related problems were reported. Most of these were to be seen in insufficient training of the data entry clerks, which were more or less left alone after the sudden dropout of the statistician. The need for further training could therefore be identified at an very early stage and was encountered as soon as possible. Starting from the sixth week of deployment, computer classes were offered for data entry clerks and other interested staff. These classes also gave a good opportunity to estimate capabilities of the potential users of the software application product, which had to be constructed later on.

Previous File	Overview	Next File
	Detailed Table of Contents

This page was constructed by Uwe Wahser (uwe@wahser.de)
Last Revision: May 1996