Sama&pen7.pmd

Martyn T. Sama and Richard Penn
Introduction
Onchocerciasis is the world’s second leading infectious cause of blindness with an
estimated 123 million people under risk, and about 18 million people in the world
suffering a grave burden imposed by the disease (WHO 1997a). In Africa, some
17.5 million people are infected with Onchocerca volvulus (WHO Technical Report
Series no.852). It is estimated that more than 6 million people are suffering from
Onchocerca skin lesions and severe itching (Remme, Murray, and Lopez 1990).
The vector Onchocerca volvulus produces millions of microfilariae worms which migrate to the skin and the eyes of the human host, causing severe itching andpigmentation. The most severe manifestations of onchocerciasis are irreversible:ocular lesions of both the anterior and posterior segments of the eye, resulting inimpaired vision and ultimately total blindness. In Africa, it is one of the leadingcauses of visual impairment and blindness. Rarely life threatening, but causingchronic suffering and severe disability, onchocerciasis constitutes a serious obsta-cle to socio-economic development (WHO 1996).
Mass-treatment of onchocerciasis is carried out in meso- and hyper-endemic areas once a year, and in hypo-endemic areas, treatment is clinic-based. The targetpopulation is persons of five years and above, with the exception of seriouslysick persons, pregnant women, nursing mothers whose babies are aged beloweight days, and very old persons.
The process of Community-Directed Treatment (ComDT) with Ivermectin (CDTI), has been adopted for onchocerciasis control in some places of Africa.
A multi-country study (WHO 2002) conducted in some of the endemic countrieshas demonstrated that Community-Directed Treatment (ComDT) is an effective Sama and Penn: Another Look at Community-Directed Treatment in Cameroon strategy for drug distribution. Those communities are deeply involved in theirown health care on a large scale. An assessment by the African Programme forOnchocerciasis Control (APOC 2000) showed that ComDT is effective also forother health and development activities like distribution of Vitamin A, Malariacontrol, Guinea worm control, and Sanitation. Despite the success, several questionsremain unanswered about community and health system interaction for sustainedcoverage of Ivermectin distribution (WHO 2002a).
Ivermectin distributors, the Community Directed Distributors (CDD), are supposed to be members of their community chosen by the community througha democratic process, trained to distribute Ivermectin and supervised by healthservices staff. Ivermectin is considered safe enough to be administered by non-health personnel.
Treatment coverage varies between contexts where ComDT has been tried.
The mean reported coverage over different ongoing projects is 70 percent. Prob-lems in the selection of CDDs, inadequate supervision by health staff and limitedcommunity participation in decision making are common obstacles. The follow-ing problems are common to all project sites, these range from poor selection ofCDDs, inadequate supervision by health staff to limited community participa-tion in decision making (APOC Technical Report 2002). The urgency of researchon ComDT is underscored by the fact that advocates of ComDT want to alsouse it for other community-based interventions.
This paper reports on a study of ComDT Ivermectin treatment in Cameroon.
More specifically we give results on coverage and the views of the CDDs, thehealth personnel and the community on the ComDT of Ivermectin. The study isan exploratory single case-embedded design, seeking to understand the factorsinfluencing the effectiveness of community-based approaches to drug distribution,in this case ComDT Ivermectin treatment.
Methods
The study was conducted in the south western part of Cameroon where the
NGO Sight Savers International (SSI) currently has a Community-Directed
Ivermectin Treatment (CDTI) programme. The South West Province lies be-
tween 5º20 and 4º N and 8º45 and 9º45 E. The study area includes the health
districts of Muyuka, Kumba, and Konye. The areas were purposely selected for
this study because they were among the meso- and hyper-endemic communities.
Two of them are hyper- and one meso-endemic, thus meeting the criteria for
mass treatment with Ivermectin. The study area has a very rich network of drain-
age systems, most of which flow from high altitude and are interrupted by nu-
merous cascades, rapids, and waterfalls. These streams provide sites for Simulium
vectors, which can be found at high altitude in the area.
A traditional chief heads each community, and links the communities and the administration. However, the influence of the chieftaincies over the communities varies from one ethnic group to the other. In some communities the people arebetter organised around the administrative authorities.
Several channels of communication are used. Each community has a town crier whose role is to transmit messages in the local languages to the community.
Churches and ‘Njangi houses’ are also commonly used for disseminating infor-mation. Other traditional channels include the talking drum, and flute.
For this study information was obtained from the study population which consisted of different actors involved in the process of ComDT with Ivermectinat the community level. These include the CDDs, the health facility staff, and thecommunity members and their leaders. At the Local Government Area/Districtlevel, the heads of the other health and developmental activities operating withinthe study communities were also studied. The team that conducted the studycomprised of a social epidemiologist, a medical doctor, an anthropologist, acommunicator, a bio-statistician (part-time) and six interviewers.
The interviewers were under-graduate students of the Department of Soci- ology and Anthropology at the University of Buea. Training was carried out fortwo weeks on the techniques of using quantitative and qualitative instruments fordata collection. Confidentiality protection was guaranteed by demanding andreceiving oral reports from all CDDs, households, community leaders and healthworkers to insure data quality before interviews.
Forty communities within which at least two CDDs were found were randomlyselected for data collection. In each community, two CDDs who covered a mini-mum of fifteen households were selected for surveys - hence a total of eightyCDDs, seventy-five male and five female, who met these criteria, were interviewed.
A semi-structured questionnaire was administered to each CDD to collect infor-mation on the different health and development activities in which he/she wasinvolved, how and when they got involved, their motivation, the number of daysspent on each activity, similarities to the other activities in their work as CDDs, andhow their involvement in the other H&D activities affect their work as CDDs.
A total of 1200 households in which the eighty CDDs worked were sur- veyed to estimate for household treatment coverage, using a pre-tested standardhousehold survey form. Fifteen households per CDD were used to collect in-formation on sex, age, treatment coverage, period of treatment, treatment effectand side-effects. Information on actual coverage was obtained from 1185 house-holds with 5812 individuals, consisting of 2919 (50.2 percent) males and 2893(49.8 percent) females.
Qualitative Data CollectionForty focus group discussions (FGDs), one per community were conducted;twenty male FGDs and twenty female FGDs. Each FGDs ranged from eight to Sama and Penn: Another Look at Community-Directed Treatment in Cameroon twelve persons. A FGD guide was used to direct the moderator, note-taker andobserver during the discussions.
Forty in-depth interviews were conducted with each community leader using an in-depth interview guide. In-depth interviews were conducted with all ninepeople (eight nurses and one DMO) from health facility staffs that supervise theCDDs in the study communities. An in-depth interview guide for health staffwas used to guide the interviewers.
EPI-info was used for entering data from the household surveys and CDDsurveys and for the questionnaire’s descriptive analysis. For more detailed analysis,SPSS was used to examine variation and correlations.
Data were transcribed from tapes to records. MSWord was used for entry andtransfer to text files. Textbase-Beta was used for content analysis.
Quantitative ResultsTwo questions in the household survey give estimates of the proportion of per-sons that report they received tablets and those that report they swallowed thetablets. The overall estimates for persons five years and older were 73.5 percentand 72.6 percent (14 percent of the persons in the household survey were underfive). Among those that reported tablets swallowed, 46.3 percent reported sideeffects and 3.5 percent reported having taken any health care action. Table 1shows the detailed estimates by age and sex. Both sexes report lower coverage inthe age groups 15–34 years. Children and older people have higher frequenciesof persons that received and swallowed tablets. Women have generally lowercoverage. The proportion having swallowed tablets was about one percent lowerthan the proportion that received tablets quite systematically. Women more oftenreported side effects than men and also sought health care more. Both theseproportions were larger for the older groups.
There is a large variation between CDD areas and between households within these areas. The coverage estimated from reported numbers in the CDDs ques-tionnaire is considerably lower, less than 50 percent. No correlations were foundbetween the coverage estimated from the household survey and the informationfrom the CDDs. We did not find any correlation between CDD age, sex andeducational level, time spent on CDTI, involvement in other activities and moti-vation. An attempt to use a multiple regression model gave the result that theCDD variables only explained 1.9 percent of the variation in coverage betweenthe areas where CDDs operated. The model was not statistically significant, i.e. allcorrelations were fully explainable by chance variation.
There were five female CDDs among the eighty that filled in the questionnaire.
The ages ran from 22 to 59 years. Ten reported having attended secondary school.
Seventy percent reported having been selected by the village leaders and most ofthem were appointed during 1999. As many as 77 reported that they had spenttwo weeks or more in a year on CDTI.
Some opinions expressed in the questionnaire answers are summarised in Table 2.
There are some doubts about CDTI working well. The question on supportfrom the health sector was mainly answered positively but a reasonable numberof CDDs did not share that opinion. The views on CDDs’ improvement, com-munity involvement and the feasibility of taking on other tasks were quite uni-form. Generally, the last four questions point to the negative view regardingother tasks. Table 2 provides data on the responses from the CDDs.
Table 1: Estimated Coverage, Proportions of Persons that Received and
Swallowed Tablets, Frequencies of Side Effects and Health Care
Seeking Actions by Age and Sex in the Household Survey
Table 2: Frequencies of Responses to CDTI
Given by CDDs in Questionnaire
Sama and Penn: Another Look at Community-Directed Treatment in Cameroon Most CDDs reported that community members selected them during gatheringsor general meetings but some reported that the procedures did not follow CDTIguidelines. A CDD could be selected by the village chief alone or with his cabinetor by a health worker. The village head could also appoint himself, by villagehead alone, or by a health worker and village chief together.
Various problems in carrying out activities were identified in the process of preparing for and implementing the distribution of Ivermectin. Transportationwas a major concern. Several CDDs complained because they received no pro-vision to cover the transportation cost needed to collect Ivermectin. Some com-plained that they spent their own money. Shortages of Ivermectin were notedcreating tension between CDDs and community members. Poor storage facilitiesexisted. Batches of supplies were often known to have expired because of bu-reaucratic delays in the system.
The issue of the fear of side effects and the impact on compliance concerned several CDDs. They noted that the issue is complicated by the lack of drugs thatrelieve side effects, for example itching. CDDs reported low morale due to thelack of incentives and compensation for time spent on distribution. CDDs ob-served that people were often absent during distribution, and this required theCDD to make repeated additional home visits. The CDDs often complainedthat ‘I have no time of my own. People can come to me any time of the day’.
Several CDDs said they faced difficulties in reporting and documenting their activities. The problems were said to be caused by the short time for training andby community members not providing towards buying or recording supplieslike notebooks and pens.
The overwhelming majority of CDDs reported they received no assistance from the health facility. Only a few reported receiving assistance, which took theform of the mobilisation of community members and making announcementsabout the availability of drugs. Training on implementation, reporting and man-agement of side effects was not sufficient. Supervision and monitoring of CDDactivities from the health sector was not at an acceptable level.
In a similar manner, CDDs often reported that community members did not play an active role in the distribution of Ivermectin. One response that summedup a common reason for low community involvement was: ‘Nobody assisted. Theydid not know about the procedure. They did not receive any information about it’.
Views and Experiences of Health Workers Selection and training of CDDs were reported to present problems. It was dif-ficult to find literate candidates; even semi-literate CDDs find it difficult to copewith the task and the training. The health staff, as well as the CDDs, face trans-portation problems in getting to the training venue. Health workers also notedthe lack of response and support from health authorities: ‘We have written aproposal on how to train CDDs but we have not received any response yet’.
Most front-line health workers were not involved in CDD or programme supervision. This was actually left in the hands of the onchocerciasis coordinatorof the District Health Service. The need for reinforced supervision was clearlyrecognised.
Management of side-effects was another important task discussed by health workers. They noted that ‘Some people fear side-effects so they don’t want totake the drug. They complain of itching, swelling of body, dizziness, and stom-ach-ache’.
A finding from the in-depth interviews with health personnel was that supervi-sion of CDDs was a major problem:‘CDTI is a very difficult programme. I am the chief nurse, mid-wife, consultant,leprosy inspector and in charge of delivery and outreach activities. The CDTIprogramme is a burden to me. It adds too much work on me’.
The DMO for Muyuka noted: ‘Some nurses are not competent enough to man-age and handle records, especially financial reporting. The programme lacks agood information management system. More so, issues of onchocerciasis ende-micity are not yet clear to the community members. They do not understand whysome people are supposed to be treated in the hospital and some through masstreatment in the communities’.
The perceived roles of the health worker were captured in the following state-ment. ‘Delivery has been regular for the past three years. The health worker spendsabout one hour to discuss health issues but they rarely talk about onchocerciasis’.
Views and experiences of community members In a female FGD it was noted that CDD selection criteria were not always observed.
‘We do not know who chose them. We only saw them with the nurse moving fromhouse to house taking our names, after which they came with Mectizan. They saidwe should take it to treat our filaria’. Some community members were of theopinion that the CDDs were chosen by the health personnel. ‘To me, I know thathe was chosen by the mid-wife to help to distribute Mectizan. We were not askedto choose them’. In an in-depth interview, a community leader admitted havingappointed all the CDDs in his village after he ‘received a letter from the chief ofpost, to select four people and send to him for training. I called three of mycouncillors and my daughter and sent them for the training course’.
A major issue raised in FGDs, in-depth, and key informant interviews with community members was that of ownership of the programme. The majorityof respondents said that CDTI belonged to the community. On why they thoughtthat the programme belonged to the community, they gave responses such as:‘we are the beneficiaries’, ‘we plan the distribution’, ‘we do the distribution’, ‘weselect the distributors’, ‘the distributors is ours’, and ‘because we are told so’.
Those who thought that the programme does not belong to the community stated that it belonged to the government or the ministry of health. ‘It is the Sama and Penn: Another Look at Community-Directed Treatment in Cameroon government that brought this programme to us but we have been told that in thenear future, it will be our own.’ Most community members saw their role inCDTI as mainly passive.
Community members also identified other problems with the programme.
These included non-involvement of certain segments of the population (espe-cially men). On the subject of absenteeism, one woman said, ‘Our men are some-times not present when distribution takes place’. Others observed poor compli-ance, including both low turnout and refusal to take the drug. Others noted thatthere was a lack of awareness of the importance of the drug. This problem iscompounded by the wrong messages which are sometimes passed to commu-nity members. Community members also pointed out organisational problems.
The major complaint of communities with no health facility is that they do nothave an opportunity to discuss their health needs with health workers.
Discussion
The study provides an overall estimate of Ivermectin treatment coverage as about73 percent - not much lower than the long term sustained 75 percent stipulated asnecessary to effectively interrupt transmission. The results from this study showthat the overall treatment coverage (68.2 percent) is low in the study area.
In order to interrupt transmission, APOC sets a minimum coverage rate of 75 percent sustained for a long period (APOC 2002). However, there are largevariations in coverage between districts as well as between households withindistricts. No single factor can conclusively be identified as responsible for cover-age variations. Some suggestions are that overall low treatment coverage may beattributed to: poor selection and training of CDDs, poor supervision of CDDs,management of severe side effects, and the distribution process. The quantitativeanalysis fails at this point. Different mechanisms might be the reason for lowcoverage in different districts. The qualitative analysis reveals some of these.
In the CDTI approach the community, as an administrative, geographical and social construct, plans their own distribution system. They make decisions onwho should distribute the drug, the mode (house to house, central location) andplace (chief ’s compound, school, or church) of the distribution system.
Communities collect Ivermectin from the collection point if it is not located farfrom them and decide when to distribute. The CDTI approach is an evolutionfrom other community-based delivery strategies. It is supposed to promote activecommunity participation as an integral part of Ivermectin distribution, to improveaccess to the drug and give a sense of community ownership of the process.
The information obtained in interviews reveals that the guidelines for selec- tion and training of CDDs were sometimes ignored by some health personneland community leaders. In addition, the health system general y did not provideadequate training skills to the health personnel. CDDs to some extent lacked theskills to conduct a household census, keep good records, maintain treatment registers, observe and identify side-effects, report severe side-effects and giveinformation to the community about side-effects. The lack of management skillsof side-effects provided a major barrier to high treatment coverage of Ivermectin.
The programme for side-effect management did not train nurses. When severeside effects occurred, nurses were not available for management. The referraland counter referral system was not working.
A lack of adequate supervision during distribution was reported, and can be one reason for low coverage. Enormous supervision problems existed at thelevel of the health system. There was a gross lack of transportation for the localhealth staff to supervise CDDs during distribution. Many of the health facilitiesare under-staffed and the health personnel are not properly trained to conductsupervision. No incentives were provided for supervision; therefore no healthpersonnel were motivated to supervise the CDDs.
The Ivermectin distribution process appears to be flawed with various sys- temic problems. There were no drug distribution plans for CDDs, communitieswere given responsibility to take decisions on the mode, time, and place of distri-bution, but were not empowered in the decision-making process. Although theentire community should decide the selection of distributors, the decision-mak-ing process that may exist in a given community prior to the commencement ofthe control programme led to village leaders in some communities selecting them-selves or relatives as distributors (Amazigo 2001).
A TDR Report identifies constraints influencing the task of Ivermectin dis- tributors as: delays in the delivery of Ivermectin from the port to the country;follow up and treatment of the members of the community who are absentduring the period of mass treatment (absentees); refusals; the house to housemode of distribution; and the complex record keeping demands. The scheduleand work load of distributors resulted in some instances in a high attrition rateamong distributors.
The success of Community-Directed Treatment with Ivermectin (CDTI) us- ing Community-Directed Distributors (CDDs) in Onchocerciasis control hasdrawn attention from other disease control programmes (Walsh 1979). How-ever, the health system is faced with quality challenges regarding their selection,training, supervision, management of severe side effects, and the distributionprocess. There are systemic issues that need to be addressed before ComDT cantake on its role as an entry point to community-based healthcare interventions ata time when there is a need to critically examine determinants of treatment cov-erage of Ivermectin distribution. The main purpose here is to propose somebasic conceptual elements that may help establish a consistent basis for policy,action, and research before CDDs can take on additional health care activities.
There have been various attempts in Cameroon to use a Community-Directed Treatment System and Community-Directed Distributors for other healthinterventions but the implications of this development for the treatment coverage Sama and Penn: Another Look at Community-Directed Treatment in Cameroon of CDTI are not clear. It is expected that the integration of additional communitylevel heath care activities into CDTI would enhance treatment coverage. However,treatment coverage may be at risk if the health system starts using CDDs forother activities without ensuring sustained high coverage. Overloading CDDswithout sustained treatment coverage can erode the health system. Opinions havebeen expressed, rather strongly, as to the negative effects of the ultimateconsequence of overloading CDDs with programmes built essentially on top-down approaches on the effective implementation of CDTI. (Brieger 2000; Zekusand Lysack 1998; Schwap 1997, Walsh and Waren 1979). What this means is thatmore evidence on the nature of the effect of involving CDDs in other health anddevelopment programmes on CDTI implementation is needed. This study pointsto some major weaknesses.
Conclusions
Community-Directed Distributors find it difficult to achieve high coverage and
sustain it due to programmatic obstacles in their selection, training, supervision,
and management of severe side effects and non-empowerment of the commu-
nities in decision-making.
Although coverage is not extremely low overall, there are large variations and pockets of the population are left without treatment.
At the organisational level, the issue of the quality of the distribution process has not been addressed, and because of this, the programme suffers from tech-nical inefficiency.
Serious systemic problems still exist, and need to be rectified before commu- nity-directed distributors can take up additional health and development activities.
The communities are not involved in decision-making regarding selection of distributors, mode, place, and time of Ivermectin distribution, hence ownershipand sustainability of the programme seems to be eroded.
The support from the health sector in terms of training, supervision and assistance is not sufficient everywhere since health facilities are under-staffed, poorlyequipped, and poorly paid.
It is necessary, therefore, for health systems to address systemic programmatic and organisational issues before undertaking large-scale implementation pro-grammes like community-directed treatment with Ivermectin.
To obtain high and homogenous coverage there is a need to use better rou- tines for the selection of CDDs and select only persons that can be effectivelytrained. CDDs must be given better training for strengthening the health systeminvolvement. It is necessary to give the CDDs information and knowledge aboutthe management of severe treatment side effects including the capacity to reportadequately and inform the community. Communities must be empowered totake decisions. The linkage between research-to-policy-to-action-to-practice shouldbe clear.
References
Amazigo, U., et al., 1998, ‘Delivery Systems and Cost Recovery in Mectizan Treatment for Onchocerciasis’, Ann. Trop. Med. & Parasitol.
APOC Partners Meeting, 2000, DIR/APOC, Meeting May-June.
APOC Technical Report, 2002, ‘Community Directed Treatment with Ivermectin’, SW1 Blas, E., and Limbambala, 2001, ‘The Challenge of Hospitals in Health Sector Reform: The Case of Zambia’, Health Policy and Planning, 16 suppl. 2.
Brieger, W. R., 1996, ‘Health Education to Promote Community Involvement in the Con- trol of Tropical Diseases’, Acta Tropica, 61: 93-106.
Frenk, J., 1993, ‘Dimensions of Health System Reform’, Health Policy, 27.
Godin, 1998, ‘Cameroon Chad: Cost Recovery’, Annals of Tropical Medicine & Parasitology, Paphassarang, C., Philvong, K., Boupha, B. and Blas, E., 2002, ‘Equity, Privatization and Cost Recovery in Urban Health Care: The Case of Lao PDR’, Health Policy and Planning,17, suppl.1.
Remme, J. H. F., 1990, ‘Onchocerciasis’, in C.J.L., Murray and A.D. Lopez , The Burden of Diseases: Global and Regional Estimates for 1990, World Health Organisation.
Schwab, M., and Syne, S. L., 1997, ‘On Paradigms, Community Participation and the Future of Public Health’, American Journal of Public Health, 87 (12):2049-2051.
Walsh, J. A., and Warren, K. S., 1979, ‘Selective Primary Health Care - An Interim Strategy for Disease Control in Developing Countries’, New England Journal of Medicine, 301: 967-976.
Wealth Health Organisation, Technical Report, Series no. 852.
World Health Organisation, 1996, ‘Community-Directed Treatment with Ivermectin: Re- port of the multi-country study’, Document TDR/AFR/RP/96.1.
World Health Organisation, 1997, ‘Twenty Years of Onchocerciasis Control Review of the work on Onchocerciasis in West Africa from 1974-1994’, Geneva, WHO.
World Health Report, 2000a, Health Systems: Improving Performance.
World Health Organisation, 2000b, ‘Implementation and Sustainability of Community- directed Treatment of Onchocerciasis with Ivermectin: Report of the Multi-CountryStudy’, Document. TDR/IDE/RP/CDTI/00.1.
World Health Report, 2002, WHO/WHR/02.1.
Zakus, J. D. L. and Lysack, C. L., 1998, ‘Revisiting Community Participation’, Health Policy and Planning, 1998, 13 (1):1.

Source: http://www.codesria.org/IMG/pdf/Chap7-2.pdf