Independent monitoring of the CDTI projects in the Southwest province of Cameroon 9th August,1999 to 8th September1999: independent monitoring report

INDEPEI\DENT N{ONITOIING OIT CDTI PROJEC'IS IN'I'HE SOIITFNVEST PRO\/INCE OF CAMEROON 9th August,1999 to 8th Septenrber',1999 INDEPENDENT MONITORING REPORT Submitted by Mr. Oka Obono(Prrncipal Monitor and Team Leader) Department of Sociology, University of fbadan, Nigeria Dr. Rrchard Ndyomugyenyi(Ext.ernal Monitor ) l.lational Onchocerciasrs Control programme, Kampala, Uganda Mr. Nam.anga Kingue ( Internaf Monitor ) Helen Kelfer fnternational Yaounde, Cameroon Mr. Desir6 Njombinr ( Internal External ) Li-ttoral 2 Oncho Pro j ect, Cameroon To Dr. A Sek6t61r Director a. i. African Prograrnrne on Onchocerciasis Control Ouagad.ougrou, Burkina Easo ,+'+qq' t: ir( '.,, a ! l l ,i)lF. ' l-a\. ,71( F,' n l,' '' I r t"+2h-3-)) I i:, : ^' i':J U" l; +./ t-, APOC / PIU T.\llLE OI,- CON'tIlN'l'S EXITCUTI\/E S Ll MNIAIIY TNl'I{OI)LICl'ION I.l. Backgrountl .. 1.2. Thc Soutinvest Provincc o1'Car-neroor-r l.i. l'canr ('onrposition ..... I .+ Ternrs o1' Rcli'rcrrce . . . . . METHODOLOGY 2.1 Prtlccdure s).2 Populatiort .. 2.3 Sanrpling ancl Selection olVillages l.+ Inst runrents RESULTS 3. I lndicators. 3.2 Accuracv and Compatibility of Records 3.3 Constraints 3.4 Sustainabilitl'. .. 3.5 CornmunitvPerception 3.6 Quality of Irnplernentatior.r of CD fl . . . .3.7 Uniciue Featurc of Projcct Area....... DISCUSSION AND CONCLUSIONS .. RECOMMENDATIONS ... _1 -5 6 l 2 J 8 9 10 13 13 20 24 26 28 29 .30 30 )-) .1. 5. 1 LISl'OF ]".\BLES -['able i Distribution of Sarnplecl Conrmunities in the Southrvest Province by District, Ilealth Arca. and Village Categorics 12 l-able 2: hrput Indicators 14 'l-able 3: Distribution of Clonrnrtrnities br N4ode of Selection of CDDs 16 fablc 4 Decision N4aking Process at the Conrmunity, l-evel in Category B Villages in I{espeot of Month of Distribution of Ivermectin .. . . . . . t7 '[able 5 Decision Making Processes at the Con-rmunity Level in Respect of the Time, Mode and Month of Distribution in Category A Villages .18 -l-able 6: Decision Making Process in Respect of Mode of Distribution . ..19 Table 7: Treatment Sumrnary by Villages 21 Table 8: Reasons for No Treatment, All Communities .L) l TiXECI ]TI\'E S T]M NI ARY The APOC independent rnonitorhrg teanr visited 6 Category A villages and 25 Category B villages. distributed atnong Lirlbe. Kutnba. and Ngtti health districts of the Southn,est Pror,ince, Catrteroon. A total of 532 people r,vere itttel-u'ierved fi'onr 110 households in the 6 Categow A villages during tlte stuve)' Tlu'ee ltturdred attd tu'elve people (5g%)hacl received treatment rvhile 220 (119'i,)had trot beeu treatecl by the tinte of the suruey. Horvever', a chrster of settletnents in Nguti ltave not received treatrnent at all and. as such CDTI has not been irnpletneuted arlong thent The CDTI progratllnte is being irnpletnented fbr the first time in the Sotrtltrvest Proyince of Carneroon. It is therefbre associated rvith problerns that catt be overr:otne rvith proper orientation of the comrnunities and the right emphases in the trailrhg of i CDDs and their supervisors. For example. no corilnunity decided on the period of i treatnlent, but underlook distr-ibutiorr ott the basis of rvhen the drug was available. There is a need in this regard to orient the drug provisions systems more suitabll'to periods that the communities detennile as convenient to thenr. The cornnrunities must accordingly be guided to take decisions for this pulpose. Less than half of the communities monitored (45%) are curently taking decisions on the nrethod of treatment, but appear to have left this crucial decision to either the CDD or the health rvorker. This implies a- loss of the eillpowerlnent that the CDTI strategy rvas deigned to foster and perpetuates the notion that CDTI is an extemal programme. Although training of CDDs in the Southwest Province rvas found to be universal, there are indications that it was inadequate and adversely affected programme implementation and supervision. The 112ining was said to be too compact and compressed into 1 or 2 days and did not promote fulI comprehension by the CDD trainees and their supervisors. On the part of the communities, although sensitisation and mobilisation took place, these processes were also rushed and have not had the fulI desired impact on community perception of programme ownership and issues of responsibility. The quality of record keeping is poor and the level of absenteeism owing to lack of information high (28.6%), as is also the refusal rate (27.7%o). Disparities were found between coverage and other rates recorded in the treatment registers and those calculated from the independent household survey. Few communities have made an 1 I 2effective transition to a perception of community ownership of the proglalillle and accept minimal responsibility for its implemetrtation, although all eucouragLtg pt'oporlion (612o/;) did decide on the criteria for CDD selection The lolv recorded propofiion of refusals trvo rrrolths after distribution (9 6%) is not cor:-oborated by our uonitoring experience. Tlie level of scepticist'I, doubt and pessimism is gelerally high. A proper ernphasis on sensitisation and mobilisation for community participation is essential for reducing the percentage of refusals in the nexl round of distlibutiols lrr 74.2oh of the oomnunities itrvestigated, a health rvorker supervised the CDD, but this means that in more than a quarter of the cornnrunities (25.8%),there rvas no supervision. ln the cases where zupervision was caried out, the ildications are that it was not effective due to defective training and motivation. Among the cornmunities 96.7oh are reporled to have received health education but because trailing was problematic, the health education t'eceived rvas also not adequate It rvas not sufficient to allay the fears and misconceptions of the community members about side effects, nor was the scope wide enough to cover the entire populations With regard to input indicators, 96.7% of the villages have trained CDDs but there is need for retraining. Drug supplies were not timed to the periods detennined by the communities. There is need to address these issues in subsequent distributions because it is one strorg way of demonstrating to the cornmunities that the CDTI prograrnme belongs to them. Not enough communities collected their drugs from a central point. Only 58.1% did so. This naturally suggests that the perception of their responsibility for the CDTI process is low. The overall solution to the teething problems of CDTI in the Southwest hovince lies in proper training, motivation, and sensitisation and mobilisation. The use of IEC materials for these purposes is highly lsc,smmended. Greater attention should be paid to sensitisation than has been the case. The CDTI programme would face sustainability problems if the community does not see it as its own. Feasible aspects and processes in the implementation design, including the provision of mectizan, should be community- directed. Such a situation would enhance the credibility with which communities may be persuaded that the CDTI programme truly belongs to them and that they should accept responsibility for it. APOC Monitoring in Southwest Province, Cameroon - 7th - gth 1999 3N{ONITORNG B}' INDE,PIINDIiN'T SCIENTISTS OF C O i\l I\ItlN I'I'\' DIREC'IIiD ]'IIEATNI ENT'S I N ;\P O (l (-' O tlNT RI E S (s o UTINtsST PROVINC Ii, C',Al\IEROON ) 7,I'II r\UGUSI.-8.h SEPTEN,IBER, I 999 INl'RODT]CTION 1.1 Background The Ai.ican Prograrnrne on Onchocerciasis Control (APOC) rvas established il Decembet', 1995. to set up effective and self-sustainable comnrunity-directed ivernrectin treatment for onchocerciasis throughout l9 endelric countries itt tropical A'fiica rvhere 50 nrillion people are at risk of the disease. APOC is baserl otr tlte Mectizar'8 (Iverniectin) Donation Progranrme of Merok and Cornpany. lttc., and arvarded its first country project grants in i996. To date. sotne 45 grants have been approved under this progratnme, about 80% of which are specifically der,,oted to field operations for Conurunity-Directed Treatnrent witlt Iverntectin (CDTI). Field proposals are based on a partnership betrveen APOC. the National Onchocerciasis Control h-ogramrnes (NOCPs), and several Non-Governmental Development Organisations (NGDOs). Each project is funded for a nraxirnurn of five years, and this significantly constitutes the community parricipation into a critical component of APOC-funded interventions. Over the five-year period, it is expected that APOC support rvill decrease in inverse propoftion to the support provided by other partners, and that the cost per individual treated will reduce by 90o/o. With this achieved, it is further expected that individual NOCPs will contfurue to manage their own CDTI activities for upwards of 15 years to ensure the effective control of onchocerciasis. As an approach to community involvement in disease control, CDTI was field tested with support from tINDPMorld Bank/ WHO Special Programme on Research and Training in Tropical Diseases (TDR). Under CDTI, communities obtain their ivermectin from the health service' select their distribution agents' or commudly-Directed Distributors (CDDs) for training, and detennine both the date and mode of ivermectin distribution in the community. Through their CDDs, communities are required to gather information about coverage in order to take necessary steps to ensure that high levels of APOC Monitoring in Southwest Province, Cameroon - 7th - gth ber,1999 1treatrnelt colrpliance are achieved ancl sustained. 'l'he TDR projects established that con,,unities are capable of managing theil o*n ivennectin distribution rvith facilitative srlppolt 1om the health sen,ices and that they achieved equal, or better, covel'age results tlran distribution activities that were centrally organised by district health setvices CDTI $as thought to offer the best prospects for sustairlable community level ottchocerciasis iltervention, and rvas adopted by APOC as its basic strategy' There is one fuldarnental area in which CDTI can be distinguished from the earlier Comrnunity-Based 'freatment with Ivermectin (CBTI) approach, or Colntntrnity- Based Ivermectin Treatrneut (CBIT), as it is sometimes interchangeably called. The latter is a procedure developed by health providers to distribute ivermectin to endemic cormnunities through systerns or methods that are decided upol by the health services. ln other rvords, it usually involved mjnirnal input fi'om the target communities, if any at all. A number of ethical and empirical issues arose that questioned the feasibility of that method to guarantee sustainable distribution in the post-APOC period. Ulder CBTI (or CBIT), for exarnple, activities rvere based m the communities but rvere not owned, controlled, or directed by lhe communities. With CDI-I, however, activities are botlr in and of the community. The approach that accords well with an emerging perspective of health as a lneans of promoting community empowerment, particularly because top-bottottt approaches appear to undermine the community's capacity for self-determination or the relevance and integrity of its indigenous klowledge and practice. GDTI promotes progressive community involvement in health management and health seeking behaviour through their joint basis in the social and political experience that has been acquired by communities in dealing with their health and social problems over time. ln the light of this, the CDTI approach was designed to emphasise and en-hance the decision making and problem solving capacity of local communities, and to render that capacity relevant to onchocerciasis control. The present report documents the experiences and findings of an lndependent Monitoring Team set up by APOC to visit the Southwest Province, Cameroon, where CDTI activities commenced in 1999. The tndependent Monitoring Team paid close attention to the transition from several years of CBIT interventions to the CDTI- L Monitoring Southwest Province, CameroonAPOC 7th ln gth 1999 51.2 'l'he Southrvest Provincc of Cameroon Solthlvest I Project Area (SW 1) is parr of the Southrvest (SW) Provitlce of Canreroon. lt lies betrveen 5o20'attd 4o N ancl 8o45'and 9"45' E. SW I includes 3 adnilnistrative divisions (Fako. I(upe Mauengouba, attd Meme). Tltese divisions are made up of about l0 subdivisions. Buea torvr il the Irako divisioti is the adntilistrative headquarlers of the SW Provilce. SW I is bordered to the Nofth by SW adrnjnistrative divisions of Lebialenr and Man1m, to the South by the Atlantic Ooeatt, to the East by the Littoral and West h'or,,inces and to the West by the I{epublic of Nigeria. Irrorn tlte rieu,point of health admjnistration, SW I is divided into g health districts (Buea, Limbe. Muyrka. Tiko, Bangem, Tombel, Nguti, Konye and Kumba). subdivided into 65 health areas. The total area of the province is approxitiately 14,300 kmz and the population is alter-nately estimated at 710,050 people living in about 780 conmunities. The predoninant vegetation is Equatorial Rain Forest rvith mangrove vegetation on the coast. The altitude ranges from 0 metre on the coastto 2,200 metres in Buea town. Mount Cameroon. the highest peak il West A-frica, is located in Buea town. The Soutlrrvest hovince has a netrvork of drailage systems that florv fron-r these high altitudes and are iltenupted by nurnerous cascades, rapids and rvaterfalls. These fast-flowing and higlrly oxygenated water sources are perfect breeding sites of Srnrulturn vectors. The rainy season is Iiom mid-March to mid-October with its peak around July and August. The dry season extends from the end of the rainy season to mid-March with the exception of Debundscha, one of the Category B villages visited by the lndependent Monitoring Team in the Limbe Zote 2 district of the Southwest Province. Located approximately 14 kjlometres from coastal Idenau and 41 kilometres from Limbe, Debundscha is reputed to be the wettest settlement in Africa, with daily rainfalls and drizzles. Farming is the main provincial agricultural activity and is mainly practised from March and April at the beginning of the rainy season. lts intemrptive influence on both distribution and monitoring activities this year are obvious and will be examined more systematically in a later part of this report. The roads in the province are impassable in the rains, a factor that may prospectively affect coverage in a district like Nguti that is notorious for the poor quality of its roads and its relative distance from Buea the provincial headquarters. The worst 3 months for driving through the Southwest are from APOC Monitoring in Southwest Province, Cameroon - 7th August - 86 September, 1999 6Jull, to September, ancl these incidentally rvere the nronths that duririg rvhioh distlibutiorr in some comnunities as rvell as iudependent tuonitoring took place [t is possible to orient these activities to the rnonths in the respective districts rvhen they lrlay be most suitable The potential negative ilfluence on progranme sustainability of selecting the \\rong rnonths for distribution camrot be exaggerated Diurlal fanning activities iliply noctumal tr.eatment rvith the implication that the CDD is frequently unable to personally obsen,e the achninistration of ivermectil. This can affect the veracity of coverage rates. Onchocerciasis control in the province is organised in a four-tier structule that conforms to the htnary Health Care (PHC) approach in Cameroon This involves a reodeltation process of the district health system at all levels. The cetttral level comprises direct inter-relationships among APOC, Sight Savers Interlational (SSl), and the Ministry of Public Health. At the provincial level, the Provincial delegation, the office of the provincial Onchocerciasis Co-ordinator, the Provincial Chief of Commulity Health. and the offrce of the Essential Drugs Programrne are the principal parties in the control programme. At the district level, the District Medical Officer (DMO). the Chief of Health, the Chief Medical Officer and the Chief of Finance and Adrninistration compose the control team. The health centre level comprises the Chief of Health Post and the nurses, while at the community level, the traditional chiefs and councillors, the village health committee, the comrnunity members and the CDDs guide control activities. 1.3 Team Composition The Independent Monitoring Team that visited the Southwest hovince comprised two external and two internal monitors. The extemal monitors were Mr. Oka M. Obono of the University of Ibadan (Principal Monitor and Team Leader) and Dr. Richard Ndyomugyenyi (National Co-ordinator of the Onchocerciasis Control Programme of Ugan<la). Messrs. Desir6 Njombini (Provincial Co-ordinator of the Littoral 2 Onchocerciasis Control Project, Cameroon) and Namanga Kingue (Country Representative of the Helen Keller lntemational, Cameroon) were the internal monitors. I Monitoring ln Southwest Province, CameroonAPOC 7th August gth ber, I999 11.4 Terms of Reference The Independent Monitoring team that was sent to thc Southrvest proviuce rvas set up rvith the following tenns of referettce. I Succinctly docutrent ltorv ivermectin treatnrents rvere undeftaken in a sarrrple of conrmunities rvith apploved the CDTI proiects in the Soulhwest Provinr:e. Carneroorr 2. Assess cotlnunity involvernent ir the seleotion of CDDs, and dates and modes of drug collection and distributior. The state of comnrunitl, rvillingness to accept and bear the responsibilities designed il the CDl'l process rvas also to be exanrired. 3. Document somrnunity perception and expectations of CDTI processes and, on these bases, deterrnine the extent of comrnuuity satisfaotion rvith programure activities and outcomes. 1 Assess the qualitl, of tlaining received by Community-Directed Distt'ibutors (CDDs) and health personnel involved in the project 5. Examine the record books of the CDDs (and health pel'sonnel associated rvith the projects), and assess the quality and accuracy ofrecord keeping. 6. Detemine the nuurber of comrmurities and eligibles treated and compare findings rvith quarter', village, health area, district, proviucial and national level records. '7. Assess the quality of participation in and srqren,ision of, ivermectiu distribution by health staff. The orientation of the health staff to the CDTI process in terms of 112ining received was also to be assessed. 8. Identi$r constraints in the distributions and recommend appropriate measures to APOC Management to be taken before the nex round of treatments. 9. Discuss prospects of CDTI sustainability in the Southwest Province APOC 7th ln g6 , I999 Monitoring Southwest Province, Cameroon 82 METIIODOLOGY 2.1 Procedures The methodologl, for the nronitoriug exercise consisted of a triangulation of qualitative and quantitative instrurnents that were first developed at a nreeting of nine Independent Scientists fi'orn l-1 - l5 September 1998 in Ouagadotrgou. and designed to rnonitor the process of ivermectin distribution itr APOC countries. Since this process is ruulti-faceted, it was inportant that the various steps or phases in the treatment process be identified for attettion The different categofies of persons ilvolved in CDTI rvere targeted for interviews or discussiorrs. Teanr tnetnbers exarnined the nine objectives in theterms of reference and compared them to a series of process indicators developed by APOC Managernent for Independent Monitoring. They then worked in srnall groups to address the tenls ofreference and to: I Develop specific indicators fbr each, ensuriug that the indicators originally developed by APOC Management .-ere included; 2 Identi! variables for each indicator so developed; 3. Design questions, observational techniques, and content analysis procedures for record reviews; 4 Determine target groups for the administration of the respective ilstruments; and 5 Propose a schema for harmonising findings from the various instruments in order to succinctly address a number of research questions presented to the Team members by APOC Management. These questions are recapitulated here as follows: ,e Does the implementation follow the CDTI approach? re What are the corutraints of the CDTI approach as proposed by APOC? FF How was training of CDDs carried out? rs Did NOTF supervise the project? FF Did health staff participate in distribution, management of cases, and supervision? Based on the experience gained in the 1998 round of independent monitoring activities, the monitoring insruments that were fust developed in 1998 were revised, again in Ouagadougou, from 1-llMarch,1999. The instruments were further reviewed APOC Monitoring in Southwest Province, Cameroon - 7th st - 8th ber, 1999 t I I L t f I l_ 9at a rvorkshop itr Katrale Districl of Ugatrda bl,National Ortchocerciasis Co-ordinators, District Co-ordinatols. represeutatives of NGDOs, ofllcials of the Ministry of Health and project Managers. 'fhis review process canle to an end at a CDTI Monitoring Instnulents Filalisation Meeting held at Karnpala, Uganda. l0-14 May, 1999. The Monitoring Teani that visited the Soutlrrvest Province of Carneroon used thcse finalised instnunents and entered its data using EPI lnfo, as recomurended by APOC. A diskette containilg the respective files accompanies this repofl. 2.2 Popul:rtion Five broad groups of persons were studied in the collrse of tlie monitorirrg exercise. These included l) r,illage leaders in both Category A and B villages or their represertatives, 2) household heads and members in Category A villages, 3) Conmrunity-Directed Distributors (CDDs) in category A arid B villages, 4) rnen, womerl and youths organised in focus groups in Category A villages, 5) health personnel involved iu the CDTI process at various levels of programrne illplementation. At the national level, personal interviews rvith World Health Organisation (WHO) Representative, Dr D Waniing (9e August. lggg), the National Co-ordinator of the National Onchocerciasis Control Ploglamme, Dr. Marcelline Ntep (9ft August, 1999), the Country Representative of Sight Savers lnternational, Dr. Rosa Befidi- Mengue (9ft-10th August, 1999), and the Director of the Community Health Department and Chailman of the National Onchocerciasis Task Force (NOTF), Dr. Basile Kollo (l0t August, 1999). The Provincial Co-ordinator of Southwest Province, Ms. Cecilia Mah and Dr. Martin Ekeke Monono (Provincial Health Delegate for Public Health) were also interviewed (l2e August, 1999) as were other district and community level health personnel involved in the CDTI process. Relevant records were examined for quality, accuracy, and consistency at these levels. The primary nature of the lndependeut Monitoring was process docuntentatrcn of the zustainability potential of community- directed efforts at onchocerciasis control, specifically with regard to ivermectin distribution. The sampling procedures adopted for the sfudy were based on this central consideration. These procedures are examined in the next section. APOC Monitoring in Southwest Province, Cameroon - 7th st-8e er, 1999 l0 2.3 Sampling and Selection of \/illages Multistage sarnpling began by stratifuirig the nine health disricts in the Southwest Ilrovince of Cameroon, using a cornbination of geographical atrd coverage criteria The box to the left of this page shorvs the nine health districts of the Soutlrrvest hovince and the district level coverage rates supplied by the Yaounde office of Sight Savers lntenrational (SSI). These districts are Bangem, Buea, Konye, Kumba Limbe, Muyuka, Nguti, Tiko, and Tombel. The first round of sampling was purposive and was designed to provide three districts that would represent the three main strata that are fourrd in the Southwest Province of Cameroon. The second stage of this process was designed to select one health area from each of these districts, bringing the total number of health areas covered to 3. As will become evident presently, this procedure was untenable in the field. The sampling process had assumed that each health area would be large enough, and that the settlements in them would be spread out sufficiently, to pennit the selection of two "Category A" villages, one of which would be over 8 kilometres fi'om the nearest health facility. This criterion was not feasible in two of the three health districts visited. Limbe was selected to represent the West Coast of the province. This region is characterised by a linear cluster of heterogeneous settlements, with a projected population of 96,346. The complex mix of ethnic and racial nationalities is promoted by the location of Idenau, a rustic port settlement, within 55 kilometres of the district headquarters. Limbe Zone 2 Health Area could not provide the two sets of "Category A" villages based on the criterion of proximity. Accordingly, Batoke Health Area was selected in addition to Limbe Zor,e 2. As noted, this deviation was the result of normal patterns of settlement in the health area and how they interacted with our sampling procedures, but on the whole, this exception proved the norm. Similar conditions as those found in Limbe were repeated in Nguti sub-division which was selected to represent the upper geographical limits of the province, where it began to border Nigeria, close to the Southwest 2 hoject territory. In this case, Nguti and Manyemen Health Areas were selected, again because District Rate Tiko Konye Bangem Buea Tombel Limbe Kumba Muyuka Nguti 74.890h 73 98o/o 73.92o/. 70.030/, 69.34o/', 69 18% 62 06% 6r.98% 44.56% t- : t. r L a Monitoring southwest Province, Cameroon I APO c 7th tn gft er 999 ; l1 the linear settlem.ents were so close together that it rvas not possible to obtail) a "categoty A.,r,illage that rvas over'8 kilornetres fi'onr the nearest healtlt facilitl'tt'tlltrtt tJrc same heallh areo. Betrveel Lirlbe clistrict and Nguti sub-divisior), \ve selected a third district, bringing to tluee the nlnrber selected to cover the geographic and socio-denrographic di'ersity of the province. This district rvas Krrnrba and it is located sorltervhat ceillrally ilr the province. lt was olly withitr Kumba that the sarttpling procedure adopted acttrally held We were able to select olly one health at'ea, Kotto Barorrlbi, large ertough to provide a "Category A" village cotrmurity that was ovet' 8 kilometres arvay liorn the nearest health facility. But even this health area was lot, however, large enough to provide tlie requiled number of "Category B" r,illages, a reality that saw us selecting this set of villages based on proximity as required, but fiorn different health areas. A case for serendipity can be made for rvhat transpired both durtrg saurpling and the period of the ercounter with its implications. In a willy-nilly fashion, the ttronitors assessed the implementation of CDTI in seven health areas spread over three healtlr districts, as agailst 3, which a perfect application of the sampling assrttnptiotts would have led to. We are therefore confident that our firidings llossess a tobust and con-rprehensive charactel than would have been impossible to obtain otherwise. In general, the sirnple random process agreed upon by the Indeperldent Scierrtiss at their various instrument-drafting and instrument-review meetings was followed to select six Category A villages and 24 Category B villages in a distribution of four Category B villages per Category A village. The step-wise process described above was followed to achieve this result, although the monitoring team eventually investigated 25 Category B villages instead of 24, owing to the dynamics of the CDTI implementation in the field. That distribution is shown below in Table 1. The fust order of communities (Category A villages) comprised the monitoring units within which extensive investigations took place. Six of the 8 instruments were administered in this category of villages and it is from them that the most in-depth information on the CDTI process was obtained. tn sum, therefore, the selection of Category A villages depended on the prior selection of the 3 districts in which they are located out of the t health districts in the Southwest Province. Category B villages were selected in relation to the selected Monitoring Southwest Province, CameroonAPOC 7th tn gth t9'99 Category A villages, in a latio of 4 per Category B village per Category A village, on the basis of proximity. Categoly B t,illages rvere nteattt to provide cotnpletrtetttary quantitative evidence on tlie exlent of CDTI itt4rleurentatiort. Tr.vo ilsttumeuts rvere administered in them (see Seotion 2.4 below). l}e selection procedttre in the Southwest hovince brought the number of Category A villages to 6 (2 eaclt in 3 selected districts and Category B villages to 25). We selected orte Category A village per district to represent'hear" cornrnunities and another to represent "fhr" ones. Mottitors had to compile and collate these figules for tltei' orm investigations because village level summaries were not available at the couununity, district, and provincial levels. Table I below shows the distribution of villages by district, health area, and proxinrity criteria in the Southwest Province. vill eC vrllage falls r:nder Ideuau Heafth Area ullage falls uder Kombmc Heahh Area t- Category BCategorv ADistrict Ilealth Area Cassava Fanns Lumpsum Towe Unrty Quarters Lirnbe Zone 2 Health Area Ivhle 2 (T.Jear) Debundscha Beachr Debundscha Centralr Isongo' Nronii' Batoke Health Area Bakngrh (Far) Lrrnbe Bai Manya2 Bai Muassa Bai Mutoko2 Bai Panya New Torvn Barombi (Near) Bakundi Foe Bai Longe Kotto Barombi Kuke Mbomo Kotto Barombi Bar Foe (Far) Kumba Mboka (Near) Afrikpwabi Betock3 Ediango Ekenge Nguti Health Area Kokobuma (Far) Babensi I Babensi 2 Ebanga Ekrta Talangaye Manyemen Health Area Nguti Heahh Arca village falls rmder Mmyeznen i APOC 7th ln ge er , 1999 Monitoring Southwest Province, Cameroon {- ,'. t- tI L I t Table l: Distribution of Sarnpled Comnrunities in the Southwest hovince by District, I f ll 2.1 lnstrunrents ln the Categorl,A villages, six instnunellts \vere admfutistered,vr:. the household surl/ey questiormaire. llle persoual interuierv rvith CDDs, the personal intervierv with rillage leaders, th'ee gloup discussious (one each rvith rnen. wor]ren and a nrale or fenrale vorrth group aged 16-24 1,ears), and persoual interuiervs rvith healtlr personnel (super-r,isors). ln the Category B villages. rve conducted intervieu's with the village leaders and also furteruierved the CDD(s). 3. RESULTS J.l Indicators Effect Indrcators E1 Percentage of target communltres that decrded on the penod or method of treatnrent Whrle no communitv decrded on penod of treatment. 45oh dectded on rnethod of treatment Number of target communrtres that decided on the cntena for CDD selectton 61.20h aulplr]lldrcalsr! E2 ol o3 o4 lnput Indrcators II 12 Nunrber/percentage of refr.rsals two months after drstnbutron 9 6% Nr,rnrber/percentage of vrllages treated: 96 1% Cost per person treated: We recor-nrnend that APOC calculate thrs iterl when all the retums are rn 05 Number/percentage of vrllages where CDDs were changed after the first treatment. 0% O6a Nurnber/percentage of vrllages where the CDD was a health worker:O% O6b Percentage of villages where the CDD was supervised by a PHC staff 74 2%. but the supervision was not adequate. 01 Percentage of villages that received health education 99 7%. but the health education received was Del3dequate to allay fears. nerther did it cover the entire populations. Percentage of vrllages wrth tramed CDD(s): 96 7%. but there is need for retrarrung, Percentage of villages that expenenced late supply of ivermectrn:No villase received late supplv in the first qonsignment. it was with the second that 100 oh complained of lateness Did prolect experience late supply of drug5? No. Number of vrllages that collected ivermectin'.58.1% Number of CDDs with measunng device: 100%. but manyhad relumed their measunng sticks to the health centres by the time of the monitoring exercise Number of vrllages with summary forms at the villages. 0% Number of health centres with records 100% I3 T4 I5 I6 l1 APO C 7th ln s gth ,ber, 1999 Monitoring outhwest Province, Cameroon I4 These indicators are disagglegated in the accompanyitrg tables. Table 2 shows input indicators according to key issues affecting programme irnplernentation. Table 2: Input lndicators Table 2 seems to give the impression that CDTI processes are proceeding smoothly, except with respect to collection procedures, the availability of summary forms at the district level, and the availability of side reaction records. Thus, all surveyed Village Trained CDDs Late Supply No Drug Sh'tage Late Funds Collect. From a Point Meas. Device Treat. Register Summ. Fornr in District Office Side Reactn Record I Bar Foe* Yes No NA No Yes Yes No No 2 Bakrngrh+ Yes No No NA Yes Yes Yes No No I Kokobumax Yes Yes No NA No Yes Yes No No -1 Mboka* Yes No No NA Yes Yes Yes No No 5 Mile 2* Yes No No NA Yes Yes Yes No No 6 NewTorvn B'brx Yes No No NA Yes Yes Yes No No 7 Afrikprvabr Yes No No NA Yes Yes Yes No No 8 Babensr I Yes No No NA No Yes Yes No No 9 Babensr II Yes No No NA No Yes Yes No No l0 Ba Longe Yes No No NA Yes Yes Yes No No 1L Bar Manya Yes No No NA Yes Yes Yes No No 12 Bai Muassa Yes No No NA Yes Yes Yes No No ll Bar N{utoko Yes No No NA No Yes Yes No No 1-l Bar Panya Yes No No NA Yes Yes Yes No No 15 Ba-lrundu Foe Yes No No NA Yes Yes Yes No No 16. Betock Yes No No NA Yes Yes Yes No No I 7 Cassava Fann Yes No No NA No Yes Yes No No l8 D'scha Beach Yes No No NA Yes Yes Yes No No 19 D'scha Cent'l Yes No No NA Yes Yes Yes No No 20Ebanga Yes No No NA Yes Yes Yes No No 2l. Ediango Yes No No NA Yes Yes Yes No No 22.Ekenee Yes No No NA Yes Yes Yes No No 23 Ektta No Treatment 24.lsongo Yes No No NA No Yes Yes No Yes 25.Kotto B'bi Yes No No NA No Yes Yes No No 26 Kuke Mbomo Yes No Yes NA No Yes Yes No No 2T.Lumpsum Yes No No NA Yes Yes Yes No No 28.Nronii Yes No No NA Yes Yes Yes No Yes 29.Talangaye Yes No No NA No Yes Yes No Yes 30.Towe Yes No No NA No Yes Yes No No 3 I.Unity Q'ters Yes No No NA Yes Yes Yes No No Monitoring Southwest Province, Cameroon I APOC 7th IN gth 999 I {. t' L ? { I l5 colltrnunities are sho\Mt to have trained CDDs. tlolte except otre received late supply of ir ernrectlt or experienced dmg shorlage. and so forth Nevertheless. rvheu data fi'otn this lable are discttssed in cortrbirratiort with infolnatiorr fi'oni other sources durtrg the field irvestigatiott. differetrt conclusions enterge. Although training of CDDs was utrivet'sal. there are strong indications that it q,as iladequate. While lrrost conrmunities repotled that thele was tto late supply of drlgs. it is ilrportaut to uote that this rvas with ref'erellce to first consigtunents only. Requests u,ere nrade fbr lrore ivermectil wheu the first supplies got exhartsted but the second round rvas usually late fur an'iving. The distribution process is hirrdered by the lull in between srpplies, during rvhich period CDDs engage in other ecottornic activities. The table also shorvs that not all the communities collected their drugs fi'orn the health facilities. Health workers brought the drugs to them. The health wotkets, rather thal the cornmunities, often compensated CDDs who collected the drugs on behalf of their' communities. This is contrary to the CDTI objective of community responsibility for the collection of its dnrgs and points to the iradequacy of their sensitisation and nlobilisation in this respect. Where health workers did not conrpensate the CDDs. the sitnation was irterpreted as discouraging to CDD rnotivation. Additional mobilisation and sensitisation rvorkshops wonld are uecessary to create the type of arvareltess that is needed to ensure the sustainability of the CDTI process in the comnunities of the Southwest Provilce. Table 3 shows that most communities (61.2%) displayed a fum ability and willingness to select their own CDDs. But it appears as though not many communities possess a comprehensive understanding of the nature of community ownership and responsibility in the context of CDTI. In analysing community participation in this regard, we excluded the four instances where the village chief (VC) made unilateral selections of CDDs. ln our view, column 5 of Table 3 does not provide a good index of c ommunity p articip ation. A degree of confusion exists over the status of Ekita with regard to treatment. The village leader alleged that this community did not receive treatment under CDTI because the CDD did not reach that village to distribute ivermectin. Ekita and Babensi II have one CDD in common, who contests the chiefs allegations. There are, however, indications that Ekita had been treated under the Community-Based Rehabilitation APOC Monitoring in Southwest Province, Cameroon - 7th August - 8e ber,1999 16 h.ograrru.e (CBR), a situation that urakes it inrperative to streatnline the distribution activities of APOC atrd other agencies working iLr the salne area. Table 3 Distribution of t onxnunrtles bv Mode Selection of CDDs ln combination with other research instruments, Table 3 indicates that generally, the communities selected their CDDs following locally identffied criteria. The focus group discussions and the results of in-depth interviews with village leaders in both categories of villages support this view. In Talangaye and Babensi II, however, the health worker associated with the programme was said to have appointed the CDDs. This situation zuggests a need to reinforce reorientation to the CDTI approach, especially t* 1' i- I L r L I Village Type VM VE VC HW VHC Vconr Other I Bai Foe A 2 Bakingrli A 3 Kokobuma A 4 Mrle 2 A 5 Mboka A 6.NewTown B'mbr A 7 Afrikpwabr B 8 Babensr 1 B 9 Babensr II B 10 Bai B I 1. Bat B 12. Bai Muassa B 13 Bar Mutoko B 14 Bar P B 15. Bakundu Foe B 16. Betock B lT CassavaFarm B 18.D'scha Beach B l9.D'scha Central B 20 B 21 Ediango B 22 B 23.Ektta B No Treatment 24.\s B 25.Kotto Barombi B 26.Kuke Mbomo B 27 sum B 28 B 29.Talarrgaye B 30.Towe B 3l.Unity Quarters B Monitoring Southwest Province, Cameroon 1 APO C 7th ln gtL 999 I t7 irr these conununities and to emphasise it in the leorieutation and retrairiing rvorkshops of the liealth rvorkers and supen,isors l'able -1 shorvs that rnany communities eltose the trtonth of distritlntion rvhen that uas nrost evidentll'not the case. 1he use rnultiple instrumentshelped identify tliis. the 1l11l lull el in B The above results refer to Category B villages and are misleading when taken on their own. As the in-depth interviews with health personnel and village leaders as well as the focus group discussions indicate, the comrnunities did not take decisions concerning the month of distribution of ivermectin because distribution followed the introduction of the programme in a swift sequence. If the communities took such decisions at all, they APOC Monitoring in Southwest Province, Cameroon - 7th - gth er, 1999 Village Meeting Village Elders Village Chief Healtlr Worker Village Herlth Comm. Village Comnr. Other-s L Afiikpwabi Village 2. Babensi I 3. Babensi II 4. Bai Longe 5. Bai Manya 6. Bai Muassa 7. Bai Mutoko 8. Bai Panya 9. Bakuridu Foe l0 Betock I l. CassavaFarn-r 12 Debundscha Beach 13. Debundscha Central 14. Ebanga 15. Ediango 16. Ekenge 17. Ekita No Treatment 18. Isongo 19. Kotto Barombi 20. Kuke Mbomo 21. Lumpzum 22. Nionii 23. Talangaye 24. Towe 25. Unrty Quarters l8 \vere rnostly overtakerl by the time rvhen the dmg rvas in fact available. Table 5 sliows that this trerd erists anrorlg Category A villages as well. Table 5 Decision Makiug I)'ocesses at the Courrnunity Level in Respect of the Time. fDi ^in Cate A VilI l=\IllageMeeting; 2=\4llage Elders; 3=\'illage Chief; 4=Health Worker; s=Vi]lage Health committee; 6=Village Committee; 7=Other According to the table, the health worker took the decision conceming the time of distribution 66.7% of the time, that is in four out of the six category A villages, although tliis might not be cornpletely accurate. What the respective community-level interviewees arrd discussants might have perceived as the health worker's decision might actually have its explanation in the availability of drugs. Once the drugs were available, the health workers informed the communities and distribution followed. This was then taken as meaning the distributisn ssmmenced at the instance of the health worker. [n five out of six cases (83.3oh),the mode of distribution was either determined by the health worker or by other means. An expatiation of the category of "Other" in the table reveals that, for the most paft, the mode of distdbution was perceived as having been determined at the CDD training workshop. In this regard, the CDD returned from the training workshop and more or less informed the community that the recommended mode of distribution was such and such. The table captures this situation succinctly well. ln many communities, the house to house method was complemented with a central place approach, usually the chiefls palace or compound as in Mile 2 oflimbe. Table 6 displays communities by their preferred mode of drug distribution. APOC Monitoring in Southwest Province, Cameroon - 7th August - 8e September, 1999 I t_ I L t i Mode MonthTime 4 5 6 7 I 2 J 4 5 6 7 I 2 J 4 5 6 7Village I 2 J I Bai Foe 2 Bakingili 3. Kokobuma I Mile 2 5 Mboka 6 Nerv Torvu Barontbt I lq l'able 6 Decision Making Process in Respeot of Mode of Distribution lllbeqI Bar Foe** VM \T VC HW VHC Vcorn Othcrs 2 Bakrngrh** 5 NLle 2*x 6 NewTorvn Barornbr ** 7 Afnkprvabr 8 Babensr 1 9 Babensr ll l0 Bai Longe ll Bar Manya 12 Bar Muassa l-l Bar Mutoko ll Bar Panya 15 Bakundu Foe 16 Betock 17 Cassava Farm l8 Debundscha Beach I9 Debundscha Central 20 Ebanga 2l Edianeo 22Ekenge 23 El<rta No Treatment 24 Isongo 25 Kotto Barombr 26 Kuke Mbomo 27 Lumpsum 28 Njonji 29 Talangaye 30.Towe 3l.Unity Quarters 3 Kokoburna** I N{boka** iThrs table rs develcped from the Vrllage Leadcr lostrument, ++Thesc re the srx Calegcry A vrllages The fust five villages in the list are Category A villages while the rest are Category B villages. In 35.5o/o of the cases, the health worker determined the mode of distribution for the communities. Community participation was found in just 38.7%o of the cases. The explanation for this trend lies in two factors. Firstly, the dfficulties associated with training and supervision led to the circumvention of some procedures. Secondly, a consensus on mode of distribution was developed at the CDD training workshops and this consensus was interpreted by village leaders as being a decision taken by the health worker. The category "Otlers" in the last column of Table 6 frequently APOC Monitoring in Southwest Province, Cameroon - 7th Au - gth 1999 20 refers to a decision an'ived at during the CDD trailing rvhich rvas tltett handed otl dorvtt to the village witliout the latter's genuitte consideration. 'Ihe conttntttlities sirr4rly did not Save sufficierrt infbrrnation conceming their role in the CD'II plocess to elnpo\\'er therl to take these decisious ott their own behalf. 3.2 Accuracy and Compatibility of Records A total of 532 people rvere ilteruiewed fiorn I l0 households in 6 Category A villages during the independent rnonitoring survey. The sampled households were distributed among the Category A villages in the following order. Bai Foe ( t 7), Bakingili (19), Kokobuma (21), Mboka (16), Mile 2 (20), and New Torvn Baronbi (17). Tlree hundred and twelve people (59%) had received treatment while ?.20 (41%) had not been treated by the time of the survey. Table 7 displays the cornparison of rates calculated fiom the household survey and those monitored from the treatment registers. A close svzrrrirration of this table shows that the household data represent nearly a fourth (72 9%) of the total registered population, brokert dorvn as follows: Bai Foe (28.1%), Bakirgili (20.goh), Kokobuma (25.2%), Mboka (68.3%), Mile 2 (12.2%), and New Town Barombi (36.5%). The respective proportions render the findings of the investigation representative, particularly rvhen cognisance is taken of the probabilistic methods used in sampling the households during the household survey. Treatment coverage from the household survey was compared with treatment coverage obtained from the village treatment registers as shown in the table. There were discrepancies in the results' ln some villages, census was improperly or incompletely conducted. ln New Town Barombi, for instance, the population aged under 5 years represented only 0.9% of the total population, meaning that there were just two children in population of 208. In Bai Foe, the register indicated that there were only 5 children in a population of 305, or 1.6%. In Bai Longe and Betock, both Category B villages, the proportion aged under 5 was O%. These proportions are ostensibly unrealistic as they de& the known demographic pattems of Cameroon, a country with a generally high fertility rate in the rural areas. High levels of completed fertility imply a large under 5 population and an accentuation of the dependency burden, as has been documented for Cameroon. Monitoring Southwest Province, CameroonAPOC 7th ln giL tem,ber, 1999 a Table 7' Treatnrent umnlarv bv Villases * Th ese re the sx Category A l'r1lages, wh:le vrllages 7 to 31 are category B ullages The mean national propofiion of the population aged less than 5 years is given as 18%. This statistic casts doubt ontheveracity ofthe Southwest provincial census carried out by the CDDs. In Bai Longe and Betock, the monitors could see little children loitering about, playing football, and being sent on errands by their parents. Quite obviously, such children and infants had not been registered for a variety ofreasons. In- depth interviews revealed that in such communities, scepticism over the programme's political innocence led many parents to prevent their children from being regisered. They did not want any evil that might befall them througj mecttzan treatment to affect their children. APOC Monitoring in Southwest Province, Cameroon - 7th August - 8e September, 1999 Village Total Population No. Treated Coverage Rate Refusal Rate Absenteeism Rate HS Rec HS Rec HS Rec HS Rec HS Rec I Bar Foe* 86 305 51 86 62 70h 28 l" 19 8% | 30,'o t0 5% 70 5% I Bakurgrli* 80 382 13 313 t6 2% 81 90 38% ?3 8% 3 Kokobuura* 111 152 53 t62 46 4% 15 89/o 28.t% 46 9y, 10% 4% -1 Mbokax 5 lvtule 2* 69 t07 101 60 '12 86 9o/o 11 2'h t4% s9% 0% 69% 875 90 697 84 10h 79 60 I90 3 l,/o 93% 58% 6 N T Barombr* 76 208 12 199 55 2"/r 95 6'/0 l0 5o/o 0% 1a .10/;La/o 2 9o/o 7 Afrikprvabr 10 3l 14 20 35 l9/o 40 0% 8 Babensi I 3t6 ttl 31 0"4 13 204 9 Babensi Il 203 80 39 4o/o 13 8y, 0% 10 Bai Longe 12 26 6t 9yo 0% 42% 1l Bai N{arrya Regrsters were either unavarlable or inaccesstble 12. Bar Muassa 280 69 24 6% 33 9'/o 15.1% 13 Bar Mutoko Regrsters were ettlier unavailable or tnaccessible l-l Bar Panya 341 178 52 to NA 12.0% i5 Bakundu Foe 587 351 60 8yo 0 1o/o 39.0% 16. Betock 89 56 62 9% 22 4o/o t4 6% 17 CassavaFann 3426 2834 82 1% 9 lYo r.3% 18 D'scha Beach Regrsters rvere ertlter unavailable or inaccessible 19 D'scha Cent'l 1907 1 573 82 4o/o 31% r.6% 20 Ebangp 232 116 50 0% 24 1' tt 2% 2l Ediango 134 .16 31 3% 35 8o/o tt.9% 22.Ekenge 163 81 19.6% t9 6% 8.5% 23 Ekrta No Treatment 24.lsongo 296 160 54.lyo 0% 36.5% 25.Kotto B'br t1t1 922 82.5yo t0% 61% 26 Kuke Mbomo 510 311 65 0% 15 9% 0.8% 2T.Lumpsum 1041 I 046 99 9% 0% 0% 0% 0% 28 Nionii 416 40'7 9r.3% 83% 22 4% 29.Talangaye 315 103 32.6% 98% 43.8% 30.Towe 2014 1318 68 4Y" tr 2% 9.3% 3l.Unity Q'ters I 189 l 009 84 8yo 06% 0.s% 22 The uuusually high coverage rates repofted in Lrunpsum (99.996) and Njonji (91 396) were obtained because, iu these conununities, only those who rvere treated rvere registered. In other rvords, registratiott was not preceded by adequate sensitisatiott and rnany people had not registered during the census. Subsequently, treatntent followed the line already traced by the census Because those who had initially registered tvere ahvays better disposed or predisposed to treatrnent, coverage then appeared to be nearly total. We carried out no household assays in these communities because tltey rvere Category B r,illages, but a careflll exarnination of the treatruent registers showed this rvas the case. Our vierv is that the coverage rates recorded ilr many villages were liigher than *'hat rvas real. The recorded coverage rates in villages like Bakilgili (81.9%) and figures calculated fiorn the household suruey (16.2%) show a wide disparity. The figure in Bakingili is suspected to be the result of inflatiorr, deliberate or inadverlent. ln this community, 35o/o of the people surveyed reported that they did not receive treatment because they were not inforrned or aware that distribution would take place. In Bai Foe, tlre household sun,ey recorded a higher coverage figure (62.7%) than the treatment register (28.1%) because a significant proportion of people were not registered prior to distribution but nonetheless went on to receive the drugs laler without registratron. That is, the denominator remained fairly constant while the numerator rose. By the time of monitoring, the proportion treated had increased because the monitoring instrument was sensitive to changes in both the population regisered and the population treated. The situation indicates a strong need to update the register in communities like this. Thus, from the general pattenxs of disparity follows one conclusion. The nature of census in some communities was a function of the nature of the emphasis laid and received on that aspect of the CDTI process during training. Conflicting figures may then show over-estimations or under-estimations in the treatment registers depending on the form that the registration process took. It would be needful for trainee CDDs to engage in field practice of registration during training workshops, with variations on the theme to cover different possible scenarios that could produce errors in the registers. Correspondingly, effective strategies should be worked out to address absenteeism and refusals, as these are factors that reduce treatment coverage 28.60/o and 27.7% of the time, respectively. This should be in the form of practical steps to be L l. t ! Monitoring Southwest Province, CameroonAPO C 7th IN gth 1999 )3 fbllorved by CDDs iu situation rvhere these factors could cripple progranul)e objectives T'able 8 displays the most comlnoll reasons that respoudertts gave fbl not receiving treatrnent at the tinre o1 distribution. l'able 8. Reasons for No Treatment, All Comnlrnities TOTAL The table reveals that nearly half as many people as rvere interviewed during the Irousehold sulvey (41.3%) had not been treated. The reporled absenteeism atnong the nor-treated population of 28.6oh is evidence that some of the constraints highlighted previously and elaborated in Section 3.3 lrilitated against repeat visits iu many communities. It is illustrative to observe that some of the villages ale sparsely populated line villages of 40 houses or less. It is therefore dfficult to envisage horv abserteeisnr could be this high rnore than two months after distribution had taken place. The level of refusal is high. Nearly 3 out of every 10 people sampled (27.7%) had refused treatment. This high refusal rate may be linked to the absence of a strong and committed sensitisation and mobilisation effort. In a related issue, therefore, the awareness rate was found to be low also. More than a tenth of the people (12.7%) were not informed of the distribution exercise or programme. There are indications that mobilisation and sensitisation, which are key gemponents of a sustainable structure, had been rushed. The monitoring team also found a conspicuously low use of information, education, and communication (IEC) materials. We consider that the high level of refusals and absenteeism could be substantially reduced through a virile sensitisation effoft. We are persuaded that rushed sensitisation and the dearlh of IEC materials are key factors in explaining the refusal rate of 27.7o/o, absenteeism of 28.6%o and a non-awareness rate of 12.7Yo. APOC Monitoring in Southwest Province, Cameroon - 7th - g'h ,1999 Reason for No Treatment Number Percent (%o) I Child under 5 veals 28 11 10/tr'- / /o 28,:6Yp2 Absent 63 : ItqpSrl! -l Refusal t2 Gh.Etfe*,tuP 6l 28 1'1 10/LI,I /O 11 10/tL. t /o5. Not hrformed 6. Sick 8 3 6Yo 7 Others 20 9t% 220 100o/o a 21 A11 the CDDs itrterviewed expressed a cletermirtation to contitrue firnctiotrirtg as distributors in the next rourd of distributions. A degree of disappoitrtnrent rvas expressed rvith the comrnunities' attitude toward incentive for tlie CDDs, but this has apparently not been strong enough to dissuade the CDDs fiom future participation in the progranlme. The retraining of CDDs is irnperative, as is the orgauisatior of multi-sectoral seusitisation meetings with the conununities in order to build a corlmon appreciation of the irnplication5 ef sonrrnunity ownership of, and responsibility for, the CD'l'l prograrnme If this is not done, cornmunities may continue to perceive the control progtamrne as an extemal initiative that does not require theil effective participation. This might adversely affect the chances of sustainability. 3.3 Constraints Inadequate fttnds at the provincial and lower levels complicated the supelvision process. The Provincial Co-ordinator observed that the only amount in the project account at the provincial level was the 100,000 Francs used in opening it. Funds are reportedly transported in cash from the national to the provincial level and they are said to sometimes arrive late and may be inadequate for implementation. We note that a potential danger exists in this form of transaction as funds may be stolen in transit from the national to the lower levels. The inadequacy of funds for zupervision at the provincial level is indicated by the somments of a health worker at the Manyemen Hospital, who reported that they received 10,000 Francs for supervision. ln describing this amount as "nothing", this health worker was drawing attention to its gross inadequacy when cognisance is taken of the scope and centrality of zupervision to the CDTI process. The poor rural road infrastructure in some areas, particularly the Nguti Health Area, does not support effective or regular zupervision of the implementation process in some villages. Some villages in Nguti health area did not receive treatment because they were allegedly inaccessible. These were the villages covered by the St. John of God Hospital, Nguti. A combination of internal misunderstandings and the dfficulty of co- ordinating mectizan distribution in the context of the newly created health district a a- t, Monitoring South' Province, CamAPO C 7th ln 8A west ) 1999 eroon t 25 contributed to the co\/erage level reported Perhaps if the conununities hacl leceivcd adequate sensitisation and nrobilisation on their role would have gone to collect the dnrg and get treatrnent, but this rvould presuppose that the institutiorral constlaints between the Hospital and the districl health office are resolvecl The District Medical Officer' (DMO) of Ktrnrba, Dr Victor Afanrvi Ndiforchu, refered to "the tirnidity of liealth stafF', u,hich for hirn oonstituted thcir inability to superuise effectively due to poor intellectual arrd furancial motivation IIe advised that health rvorkers be trained rrore intensively and exlensively, and be adequately renlunerated ilr order for thenr to perfonl their hrnctions creditably,. On training, he noted one could not supervise auotlter person, if the superuisor does not possess the intellectual security that emanates fi'orn superior knorvledge. For the District Medical Officer of Nguti, Dr Morku Fo{n, on the other hand, "It [CDTI] is a nerv prograilrme so it has problerns. The comrrrunities are ltot cottvinced." This lack of "cottviction" has many underlying political, cultural, and econornic causes and gives rise to an equally complex varietv of behavioural pattenrs that could militate against sustainability. The lack of adequate ittcentive for CDDs rvas identified as constrailt that is related to the perception of CDTI by urany comnrunities as a goverxrlent or WHO prograrnrne. They thus t'eason or exlect that the CDDs were either rernunerated by the govemment or should be. At Limbe, the DMO, Dr. Matilda Akoh-An'ey informed us that 'the delay between consignments led to laxity on the part of some CDDs because they then got involved in other activities and were slow to collect their supplies". She drerv attention to the inadequacy of funding provided for CDD training. In her view, the effective training of CDDs lies at the core of the CDTI implementation process, and greater effort should be put into it. The work of supervising CDDs was found to be time-consuming and tedious. Supervision takes place after official hours because CDDs are constrained to distribute in the evenings uzually under very difficult and hilly terrain. The dfficulty of supervision is complicated further low staff strength at the district levels. At Nguti, the district health office had only three personnel, including the DMO. The process of integration of primary health activities has therefore involved a reduction of the amount of time that the health workers can devote during official hours to the control of APOC Monitoring in Southwest Province, Cameroon - 7th - gth ',1999 )6 onchocerciasis It is in this respect especialll, that the poirtt rvas repeatedly rnade tltat health workers involved in superuision should adequately rnotivated for that pulpose. I'he generally low rate of registration rvas found to be partly attributable to the lulnorlrs that qrread at the time of census to the effect that the ruling party was behind the distribution proerarnnle. The runours specified that the govemment was collecting llames to "sell to secret societies" that rvould destroy them, and that the drug was an altemative instmment of the ruling pafty to either steri]ise or kill them. Fears such as these constitute sensitisation and mobilisation into crucial priorities before the next distribution exercise Tlre perceived link between the prograrmre and party politics must be broken as a matter of urgency. In communities like Gardens and Bende, progress was hildered by the suspicion of quarter heads that money meant to accrue to them was being rvithheld. These conmnurities did not receive treatment as a result. Urgent sensitisation is required. Among the expatriate comrnunities, where it is feared that the registration was a ploy meant to fish out those persons who did not have valid resident pennits, the use of trusted representatives would be a step in the right direction. One Chief of Post wryly noted that 'the programme is very boring. The entire population doesr't like the programme and need to be educated. The duration of the training was too short and not everybody understands things at the same speed especially as it is a newly introduced idea". This calls for an extension of the number of days set aside for the various training workshops, and an intensification of the contents of training. On the whole, the constraints faced in the implementation of CDTI can be broadly summarised into geo-physical, political, logistic, traininglknowledge-level and other factors. The topography and climate of the Southwest render most of the roads impassable during the rainy season. This reduces the effectiveness or frequency of supervision. The hilly terrain also probably discourages repeat visits among improperly motivated CDDs. This could contribute to the high absenteeism rate more than two months after distribution. The volatile political situation promotes a view among the populace of ivermectin as a drug produced or procrued by the ruling party to either sterilise e1 sliminate the opposition. This reduced the number of persons who regisered and who got treated thereby distorting the reality of coverage in the treatment registers. APOC Monitoring in Southwest Province, Cameroon - 7th August - 8e September, 1999 , I t t i t a_ { 27 'lhe absence of integt'ation between CDTI and other distribution l)rogrammes could lead to duplication of effofts with tragic lesults if sottre people at'e overdosed. There is the allegatiot that mectizan is fi'eely sold in the open market. Such logistic and adrlinistrative issues should be adclressed institutional with nrinimurn delay. The quality of training and conununity sensitisation rvas uot adequate and treeds to be strettgtherred. l'he comrmrnities do not seeur to larow rvhat is expectecl of tlierrt and CDDs are approaching registration rvithout a cornprehertsive idea of its centrality to the entire CDTI process. The corulurrities in particular require in-depth sensitisation and rnobilisation to the duties and obligation s that are expected of thern in the CDTI process 3.4 Sustainability The prospects of sustaining the CDTI programme in the Southwest h'ovince of Caureroon al'e vely good. Health personlel iltervierved were optimistic and regarded the problems currently being faced as teething problems. Along this line of expectation, the DMO of Nguti reasoned that 'With sensitisation and mobilisatiorr, I thilk they [the communities] will take mectizan". Nevertheless, the high refusal rate due to tlie fear of side effects needs to be addressed urgeutly, for better coverage in the Iirture. Many courmunity mernbers dread the side effects of the drug and probably do not believe that the real benefits outweigh them. The Government Delegate of Kurnba Town alleged that 3 people died after taking ivermectin and 15 people (including his own relative) were hospitalised. He threatened to stop the programme in his area. There is a need for advocacy visits and meetings to leaders like this who are not so favourably disposed to the distribution programme. Their negative intervention could jeopardise programme sustainability if they mobilised the sentiments of their constifuencies and other collaborators. As an encounter with youth at Nguti revealed to us, 11 i5 important to adopt a multi-sectoral approach to both mobilisation and sensitisation that extends beyond the formal leadership structures of the province. When a group of youths in this area observed our presence, they began to speak loudly among themselves, stating repeatedly that "These mectizan people have come again. They are the ones giving out drugs which kjll people". We invited the more vociferous ones among them to discuss details of their APOC Monitoring in Southwest Province, Cameroon - 7th August - 8th September, 1999 asserlions We souglrt to know horv many had been treated ar)otlg the 9 of theln gathered arouncl us. There rvere 4 in the group who had received treatlnent. WlteIt r.ve specifically asked the one who appeared to be the opiniol leader among them if he had received treatment, he resporrded this way: I dey craze? I dey see as all people jus' dey die like forvl, I go come take the rnectizan? Na stupidpeople de1' 61o that kin' of thing' rnassa As I dey so, nothitrg fit make me take tnectizan. I dey craze? lf this statement and the attitude of the youth we spoke rvith are representative of this segment of the Nguti population, then the level of sensitisatioil was low in this community. It rnight be traced to the acrirnonious relations betweett the district health office and the St. Jolm of God Hospital, as well as the intemal misunderstanding that the latter establishment rvas having right up to the distribution period. However, because Nguti where this interesting encourter took place, was not one of our sampled rnonitoring sites (see Table 1). we gave impromptu health education and came away with the irnpression that the education needs of these populations may not be as enormous as one might suppose them to be. By the time we left Nguti for the first sites in the district, the same youth opinion leader was saying, "If dem dey tell us everything like dis, we for know". The need for health education, sensitisation and rnobilisation in the Southwest Provilce can sirnply not be overstated. In response to a question conceflling if he will accept treatment il the next round, he replied " Yes, if I know everything they are doing". 3.5 Community Perception It is expected that a community will sustain the distribution of ivermectin in the post-APOC period if it currently perceives the programme as its own. If a community persists in expecting that incentives or remuneration for CDDs should be generated externally, or are indeed so generated, then prospects of sustainability are to that extent compromised. In general terms, few communities in the province have made an effective transition to a perception of community ownership of the programme. The amount of responsibility they accept for programme implementation is minimal. They frequently thought that they programme was either a goveflrment programme or a World Health 1 , t t- I L I I Monitoring Southwest Province, CameroonAPO C 7th ln gth 1999 i2() Orgarrisation (WHO) initiative. Sensitisation and orientatiolt rneetittgs Ireed to be iutensified ancl in a waythat will ellsure that the requirecl awareness is ntore broad-based. Such sensitisation nreetiugs should not be rushed its tlre cttrrent ones wct'e None of the CDDs received any supporl fi'om the corntnunities during their rvork as CDDs. Sustainability of CDTI rvill depend ou the acceptance of cotttmuttity mentbers to coritinue taking ivermectin aud the rvillinguess of CDDs to contirtue distributing the drug to other community members. In line rvith this, au ilrcrease in the percetttage of tlte aurount meant for CDDs, obtained through the cost recovery programme, is recomnrended. 3.6 Quality of Implementrtion of CDTI hrteryiews with individuals invoh,ed in CDTI at different levels indicate that the training received by the various cadres of CDTI operatives was inadequate, especially with regard to issues of community reqronsibility, distribution plocesses, measurement and eligibility cliteria. Accordirg to Dr. Matilda Alioh-Arrey, District Medical Officer (DMO) of Limbe, funding for CDD training was grossly inadequate and is a cardinal factor militating against CDTI implementation in this district as demonstrated by the poor quality of record keeping. There is need to lay more emphasis on the relevance of good record keeping. Speaking on the quality of training received this year, Dr. Morku Forju, DMO of Nguti, expressed her disappointment in the following words: We all need to be retrained, from my level down, because the initial training we all received was inadequate and rushed. It compressed into other activities we had and we could not easily absorb the contents within the available time. We certainly have to be retrained. In the same vein, Dr. Ndiforchu Afanwi Victor, the DMO of Kumba, observed that'the poor quality of training led to poor quality of supervision". ln zurL therefore, the quality of implementation of CDTI was poor and rushed. There was inadequate sensitisation of the communities on their role in CDTI and APOC Monitoring in Southwest Province, Cameroon - 7th - gth , l99g .t i) ownership of the progralnr.rle by the community menrbers was ilot enrphasised. The trairring of CDDs rvas i:radequate and it ranged fi'onr l-2 days as opposed to the recornmended 3 days. This explails the very l)oor state of record keeping and the high rates of absertteeisltt 3.7 Unique I.eatures of Project Area Nine people allegedly died after taking ivermectiu in the 3 health districts where rnonitoring rvas done. According to the Govemrnent Delegate of Kuniba Town, 3 were from Kumba, 3 fi'om Nguti, and 2 from Batoke. One person was repolled to have entered coma four days after taking iverrnectin and died at the Manyeuen Hospital. Co-infection rates of loa loa and onchocerciasis in this area need to be established as rvell as the other factors that are contributing to this high death toll. 4 DISCUSSION AND CONCLUSIONS The CDTI programme is being implernented for the fust tirne il the Southwest h'ovince of Cameroon, in 1999. Before now, disttibution had been taking place through a variety of strategies, prorninslt among which was the Community-Based approach. In addition, 1[s Qornmunity-Based Rehabilitation Programme (CBR) has been distributing in the province for several years. The parallel nature of disribution in this province especially makes it imperative to streamline the activities of these distribution programmes in order to enhance the benefits accruing to the communities from the exerclse. The CDTI approach is new and is therefore associated with several problems that can be overcome with proper orientation of the communities and the right emphases in the training of CDDs and their supervisors. For example, no community decided on the period of treatment, but underlook distribution on the basis of when the drug was available. There is thus a need to orient drug provision systems more suitably to periods that the communities dslslmine as convenient to them- The communities must accordingly be guided to take decisions in this regard. Less than half of the communities .l a Monitoring Southwest Province, CameroonAPOC 7th ln gth 1999 t f I al1 rroritored (459,o) took clecisions on the rnethod of treatnrent. but appear to ltave left this cmoial clecision to either the CDD or the healtlr rvorker. This irnplies a loss of the entpowenlletrt that the CDTI strategy rvas desigued to prourote. h tlre nexl round of serisitisation. talks rvith corulunities these elernents of the CDTI process shotlld be ernpha sised. An elcouraging propoftion of cornrnunities (61.27") rlecided on the criteria fbr CDD selectiol, but there is roorn for improvernent. A close look at Tables 2-6 provides irrdications as to rvhich comrnunities requile stretgtheriing in these respects. Sensitisation and rnobilisation should be seen urrder CDTI as continuous processes that are irterwoven rvith the superuisory activities of the health workers. In this way, the iiealth tvorkers *,ould have a personal schedule of the rurique featut'es within their' respective jurisrliction that they rnust pay close attention to. Hopeflilly, with the right eurphasis, the next rourd of rnobilisation meetings will produce a higher proportion of courtrunities that took decisions on the criteria for selecting CDDs. The low propofiion of refusals fwo months after distributiorr (9.6%) is not corr.oborated by our rnonitoring experience. The level of scepticisrn, doubt and pessimism is generally high as a result of a range of factors previously highlighted in Section 3.3. Once again, a proper emphasis on sensitisation and mobilisation for community participation is essential for reducing the percentage of refusals in the next round of distributions. Nearly all villages (96.1%) have been treated. However, attention must be drawn to the cluster of communities in Nguti that were the responsibility of the St. Johl of God Hospital that were not treated owing to transportation, administrative and institutional problems affecting the hospital and the district health office at the time of distribution. The treatment registers for this group of villages have since been returned unfilled to the district health office. There is need to integrate the control operations of the hospital and the district health office's in the interests of the communities. Those communities that have not been treated require urgent attention, as this implies that the CDTI process has not been implemented in them for now. No CDD has been changed since selection in a1l the surveyed locations, although h a couple of cases, it was reported that not all the selected CDDs have been active. The APOC Monitoring in Southwest Province, Cameroon - 7th August - 8th ',1999 I 3? CDD in Mboka, a Categorv A village, died a ferv weeks after distr-ibution and no replacement had been ntade at the tirne of the monitoring team's visit Tlte contruttity has not addressed the need to replace him, possibly because they do not appreciate that CDDs nray still be uselirl in terms of health edttcation, even outside distribution periods. h74.Zoh of the commrutities ilvestigated, a health rvorker sr4lervised the CDD, but this mealls t|at i1 more thau a quafter of the conrnunities (25.8%), there was lto superuisiol. In the cases where superwision was can'ied out, the indications are that it rvas not effective Similarly, 96.7o/o of the comrnunities are reported as leceiving health educatiotr but because training was problematic, the health education received was also ruot adequate It rvas not suflicient to allay the fears of the community rnetnbers, nor was the scope wide enough to cover the entile populations. Accordingly, with regard to input indicators, 96.7% of the villages have trained CDDs but there is need for retraining. Drug supplies were not tirned to the periods determirred by the communities. There is need to address this issue in subsequent distributions because it is one strong way of demonstrating to the communities that the CDTI programnle belongs to them. Not enough comrnunities collected their drugs from a central poi1t. Only 58.1% did so. This naturally suggests that the perception of their respolsibility for the CDTI process is low. The transition to the new system does not seem to have been successfully or effectively made. We did fild, however, that all the CDDs had measuring sticks, although we did not think it was necessary for these sticks to be retrieved from them after each distribution to the central point along with the balance oftablets. We found as summary forms at the villages, and learnt that the CDDs were not the ones who filled in these summary forms because it was considered that they would not do a proper job of it. Our view is that proper training in fi.lling the summary forms would empower CDDs to do so competently. The lack of confidence in the ability of communities to handle this aspect of CDTI negates some fundamental principles of the CDTI approach and calls for a reorientation on the part of programme managers. The correrstone of the CDTI approach is that, through their CDDs, communities can gather, collate, summarise and interpret information about coverage in order for them (the communities) to sustain treatment. Communities are believed capable of managing a t , t l,' t ! Monitoring ln Southwest Province, CameroonAPO C 7th August gth er, 1999 I l ll3 tleir oy,r iverrnectin clistributior with facilitative supporl fi'onr tlie health set-r'ices and that they achieverl equal. or better, coverage results than distribution activities tltat were certrallv orgalisecl by clistrict health services. CDTI thus presents tlte best prospects for sustainable cornrnulity level onchocerciasis itttetvetltiott 'fhese underlyirtg colsiderations a1'e being undermiled by the usurpation of theil role in these areas, whiclt often takes place as acts of necessity. ln other words. rvlten exlt'aneotls (e.g. NGDO) deadliles are superitnposed on the normal routine of the reporling process, it underlrines capacity building if the situation ilvolves the trausfer of responsibilitl' for sumnrarisation to already over-laboured health rvorkers. The solution lies in traitring and a recogrrition that the CDTI plocess, properly understood, must go beyond a rvholesome degree of facilitative ilvolvement from the health seruices to include a built-in respect for the normal patterls and rhythm of comnrunity life. The communities or.rght not to be pushed to meet exlraneous deadliues because the distribution had ttot beett sensitive to the period they deterlrined in the first place. Such a situation would redttce the credibility with rvhich the con-rnunities may be persuaded to accept the proposition that the progratnnte belongs to them. 5 RECOMN{ENDATIONS d Timely and adequate transfer of funds from the national to the provinciallevels should be facilitated following proper financialand lower procedures, More firnds should be eannarked for community mobilisation and sensitisation to allay fears of side reactions and create a sense of ownership of the proglamme, which is currently being perceived as World Health Organisation/Government pro gramme. . J. To motivate CDDs and reduce running costs of the programme incurred by them, the amount of money reserved for CDDs under the cost recovery programme should be increased. APOC 7th tn s gth €rr 1999 Monitoring outhwest Province, Cameroon 434 Those contnrt. u.ities il Nguti that have not been tleatecl leqtril'e urgellt affention, as this itrplies that the CDTI process has not beelr inlplemetrted within tltern for now. Health education and training fi'onr the district to lorver levels including re/trainitrg of CDDs shottld be intensified before the secortd roruld of treatment Census should be identified as a cardinal colltponeut of CDTI implementation and addressed as such during training r'vorkshops. Village registers should be updated so that all village members including children are registered for proper calculatioli of treatnletlt coverage. CDDs should be empowered to filIthe village surxrnaly fbrrns in line with APOC guidelines. Simple analgesics should be tnade available at the conur.tunity level, for the treatment of minor reactions as a means of reducilg the refusal rate. The conspicuously low use of information, education, and communication (EC) materials should be improved upon to reduce refusal rate and comrnunity misp erception of pro gramm e obj ectives. Sensitisation and mobilisation should be seen trnder CDTI as continuous processes that are interwoven with the supervisory activities of the health workers. 5 6. 7 8 8 ,/1 /"-v6 10 -/. --..\ There is nee6l to streamline ploglammes health office. activities of all parallel distribution with the operations of the district I a t t APOC Monitoring in Southwest Province, Cameroon - 7th August - 8e ,1999 t i l t L t, 1 I t 1