Assessment of the effectiveness of different delivery channels for rapid monitoring of treatment coverage using school children as proxy

0ASSESSMENT OF THE EFFEGTIVENESS OF DIFFERENT DELIVERY GHANNELS FOR RAPID MONITOR!ilG OF TREATMENT GOVERAGE USIiIG SGHOOL GHILDREN AS PROXY AFRICAN PROGRAMME ON ONCHOCERCIASIS CONTROL oUAGADOUGOU, BURKINA FASO JUNE 2OO4 ( ACKNOWLEDGEMENT We are grateful to Dr A. Seketeli, Dr. Uche Amazigo, Dr. Hans Remme, Dr Richard Ndyomugenyi for their technical input and personal commitment to the successful conduct of this study. We are also grateful to our research assistants who helped with the collection and organization of data. To the health personnel, the policy makers of the primary education board of the study areas in Enugu and Kaduna State, Nigeria as well as the transporters, who assisted with the delivery of the forms to schools in Sudan. To the CDDs and members of the study communities, we say thank you. We cannot thank the coordinators enough for their support and encouragement, which saw us through the study. I 2ACRONYMS APOC African Programme on Onchocerciasis Control CDD Community Directed Distributors CDDCOV Coverage in Community Directed Distributor's Register CDTI Community Directed Treatment with Ivermectin HHCOV Household Coverage LGA Local Government Area OCP Onchocerciasi s Control Programme STANDSCO Standardized School Coverage SPEB State Primary Education Board JTEAM COMPOSITION ENUGU STATE, NIGERIA tr Joseph Chukwudi Okeibunor [Principal Investigator]tr Nkechi Genevieve Onyeneho KADUNA STATE, NIGERIA tr Kayode Ogungbemi [Principal Investigator]tr Oluwatosin Adekeye tr Philip Sankwai tr Femi Daramola ABU HAMAD PROVINCE, NORTHERN SUDAN tr Nazar A. Mohamed [Principal Investigator] COORDINATORS O Adenike Abiose O Obinna Onwujekwe 4CONTENTS ACKNOWLEDGEMENT ... ACRONYMS TEAM COMPOSITION ...... CONTENTS EXECUTIVE SUMMARY l.O INTRODUCTION ,... 1. 1 RationaIe ............... 1.2 Research Questions 2.0 STUDY OBJECTIVES 3.0 STUDY AREA 3.1 Enugu State, Nigeria ............ 3.2 Abu Hamad Province, Sudan 3.3 Kaduna State, Nigeria .......... 4.0 METHODOLOGY 4.1 Study Design ..... .1 ,2 .J ,4 .5 .7 .7 .8 .9 4.2 Description of Alternative Rapid and Reliable Methods of Delivering Monitoring Tools 4.2.1 Enugu State, Nigeria ............. 4.2.2 Abu Hamad Province, Sudan 4.2.3 Kadlna State, Nigeria ..... 4.3 Study Population 4.4 Sampling and Data Collection 4.5 Methods of Data Analysis ..... 5.0 RESULTS 6.0 DISCUSSION 7.0 REFERENCES ANNEX 1: HOUSEHOLD COVERAGE SURVEY FORM 10 10 1l 11 t3 13 13 13 13 t4 t4 18 28 30 ANNEX 2: CDD RECORD FOR RAPID MONITORING STUDY ......30 31 5EXECUTIVE SUMMARY Community Directed Treatment with ivermectin in which communities design and implement treatment of its members, has been identihed as an effective strategy of meeting the principal challenge of sustained delivery of ivermectin to all endemic communities. However, the ONCHOSIM simulation model for onchocerciasis estimates that, depending on the prevailing level of endemicity, an annual ivermectin treatment coverage rate of at least 65Yo of the total population of the community, for a long period ranging from 15 to 35 years, is required to eliminate the disease as a public health problem. However, the review of the prograrnme performance data from OCP areas suggest that elimination of transmission is unlikely indicated that the goal of reducing onchocerciasis as a disease of public health in most situations, and that prograrnmes may be required for the foreseeable future. This calls for regular monitoring of the implementation of the APOC CDTI. Treatment coverage needs to be monitored regularly to identify communities with poor or insufficient coverage in order to institute prompt and appropriate interventions to improve coverage. Household treatment survey is a method, but this is very expensive in terms of the financial, human and time involvement. The review of treatment registers of CDDs, another approach to monitoring treatment has not proven very effective because of the low level of education of many CDDs. The school based monitoring method has been identified in a study in Uganda as an effective and low cost approach to monitoring community treatment. The present study was designed to validate the Uganda study and evaluate the correlation of the school based monitoring of ivermectin treatment coverage with the household treatment coverage. The study was conducted in three sites, namely Enugu and Kaduna States in Nigeria and Abu Hamad Province in Northern Sudan, using cross-sectional design to collect data from primary school pupils, households and CDD registers on ivermectin treatment coverage rates, between 5-6 weeks after the distribution of ivermectin in the study sites. The level of correlation of the methods was tested using Pearson's correlation coefficient. The school monitoring form was delivered to the schools through different media. In Enugu, it was delivered to the head teachers through the Local Education Secretary. On the other hand, the head teachers in Sudan received the forms from public transporters, while the health system was used in delivering the form to head teachers in Kaduna. The researchers trained the head teachers in Kaduna and Sudan on how to complete the form. The correlation of household survey and school-based strategy was 10%o, 49oh andT2oh in Kaduna, Enugu and Abu Hamad respectively. The combined data from the three study sites gave a correlation coefficient of 68% (p<0.0001). The Pearson correlation coefficient for household survey with CDD register, was low and not significant in Abu Hamad and Kaduna but high and significant at less than 0.0001 level in Enugu. However, the combined data revealed that the treatment coverage rates from the CDD registers correlated highly with household survey (86%). All the same the results identified the school-based monitoring provided effective, simple and a low cost approach to providing reliable data on community treatment coverage. 6No matter the channel used for the delivery of the forms to the schools, substantial success was recorded in relation to the valid return of the forms sent to the schools. One hundred percent valid return of the school forms was recorded in Sudan and Kaduna respectively, while 70%o valid return was recorded in Enugu. The periods between delivery and return of the school forms ranged from one to three weeks. It is thus recommended that the school-based monitoring approach be further validated on a large scale by programme managers and adopted for low-cost rapid monitoring of coverage with ivermectin in the different CDTI projects. 7l.O INTRODUCTION 1.1 Rationale The African Programme for Onchocerciasis Control (APOC), a health partnership, has adopted the community directed treatment with ivermectin (CDTI) as its main strategy for ensuring mass distribution of ivermectin in the Onchocerciasis endemic communities in the nineteen countries that were not covered under the Onchocerciasis control prograrnme (OCP). In this strategy, community itself designs and implements treatment of its members. This has been demonstrated in a multi country study to be effective in ensuring sustained mass treatment of Onchocerciasis endemic communities with ivermectin. The elimination of onchocerciasis as a disease of public health importance in Africa is a longer-term goal and there remain serious questions about its feasibility. The ONCHOSIM mathematical model suggests that periods of over 35 years will be required for areas with medium or greater levels of endemicity, significant heterogeneity in prevalence, and annual treatment with modest coverage of 65%o. However, there are indications from the review of the OCP areas that elimination of transmission is not likely in most situations and that prografirmes may be required for the foreseeable future (Borsboom, et al, 2003; Tielsch and Beeche ,2004). For these and other reasons, treatment coverage needs to be monitored regularly to promptly identify communities with poor or insufficient coverage in order to i nstitute timely and appropriate interventions to improve coverage. Asking randomly selected community members if they swallowed ivermectin, in periodic surveys was adopted for monitoring the CDTI performance of endemic communities and validating the reported treatment coverage. This has been effective but involves enorrnous cost in terms of personnel and logistics, which keeps it out of reach for use by projects. Another monitoring strategy, which combines simplicity, rapidity and cost efficiency, is the use of reported treatment by school children interviewed by the teachers. This was identified and demonstrated to be a low cost and reliable method in Uganda. Another Ugandan study (Ndyomugenyi and Remme 2002) and similar multi-center studies on monitoring conducted in Nigeria in 1994 used public mailing method to deliver monitoring tools and collect monitoring data from primary schools. However, mailing system in African countries appears not to be fast and reliable enough to be used as a medium for communicating and reporting ivermectin coverage data. In the light of the above, there is need to identify and assess effectiveness and efhciency of other altemative rapid and reliable delivery systems for collecting the information from schools. The study was designed to validate, in four sites, namely Cameroun, Enugu and Kaduna State in Nigeria and Sudan, the findings of the Ugandan study, which identified reported treatment of school children as a reliable alternative approach to monitoring community treatment coverage. Italso aimed atassessingthe effectiveness of differentmodes of 8questionnaire delivery and collection from schools. However, the result presented is on three sites. Data for the fourth site is not readily available. 1.2 Research Questions l. Will the Ugandan experience in rapid monitoring of treatment coverage using school children as proxy be replicated in other sites? 2. What are the other alternative rapid and reliable methods of delivering monitoring tools and collecting data on treatment coverage from school? 3. What will be the efficiency rate of various methods of delivering monitoring tools to schools and collecting data for monitoring treatment coverage? 92.0 STUDY OBJECTIVES 1. To validate the Uganda result on the relationship between treatment coverage of school children and community treatment coverage. 2. To identify and describe other alternative rapid and reliable methods of delivering monitoring tools and collecting data for treatment coverage from schools. 3. To assess the effectiveness of various methods of delivering monitoring tools to schools and collecting data for monitoring treatment coverage. 10 3.0 STUDY AREA The study took place in four sites, in three countries, namely Enugu State and Kaduna State in Nigeria, Abu Hamad Province in Northern Sudan as well as Southwest Province Cameroon. In all sites ivermectin distribution has been going on for many years and five years ago they all adopted the CDTI approach for ivermectin distribution. However on data on three sites were available for analysis and are presented below. 3.1 Enugu State, Nigeria Enugu State is located in Southeastern Nigeria, and lies between latitudes 50 55' and 70 10' North and longitudes 60 50' andJ0 55' East. It is bounded in the north by Kogi and Benue States, in the East by Ebonyi and Abia States, in the South by Ebonyi, Abia and Anambra States and in the west by Anambra state. It is in the rainforest zone. Administratively, the State is divided into seventeen (17) Local Government Areas all endemic for onchocerciasis. A ccording to the 1991 population census, the population was put at 2.7 million with a 2004 projection figure of 3,568,260. The state is predominantly rural and agrarian, with 78.3% of the population living in rural local government areas and 65Yo of the working population being farmers. The agelsex structure of the population of Enugu S tate is broadly similar to the overall pattern for Nigeria, with 50.7Yo being males and 49o/o females. The population is typically youthful with <10% under the age of 5 years. Epidemiological assessment for level of endemicity using the Rapid Assessment Method (Remo) show that out of 1,683 villages identified, 1,120 are hyper endemic for onchocerciasis, 211 are meso endemic, 349 are Hypo endemic. Only three villages had no signs of the disease. Distribution of Ivermectin in the State started in January 1995 with hyper and meso endemic LGAs being actively treated while hypo-endemic LGA was treated passively through the community based ivermectin treatment strategy (CBIT) supported by the Lions Club International Foundation. In 1998, the community-directed treatment with ivermectin (CDTI) project was established with APOC support. Udenu LGA was randomly selected for this study. This LGA is about 120 km from Enugu the State capital. Benue State bound Udenu on the North, in the east by Nsukka Urban LGA and Igbo Eze North and South in the West. Udenu LGA is made up of a number of communities each of which has at least one public primary school. Some of the communities have post primary schools. Education system in Udenu LGA, as in other parts of Enugu State and Nigeria is divided into primary, post primary and tertiary. The primary education is considered basic and every child between the ages of 5-14 is expected to be in school. In other words, enrolment into the basic primary education starts at 5 years. Though this is not compulsory, in Udenu LGA there is almost one hundred percent school enrolment at this level. 11 The schools, both public and private primary schools come under the coordination of a local education authority in the LGA headquarters in Obollo-Afor. 3.2 Abu Hamad Province, Sudan Sudan is one of the countries in Africa that is affected with Onchocerciasis. The estimated number of cases is 2 million in and 80% of cases are in Southern Sudan (war- affected area) where endemicity ranges from meso- to hyperendemic. CDTI project covers nine large areas or zones, the majority in southern Sudan, where mass distribution of ivermectin took place. Abu Hamad province was selected for this study because it is among the most stable areas and there is infrastructure for basic education, which is needed for this study. Abu Hamad province is the only area in the far northern part of Sudan that is affected with dermal onchocerciasis and it is mesoendemic. CDTI started in this area in 1998 by community organization, selection and training of CDDs and supervision and ivermectin distribution. The province is divided into three districts namely Al Kap and Sheeri, Abu Hamad and Sheraik. Abu Hamad town is the capital of the province, which is linked to the Khartoum (capital of Sudan) through a railway and unpaved tough road (500 Km). The onchocerciasis sub-office is located there where all the operations took place. The total population is estimated to be 88,429 in May 2003. The major means of communication in the province is either through telephones, which are a bit expensive compared to the standard of living or public transport, which is frequently used especially on markets days. There is no functioning postal system within the Province. 3.3 Kaduna State, Nigeria Kaduna State is located in the north central part of Nigeria. The State shares boundaries with Plateau State to the east, Nasarawa to the southeast, Niger to the west, Federal Capital Territory to the south, Katsina and Kano State to the north. The State has 23 LGAs and a projected population of 5.5 million in the 2004. The major rivers are river Kaduna, Galma, and Gurara. The vegetation is mainly Savannah grassland with pockets of forest mosaic. The climate consists of the dry season (November to March) and the wet season (April to October). Kajuru LGA, which is hlper-endemic for onchocerciasis was randomly selected for the study. It is located 55 Km from Kaduna, the capital of the State. The main occupation of the people is farming, fishing and trading. The people are mostly indigenes living in permanent settlements. The major ethnic group in the area is Jaba. However, there are minor tribes like Hausas, Fulani's, Gbagyis, Koro, Bajjus, and Kadara. t2 Ivermectin distribution started in Kaduna State in 1988 in Lere and Kauru LGAs as part of a research project. Distribution was expanded to Bimin Gwari LGA in 1990 and Kachia and Zango Kataf LGA in 1993. The APOC CDTI project proposal was approved in May 1997 and funds were received in October 7997. In the first year (1998) of CDTI implementation, the five LGAs, namely, Lere, Kauru, Birnin Gwari, Kachia, and Zatgo Kataf were covered. Since then, the project has expanded to all fifteen endemic LGAs. The study was carried out in Kajuru LGA with 53 communities. Each community has a primary school. The State Primary Education Board (SPEB) supervises the primary schools. 13 4.0 METHODOLOGY 4.1 Study Design This study used the cross-sectional, comparative study design. It compared the ivermectin treatment coverage rate from CDD record, household coverage survey and treatment coverage of school children collected by teachers. The unit of analysis is treatment coverage rate. The study also compared the efficiency of different alternative methods of delivering and collecting monitoring tools from schools. 4.2 Description of Alternative Rapid and Reliable Methods of Delivering Monitoring Tools 4.2.1 Enugu State, Nigeria A local Education authority serves Udenu LGA, like other LGAs in Enugu state, which is under the control ofan Education Secretary. For different reasons the head teachers ofthe widely dispersed schools in Udenu LGA have regular contacts with this office. Some of the reasons include collection of salaries, which is expected to be monthly and submission of reports of different types. They also come to this office for briefing on policies of the government. The head teachers thus make at least one contact in a month with the office of the Education Secretary. This study therefore used this office for the delivery of the school forms to the schools through the head teachers who report to the office for sundry reasons. The same channel was used for collecting the completed forms from the schools. 4.2.2 Abu Hamad Province, Sudan The school forms were deposited at the provincial education office and the Director was asked to deliver the forms through public transport to each of the selected schools. Each group of villages has a common day per week for marketing where they gather together, exchange their goods and buy their necessities, except for Abu Hamad town, which has two days. For the study, the public transport used for market gathering was used to disseminate and collect the information and treatment coverage data. Drivers and passengers do not mind to take the reports and hand it to the concerned person or office, at no cost. 4.2.3 Kaduna State, Nigeria The local primary education board supervises Kajuru LGA, like other LGAs in Kaduna State, which is under the control of the Education Secretary, who suggested delivery of the forms to the onchocerciasis team leader in Kajuru LGA. The Kajuru LGA Health department and Education departments share the same office complex in Kajuru, where head teachers collect their monthly salaries. The head teachers are thus in regular contact with the local government health department, including the onchocerciasis team leader. The school forms were deposited with the onchocerciasis team leader in the LGA, who delivered them to head teachers during their routine monthly visits. The same channel was used to collect the completed forms from the school. t4 4.3 Study Population The study population consisted of members of the study communities. These include village/community members and the CDDs they have selected to distribute the drug. In the schools, the study population consisted of primary school children aged 5-14 years. 4.4 Sampling and Data Collection A multi stage sampling technique, which involved a sequential sampling in stages to get the schools, school children, study communities and study households was adopted. First, one CDTI operational unit, in the case of Nigeria LGA and Province for Sudan, was randomly selected (by balloting). The list of schools in the LGdprovince and the communities they serve were collected. From the list those that serve at least two discreet communities were listed and ten were randomly sampled. Table 1: Schools and Communities Served in the Stu Study Site Schools Communities Served Amutenyi Primary School Aguajaka ana Amutenyi Izugu Primary School Amachara and Umurama Umuezei or Primary School Umuonah and Umuamoke Umuosigide Primary School Umuosigide and Ada Central School, Obollo Afor Ogu and Umugbade Ohulor Centraal School Ulluno-Obollo and Ohulor Iheakpu Primary School Amuzu and Amaedem Udun Edem Primary School Obinazu and Obollo-Eke Imilike Primary School Imilike-Uno and Umundu Oebodo-Aba and Imilike-Agu Udenu Enugu Nigeria LGA, State, Ogbodo Primary School Block 1 El Qalaa and Block 1 El Sleam El Qoaz Gharb and El Sleam El Ghaba El Ghaba and Um Areif El Sultaib and El ShamkhiaEl Sulraib Kadaita Kadaita and Shutak Gharb El Deifalab Gharb El Daifalab and Jarf El Luba Magal and KiiiMagal El Sanjrab El Sanjrab Shmal and El Sanjrab Janoub Kalaseikal Kalaseikal and El Gerief Abu Hamad Province, Northern Sudan El Seihan El Seihan and Jarf El Doma Uduro primary school Uduro and Uduro campani Rubu primary school Rubu and Une Aku 1 Hanyin Gada Primary School Hanyin Gada and Gindi Dutse Issabe Primary School Issabe and Sabongida Rimau Kuffana 2 Primary School Kuffana 2 and Ugoh Danbagudu Primary School Danbagudu and Essab Turuku Kajuru LGA, of Kaduna state Ibigoro Primary School Ibigoro and Sabon Kaura 15 School Survey From each school a minimum of one hundred pupils was recruited for the study and the schoolteachers interviewed at least 50 from each of the two communities served by the school. The form for the rapid school monitoring was pre-tested in L.A. primary school in Babasaura, Kajuru LGA of Kaduna State. Modifications were made on the introduction for easy comprehension b efore it was used for data collection in this study. The rapid monitoring forms were delivered to the teachers through the head teachers who got the forms from different channels (see section 4.2). Instruction to the Headmaster/Class teacher Dear Sir or Madam The researcher has obtained permission from y our LGA Primary School Board to get your pupils interviewed on whether they receivecl and swallowed ivermectin (Mectizan@) this year. (Ivermectin is a small white tablet given to people once a year, according to height for the control of river blindness). The result will help the government and community to better plan ancl distribute ivermectin (Mectizan@) Please, identifu two villages where most of the pupils in your school come from. Then, get the class teachers to interview all pupils in their classes that comefrom the two identified villages on weather they received and swallowed ivermectin this year using the attached forms. I shall be grateful if the completed forms for your school can be returned to the same office where tltey were sent to you, as soon as they are completed. Yours Sincerely Coordinator, Onchocerciasis Control Programme Badayi Primary School Badayi and Sabon Gida Idon U/Sarki Hausa Primary School U/Sarki Hausa and U/Wahala Magunguna Primary School Magunguna and U/Dutse t6 SCHOOL STTRVEY FORM FOR ASSESSMENT OF IVERMECTIN TREATMENT COVERAGE AMONG SCHOOL CHILDREN. LGA Name of the school Location of the school Teacher's name Class of the In Enugu State, the rapid monitoring forms were delivered to the teachers through the head teachers who got the forms from the Local Education Authority office in the LGA Headquarters. The forms were earlier deposited with the Education Secretary by the researcher. Two w eeks later the forms were collected through the same channel. The Serial No child's Name Age Sex child's Village When the small white tablet given to people once a year, according to height was distributed in your village this year, did you receive it? (Yes or No) How many tablets did you receive?(specify the number of tablets) Did you swallow the tablets? t7 research team conducted no training and there was no contact with the schools while the forms were being completed. In Sudan, the rapid monitoring forms were deposited at the provincial education office and the Director was asked to deliver the forms through public transport to each of the ten schools. The researcher trained school headmasters in the provincial education off,rce for one day while they were collecting their monthly salaries. The forms after being completed by the teachers were returned to the provincial education office by the same channel. The forms were returned to the researcher within eight days of the delivery at the provincial education office. In Kaduna State, the school monitoring forms were delivered to the team leader of the Kajuru LGA Onchocerciasis control team, who then delivered the forms to the head teacher. The researcher trained the head teachers, who then trained the schoolteachers on how to complete the forms. The forms were duly completed and returned to the researcher through the same channel. Household Treatment Coverage Household treatment coverage survey was conducted in all the communities of the pupils enlisted for the study. A random sample of 15 households was taken in each community and all the members of the household interviewed with the standard household treatment coverage survey schedule. (See Annex 1) C o m m unity T reatm e nt Re g ister s Treatment registers completed by CDDs for each of the sampled communities were examined and information on treatment coverage collected. Table 2: Objectives of the Study in relation to Instruments and Sample size/method Obiectives Instruments Sample size t.To validate the Uganda result on the relationship between treatment coverage of school children and community treatment coverage. -School children interview guide -Household survey -CDD treatment coverage form Twenty randomly selected communities 2.To identify and describe other alternative rapid and reliable methods of delivering monitoring tools and collecting data for treatment coverage from schools. Description of alternative rapid and reliable method of delivering monitoring tools to schools 3.To assess the efficiency rate of various methods of delivering monitoring tools to schools and collecting data for monitoring treatment coverage. Simple tools for assessing efhciency of different methods of collecting treatment coverage. Ten schools and commtmities. twenty 18 4.5 Methods of Data Analysis Simple descriptive statistics such as percentages, and ratios were used in describing the characteristics of the study populations in relation to ivermectin treatment coverage. Pearson's correlation test was used to examine the association between the treatment coverage rates recorded in CDD treatment registers as well as those from the school rapid monitoring form and the household treatment coverage rates respectively. To compare the coverage rates obtained from the different sources, namely school survey, household survey and CDD treatment registers the following steps were followed. We i. Assumed the household survey coverage is the most accurate ii. Obtained the total coverage from the household survey and divided by coverage of school age (5-14 years) children also from household survey to get a ratio. iii. The ratio obtained was multiplied by the coverage obtained from the school forms to get a standardized school coverage rate. The effectiveness of the alternativemethods ofmonitoringwere also evaluated on the basis of the proportions of correct returns of the data collection instruments distributed and the time interval of the returns. The simple indicators for the analysis of effectiveness in the study include: Duration : The number of days it takes to return validly collected school children interview forms to the policy makers/ manager from the day the tools were delivered to the policy makers/manager. Proportion of validly completed forms received : The number of validly completed forms received divided by the numbers of forms delivered multiplied by 100. o a t9 5.0 RESULTS Effectiveness of Methods of Delivering and Collecting monitoring tools Data from all the schools in Sudan and Kaduna State and from 70% of those in Enugu state were available for analysis (Table 3). The Table also shows the number of correctly completed monitoring tools that were returned and hence were used for the data analysis. Table 3: Return of Monitori Tools Site Coverage Rates in the Study Area using the Household Surveys, CDD Registers and School-based Surveys: Table 4 shows that while the overall coverage rates with ivermectin in Sudan and Kaduna as elicited from all the study tools were high, those from Enugu State were very low. All the school-age children in the household were assumed to be between ages 5 to 14 years, justifyng why other age groups are not presented. Table 4: Treatment from CDD School and HH S School Forms Household Survey CDD RegisterTools Sites Sample of schools Correct Returns Sample of households Correct Returns Sample of registers Correct Returns Enugu State, Nigeria 10 7 (70%) 300 300 (100%) 20 20 (t00%) Abu Hamad, Sudan 10 10 (r00%) 300 28s (9s%) 20 20 (100%) Kaduna State, Nigeria l0 10 (r00%) 300 300 (100%) 20 te (9s%) Study Sites and Data Source Population Number Treated Coverage rate Treatment coverage from CDD register 9096 18326 14196 2839 I 5008 I 2008 31.20% 81.9% 84.6% Total Population Enugu State, Nigeria Abu Hamad Province, Sudan Kaduna, Nigeria Population aged 5-14 Enugu State, Nigeria Abu Hamad Province Sudan Kaduna State, Nigeria 3621 5395 4788 1436 4457 4006 39.7 82.6 83.7 Treatment Coverage from School Forms Enugu State, Nigeria Abu Hamad Province, Sudan Kaduna State, Nrgeria 700 1 020 3296 225 781 2021 32.14% 76.6% 61.3% Treatment from Household Survey 1881 t] 51 2853 559 1213 2381 29.1% 12.1o/o 835% Total Populatiott Enugu State, Nigeria Abu Hamad Province, Sudan Kaduna State, Nigeria Population 5-14 Enugu State, Nigeria Abu Hamad Province, Sudan Kaduna State, Nigeria 839 448 718 214 364 614 25.5 81 .3 93.9 20 A close examination of the raw coverage rates showed that the coverage rates that were elicited by the different methods mirrored each other within specific study areas. In each study site, similar trend in coverage levels were recorded using CDD register and school forms. Generally, data from all sources reveal that while the overall coverage levels with ivermectin in Sudan and Kaduna states as elicited from all the study tools were high those from Enugu state were very low. Figures I to 3 below show that while the Sudan and Kaduna sites in the study had high mean scores of 72.3 and 80.1 respectively. Enugu state recorded a mean coverage of 29.7 from household covereage. These ranged from 16.7 to 42.7 for Enugu State, 61.7 to 88.9 for Kaduna and 41.3 to 83.6 for Sudan in the same household surveys. The boxplot for household coverage in the three sites further revealed that more than 75oh of the communities in Kadunahadtreatment coverage of more than & 5%o in the household survey. Similarly, 75Yo of the communities in Sudan had treatment coverage of more than 70%o. On the other hand, almost 100% of the communities in Enugu recorded treatment coverage of 40%o with an outlier commurity at slightly above that coverage rate. The boxplots on the data from other sources namely CDD register and school-based survey approximates closely to the results of the boxplot on the household treatment coverage rates. For instance, 50%o of the communitiesin Kaduna and Sudan recorded treatment coverage of more than 75o/o in the CDD register. In Enugu, there was a concentration of the treatment coverage of the communities at a very low level. The school-based survey data showed that Sudan recorded a mean coverage of 67.9, with a range of 40.5 to 99.1, Kaduna recorded a mean of 58.2 with a range of 22.4 to 85.3 while Enugu had a mean coverage rate of 22.6 with a range of 26.0 to 41.1 treatment coverage for the communities studied. 21 o uJtro o o 120 100 80 60 40 20 Fig. 1: Boxplot of Coverage from CDD Registers By Study Sites 14 19 ENUGU KADUNA SUDAN Fig.2: Boxplot of Household Coverage By Study Sites 20 KADUNA 19 100 80 60 40 SITE SITE ENUGU o OI -0 N 14 SUDAN OB9 *€gz - - G - G3 l€9E --&'- 22 Fig. 3: Boxplot of Standardised School Coverage By Site 20 KADUNA 19 ENUGU SUDAN SITE Table 5 presents the results that were ultimately used to compute the level of correlation of the three monitoring methods. They represent percentage coverage levels. The average ratios that were used to standardize the coverage of school children in Sudan, Kaduna and Enugu were 0.87, 0.89 and 1.08 respectively. Hence, "standard school" represents the standardized school coverage levels. Correlation of the Coverage Levels from the Monitoring Tools Table 6 presents the level of correlation of coverage rates of both school-based method and the CDD register with coverage rates of household surveys. It is seen that in Sudan, the correlation of household survey and school-based strategy was high at 72o/o, while it could be termed low to moderate in Enugu, Nigeria at 49o/o though significant. The Kaduna State data on this showed that there is no statistically significant correlation between the standardized school treatment coverage and the household treatment coverage rates. On the other hand, the correlation of household survey with CDD register gave a Pearson correlation coefficient, which was very low in Sudan and Kaduna, Nigeria and statistically not significant, but a high 89% correlation between the CDD records and household treatment survey coverage rates in Enugu State, Nigeria. The difference here could be attributed to the level of education and skills in record keeping among the CDDs in Enugu State. Furtherrnore, Enugu State is relatively stable when compared with the other sites politically. The political strifes in the other sites are often responsible for the constant displacement and mobility of persons across the borders of the study areas and possible distortion in the treatment registers. o Oaoz Fa 120 100 80 60 40 20 0 14 c25 23 Table 5: Comparison of household, CDD register and school coyerage rates in the three areas Table 6: Correlation coefficients of combinations of the various However, the combined data from both countries, which has higher statistical power because of the increased number of communities showed that both school-based survey and CDD registers correlated highlywith household survey (p<.0001). Figures 4 to 11 further illustrate the correlation of the CDD registers and standardized school coverage rates with the household treatment coverage rates for the different sites 87.5 84.0 94.2 Study Site Village Serial No Enuqu State, Nigeria Abu Hamad Province, Sudan Kaduna State, Nigeria HH Treated Standard School CDD Register HH treated Standard School CDD Register HH Treated Standard coverage CDD register I 28.4 32.96 29.4 79.1 66.4 78. I 17.5 64.6 73.5 2 32.0 3l .68 33.2 I 5.3 66.1 18.2 18.6 85.3 78.2 J 36.1 34.80 45.1 11.8 14.0 83 82.2 70.0 95.4 4 42.1 38.00 42.7 77.2 65.5 88.2 84.4 64.9 84.9 5 33.9 27.90 34.0 78.9 99.1 80.7 61.7 33.7 16.6 6 33.3 33.2 33.0 42.5 92 79.6 22.6 87.5 1 25.3 28.8 29.1 75.8 64.1 78.5 88.9 83.0 64.9 8 30.9 41.1 31 .9 7t.8 56.9 78.4 77.3 69.0 9 29.6 26.00 29.6 82.1 85.0 81.8 86.2 31.6 47.6 l0 23.9 29.48 26.1 73.8 59.2 75.9 86.7 66.9 t1 31 .8 34.84 30.9 l5 64.2 79.7 78.4 52.9 t2 30.3 29.38 30.5 82.4 71.0 90.6 78.4 49.3 l3 30.5 29.44 31.3 73.6 76.8 83.9 83.s 34.1 83.6 t4 32.6 31.14 33.6 63.2 74.1 92.8 79.8 74.8 86.6 t5 28.0 28.0 83.6 13.9 83.6 84.9 78.7 90.5 l6 25.0 25.0 82.6 67.9 82.4 t1 33.0 33.0 75.6 61.6 81.6 75.5 s6.4 17.4 l8 16.1 2t.1 4t.3 40.5 10.4 71.'1 78.1 95.2 19 18.3 23.0 82.9 85.1 82.2 84.6 22.4 86.0 20 27.0 27.0 63.2 60. r 63.6 18.9 58.3 84.5 Average 29.1 21.6 30.9 12.3 61.9 8l.s 80.1 58.2 50.0 Pearson coefficient p-value Number of communities Study Site and Coverage Data .49 .89 .08 .0001 14 14 Nigeria (Enugu) . Household survey and school-based strategy o Household survey and CDD register .10 -.1 I .11 .65 20 t9 Nigeria (Kaduna) o Household survey and school-based strategy . Household survey and CDD register Household survey and school-based strategy Household survey and CDD register Sudan a a 12 l5 0001 55 19 19 Combined data (Sudan and Nigena) r Household survey and school-based strategy e Household survey and CDD register 66 86 000 r 000 l 53 52 46.3 24 100 90 80 70 60 50 40 830Ir20 Fig. 4: Correlation of Household Coverage and Standardised School Coverage 40 60 80 10020 120 STANDSCO 20 30 CDDREC Fig. 5: Correlation of Household Coverage and Coverage from CDD Registers o OII 100 90 80 70 60 50 40 30 20 50 60 70 80 90 100 n o o 0g o 6O Cl tr- o o' tr o E o m o tr Do n -oo ooo o o B D n dB f I] otrn -nU D D o nn o o otr o on rl tr 25 90 BO 100 70 60 100 70 bU Fig. 6: Correlation of Household Coverage and Coverage from CDD Register in Kaduna 40 50 60 70 BO 90 100 CDDREC 90 o OII 90 BO Fig. 7: Correlation of Household Coverage and Standardised School Coverage in Kaduna 20 30 40 50 60 BO STANDSCO oo :E = n o tr nn o0 tr tr E o n utr tr tr t3g n nC] n c) trg 0 0 n n D 26 Fig. 8: Correlation of Household Coverage and Standardised School Coverage in Sudan 90 80 70 60 o OI r40 90 80 70 60 40 50 60 70 80 90 100 100 STANDSCO 60 CDDREC Fig. 9: Correlation of Household Coverage and Coverage from CDD Registers in Sudan 90 ooI -40 tr o # o o c, D o B B tr o o# D 0 n n tr tr B o o 80 0 B n 27 o OII 50 40 30 20 Fig. 10: Correlation of Household Coverage and Coverage from CDD Registers in Enugu 40 Fig. 11: Correlation of Household Coverage and Standardised School Coverage in Enugu 20 CDDREC STANDSCO 30 50 50 40 30 oo -r20 24 26 28 30 32 34 36 38 40 42 n 0 tr o rl o o n n o tr 0 tr u o 28 6.0 DISCUSSION School-based survey is a credible and effective means of monitoring community coverage with ivermectin. Hence, it represents a rapid monitoring method that could be used by CDTI prograrnmes to know the ivermectin requirements of their communities and more importantly to rapidly determine the coverage of their communities after a round of ivermectin distribution. It was observed that in one of the sites, the school-based method did not perform well relative to the household survey. Nonetheless, the method has potential to be used to rapidly monitor the coverage and outputs of various healthcare interventions, such as level of ownership and use of insecticide-treated nets (ITNs) within communities, ownership of filters for control of guinea worrn infestations and possible others The results of this study compares favourably with the findings of the study in Uganda (Ndyomugyenyi and Remme, 2002), where school-based method and household survey was well correlated. Therefore, it could be argued that the school-based method is a valid and reliable method of assessing community c overage with ivermectin. It is worthy to note that school-based methods have also been used as proxy of other outcome measures in CDTI. For instance, Shu et al (1999) showed that delivering health education about onchocerciasis in schools led to a cofirmunity-wide improvement in knowledge about onchocerciasis and that a school-health club improved community-wide coverage and adherence to ivermectin (Shu et al, 2000). Thus, schools present a low cost and effective avenue that could be explored for strengthening CDTI as countries and communities assume full control of the prograrnme. Unlike the Ugandan study, it was found that CDD register and household survey coverage levels were highly correlated. The level of education of the CDDs could have played a role in the high level of correlation between household record and the treatment registers. As noted by Burnham and Mebrahtu (2004), although literacy is not a requirement for selecting CDDs, numeracy greatly facilitates record keeping. Therefore treatment registers of CDDs should not be dismissed as an inaccurate source of information of treatment coverage. Programme managers should find ways of improving thequality of recordsthatare keptby CDDs. Whensupervisorswere abletoprovide support for community distributors, the quality of the distributors' work and record keeping improved (Burnham and Mebraht\, 200 4). An assessment of the various channels that were used to deliver and collect the forms from the schools showed that no matter the channel used, reasonable returns of correctly completed forms were achieved. However, it appears that channels that lead to some initial interactions by researchers (or prograrnme managers) with teachers led to a higher percentage return of forms. This is because in the case of Kaduna and Sudan where the head-teachers were trained on how to administer the tools, more schools returned their forms and promptly. Nonetheless, this extra interaction involves costs and it is a decision better left to the programme managers to decide whether to undertake such interactions. Some initial training of the teachers o r their representatives may b e required, but this could be integrated with other routine training prograrnmes that the teachers undertake. 29 The school-based method for monitoring ivermectin treatment coverage could be adopted by programme managers for the accurate and low-cost rapid monitoring of coverage with ivermectin in their areas of operation. Rapid monitoring would enable them to quickly identify areas with low or sub-optimal coverage levels and intervene immediately to increase coverage. S ub-optimal ivermectin coverage levels that are undetected because the CDD registers are inaccurate, inaccessible or because there are no funds to conduct household surveys to estimate coverage would be inimical to the public health objective of ivermectin distribution. Recommendations: o Programme managers should adopt the school-based method as an alternative effective rapid means of monitoring community treatment coverage with ivermectin after 4 - 6 weeks from last distribution. o The tool should be more widely to determine the effectiveness/efficiency of alternative channels for delivery monitoring forms to the schools. o Other innovative cost-effective should be channels for delivery of monitoring forms to schools. 30 7.0 REFERENCES Burnham G and Mebrahtu T (2004), "Review: The Delivery of Ivermectin (MectizanR;". Tropical Medicine and International Health, 9(4): A26-A44. Ndyomugyenyi R and Remme J (2002), "Using Ivermectin-Treatment Coverage Among School Children Monitored by School Teachers as a Proxy of Population Coverage in Areas of Uganda where Onchocerciasis is Endemic". Annals of Tropical Medicine and Parasitology, 96(1):53-60. Shu EN, Okonkwo PO and Onwujekwe OE (1999), "Health Education to School Children in Okpatu, Nigeria: Impact On Onchocerciasis-Related Knowledge". Public Health ; | 13:21 5 -218. Shu EN, Onwujekwe OE, Lokili P and Okonkwo PO (2000), "A Health Club for a Community School in Southeastern Nigeria: Influence on Adult Perception of Onchocerciasis and Compliance with Community-Based Ivermectin Therapy". Trop ical Medicine and International Health, 5 (3): 222-226. Tielsch and Ariyne Beeche, (2004), "Impact of Ivermectin on Illness and Disability Associated with Onchocerciasis", Tropical Medicine and International Health, 9, (4):,A.45-A56 (, z u, @ u. olr UJ oo z o- o t-Ia L o !6 @ illz .9 o tf, il!$o)ll Eu, I b.0 .CC6; -oCDCil-oscil 'oE Eoodtro) E? o.9 *?= EhigEElEoN o- o- llo oNoElt ^.() e =E o;P# goir e Evx tE =r -5' gE zq9 0 s369 0UJ 3*eEITO (,) .c 0)q) t-o llac .o o o o (s E (oo il o) E E o o oo oc lr)o 5c: lo) =lant_lslo ll oE o(5 E oo -c. COo iloao .E NN E c!o LillotIC l-c| (.)t:t_l-oI c .9 (-) o oL oc oo il E o (g oL oc g, o,E' o o L-oY €R .e6 -L sE l- : EtaEhoo5E EoU'LISIf oEl- ^-J: EcrLll?,=S<)i o 1o7 ..:5r<Fo i4El z EiHSP E HE;eFN9bi; b: ii irD O) d e36 or;,.. -c c X^o'=E=o=^) EE 19 e,P;o - - -c a\ : B;=Q ooll{-lrEt- h 6 81rr =-!t' .d3 r- I E.E2 l-c.= ^)-d)C'x >q I:/ " d= :- Etr,26 se P tl C v- ..H OgPlx -,r, 3 g.E q ra o EE1;i = oE-.,i\J = c 6JA ;iE>kl 9-e a .ooLO 6= Er\ (D v=v =aouoi PE€V>EL!i 6q* E ^! =a E;HD 9EOO EBP -=rr ' I - ..O -2.= L -E=ur--il €EH;Z ; gsE -4 C O-cr-T -LeU co N z(J tra uJl o -gE, rs#-E .E- EETEE+poqo6lr u= >=E)- Fat! =o F oz LL o c.. f o 6y'5 o€ o c;i o! O: O-N 6Y! c il llI-g.go>z E- -. c'E9y e >s;i o) 5 E.EEE z tr U) uJ f oF U) uJ E o=i c .-Y ! EEgliE qE ;= OG.Ep€ BIE (E >, B: T !;Efirl gEE*g f oo>' =36e8o Go o > -ESI]'o, z o F U) ull a oc.E Zltt -tiXo) EEEEET7d,s;926*6.9o: 5 [a:9 E ir o)g) x'!: Noll I(,>LL a o -o Eo E E oE o)oJ oE o U) UJ =z Rapid Monitoring Instrument 3 ANNEX 2: CDD RECORD FOR RAPID MONITORING STUDY l. Total 2. Age composition of people: Below 5 years 5 years and above_ 3. Sex composition of the population: Male Female 4. Number of persons treated Male Female 5. Number of persons under-5 years who received treatment 6. Number of refusals 7. Number absent during last treatment_ 8. Number with severe side effects 9. Number of tablets received 10. Number of tablets used I 1. Number of tablets left in the drug kir 31