Feasibility of measuring compliance to annual ... - Wiley Online Library

Tropical Medicine and International Health

doi:10.1111/j.1365-3156.2006.01796.x

volume 12 no 2 pp 260–268 february 2007

Feasibility of measuring compliance to annual ivermectin treatment in the African Programme for Onchocerciasis Control William R. Brieger1, Joseph C. Okeibunor2, Adenike O. Abiose3, Richard Ndyomugyenyi4, William Kisoka5, Samuel Wanji6, Elizabeth Elhassan7 and Uche V. Amazigo8 1 2 3 4 5 6 7 8

Department of International Health, The Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD, USA Department of Sociology/Anthropology, University of Nigeria, Nsukka, Enugu State, Nigeria Sightcare International, Ibadan, Oyo State, Nigeria National Onchocerciasis Control Programme, Kampala, Uganda National Institute for Medical Research, Dar es Salaam, Tanzania Research Foundation in Tropical Diseases and Environment, Buea, Cameroon Sight Savers International, Kaduna, Nigeria African Programme for Onchocerciasis Control, Ouagadougou, Bukina Faso

Summary

The African Programme for Onchocerciasis Control (APOC) sponsors annual distribution of ivermectin to control onchocerciasis. Ivermectin should be taken annually by 65% of community members for a number of years to eliminate the disease as a public health problem. While many community coverage surveys have been undertaken during project monitoring, individual compliance could not be studied until several annual rounds of distribution had occurred. This paper reports on the efforts to determine whether adequate records could be found to enable a compliance study. A step-down process from project to district to community level was used to identify project sites where continuous ivermectin distribution up through 2004 had occurred. The first step consisted of selecting 17 of 25 projects by APOC. The second step determined adequacy of districts where distribution had occurred on a regular annual basis. Among 121 districts 58.6% undertook distribution in all 7 years. A total 852 villages were visited and community level records were found in all but three. Records showed that distribution had occurred for a minimum of five consecutive times in 429 villages, and ultimately 10 projects. While the feasibility study found an adequate number of villages to study compliance, the large number of projects, districts and villages that did not qualify for the follow-on compliance study should lead National Onchocerciasis Control Programme managers to strengthen the overall coverage and consistency of their efforts. keywords onchocerciasis, records, compliance, ivermectin, community-directed treatment

Introduction The African Programme for Onchocerciasis Control (APOC) was launched in December 1995, with the plan to implement onchocerciasis control by distributing ivermectin through full community participation using a well tested community directed treatment with ivermectin (CDTI) approach (UNDP/World Bank/WHO 1996). Simulation models have been developed that shows a need to run CDTI continuously for at least 15–25 years to eliminate onchocerciasis as a public health problem, assuming annual population coverage meets or exceeds 65% annually (Plaisier et al. 1997; Winnen et al. 2002). 260

While annual coverage figures are available for all project sites, they do not provide a direct insight into why individual community members do or do not take ivermectin and thus provides a basis for understanding how to sustain community control efforts over a long period. The first 25 project sites were funded to implement CDTI during 1996–1997 and began distributing by 1998. A project may consist of a country or some larger subdivision or a country such as states, provinces or clusters of districts. Projects are co-managed by a government entity, such as a health ministry and a Non-Governmental Development Agency. Although 108 CDTI projects have been funded in 16 countries as the beginning of APOC, it is

ª 2007 Blackwell Publishing Ltd



1 W. R. Brieger et al. Compliance to annual ivermectin treatment

those early 25 projects that offer the maximum opportunity to study long- term compliance with annual ivermectin dosing and thus provide insights into how to enhance compliance. Compliance, also known as adherence (Weiden & Rao 2005), lies at the heart of any medical treatment or public health intervention. It does little good when medical or public health workers set out a course of action if the client does not follow through with that action in an appropriate manner. Compliance has been defined as, ‘the extent to which a patient acts in accordance with the prescribed interval and dose of and dosing regime. The unit of measure for compliance is administered doses per defined period of time, reported as a proportion (%) of prescribed doses (D) taken at the prescribed time interval (T) as measured by the period of time, i.e. percentage of TD, measured by percentage’ (IPSOR Medication Compliance and Persistence Special Interest Group 2005). Compliance is often discussed in the context of chronic, long-term regimens such as for HIV and hypertension or acute, short-term treatments for malaria. There are, of course, public health programmes that require an annual or semi-annual performance of a health action, and a prime example is the dental checkup and hygiene visit (Miyamoto 2 et al. 2006). A wide variety of patient, provider and drug factors influence compliance. Patient factors include social and demographic characteristics (Cramer & Pugh 2005) such as ethnicity (Yuan et al. 2006), educational level (Iliyasu 3 et al. 2005), social support (Elliott 2003) and work/ personal schedules (Rubio et al. 2005). Provider factors can include bias towards patients with certain characteristics (Ikeda et al. 2005; Ragot et al. 2005; Stone 2005), poor communication concerning the regimen (Depoortere et al. 2005), example of drug factors include side effects and adverse reactions (Trotta et al. 2002; Pulliam et al. 2003; Engelhard et al. 2005; Weiss et al. 2005; Yuan et al. 2006), perceived efficacy (Weiden & Rao 2005) and complexity of the regimen (e.g. timing, dosages) (Gardner et al., accessed 2006; Yuan et al. 2006). According to Van Wijk (2005), adherence could be measured through four broad methods: (i) observational methods such as pill counts; (ii) reported behaviour as obtained through medication diaries, patient self-report and provider report; (iii) recorded data found in clinic and pharmacy records and (iv) biomarkers found serum and saliva samples. Clinical records include data on timing and frequency of refills (Choo et al. 1999; Sikka et al. 2005). Compliance measures such as patient, family member or physician reporting when used alone present problems of validity (Nichol et al. 1999).


Although CDTI approach is effective in reaching hundreds of previously neglected communities or those with limited access to health services (UNDP/World Bank/WHO 2003), the rate of compliance of individuals who have been on ivermectin treatment since the inception of the APOC sponsored community-directed treatment with ivermectin programmes has not been documented. The feasibility of these compliance measures must be considered in the context of CDTI, which takes place in rural villages scattered throughout much of sub-Saharan Africa. To date published reports of CDTI interventions, including independent project monitoring sponsored by APOC, have focused on population coverage. Only two published CDTI studies have recorded coverage on a small number of villages over at least 5 years (Emukah et al. 2004; Ndyo4 mugyenyi et al. 2004). It may be possible to infer individual compliance from long-term coverage, but unless longitudinal data are available on community members, one cannot learn about the actual factors associated with or effects of individual compliance. As Emukah et al. (2004) observed, ‘we did not observe a relationship between community coverage and community morbidity reduction. This is probably because morbidity (therapeutic) impact is related to individual compliance, rather than community coverage, and questionnaire data from the cohort suggested high cohort compliance with therapy’. Coverage surveys have been useful in identifying factors that might potentially influence compliance such as gender (Brieger et al. 2002; Maduka et al. 2004) and minority ethnic status (Brieger et al. 2002). Social support and drug perceptions are the other factors that have influenced coverage, and hence may impact on compliance (Nuwaha et al. 2005). In fact, although the study by Nuwaha et al. (2005) refers to ‘compliance’, it in fact is a cross-sectional study of one annual distribution. Oyibo and Fagbenro-Beyiku (1998) reported on 5-year compliance in one Nigerian village, and found that while on average 53.5% of eligible villagers took ivermectin each year, individual compliance was only 1.9 times on average, which they blamed on high levels of absenteeism due in part to poor mobili5 zation. Newell (1997) examined onchocerciasis infection rates in a Burundian village, and found through recall that a small sample of villagers complied with annual ivermectin treatment an average of 2.4 times over a 4-year period. There are no recent large-scale and long-term studies of annual compliance with ivermectin. The question arises as to whether it is feasible to measure compliance rates over several years and across several projects to determine more accurate and current compliance rates for the programme 261




we use the term village. In the case of annual ivermectin distribution, individual compliance is defined as taking ivermectin at each annual distribution as the inception of CDTI in the village, unless the person does not meet eligibility requirements in a given year (over 5 years old or 90 cm tall; no serious health problem, e.g. asthma, not pregnant). A village has complied with CDTI if the village members have arranged and held ivermectin distribution consistently on an annual basis since the inception of the project. If a village does not comply in a given project year, it would be difficult for the individual residents to comply. Specific objectives included:

as a whole as well as the factors that may influence compliance. Community-based records and annual surveys have been the foundation for measuring coverage in CDTI programmes to date. A sample of one of the many formats used is shown in Figure 1. As community members organize their own annual ivermectin distributions with support from the district health service they also are responsible for conducting a village census, maintaining the village’s register, and recording the annual doses of ivermectin taken by each eligible community member. While record keeping is a village responsibility, village records are usually maintained by volunteers known as ‘community-directed distributors (CDDs)’ who are selected by the villagers and trained by the district health staff (Oladepo et al. 1999–2000; Katabarwa et al. 2002). CDDs, like most rural residents, have limited literacy skills, but although various experiments using tally sheets have been tried, CDTI programmes ultimately have found the village’s register to be the most useful format for keeping track over time. These village records form the basis for APOC monitoring activities (Amazigo et al. 2002).

• Ascertaining the existence of CDTI records at all levels of a project. • Determining whether an adequate number of accurate records exist at community level to measure compliance to treatment with ivermectin. • Calculating village compliance as a foundation for future studies on individual compliance. Village-based registers are the foundation from which individual compliance with annual ivermectin treatment is documented. Summary data forms are forwarded to nearby health facilities, and then onwards to district health authorities. District summary information is sent to the next level which is the project. The project is the entity receiving APOC financial and technical support. Data are ultimately compiled by the national onchocerciasis control programme. To trace the feasibility of studying individual compliance, the researchers essentially traced the ivermectin documentation process from the project level back to the village.

Methods The general aim of the study was to determine the feasibility of measuring long-term compliance of individuals and villages with annual treatment of onchocerciasis with ivermectin. Several terms have been used to define the smallest socio-political unit that organizes CDTI, for example, village, community or settlement. For simplicity, 17

17

85 153

37 32 12 sick 10 0 6 0 abs 62 3 17 4 4 2

13

15

165 142 123 105 142 160

Remarks

Tablets

38 33 13 11 7 63 18

Height

4 4 2 1 1 3 0 abs

Tablets

Height 165 160 138 118 92 142

Age

165 160 130

Year 2002 Remarks

36 31 11 9 5 61 16

Age

4 4 2 1 0 3 3

Remarks

165 160 121 95 65 142 148

Year 2001

Tablets

35 30 10 8 4 60 15

Height

Tablets

M F F M M F F

Remarks

Height

Name Isaac Esther Camille Francls Samuel Anna Mary

Age

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Total

Year 2000

Sex

Serial No.

Year 1999

Age

Household number

4 0 preg 3 2 1 3 4

17

Figure 1 Sample village record, register page.

262





Village compliance rate was the proportion of times CDTI was organized annually since the inception of the project. Individual compliance rate was defined as the proportion of times a person took ivermectin at each annual distribution since the inception of CDTI in the village, unless the person exempted because of ineligibility. It was necessary to interview only those who were exposed to the programme since inception, i.e. people at least 5 years old at inception and thus, at least 12 at the time of the study. It was expected that the maximum denominator for either village or individual compliance would be seven distributions. The challenge of recall was recognized as a limitation. As compliance is best measured over time instead of at one or two individual distributions, it was decided to base the study in project sites that afforded the maximum potential opportunities for individuals and villages to take ivermectin. Altogether 25 CDTI projects started in late 1997 and 1998, giving a potential of seven annual distributions by the start of this study. The period also allows adequate time for positive dermatological and clinical benefits to be observed and perceived (Kennedy et al. 2002, Akogun et al. 2000), thus enabling researchers to address the important issue of how perceived benefits might influence compliance. Study instruments were developed for the various steps in the CDTI reporting process. These steps are outlined below. • The first step was review by APOC management of all 25 project sites and the selection of those known to have annual distributions for the period under study. • The second step was used at the project level to obtain information from staff on districts that had actually undertaken ivermectin distribution from the inception of the project. Districts on this list that had the required number of annual distributions were used as a frame for sampling five districts (also known as divisions, local government areas). • The third step repeated the ‘screening’ process for villages within the chosen districts. Ten villages where records indicated that distribution had taken place over the requisite number of years were chosen from each district, yielding 50 villages per project. • The fourth step had two instruments, the first to review the CDD records and the second to obtain information from villagers on number of annual treatments received that was compared with CDD records. The fourth step was crucial for answering three questions. First, were there any records that could be reviewed? Secondly, if records were available, did they indicate adequate consecutive annual distributions? Finally, was the quality of records adequate to determine individual compliance over time?


Field work was undertaken by a site principal investigator (PI) who hired one or two research assistants. In the field it became apparent that there was also need to bring on board members of the local health team as guides. The site PI was ultimately responsible for all aspects of field coordination and data management. Each site team compiled the daily field experiences into a diary/report from which logistical and programmatic lessons were extracted to guide further study. 6 We developed data entry templates in EPI-INFO 6.04d (Centers for Disease Control and Prevention, Atlanta, GA, USA) and circulated them to each site PI. Data were then entered in the field and returned for compilation and analysis. Each field team also maintained a detailed daily field notes, which were submitted to the coordination team for review of factors that might affect the data quality and help explain the treatment gaps. Results Table 1 shows the project sites included and dropped at each of the four steps in the assessment process. At first step, seventeen project sites were selected by APOC management in five countries and were visited between August and November 2005. Reasons for the exclusion of projects at the first step included political instability, known break in distribution, known poor performance as reported during monitoring activities, small size (that would not yield enough communities) and over-representation by a country/ 7 NGDO (Non-Governmental Development Organization). The 17 selected sites/projects included Chad, Ruvuma in Tanzania and the first three phases of districts in Uganda. Cameroonian sites included Adamaoua II, Centre III, North Provinces and South West I, while Nigerian sites were based in Abia, Cross River, Kaduna, Kogi, Imo, Osun, Taraba and Zamfara states. In Nigeria, Imo and Abia states were initially managed as one project, but each State Onchocerciasis Task Force currently ran them as independent operations. Thus, it was decided to treat them as two projects. The presentation of results begins with the examination of adequacy of district and village level distributions over the period of study. Within the 17 project sites, researchers visited 121 districts and enumerated 984 villages. The definition of compliance at all levels, district, community and individual was based on consecutive distributions occurring from the most recent (2004) backwards. Figure 2 shows that in the period under study 58.7% of districts undertook distribution in all seven possible years, while 34.7% of districts had at least six consecutive distributions including 2004. None had five consecutive distributions. A minimum of six consecutive distributions was considered acceptable (there 263




Table 1 Step-down process of identifying feasible sites to study compliance

Cameroon South West I Cameroon Littoral II X Cameroon Adamaoua II Cameroon North Province Cameroon Centre III CAR CAR X Chad Chad EQ. Guinea EQ. Guinea X Malawi Thyolo and Mwanza X Nigeria Taraba Nigeria Kaduna Nigeria Cross River Nigeria Kogi Nigeria Plateau/Nassarawa X Nigeria FCT X Nigeria Osun Nigeria Enugu, Anambra, Ebonyi X Nigeria Imo and Abia Nigeria Zamfara Sudan Northern Sector X Tanzania Mahenge X Tanzania Ruvuma Uganda Uganda Phase I Uganda Uganda Phase II Uganda Uganda Phase III

Village

58.7

49.6

40.0

34.7 27.2

30.0 20.0 10.0

X

X

X

X

X

were none with five), and 15 projects scaled through second step. Among the 984 villages on record at the district level, only 29.0% had seven consecutive distributions, followed by 21.7% with at least six. At the third step, nearly twothirds of villages (65.4%) had at least five consecutive distributions as seen in Figure 2. This is a measure of ‘village compliance’. Table 1 again shows consecutive distribution by project site. At this level, three of the 17 projects lacked adequate villages to be included in the further study of compliance.

13.8

9.3

6.6 0.0

0.0

X X

*Step 1: clearance by APOC management. Step 2: >50% districts at project level distributing ‡5 consecutive times. Step 3: >50% communities at district level reported to have distributed ‡5 consecutive times. Step 4: >50% community records showing distribution ‡5 consecutive times. X: project site dropped at step; shaded block: project site continues to next step. APOC ¼ African Programme for Onchocerciasis Control; CAR ¼ Central African Republic; CDTI ¼ community directed treatment with ivermectin; FCT ¼ Federal Capital Territory (Nigeria).

264

Percent

Step 1 Step 2 Step 3 Step 4