Delivery mechanisms.

A number of factors related to details of how drugs were delivered affected compliance in the reviewed studies. These included the identity of the drug distributor, the time of day MDA was carried out, the use of community organizations, the training of the health workers and drug distributors, motivation of health staff, and the use of DOT. These elements were repeatedly cited in reviewed studies as having an impact on compliance with MDA.

Drug distributors are at the interface between MDA programs and their target populations. Characteristics of these frontline workers were identified as a key factor regarding compliance with MDA. Compliance tended to be positively affected when the drug distributor came from the area where he/she was distributing the drugs [16] or when recipients personally knew the drug distributor [26]. Conversely when drug distributors came from a different group or caste, individuals were reluctant to comply [36]. The positive motivation provided by drug distributors was often cited as an important factor that affected compliance with MDA [22], [24], [25]. For example, compliance was positively affected when drug distributors took the medication directly in front of the community [16]. Similarly, if drug distributors visited the household at least one time (or more) prior to MDA, this increased the probability of drug uptake in those households [13].

Some papers reported that direct participation of health personnel in drug distribution positively influenced MDA [36], [37]. In the Philippines, surveyed individuals reported that they complied because they had accepted the health worker's advice [33]. Similarly, in one study from India, compliance was 5.6 times (95% confidence interval (CI) 3.4–9.1) better when MDA was distributed by staff from a local health center [38]. Interactions between health workers and the community sometimes motivated village leaders to participate stressing the “necessity of their cooperation” [16].

The involvement of community organizations and other preexisting networks in MDA distribution sometimes had positive or negative effects on compliance. In several sites, the involvement of community groups or networks in the distribution of drugs improved compliance, because they were able to reach community members efficiently and effectively and thereby enhance participation [11], [39], [40]. The Department of Health in American Samoa evaluated the territory's MDA campaign after inadequate compliance rates (49%) were reported. Following the department's recommendations, subsequent MDAs incorporated churches in the promotion and distribution of the tablets, and this significantly improved compliance rates to between 65% and 86% over the subsequent four years of the program [41]. In contrast to this experience, Aswathy et al. reported that people's fears were exacerbated when members of local self-help groups in India acted as drug distributors, because they were too hurried to answer people's concerns about adverse events [42]. Low motivation by drug distributors was also mentioned as contributing to low compliance in a report from another area in India, where overburdened local health workers considered the MDA program to be an unwelcome additional duty [16]. Similarly in a study from Kenya, drug distributors cited lack of supervision from health staff, short training, lack of incentives, and delays in supplies as reasons for low motivation [43].

Inadequate training of health workers and drug distributors can have negative effects on MDA compliance. Some reports mentioned that pills were sometimes dropped off at the house [11] or left for family members to distribute without an adequate explanation [12], [14]. In a study from Sri Lanka, individuals reported that they did not like receiving drugs from volunteers, because the volunteers lacked information about the drugs. Also, in some cases volunteers were considered to be too young to be distributing drugs. In the same study, volunteer distributors might have had more credibility if they had arrived with a badge or an official letter [26]. In an Indian study, drug distributors' poor communication skills were credited with contributing to poor compliance as people doubted their ability to assess eligibility for DEC [38]. While some of these issues may not have been directly related to poor training, improved training and supervision may have improved performance.

The private health sector can have an important influence on compliance with MDA. One study from India reported that private practitioners exaggerated the possible adverse events of MDA, and this discouraged people from taking the pills [44]. Other reports mentioned that when medical practitioners were not involved in the MDA process, they remained unaware of the benefits and necessity of MDA and sometimes advised their patients against compliance [45], [46]. In other cases, community participation was positively affected when practitioners took the pills in public and actively supported the program (authors' unpublished observations).

Although directly observed drug administration is recommended for MDA programs (DO-MDA), it was infrequently used in the studies reviewed for this paper. MDA programs tended to be highly successful when DO-MDA was systematically employed [34], [41], [47]. One of the difficulties of ensuring DO-MDA was the perceived challenge of reaching individuals after meals so they could avoid taking the medications on an empty stomach [13], [14], [37].

MDA medication issues.

Many studies reported concerns from individuals about the pills provided during MDA as reasons for non-compliance. The large number of pills to be swallowed was cited most frequently, and this was followed by complaints about the size and taste of the pills [13], [15], [16], [48]–[50]. There were also concerns about the quality of the drugs. One report from Haiti mentioned that the pills were “bad” [23], while one paper from India mentioned that the pills had disintegrated into powder [30]. The distribution of “loose tablets” was also cited as a reason not to comply with MDA [17], [22].

Awareness and knowledge.

One of the most prominent factors associated with motivating compliance and increasing coverage was advance knowledge of the MDA, as demonstrated in studies from India, Sri Lanka, the Philippines, Sierra Leone, and Vanuatu [15], [18], [25], [26], [33], [51], [52]. The inverse of this was also observed, when individuals did not comply because they were unaware of the MDA or the LF elimination program [13], [19], [22], [53].

Across the literature, knowledge associated with compliance was primarily related to three specific topics: transmission of LF, knowledge that MDA protects against LF, and knowledge of lymphedema management techniques. Knowing that mosquitoes are responsible for the transmission of LF was shown to be associated with compliance in several studies [24], [26]–[28], [51], [54]. Ramaiah et al. described the inverse of this phenomena whereby people were unwilling to comply because they reported that they controlled mosquitoes in their own living environments and therefore were not at risk for LF [55]. Those individuals who reported knowing that MDA medications prevented LF were also more likely to comply [13], [15], [16], [22], [24]–[26], [33], [39], [42], [53].

In studies that performed multivariate analysis, knowledge that MDA was for LF was a key factor associated with compliance [24], [25]. One Indian study reported a strong association between compliance and the interaction between knowing about MDA in advance and knowing that everyone was at risk for LF (odds ratio (OR) 16.1; 95% CI 8.8–29.3) [25]. Another Indian study demonstrated that the combination of advanced knowledge of MDA and knowing that mosquitoes transmit LF dramatically increased the probability of a person complying with MDA [24].

Finally, there was evidence that knowledge of lymphedema management techniques and the availability of case management programs for diseased individuals (lymphedema management programs and hydrocele surgery) improved compliance with MDA [25], [35], [51]. For example, establishment of and support for Hope Clubs in Cap Haitien, Haiti helped to demonstrate that the LF elimination program had a long-term commitment to help those affected by LF and that it was not simply a short-term drug administration program (L. R. Carpenter, personal communication).

Misconceptions and misinformation about the treatment inhibited people's uptake of MDA medications. For example, some people in the Philippines believed that the treatment might cause sterility, fainting, or even death [33]. Uptake of MDA increased after these misconceptions were corrected. In Tanzania, misperceptions about MDA resulted in lower compliance rates; specifically, some people believed the drugs were a contraceptive while others thought that the pills inhibited libido [56]. Similar fears that the drugs were contraceptives were expressed in Kenya [34]. In Vanuatu, some believed that the pills would cause LF and that the government was trying to sterilize the communities [15]. Compliance also suffered in areas where people were misinformed about the etiology of LF (believing it to be due to heredity or caused by a curse) [13] (A. Viall, personal communication, and authors' unpublished observations).

Aside from the three primary knowledge indicators discussed above, there was scant information on further associations between knowledge and compliance. Detailed knowledge regarding LF and MDA was not always associated with high compliance with MDA in some areas (Els Mathieu, personal communication, and authors' unpublished observations [57], [58]).

Understanding the personal risk of LF and the benefits of MDA.

hose individuals who were able to express their own personal risk for LF were more likely to comply with MDA. In some papers, this was described as having seen someone with LF or having someone in the family with LF, while in other cases it was expressed as an understanding that “everyone is at risk for LF” [25], [26], [51]. One study from Sri Lanka demonstrated higher compliance in urban areas for those who had seen someone with clinically evident LF [26]. The inverse of this perception was also demonstrated. That is to say, those who did not perceive LF to be a problem in their community were less likely to comply with MDA [12], [59], [60].

Those who were able to personalize the benefit of the MDA for their own individual health were likely to be compliers [16], [22], [39], [42]. Perceived benefit was positively associated with compliance in studies from India and the Philippines [33], [42]. Understanding that the MDA was beneficial was similarly associated with increased compliance [26]. Note that perceived benefit of the treatment is not the same as understanding that the MDA is for LF.

Adverse events and MDA.

The potential for adverse events following MDA was responsible for discouraging as well as encouraging uptake. Many papers cited individuals in their surveys who were afraid of adverse events from the treatment and as a result did not want to comply with MDA [11], [13]–[18], [22]–[26], [30], [31], [33], [37], [50], [59], [61]–[63]. Some of this fear arose from inadequate explanations about possible adverse events and a failure to understand that severe adverse events are rare [42].

Rumors regarding adverse events associated with MDA can seriously affect compliance by individuals. Three studies from India and one from Haiti mentioned erroneous reports of deaths due to MDA that negatively affected compliance [11], [28], [39], [44]. When these rumors were not addressed, they persisted and negatively affected MDA in neighboring areas as well [12], [64].

When communities did not receive correct and adequate information about adverse events, the consequences were detrimental to the success of the MDA. In Orissa, India, insufficient information about the MDA and its possible adverse events led to an erosion of confidence in the community, and as a result, the MDA was postponed for three years [65]. In East Godavari district, India, MDA was suspended due to adverse events until health workers intervened and addressed community concerns [16]. On the other hand, in areas where the management of adverse events was monitored and well coordinated in communities, individuals cited this as positively influencing their acceptance of MDA and suggested that the possibility of adverse events did not inhibit their participation in MDA [16], [66]. Similarly there were some instances where those who experienced adverse events following MDA in one year were more likely to comply the following year [67], [68]. Anecdotal evidence from Alor district, Indonesia suggested that some individuals equated adverse events with evidence that the treatment was effective [57]. However, adverse events following MDA in one year sometimes deterred individuals from compliance in subsequent years [27], [67].

The elimination of intestinal helminths was cited in several papers as a positive consequence of MDA [69], [70]. This deworming effect was perceived to be a community benefit, and this contributed to overall community acceptance of the MDA program. For example, Talbot et al. found that Haitian subjects who did not know that the pills contained albendazole (ALB) were five time more likely to be non-compliant than those who did know about ALB, suggesting that the people were aware of the collateral deworming benefits of this medication [71]. Anecdotal reports from Indonesia, Papua New Guinea, and Haiti described individuals who were grateful or even “excited” when they or their children passed intestinal worms following MDA (L. R. Carpenter, personal communication, and authors' unpublished observations).

Recipient's personal situation.

Several publications mentioned that personal characteristics influenced compliance with MDA. Characteristics cited included the individual's personal health at the time of drug distribution, whether or not they were taking other medications, their gender, age, household income, or residence in urban or rural areas, and as mentioned above, their understanding of their own personal risk for infection with LF. Individual health status and current use of medication heavily affected people's compliance with MDA. In many instances, papers cited “healthy,” “have no LF disease,” or “drugs are unnecessary” as reasons for non-compliance among their study populations [11], [13], [14], [16]–[19], [22], [25], [26], [30], [37], [39], [53], [61]–[63]. Being ill at the time of MDA was also cited as a reason for non-compliance [11], [16], [22], [24], [25], [27], [37], [39], [42]. Furthermore, it was also commonly expressed that people refused MDA because they were presently taking other medications (including contraceptives) and that they were concerned about possible interactions with the LF drugs [18], [23], [26], [28], [30], [37], [61]. Treatment for other illnesses was the primary reason given for systematic non-compliance in one study [39].

Female gender was often associated with lower rates of compliance with MDA. For example, pregnancy was often cited as a reason for non-compliance, as pregnant women were not eligible for MDA, and lactation was considered to be a contraindication to MDA in the early years of the program. In Haiti, women of childbearing age were excluded at the beginning of the MDA campaign due to fears of using ALB in early pregnancy. Although this policy was reversed prior to the third round of MDA, it had a lasting effect, because women continued to have low compliance rates [71]. Other reasons cited for low MDA compliance in Haitian women included a rumor that women should not comply with MDA if they were taking contraception, fear of miscarriage, and a desire to avoid embarrassment at MDA distribution posts where they would be asked if they were pregnant or not (N. Barkey, unpublished report). In some household situations, women also were prohibited from seeking or taking MDA, because of their husbands' negative beliefs regarding the treatment (authors' unpublished observations).

Age was closely associated with MDA compliance in some studies. In most cases, the youngest (under 5 years of age) were the group most frequently cited as non-compliers, because their parents feared how the treatment might affect their children [14], [23], [30], [72]. In one study from Egypt, children aged 2–4 years showed lower rates of compliance than adults. However, this may have been due to blood tests conducted at the time of the survey [73]. In a few studies, the elderly were also reported to be systematic non-compliers [39], [40] as were individuals 15–34 years of age in a Ghanaian study [40] and people aged 16–30 years in an unpublished report from Haiti (N. Barkey [2009]). Similarly, Wynd et al. discovered that the younger generation in Papua New Guinea was becoming resistant to continuing compliance in some areas, because they thought that LF was no longer a threat [74].

Area of residence and household income often affected compliance. Living in a rural environment was associated with higher compliance in several studies [11], [13], [37], [75], while achieving compliance in urban areas was more challenging. Inadequate coverage and compliance in urban areas were attributed to some of the following factors: lack of a specific urban strategy [11], [72], fewer peripheral health workers [72] and volunteers [26], the dominance of private health care providers [22], [55], poor health care infrastructure [22], [55], and the presence of unorganized settlements and large numbers of migrants [55]. Individuals with lower incomes were also more receptive to MDA in some urban areas [55], while those with higher incomes were more difficult to reach and to convince to take the medications [39], [48].

Concepts of coverage and compliance.

Since definitions and equations for coverage and compliance varied in the papers reviewed for this study, it was challenging to compare results and experiences across different papers. It would be helpful if researchers and national programs could agree to a uniform set of definitions. The definitions in the most recent guidelines from WHO [84] (reprinted in Table 1) are a good starting point. However, we believe that the definitions in this document should be modified to clarify the distinction between coverage and compliance. Most of the publications reviewed for this paper use the term “compliance” to describe what is called “drug coverage” in the guidelines. As this contributes to confusion, the authors propose that the term “compliance” should be included as another term for “drug coverage” in the guidelines to improve clarity and understanding.

A “tool box” for difficult areas.

LF elimination is progressing well in many countries. However, special approaches and innovative methods may be required for countries or areas that are not meeting expectations. These often face special challenges such as difficult access, post-conflict health systems and populations, post-natural-disaster environments, large urban settings, and fragmented societies. They may include areas that have failed transmission assessment surveys (TAS) despite many years of MDA. New tools are needed to help LF programs to quickly assess the characteristics of these challenging environments and plan MDA or other responses accordingly. When standard approaches are not achieving the goal, other resources (private sector or other donors) and implementation partners (faith-based organizations, NGOs) may tip the balance toward success. By approaching these more challenging areas intentionally and differently, a troubleshooting guide may be helpful to guide assessment, execution of MDA, monitoring, and evaluation. A basic troubleshooting guide is presented in Table 2 with suggested tools. This can be adapted for local use, and it should be modified over time based on feedback regarding its use.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. A basic troubleshooting guide for commonly encountered problems in MDA programs.

https://doi.org/10.1371/journal.pntd.0002447.t002

Gaps in understanding.

Areas that have been identified that require additional research fall into four categories: service delivery, strategic issues for NTD programs, and compliance at the level of the community and the individual (Table 3). This review has identified gaps in understanding related to service delivery and program strategy. Some of these are related to the recent move to integrate programs for diverse NTDs that vary with regard to intervention tools, target populations, and goals (control versus elimination). More work is needed to improve understanding of the impact of this integration on LF elimination efforts so that the benefits of integration can be harvested while minimizing the risks that integration may pose to GPELF.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Research needs and gaps in understanding.

https://doi.org/10.1371/journal.pntd.0002447.t003