Search strategy.

We used three specific information sources, as described below, to identify F&E-related grey literature. We conducted searches in English, Spanish, Portuguese, and French. However, we attempted to identify all relevant literature, regardless of publication language and location (i.e., we included literature published outside of peer-reviewed journals), produced between January 1965 and August 2016.

Information source 1. Electronic database searches: We used a set of Boolean search terms (S1 Table) to search Open Grey and New York Academy of Medicine Grey Literature Report databases.

Information source 2. Systematic review of key websites: We performed a standardized search, using pre-determined keywords (S2 Table), to identify relevant literature published on or housed within the websites of key organizations implementing trachoma programming or conducting related operational or academic research. Key organizations included those participating in the International Coalition for Trachoma Control (ICTC) and WHO Collaborating Centers for Trachoma (WHO-CCT). See Supplemental Material for a full list of the member organizations targeted for this search. In addition to these network websites, reviewers systematically searched the websites of each partner or affiliate participating in these networks.

Information source 3. Literature solicitation from implementers and researchers: Reviewers approached individuals working in the trachoma sector, commencing with members of staff of WHO-CCT and ICTC member organizations, requesting that they submit relevant literature for review. We engaged these stakeholders to identify documents that were currently in process or previously unidentified through our searches of other information sources (see Supplemental Material for a list of member organizations engaged via ICTC and WHO-CCT). We completed the process by emailing the ICTC listserv twice, and through personal communication during a meeting of WHO-CCT.

Screening and literature selection.

Once F&E grey literature documents were identified, we screened them to determine which were eligible for inclusion in our review, as per our PICOT inclusion and exclusion criteria.

Participants: We considered literature referencing F&E interventions and outcomes implemented and detected amongst children and adults. Settings included areas that were endemic for trachoma (hyper-, meso-, or hypo-endemic for trachoma), and in which trachoma elimination programs were operating at the time of publication. Our operational definition of trachoma elimination program aligned with WHO’s Guide to Trachoma Control Programs for the Prevention of Blindness [26], which specifies that programs to control trachoma should include the following elements:

1. Assessment of the problem and establishment of priorities;
2. Allocation of resources;
3. Chemotherapeutic intervention;
4. Surgical intervention to correct lid deformities;
5. Training and utilization of local health aids and other non-specialized health workers;
6. Health education and community participation; and
7. Evaluation of intervention programs.

Interventions: The interventions of interest for this review were broad, yet reflect those addressing facial cleanliness and environmental improvement in the context of trachoma elimination programs, including but not limited to WASH interventions. Literature describing interventions aimed at improving at least one of the following were eligible for inclusion in the review:

• Facial cleanliness;
• Personal hygiene more generally (e.g., hygiene/health education, promotion of frequent bathing, promotion of handwashing, construction of washing facilities, promotion of clothes washing or activities to reduce transmission by other potential fomites);
• Environmental improvement, such as (see S1 Appendix for further details):
  1. ○. Provision of water supply, under the auspices of trachoma elimination programs (i.e., standalone water schemes were not eligible), to support the facilitation of facewashing, facial cleanliness, handwashing, and personal hygiene, more broadly;
  2. ○. Control of Musca sorbens populations, including:
     1. ■. Safe human feces disposal (e.g., interventions aimed at improving latrine coverage [provision of sanitation hardware: latrines, child potties] and utilization [progression along the sanitation ladder, cessation of open defection through demand-side or other sanitation programming, or improved sanitation services]), including feces disposal for infants and young children as well as disabled and elderly individuals;
     2. ■. Insecticide spraying;
  3. ○. Reduced crowding (e.g., limiting crowding in sleeping quarters).

We reviewed documentation of F&E-related interventions from grey literature that endorsed the interventions either in isolation or in combination with other interventions.

Comparators: We did not limit inclusion to literature that made comparisons between specific interventions and counterfactuals (e.g., alternative interventions or no interventions).

Outcome measures: Grey literature describing interventions aimed at modifying at least one of the following F&E outcomes (and behavioral proxies thereof) were eligible for inclusion in the review:

• Facial cleanliness outcomes or proxies thereof (e.g., researcher-determined ‘clean face’; absence of ocular discharge, nasal discharge, flies on face; fly-eye contacts; ‘any dirt’ on face; proportion of collected water allocated to facewashing; reported or observed facewashing; reported or observed facewashing practices [e.g., use of soap and/or towel, wiping practices]). Some evidence suggests that such measures have low inter- and intra-rate reliability [27]. However, these metrics have been used in trachoma elimination programs, and in the absence of more reliable measures, these outcomes were incorporated in order to include interventions addressing proxy measures related to facial cleanliness;
• Environmental improvement outcomes (e.g., water availability [e.g., access to and sustainability of water supply; distance to water source, often indicated by time to source or collection time; amount of water collected; proportion of collected water allocated to personal hygiene, handwashing], latrine coverage, latrine utilisation, prevalence of open defecation, fly densities);
• Hygiene outcomes (e.g., bathing practices, washing practices related to potential fomites, handwashing practices);
• F&E-specific behavioral change outcomes or proxies thereof, including:
  1. ○. Changes in perceived risk/vulnerability and severity (including consequences of disease/infection, causal determinants of ill/good health), knowledge (including health/hygiene knowledge–disease transmission and prevention, action knowledge), attitudes (cost and benefit beliefs, affective feelings), beliefs regarding improved practices related to any of the following: facewashing, personal hygiene (bathing, handwashing), sanitation (including adoption of technologies), crowding, and animal husbandry
  2. ○. Changes in availability of resources and/or skills required to perform improved F&E practices (e.g., action capacity)
  3. ○. Changes in F&E-related norms (i.e., social and non-social normative and non-normative beliefs), including [28, 29]:
     1. ■. Empirical expectations: (i.e., descriptive norms) social, non-normative beliefs about what other people actually do;
     2. ■. Normative expectations: (i.e., injunctive or “social norms”) social, normative beliefs about what other people think one ought to do (or ought not do);
     3. ■. Factual beliefs: (i.e., collective habits) non-social, non-normative beliefs about the world; and
     4. ■. Moral norms: (i.e., personal [i.e., non-social] normative beliefs) personal perceptions about what ought to be done (or what ought not be done)
  4. ○. Changes in confidence in performance of improved F&E-related practices at the individual (i.e., self-efficacy) and community (i.e., collective efficacy) levels [30]
  5. ○. Individual- or household-level characteristics associated with adoption of improved F&E behaviors
  6. ○. Outcome or process evaluations of interventions that include behavioral outcomes such as adoption of improved F&E practices.

Types of eligible literature: Grey literature, including technical guidelines and resources, implementer reports and exchanges, formative research reports, program documents, policy and advocacy pieces, progress reports, and white papers were eligible for inclusion.

Exclusion criteria.

Literature was excluded if it was: 1) not published in targeted grey literature databases or on targeted websites, as outlined above, and 2) not provided to the authors via the literature solicitation from ICTC and WHO-CCT member organizations. Routine surveillance reports were not eligible. Papers that were published as conference abstracts or proceedings were also excluded from the review, as the data presented in these types of publications are subject to change when published in peer-reviewed journals and/or as final programmatic reports [31].

Synthesis activity 1: Thematic analysis of endorsed F&E-related intervention activities.

To categorize intervention activities and identify emerging themes related to interventions promoted in the grey literature, we performed an applied thematic synthesis [34]. First, we catalogued all intervention activities endorsed in documents selected for inclusion in the review. Once we listed all intervention activities, we identified emergent intervention themes (i.e., categories of intervention activities). We then developed a matrix in which each column represented a single included document, and each row represented one emergent intervention theme. In order to visualize the results of our thematic analysis, we transformed the matrix into a cluster heat map [35] with each cell representing whether, and to what extent the respective literature endorsed the theme in question. We used a stop-light color system to indicate whether interventions described in the reviewed literature explicitly endorsed the emergent F&E intervention categories (green cells), implicitly endorsed the intervention categories (yellow cells), or did not mention the intervention categories (red cells). Rows of intervention themes were placed in order, according to their level of influence [17], with those least influential in changing behavior (i.e., ultimate/distal influencers) situated at the top of the matrix, and those most influential in changing behavior (i.e., proximal influencers) situated at the bottom.

Synthesis activity 2: Synthesis of documented behavior change determinants (i.e., barriers and facilitators) mediating the adoption of improved F&E-related practices.

Determinants of improved F&E practices are context-specific; however, many barriers and facilitators transcend contexts. We used the Integrated Behavioral Model for Water, Sanitation, and Hygiene (IBM-WASH) framework [36] to synthesize information regarding behavior change determinants that were indicated as known mediators of F&E practices. While we did not track information against the various hierarchical dimensions of the IBM-WASH framework, we did organize barriers and facilitators into the framework’s three overarching dimensions: contextual, psychosocial, and technological determinants.

Synthesis activity 3: Keyword exercise identifying behavioral factors and antecedents addressed through endorsed F&E-related intervention approaches.

Further to the categorization of intervention activities, we classified the types of intervention techniques endorsed as well as the behavior change factors addressed through these techniques. In order to do so, we used the Risks, Attitudes, Norms, Abilities, and Self-regulation [RANAS] approach to systematic behavior change [13]. This framework is grounded in theories of change, explores a variety of behavior change antecedents, or change mechanisms, and tracks them against specific categories of intervention techniques. By developing keywords that aligned with our adaptation of the RANAS framework’s behavioral factors and related domains and sub-domains, we tracked the attributes of F&E-related intervention components against the RANAS framework. We slightly adapted the framework by introducing the creation of enabling environments, and shifting behavioral reinforcement from the infrastructural and ability category to the planning and relapse prevention category, as we postulate positive reinforcement can help with action planning and prevent relapse. Subsequent to keyword creation, we searched each included document for targeted keywords, and recorded whether the behavior change factors and techniques indicated were explicitly, implicitly, or not at all endorsed. Although authors of the grey literature may not have explicitly mentioned some behavior change dimensions, we interpreted and/or inferred whether the F&E-related interventions described did, indeed, attend to certain behavior change factors related to the RANAS framework. We used this approach because we recognized that those designing and implementing trachoma elimination programs may not be using the RANAS or other behavior change frameworks as foundations to guide their interventions. Therefore, they would not necessarily use related terminology when describing intervention content and delivery. For ease of tracking the total number of documents describing F&E-related interventions that touched on the respective behavioral factors and accompanying change mechanisms, we tabulated explicit and implicit results (S3 Table). Also see S3 Table for the operational definition of each dimension that taps to the five RANAS domains.

In order to elucidate the conceptual and behavioral frameworks the trachoma community is using to inform the design of its F&E interventions, we catalogued this information in a table (S4 Table).

Characteristics of endorsed F&E-related intervention activities.

The heat map (Fig 2) suggests that existing interventions address enabling environments, introduce behavior change messages and promotional activities (20 explicitly, 5 implicitly), and focus on improving knowledge (21 explicitly, 2 implicitly) regarding trachoma transmission and prevention. However, in much of the literature, such interventions are not coupled with activities that address more influential (i.e., proximal) behavior change antecedents known to facilitate successful and sustainable change. This is evidenced through the relatively high proportion of green cells in rows 1 to 9 compared to the relatively low proportion of green cells in rows 10 to 16. For example, many documents focused on introducing knowledge related to trachoma (row 9), but few also gave focus to perceptions regarding the amount of water required to wash one’s face and hands (row 12), or self- and collective efficacy (row 16). The intervention activities presented in rows 10 to 16 of the table refer to those intended to:

• improve attitudes toward improved F&E-related behaviors and the skills and opportunities required to carry them out;
• address perceptions regarding: risk of trachoma, self- and collective efficacy, the amount of water required to wash one’s face and hands;
• change normative beliefs (referred to generally in the grey literature as “social norms”), such that they are consistent with improved F&E practices; and
• reinforce improved F&E practices (e.g., modelling of or praising someone for improved practices, use of role models for promotion, recommending to others that they carry out improved practices).

Download:

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

Fig 2. Heat map visualizing emergent themes related to F&E intervention activities.

https://doi.org/10.1371/journal.pntd.0006178.g002

Documented behavior change determinants mediating the adoption of improved F&E practices.

Across the three IBM-WASH dimensions, psychosocial determinants received greatest focus when consideration was given to uptake of improved F&E practices (Table 3). The majority (76%, n = 22) of the 29 categories of cited behavioral determinants portray barriers of improved F&E adoption.

Download:

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

Table 3. Behavioral determinants of improved F&E practices presented in the grey literature.

https://doi.org/10.1371/journal.pntd.0006178.t003

Behavioral factors, antecedents addressed through endorsed F&E-related intervention approaches.

F&E interventions often focus on addressing certain risk and ability factors (Fig 3). Health and hygiene knowledge was the most frequently cited behavior change factor addressed by F&E-related interventions highlighted in the grey literature. Twenty-one (78%) reviewed sources cited these types of information-based components as aspects of their F&E-related interventions (16 sources explicitly, 5 sources implicitly). Action capacity was the second most commonly cited (aggregate) behavioral factor, with 20 (74%) reviewed documents endorsing the incorporation of interventions that address this factor. This is encouraging, as action capacity addresses a more proximal factor of behavior change than knowledge improvement by working to enhance the skills and/or resources necessary to put improved F&E behaviors into action. Perceptions regarding practices typically approved or disapproved of by others, otherwise referred to in the literature as “social norms” (52%, n = 14), and benefit beliefs (48%, n = 13) were other commonly cited proximal factors.

Download:

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

Fig 3. Proportion of endorsed F&E interventions addressing antecedent behavior change factors.

https://doi.org/10.1371/journal.pntd.0006178.g003

As indicated in Fig 3, none of the included literature endorsed F&E interventions that address two essential ability-related behavioral factors–those related to confidence in continuation and confidence in recovering. Related content and delivery would address one’s perceived ability to re-introduce improved F&E practices into daily life after a disruptive event/stimulus or personal setback.

Summary of conceptual and behavioral frameworks indicated in the F&E grey literature.

Seven (26%) of the documents included in the review cited behavioral frameworks upon which their interventions and/or assessments were grounded. S4 Table indicates the frameworks presented, and provides a brief summary of each. Two documents each drew on two different frameworks. Consequently, a total of eight frameworks were indicated. While many of the frameworks present different conceptualizations of the behavior change process and antecedents thereof, nearly all of the frameworks acknowledge that individual and community level knowledge and perceptions are important drivers of the change process. Similarly, several of the frameworks stress that the change process should commence at the grass-roots (i.e., community) level, and be driven or led by the community (e.g., PHAST, CLTS, Positive Deviance approaches).