Using Data to Guide Action in Polio Health Communications: Experience From the Polio Eradication Initiative (PEI)

Abstract

Health communication is increasingly considered a priority element of investments and interventions intended to improve personal and public health (Piotrow et al., 1997). But a prevailing focus in health communication on information, education, awareness, and knowledge—and their assumed relation to changing behaviour among target individuals or households—can underestimate the complexity of wider ecological conditions that influence and limit individual, household, and even community choices and capacity to choose. Experience from the Polio Eradication Initiative (PEI)—drawing on evidence from the India and Nigeria country programmes—provides some insights into how the health communication interventions can be strengthened through the adoption of a more holistic ecological model of people and their health-related behaviours analysed in the context of larger social, economic, political, and cultural forces (see, for example, Kelly et al., 2008). In particular, polio eradication health communication offers useful lessons in the importance of generating and using data of sufficient quality to enable a more ecological analysis—combining and measuring specific communication inputs and epidemiological “outputs.”

Notes

¹For the purposes of this article, the main elements of “health communication” are the following: mass media, advocacy, social mobilisation (defined as the encouragement of popular or community participation for collective benefit towards a common goal), programme communication, and interpersonal communication. Except where more specifically distinguished, the term “health communication” (or “communication”) will be used throughout to refer collectively to these activities, recognising the distinct purposes and contributions of each type of activity.

²Social marketing defined as the application of marketing principles to programmes design to influence human behavior for the purposes of social benefit rather than commercial profit.

³Some studies have claimed considerable impact in health communications (e.g., Porter et al., 2000; Seidel, 2005). Meta-analyses have, by contrast, shown negligible or small effects of health communications (e.g., Hornik, 2002; Huang, Hui, & Kahn, 2007; Snyder et al., 2004; Snyder & Hamilton, 1999).

⁴Methodologically, the choice of India and Nigeria is influenced by their continuing prominence among remaining polioendemic countries, the prominence of emerging health communication programmes in both countries, and the availability of qualitative (direct author or reported observations) and quantitative (documented) data on PEI.

⁵The knowledge-attitude-behaviour model; the behavioural learning theory model; the health belief model; the social cognitive theory model; and the theory of reasoned action (or theory of planned behaviour) model. A central problem, however, is the lack of empirical support through research-derived data that validate accuracy and impact at higher levels.

⁶Additionally, it would be important in applying and assessing these models to distinguish between “one-off” behaviour changes (such as inoculation) and sustained lifetime changes (such as diet).

⁷For example, engagement of high-profile celebrities; use of mass media, TV, and radio spots; and print materials such as posters and leaflets.

⁸As earlier, communication here includes activities also described as social mobilisation. Social mobilisation often involved somewhat formulaic meetings and public ceremonies (for example, in Nigeria, rallies, public address systems, football matches, fanfares, flag-off ceremonies). Interpersonal or programme communication activities were increasing as a proportion of total activities, but they remained relatively smaller scale.

⁹In this article, the period 2000–2005 is described as a period of crisis for the PEI. This period is epitomised by the suspension of oral polio vaccine (OPV) delivery programming in Nigeria between 2003 and 2004, which resulted in increased cases in Nigeria and exportations to countries that previously had interrupted indigenous WPV transmission.

¹⁰Local spread occurred elsewhere in India, for example, in West Bengal, Rajasthan, Gujarat, and Haryana.

¹¹The six high-risk Nigerian States in 2006 were Kano, Katsina, Jigawa, Bauchi, Zamfara, and Kaduna.

¹²For example, given the size of birth cohorts and the target population, the 0.7% of children unreached, unvaccinated, or inadequately vaccinated could hold up India's eradication goal and set back the enormous progress made (UNICEF, 2003).

¹³In April 2002, the Technical Consultative Group (TCG) of the global polio programme had called for an urgent review of communication activities and evaluation of progress in addressing social mobilisation needs for the PEI.

¹⁴The Pakistan PEI programme, for example, employed Gallup to conduct such surveys regularly through the late 1990s.

¹⁵Most [communication] training seems to be related to polio facts, data, and “convincing” strategies, but for the most crucial aspects of communication—the difficult aspect of negotiating and reasoning and positively engaging and facilitating group processes, however, there appears to be little training (UNICEF, 2007).

¹⁶Including “unaware of the polio programme,” “out of town,” and “too busy to take child to vaccination booth.”

¹⁷In 2007, “refusal” as percentage of missed houses was recorded as 0.17% (Polio Communication TAG, 2007a). In 99.44% of cases of missed children, the causes were either that the house was locked or that the family was out of the village.

¹⁸It should be noted that progress towards a more ecological approach to data gathering and analysis in and for health communication in PEI, whilst notable and welcome, is by no means perfect. A variety of wider ecological variables, such as political context—which are more difficult to capture in quantitative data, and more difficult to thus correlate robustly with objective changes in vaccine delivery and immunologic status of target groups—are starting to be included in surveys and other ground-level analyses, but they continue to be developed.

¹⁹It is now suggested that suspension of polio programme activities in 2003–2004 in several northern Nigerian states should be understood as the product of a complex interplay of factors and political objectives (Clemens, Greenough, & Shull, 2006; Jegede, 2007; Obadere, 2005). Rather than indicating a religious nature in noncompliance, high rates of missed children can be interpreted as the result of systematic exclusion of communities from a range of political, social, and economic resources.

²⁰For example, India PEI programme data disaggregate respondent households into those who accept polio drops and would go to the booth; accept polio drops but whose interest in taking children to the booth might wane; are indifferent about polio drops and unaware of date, time, and venue for booth vaccination as well as home visits; reject polio drops because of misconceptions, mistrust or rumors; or all of these.

²¹Such as, “Why the need for NIDs again?; Why the need for repeated doses?; the concern and doubts of parents with children below 3 months (e.g., my child is too young to be immunized, and there might be side effects); the concern of parents with children above 2 years old (e.g., My child has already received greater than 5 doses, so why the need to get more doses?); misconceptions that polio drops would cause impotency/sterility in children; and rumour that poor quality vaccine was administered to minority groups” (UNICEF, 2003). Beyond the analysis of concerns relating to missed children, noncompliance, or both, the India programme sought to develop methods of understanding “conversion factors”—that is, assessing the negotiation approaches of community mobilisers at the household level to identify what specific aspects of an approach appear to provoke a positive shift in the household response (Polio Communication TAG, 2004b).

²²Children who previously had not received a dose of OPV.

²³In cases where attempt was made to assess the impact of mobilisation activities, methodological approaches frequently were limited to “before-and-after” tests, with little attention to possible confounding factors of secular trends or controls.

²⁴Early approaches to impact measurement for the India social mobilisation network in Uttar Pradesh were critically appraised on these grounds: “As with many communication and social mobilization interventions in developing countries, the SM Network primarily measured its output in terms of activities or processes. The drawback from its initial year of operation was the lack of systematic monitoring and evaluation of impact” (UNICEF, 2003).

²⁵Polio programmes have recorded high rates of awareness and supportive intention and yet continued to fail to eliminate circulating WPV.

²⁶Possible indicators included the following “percentage of caretakers of infants under 1 year who know correctly when the next immunisation is due; percentage of caretakers of infants under 1 year who know the number of visits needed to complete childhood immunization; percentage of vaccinators who know how to recognise AFP and where such a case should be reported; percentage of caretakers of infants under 1 year who know the number of visits needed to complete childhood immunisation; percentage of caretakers of infants under 1 year who know where to take their baby for routine immunisation; percentage of caretakers at NIDs knowing that NIDs do not replace routine immunisation; percentage of caretakers at NIDs who are advised about routine immunisation during NIDs; percentage of vaccinators who know how to recognise AFP and where such a case should be reported; percentage of vaccinators who can correctly explain how to interpret and use vaccine vial monitors (VVMs) on polio vaccine vials; percentage of district/subdistrict plans that map resistant or difficult groups, including “zero dose” children, and propose strategies for reaching them.”

²⁷In programmatic terms, “X” marked houses were where an eligible child or children did not receive a dose of OPV during the polio round. A follow-up communication/vaccination team—known as a “B” team—then visited the X-marked house to encourage OPV acceptance. If the child/children received OPV, the house marking was changed to “P.”

²⁸Temporary fixed site location for vaccination scheduled in conjunction with the OPV campaigns each round.

²⁹Community mobilisers (CMCs) were placed in the highest risk and often lower coverage blocks in districts with circulating WPV and large numbers of missed children.

Additional information

Notes on contributors

Lora Shimp

Lora Shimp is currently working for the USAID-funded Maternal and Child Health Integrated Program, managed by JHPIEGO and John Snow, Inc., Washington, DC, USA.