# Lesson 9: Global Health Applications *Companion-podcast transcript • Sarah & Kiffer* *Office Hours episode to listen to after working through the lesson* --- **Sarah:** In the handwashing case from the lesson, soap was kept for laundry. Asking families to use it on their hands asked them to choose between clean clothes and clean hands. **Kiffer:** So the negotiated behaviour became washing with ash or soap. Ash is free, and the households helped choose it. **Sarah:** Which I think is the method working as it should. A behaviour people can keep up does more good than an ideal one they can't. **Kiffer:** I agree about feasibility. My worry is that a Cochrane rapid review on ash, published in 2020, left its authors uncertain whether ash works as well as soap. **Sarah:** Welcome back to Office Hours. I'm Sarah. **Kiffer:** And I'm Kiffer. We're on Lesson nine, Global Health Applications, and Sarah has the first question for us. **Sarah:** Question one. What does it take for a theory to travel? I suspect a lot of students would say you translate the materials into the local language and you're done. **Kiffer:** Translation matters. In communication terms it's part of encoding, packaging a message so that a group understands it as intended. There are two further adjustments from the lesson. First, you fill in each construct with local content. Second, you check whether the individual is the unit that makes the decision. **Sarah:** For the first one, the condom example from the lesson is a good reminder. A barrier in one place is price, and in another it's what people believe a condom does to the body. For the second, many of our theories picture one person weighing costs and benefits, and in plenty of places a couple, a household or the elders make the call. A lot of that local content already lives in the ethnomedical system, meaning people's shared ideas about what causes an illness, how it's treated, and who the right healer is. **Kiffer:** Right. Let's try it on a made-up case. A program wants more young children in a rural district to sleep under insecticide-treated bed nets in the rainy season, and it's built on the Health Belief Model. Formative research with a local partner turns up several details. Here's the first. Many families say that fever in the rainy season comes from getting chilled by the cold rain. Which construct does that belong to? **Sarah:** My first thought is perceived susceptibility. But these families expect fever every rainy season, so they already feel quite susceptible. What's missing is any reason to think a net would help, because a net does nothing about cold rain. So I'd put it under perceived benefits. **Kiffer:** That's how I'd place it too, though you could argue it also lowers their perceived susceptibility to malaria as a disease carried by mosquitoes. Either way, it's a belief about cause from the local ethnomedical system, and the program has to work with it. A message might connect the two ideas. The rains bring the mosquitoes, and the mosquitoes bring the fever. Here's the second detail. Parents say the nets feel hot and stuffy on humid nights. **Sarah:** That's a perceived barrier. And I'd guess the program also needs a cue to action, something like a reminder on the radio at dusk. **Kiffer:** Good guess. In this made-up district, the community radio does play a reminder at sunset. Now here's the detail I'd like everyone to place. One household has two nets and six people, and the grandfather decides who sleeps under them. Which construct of the Health Belief Model does that belong to, if any? Take a few seconds. *(Pause)* **Sarah:** The obvious answer is a perceived barrier for the mother. She'd like her toddler under a net, and she can't arrange it. **Kiffer:** That's a common first answer, and it misses the more important point. The model describes one person's beliefs, and this decision belongs to somebody else. **Sarah:** So the mother could hold every belief the model asks for, and her toddler could still sleep without a net. That's the second adjustment. Here the household is the unit that decides, and the grandfather speaks for it. **Kiffer:** Right. So the program has to reach the grandfather, and it needs theories that work at that level, such as social norms or gender roles, or a community approach through a meeting of elders. Part of the fix is structural as well. Two nets for six people is a supply problem. **Sarah:** It would be easy to hear the details of cases like this one as a list of odd beliefs that other people have. And Western biomedicine is one ethnomedical system among many. Canadians hold plenty of their own ideas about what causes illness. **Kiffer:** That's the warning from the educational and ecological assessment. The details in our case are invented. In a real district you'd find them by doing the research with local partners, and you'd find a lot of variety inside the district. Cultural knowledge is also a strength to build on. **Sarah:** Question two. Is the epidemiologic transition a rule that countries follow? I think many people picture heart disease and diabetes as problems of wealthy countries, and infections as the problems everywhere else. **Kiffer:** Abdel Omran described the transition in 1971, and it's a useful general pattern. As a rule, it doesn't hold up. The World Health Organization estimates that noncommunicable diseases killed at least forty-three million people in 2021, about three quarters of all deaths worldwide outside the pandemic. Eighteen million of those people died before age seventy, and eighty-two percent of those early deaths were in low- and middle-income countries. **Sarah:** So chronic disease is already a large share of the burden in poorer countries. And a single country can sit at two points on the transition at once, with heart disease rising in its cities while rural districts still lose children to diarrhoeal disease. **Kiffer:** Yes, and sometimes the two patterns meet in the same family. For a planner, that means a program can't assume a single disease profile for a whole country, because the national figures may describe nobody's district very well. **Sarah:** And a health system built for short episodes of acute care will struggle with conditions that need years of follow-up. Migration carries the transition across a border, too, as in the refugee and migrant example. In Canada that connects to what's called the healthy immigrant effect. **Kiffer:** Statistics Canada research has found that, on average, immigrants have lower death rates than the Canadian-born, and that the advantage shrinks the longer they've lived in Canada. The newcomer family in Surrey is one version of that story. **Sarah:** In that case, the grandmother says the children finally look well fed, which in the village was a sign of a family doing well. Students tend to grab that detail as the explanation. Is that a fair reading? **Kiffer:** It's the most memorable detail, which is part of the risk. The same case has two parents on low-wage shift work, a convenience store as the nearest place to buy food, and a supermarket a bus ride away. The ecological point is that a cultural meaning and a new environment interact. My own view is that a program should start with the conditions it can change with the family, such as food access and time, and treat the grandmother's view as something to understand and work with. A program that treats her view as the cause can end up blaming the family for conditions they didn't choose. **Sarah:** Question three. How do you tell where a person is in the AIDS Risk Reduction Model? I'd like to try this one myself. **Kiffer:** Here's a made-up case. Peter is twenty-six. He works at a mine several hundred kilometres from his home village, where his wife and young daughter live, and he gets home twice a year. In town he has a girlfriend. At a peer education session he named the main ways HIV is passed on, and he said HIV is common in the town and that it's a serious illness. He also said the men who get it are the ones who go to sex workers, and that he's safe because he has one regular girlfriend. **Sarah:** Okay. The model's first stage is recognition and labelling. Peter knows how HIV spreads, and he says it's common and serious, so he's recognized the risk. I'd put him at stage two, commitment, and I'd have the program work on his intentions and his self-efficacy. **Kiffer:** Go back to the definition of stage one before you move him on. What exactly does a person have to recognize and label in that stage? **Sarah:** Their own behaviour as high risk. Oh. Peter has labelled other men's behaviour as risky and put himself outside that group. He has the knowledge, and he has perceived severity, but he doesn't see his own behaviour as risky. So he's still in stage one. I treated knowing about HIV as if it were the same thing as labelling his own behaviour. **Kiffer:** That's the slip, and it's a common one, because knowledge is the easiest thing to measure. In the language of the Health Belief Model, what's missing is perceived susceptibility. If formative research found that many men in the town believe one regular partner keeps you safe, that belief is the local content of the construct, and the messages would need to address it in the words those men use. His risk depends on his partners' histories too, along with his own choices. **Sarah:** So knowing the stage tells the program what to work on first. **Kiffer:** It does. Let's move him along. A few months later, Peter says he's decided to use condoms with his girlfriend from now on, but he hasn't raised it with her yet. Which stage is he in now, and what's the closest parallel from an earlier theory? Take a few seconds. *(Pause)* **Sarah:** He's at stage two, commitment. The closest parallels are intention in the Theory of Planned Behavior and the preparation stage of the Transtheoretical Model. **Kiffer:** That's right. And the next step shows what this model adds to the theories it borrows from. Stage three is taking action, which includes enacting solutions, and for Peter the solution involves another person. He may worry that bringing up condoms will sound like an accusation, or like a confession. That's emotion and the interpersonal process, the two pieces this model adds to the Health Belief Model and self-efficacy. A program that helps him at this point might practise the conversation with him, which builds skills and self-efficacy in the sense of Social Cognitive Theory. **Sarah:** And the same kind of negotiation applies with his wife, which takes us back to the couple-level decision in the India case. **Kiffer:** Yes, and it's a reminder that the model's stages describe one person, while the action often belongs to a couple. **Sarah:** Question four. When a global program reports its results, what has it actually shown? Students often read high recall of a message as proof that a program worked. I find it helps to line the evidence up from weakest to strongest. Some programs report what they did, the channels and the materials. Some report reach and recall, meaning how many people saw a message and remembered it. Some report knowledge, like the score in the cell phone games. And a smaller number report behaviour. **Kiffer:** That's a useful ladder. Each step up is closer to health, and each step is harder to measure. Of the four programs in the lesson, the Cameroon campaign is the one that reports behaviour. A published evaluation by Plautz and Meekers in 2007 used three surveys of unmarried young people aged fifteen to twenty-four in Yaoundé and Douala, in 2000, 2002 and 2003. Exposure was measured by asking young people how much contact they'd had with parts of the campaign, such as the radio drama, the call-in show and the peer educators. Those with more exposure reported more consistent condom use with regular partners. **Sarah:** And as I remember it, young people with low exposure improved as well, so the authors concluded that factors besides the campaign had also contributed. There's another catch with measuring exposure by what people remember, isn't there? Someone who already uses condoms might pay more attention to condom ads. **Kiffer:** That's a fair worry with any exposure measure built on recall, since the arrow can run both ways. I'd still call this evaluation more informative than a recall survey alone, and the authors' note about the low-exposure group shows them stating the limits of their own design. **Sarah:** Can we do a harder one, where the actors and the constructs both matter? **Kiffer:** Let's do one. This one is made up. A measles outbreak hits a rural district, mostly among young children who missed their routine vaccination. An international medical NGO and the national Red Cross society run a vaccination campaign, paid for with emergency funds from a bilateral agency and a regional body, and the ministry approves it. Teams go door to door, and in two weeks they vaccinate nine in ten children aged six months to five years. Six months later the emergency money ends. Which configuration of actors should take over for the long term, and who delivers the vaccine then? Take a few seconds. *(Pause)* **Sarah:** The campaign was the crisis configuration, with NGOs delivering on bilateral and regional money. Keeping measles away is long-term prevention, so the World Health Organization coordinates with the ministry, and a public-private partnership helps with access to vaccines. A real example is Gavi, the Vaccine Alliance, created in 2000 to improve access to vaccines for children in the poorest countries. Delivery moves to the ministry's own clinics, with NGOs supporting the demand side. **Kiffer:** That's the configuration I'd expect. Now here's the trap. The district health officer says the campaign reached nine in ten children, so the routine program should keep the campaign's message for the next five years. The message says, measles is killing children in our district, so bring your child to be vaccinated. Using the Health Belief Model, what has changed about the content of the constructs? Take a few seconds. *(Pause)* **Sarah:** Nearly all of it has changed. During the outbreak, families could see sick children, so the outbreak itself supplied perceived susceptibility and severity. The team at the door was the cue to action, and the campaign removed most of the barriers, since the vaccine was free and came to the house. Routine vaccination means walking to a clinic on clinic days for two doses months apart, while the outbreak fades from memory. That message describes a threat people no longer see, and it says nothing about the barriers they now face. **Kiffer:** And there's a twist. The better routine vaccination works, the rarer measles becomes, so perceived susceptibility keeps falling as the program succeeds. The constructs need new local content from formative research for the routine phase: the new barriers, a new cue such as a reminder from a community health worker, and a new messenger, since a clinic nurse has replaced the campaign team. Here are some numbers. Suppose the district has six thousand births a year, and routine first-dose coverage settles at seventy percent. Before the outbreak, about five thousand young children had no dose. **Sarah:** Thirty percent of six thousand is eighteen hundred children a year with no first dose. Counting only children born after the campaign, that's five thousand four hundred in three years, which is more than the five thousand the district had before the outbreak. So the pool is back in under three years. And that's generous. It leaves out the one child in ten the campaign missed, and one dose doesn't protect every child. **Kiffer:** The World Health Organization says coverage of ninety-five percent or more with two doses is needed in every country and community to prevent outbreaks. Even at ninety-five percent for the first dose, the district would add three hundred children a year with no dose. So the campaign's nine in ten shows what the crisis configuration achieved in two weeks. Long-term protection depends on routine coverage of every new birth cohort, with actors and constructs the campaign never had to deal with. **Sarah:** So, back to question four. The campaign figure measures behaviour, and it's still the wrong evidence for the officer's decision about the long term. Now the question where we come down in different places. In Trials of Improved Practices, the behaviours a household tries can differ from the ideal recommendation. In the handwashing case from the lesson, the negotiated behaviour was washing with ash or soap. I think that's the method working as it should. **Kiffer:** Go ahead and make the case for it, and then I'll push back. **Sarah:** A behaviour people can keep up does more good than an ideal one they can't. In that case, soap was kept for laundry, so asking families to use it on their hands asked them to choose between clean clothes and clean hands. Ash is free, and the households helped choose it. That's participation in Freire's sense, with the community helping define the solution. **Kiffer:** I agree about feasibility. My worry is that when you change the behaviour itself, the evidence for the original doesn't come with it. A Cochrane review in 2021 found that promoting handwashing in communities in low- and middle-income countries probably prevents about a quarter of diarrhoea episodes, and in seven of the nine trials behind that estimate, soap was handed out free. A Cochrane rapid review on ash, published in 2020, found fourteen studies, and the authors were uncertain whether ash works as well as soap for stopping the spread of infection. **Sarah:** An uncertain result still leaves room for ash to work. And for these households, the real comparison is ash compared with water alone, because the soap goes to the laundry. **Kiffer:** That's fair, and it's the strongest version of your argument, though the review was uncertain about ash compared with water as well. There's a second problem, too. The global monitoring system for hygiene, run by the World Health Organization and UNICEF, doesn't count ash as soap. Its definitions describe ash as less effective than soap and count a household that uses it as having a limited facility. So a district could change its behaviour a great deal and still show no gain in the share of people with a basic hygiene service, which is the figure the global targets track. **Sarah:** That sounds like a reporting problem for the ministry. The family's health is a separate question. **Kiffer:** Partly. The reporting influences where money goes next, so it matters to families too. Guidance on Trials of Improved Practices also says the team needs a thorough understanding of the technical basis for acceptable practices before it starts, so it can judge what families suggest. My position is that I'd accept ash as a step, say so plainly in the program's reports, and keep the soap barrier on the list of problems to work on with the women's association and with whoever funds the next phase. **Sarah:** I'd still defend the household's choice as a real result of the method. But here's a rule we can both accept. Adapt the content of the constructs and the channels as far as the setting needs. When an adaptation changes the behaviour itself, say so, check what evidence exists for the substitute, and measure something as close to health as you can. **Kiffer:** I'm happy with that rule, because it keeps what the method does well and keeps the program honest about what the evidence covers. Let's pull it together with three things to take away. **Sarah:** First, a theory travels when you fill in its constructs with local content and check who makes the decision. Translation is one part of the first step. **Kiffer:** Second, go back to the definition of a stage or a construct before you place something in it. Knowing the facts about HIV is different from labelling your own behaviour as risky, and the Health Belief Model describes one person's beliefs even when a household decides. **Sarah:** And third, read a program's evidence for what it measured and when. Recall and knowledge sit lower on the ladder than behaviour, and a crisis campaign's coverage says little about routine coverage once the actors and the constructs change. **Kiffer:** If you'd like more practice, go back to the measles district from this episode. Change the births or the routine coverage, work out how fast the pool of unvaccinated children grows, and say what new content each Health Belief Model construct needs for the routine phase. **Sarah:** Next time, it's Lesson ten, High-Risk Populations, Culture, and Health Disparities, where questions about culture, decision makers and local knowledge come home to health promotion within Canada. **Kiffer:** Take care, everyone. **Sarah:** See you in Lesson ten.