Differences between groups based on reach.

We divided the existing programs into two groups based on their reach to assess differences between the programs (above-average reach [n = 44, 60%] versus below-average reach [n = 29, 40%]). In general, the groups had similar characteristics. One eye-catching distinction is that the average costs per ice cleats from the below-average programs are almost twice as large as in the above-average programs. Also, the duration of the programs seems to be longer in the below-average municipalities (on average, almost five months longer). Distribution points also differ somewhat between the groups; distribution via retailers and “other locations” seem to be more frequent in the above-average programs (according to free-text responses, it appears that the “other” category in this group are mostly health care centers and country stores). However, the most striking difference between the two groups of programs is their scale; programs that reached a larger share of the population purchased many more ice cleats, which implies that they had the ambition or resources to cover a larger share of the population. Importantly, we note that data from both groups indicate that about 90 percent of all purchased ice cleats were eventually distributed (Table 1).

Risk-aware and in control.

It was evident that there was a common concern about belonging to a vulnerable group and how falls could lead to serious injuries.

“When you are younger, you have so many margins in life. If you are careless with one thing or another, you will not notice much. If you are careless now, then there can be consequences.”

(Gym group).

At the same time, they felt confident that they could deal with the risks of falling.

“We are getting wiser and wiser. You understand that you are past the age of 25. I mean, when I was 25, I could fall on the ice, and nothing happened. I just got up again. If I fall on the ice now, I will break myself!”

(Gym group).

Their fear of falling would not lead them to live a sedentary lifestyle, because a sedentary life was considered harmful. Instead, they shared a strong desire to stay active, which could contribute to a healthier and longer life.

“You need to stay informed about how important it is to be active and how dangerous it is to be sedentary.”

(Bridge group).

Ice cleats—A necessity.

Belonging to a vulnerable pedestrian group, together with the desire to uphold an active lifestyle, were strong motivational factors for ice cleat use. The occurrence of snow and ice would not hinder them from daily walks or running other errands. They do not have the time to wait for general countermeasures like snow removal, gravel, or road salting, which motivates the use of personal anti-slip equipment such as ice cleats.

“I am so scared to fall! Because where I live, these big streets, spreading gravel is the very last thing they do.”

(Bridge group).

The participants agreed that the use of ice cleats made it possible for them to cross icy roads and walkways, and they had the daily habit of assessing the environment before exposing themselves to ice and snow.

“Of course, you look at the thermometer every morning and assess the risk before you pick up the newspaper.”

(Gym group).

They noted that user-friendliness is a key determinant for continued use. However, they also had poor experiences with ice cleats that deteriorate quickly over time.

“… but when you walk on gravel roads, the rubber gets worn down after three or four weeks. Then you have to go and buy new ones…”

(Bridge group).

Appreciates municipal ice cleat programs.

In general, the interviewees welcomed the idea of a municipal ice cleat program. They estimated that the municipal costs would be low and especially appreciated the signals that a program of this kind would send.

“The municipalities do not have to hand out the best of ice cleats. The programs can instead create awareness; “Now we think of you and your well-being, take the chance!”“

(Gym group).

However, they saw some potential barriers. Free ice cleats would probably trigger people to pick up a pair, but they are probably going to be left unused. There were also reservations of a more ideological nature.

“I also think it is a bit of a shame to take responsibility away from people just because they are a little older… I am hesitant about such programs.”

(Bridge group).

The main hesitation to program introduction was the distributed ice cleats may not be user-friendly. They opposed programs that determine which ice cleats to use. A few participants also noted that they would prefer shoes with studs over ice cleats.

“There are so many different types of ice cleats, so you want to choose the ones that suit you best.”

(Gym group)

Inputs.

A formal decision to implement the program is required at the municipal level, as governed by the Swedish Local Government Act [18]. According to our survey, 57 municipalities had considered implementing an ice cleat distribution program but decided not to do so. The most common reasons provided for not introducing programs was that such programs were not considered to be within the scope of the municipalities responsibilities (n = 19), that economic ventures for the municipality were used elsewhere (n = 13), that the municipality did not have the resources required to introduce such a program (n = 9), or that a motion was put forth but rejected politically (n = 9). There were also municipalities that pointed out there is a lack of evidence for the effectiveness of ice cleat programs (n = 7). Some municipalities also referred to a paragraph in the Local Government Act that states that municipalities and regions must treat all citizens equally unless there are objective reasons to provide special services to certain groups (n = 6). Thus, many potential programs may fail even before they are started if these issues are not properly navigated.

A municipality that decides to implement an ice cleat program will also require a sufficient amount of ice cleats to distribute to their target population and other resources [8,9] (e.g., allocation of time to staff to perform the program activities, sufficient funding to complete the activities, appropriate knowledge, devise communication and distribution plans, etc.). One may assume that such resources are appropriately allocated upon a formal decision to implement the program. However, external factors may hamper implementation. For instance, one municipality reported that their ice cleat order was stuck in customs for several weeks (U. Stefansson, personal communication, August 21, 2019). Such delays may affect the program’s success, given that the timing of the distribution efforts may be crucial to ensure a sufficient reach before the beginning of the winter season.

Activities.

Assuming the required inputs are sufficiently secured (Fig 1), the program staff must implement the distribution plan, including when, how, and where to distribute the ice cleats to the population. According to our survey, most intervention municipalities distributed ice cleats at municipal-owned buildings (such as libraries or city halls; Table 1). Others relied on partnerships with retailers such as stores, pharmacies, and other businesses that provide ice cleats. The rest relied on partnerships with pensioner organizations or distributed ice cleats at other locations (e.g., during an annual “falls prevention week”). As noted above, it appears that municipalities with above-average reach tended to partner with, for instance, retailers and health care centers.

Another necessary condition for a successful distribution is that the target population is aware of the program’s existence [8,9,19] and that they are eligible for a free pair of ice cleats. A majority of the intervention municipalities communicated the program using multichannel campaigns [19] like traditional media, social media, mail, personal communication, etc., to create knowledge and spread awareness about the program. Proper communication about how and when to use ice cleats may encourage more individuals to use them (e.g., how to attach them to regular footwear, disassemble when entering indoors, etc., which can reduce perceived barriers to use [20] and improve self-efficacy [21]). If the communication about the program is supplemented with convincing messaging about the problem (e.g., lack of mobility, risk of falling) and how ice cleats may help combat these issues, communication theory and the Health Belief Model would suggest that the program may persuade more individuals about the utility of ice cleats [19,20].

Outputs.

Assuming the activities are performed successfully, the expected output is that the ice cleats are distributed to a meaningful share of the target population. As noted above, we found that the average intervention municipality distributed ice cleats to about 40% of the eligible population residing in the municipality. One factor that appears to increase reach is the availability of many distribution sites around the municipalities combined with a communications plan covering the entire target population. For instance, Gothenburg reached approximately 60% of its older citizens by mailing coupons to all eligible citizens and spreading the pick-up locations throughout the city [7]. On the other hand, one municipality limited its distribution in a one-week falls prevention festival and reached as little as 1 percent of the target population. Overall, our survey data suggest that the main predictor of reach is the number of purchased ice cleats per population size (9 in 10 ice cleat pairs will eventually be collected; Table 1). While factors related to communication and distribution may play an essential role in increasing the program’s impact, it appears that the main barrier to widespread distribution is a lack of scale; shortage of resources, and/or the lack of ambition to reach the entire population.

Effects on initial behavior change.

At the population level, the intervention impact is a product of the effect on the ice cleat use and the causal effect of ice cleat use on relevant health outcomes (e.g., injury rates, increased walking). Thus, a key component that will affect the magnitude of the intervention effect on all outcomes is the increase in new ice cleat users. This increase is not necessarily equivalent to the number of people who obtain a free pair of ice cleats.

The Health Belief Model suggests that individuals who identify themselves as fall-prone are assumed to be more likely to search for risk-reducing measures to prevent a fall, suggesting that fall-prone individuals may be more easily convinced by the benefits of the intervention and thus act in line with the program [20]. Another key motivational driver for the use of safety equipment is the subjective perception of its efficacy [22]. However, individuals who share these perceptions may also be more likely to use ice cleats without a distribution program. Ceiling effects are also a potential concern, as a program cannot possibly have an effect if everyone in the population already uses ice cleats. According to a national survey from 2014, about 40% of the Swedish population above the age of 65 report seldom or never using ice cleats [23], which implies that there are still individuals who may benefit from a program. However, usage rates tend to be higher in municipalities with more snow days per year (closer to 75% self-reported use) [23], which may limit the potential impact in high-risk municipalities.

In addition to these conceptual uncertainties, there is also no direct evidence to support the claim that a meaningful proportion of current non-users will start using ice cleats as a consequence of a distribution program. Indirect evidence can, however, be found in a quasi-experimental evaluation from Gothenburg, where the rate of pedestrian fall injuries among older adults appears to have decreased as an effect of the program [7]. Still, several of our interviewees noted that ice cleats could be challenging to apply and pointed out that decision-makers should focus on distributing user-friendly devices. Otherwise, there is a risk that the collected devices will be left unused.

Maintained behavior change.

Another key aspect of behavior change is its maintenance (i.e., longevity). Upholding a behavior change in the long term after an intervention appears to be difficult in general [24,25], and the behavior change associated with ice cleat programs may therefore be short-lived [19]. Our interview data support this notion, especially if the ice cleats are of poor quality or difficult to apply. On the other hand, our interviewees also noted that ice cleats could enable an active and safer lifestyle during the winter, which points to a potentially strong motivational driver for continued use. A free-text response to our survey from one municipality also showed that recipients came back asking for a new pair of devices in exchange for handing in the old broken ones. The demand for ice cleats led to the municipality distributing more devices than there were citizens in the eligible age group (approximately 1.1 pairs per older citizen). In addition, evidence from Gothenburg suggests an effect on injury rates may be limited to the first implementation year [7]. Together, these pieces of evidence indicate that individuals may change their behavior initially and then return to their normal behavior or that the ice cleats tend to break after a while and need to be replaced.

The perceived quality of the distributed ice cleats and their user-friendliness may play an important role in maintaining any initial changes in ice cleat use. Previous research suggests that ease of use and perceived utility may influence the perception of ice cleats [26,27]. A good initial experience is also likely to increase the chances of prolonged use [24], which may lead to a long-term behavior change in subsequent winters. It was also stressed by our interviewees that no matter what type of device, the ice cleats must be user-friendly and easy to apply. According to our survey, most municipalities distributed whole foot devices (72.9%; Table 1), which appears appropriate given evidence from previous preference studies that suggest that whole foot and heel devices are the preferred types of ice cleats [27–29]. Despite this apparent alignment between program designs and preference data, our interviewees advised caution against this type of design as it deprives individuals of the opportunity to freely choose devices of their choice, which can hinder new users from collecting and using the ice cleats. However, they noted that they would recommend them to friends if they found an appropriate device. These results are consistent with a previous survey of ice cleat users in northern Sweden [30]. Behavior change theory also suggests individuals are more likely to follow a recommended behavior if it is suggested by trusted peers [21]. Thus, providing high-quality and user-friendly ice cleats may increase the magnitude of a program’s impact on ice cleat use. It may also have other spin-offs, such as increased reach and use via communication between peers.

Effects on health outcomes.

A key hypothesized impact of the ice cleat programs is a reduction in the risks of ice and snow-related fall injuries. There is a strong theoretical argument in support of a causal effect on fall risk as long as ice cleats are worn, given that increased friction decreases slipperiness [27,31,32], which is the key mechanism behind ice-related falls. We also identified two empirical studies supporting this claim. One was a randomized trial on a northern US sample of older, fall-prone adults, which indicated approximately a 50% decrease in fall risk and a 90% decrease in risk of injurious falls. The other was an observational study from a general adult sample in northern Sweden, which showed a 37% lower fall risk per kilometer among ice cleat users. Overall, this evidence supports the assumption that the use of ice cleats decreases the risk of fall-related injuries during winter conditions [29,30].

Another potential effect is an increase in walking during the winter months. Our interviewees expressed an understanding of the positive health effects of upholding an active lifestyle. If the ice cleat programs increase walking, they may also affect general health and well-being [33,34]. However, the empirical evidence of the effects of ice cleats on walking is not as strong as the evidence for an effect on injury risks. We found one study that indicates that ice cleat users walk more, but it is unclear whether ice cleats lead to additional walking or if active people are more interested in using ice cleats [30]. It is worth noting that an increase in walking means an increase in risk exposure, which may offset the number of injuries prevented at the population level. However, the randomized trial reported in reference [29] used diary-days, not distance walked, as their exposure denominator and still showed a considerable impact on injury rates, which implies that the potential impact on walking does not offset the effect on injuries completely. Overall, our analysis of the evidence on health effects suggests that ice cleats are likely to be beneficial if worn.