This paper explores the outcomes of equity-oriented, web-based, consumer health information from the perspective of young children’s mothers who have a low socioeconomic status (SES). Equity-oriented interventions attempt to move toward equity, that is, reduce inequalities (systematic differences among social groups), though full equity may never be achieved [1]. Education and income are among the most important indicators of SES and are strongly associated with individual and population health status [2-4]. With respect to web-based consumer health information, the literature suggests the following stereotypical inequality: although information in general can help people of low SES, it mainly offers benefits to people of higher SES, that is, those who are the most educated and wealthiest. It is unclear whether equity-oriented information, that is, trustworthy information provided in a format that is easy to read, listen, or watch, can offer equal benefits to all, including low-SES mothers.

Regarding general outcomes of web-based information at the population level, the literature suggests mixed positive and negative outcomes. First, health services research suggests that browsing the internet leads to a decrease in unnecessary calls and visits to health professionals and helps to optimize services utilization [5,6]. Specifically, consumers’ use of trustworthy sources of web-based health information is associated with improved knowledge, empowerment, self-care, engagement in health care, health outcomes, and quality of life [6-12]. Second, exposure to web-based information can lead to negative outcomes such as worries and anxiety, for example, cyberchondria, deterioration of the patient-provider relationship, and unnecessary visits to the emergency room [13-18].

However, 4 literature reviews have shown that specific outcomes of information are rarely researched [19-22]. Typically, information outcomes are diluted in the outcomes of educational programs and communication with professionals [23-25]. With respect to the health of the mother and child, studies have examined the outcomes of lay pediatric information in the context of the effectiveness of parent-child professional communication and patient education programs. The studies concluded that the parents’ level of health literacy is positively associated with the quality of pediatric care, compliance with medical interventions, and child health outcomes but negatively associated with medication errors and visits to emergency rooms [26,27]. This led to the inclusion of the Low Health Literacy Universal Precaution Principle in medical education and continuing professional development programs [28-30], which recommends the universal provision of clear lay information in all patient-clinician encounters.

Such principles are essential for web-based consumer health information because the proportion of people with a low level of literacy is substantial; for example, more than 50% of Canadian adults have a low literacy level [31]. People with a low literacy level can read, listen to, watch, and understand one-idea sentences in plain language, but they arguably face difficulties in finding trustworthy web-based information, critically appraising it, and understanding nuanced ideas or specialized language [31]. It follows, therefore, that their web-based health information literacy level is also low, which comprises computer literacy, information literacy, and health literacy. This is particularly important for parents because their low literacy level is detrimental to child health education, healthy behaviors, health, and medication [32-34]. Parents with low education and low income, hereafter referred to as low-SES parents, typically have a low literacy level, yet they have greater information needs compared with their higher-SES counterparts [20].

In the information sciences, we found 5 studies that focused on parenting information needs and information-seeking behavior and included low-SES parents. Two surveys, one in Australia and one in Switzerland, found no statistical difference in internet utilization between low-SES parents and other parents [35,36]. Two qualitative studies on mothers’ information needs (mostly middle-class mothers) suggest that practical information on mothering produced by professionals is valuable and helpful [37,38]. One study found that all low-SES mothers search for health information on the internet [39].

In addition, little research has specifically focused on outcomes of web-based parenting information reported by low-SES mothers [40,41]. We found little evidence (and some of it contradictory) on parental perception of outcomes of web-based parenting information, namely information on child education, development, health, and well-being. One study suggested that all parents reported a similar degree of satisfaction with web-based parenting information in general [42]. Five studies suggested that trustworthy web-based information, that is easy to read, listen to, or watch, can improve the quality of life of parents of low SES (including refugee and homeless) and have positive family, economic, and social impacts [43-47]. In contrast, 4 studies showed that barriers persist regarding the acquisition, cognition, and application of web-based parenting information, which are associated with information-related inequalities, perpetuating the digital divide between low-SES and high-SES parents [40,48-50].

However, no study has compared the low-SES parents’ perception of outcomes of equity-oriented web-based information with that of other parents. In light of this, this study aims to (1) compare the low-SES mothers’ perception of outcomes of web-based parenting information with that of other mothers, and (2) explore the perspective of low-SES mothers on contextual factors and information needs and behavior associated with these outcomes.

This study is based on a partnership between Naître et grandir (N&G) and McGill University (Information Assessment Method, IAM). N&G is funded by the Chagnon Foundation, a philanthropic organization that seeks to prevent poverty. N&G produces a magazine and a website with an email newsletter (in French) for parents of 0- to 8-year-old children.

N&G seeks to raise societal awareness of the importance of early child development for enabling conditions of educational success. It provides free, independent, and trustworthy parenting information to valorize, inform, educate, and equip parents and families of young children, especially among vulnerable populations, for example, tips for parents, dos and don’ts validated by experts. Specifically, N&G provides web-based parenting information content that is easy to read, listen to, and watch, specifically webpages with corresponding videos, podcasts, and computer-audio-assistant highlighting sentences read. The website content is organized by age groups and topics.

N&G is widely read by French-speaking parents across Canada, the United States, and more than 100 other countries. In the 2018 calendar year, 61.6 million N&G webpages were viewed during 35.3 million visits to the N&G website by 15.2 million unique internet protocol (IP) addresses across the world. Among those, 20.2 million webpages were viewed during 11.2 million visits to the N&G website by 3.3 million unique IP addresses in Quebec. More than 213,000 N&G weekly newsletters were emailed to parents in Quebec. SOM Recherche & Sondages (personal communication, 2015) conducted a survey of a representative random sample of the population of parents of young children in Quebec, which showed that 82% of respondents knew N&G and that 76% consulted it.

We followed an organizational participatory research approach that blends action research and organizational learning to engage organization members to improve practice [55,56]. Researchers and N&G editors partnered throughout the study, developing the research questions together, and making other key research decisions jointly regarding the collection and analysis of data and the interpretation and dissemination of results.

We used a mixed methods convergent design [57,58]. Mixed methods are crucial for unpacking complexity in research on poverty and vulnerability [59]. In our study, mixed methods were justified because our quantitative data collected with the McGill IAM questionnaire reflect the importance of the information outcomes perceived by low-SES parents, and qualitative data were required to gain insight into the mechanisms underlying these outcomes. Thus, to respond to our first research question, we conducted a 3-year longitudinal observational quantitative component of the mixed methods study. To answer the second research question, we conducted a qualitative interpretive study on parental characteristics, context, information needs, and information-seeking behavior (the qualitative component of the mixed methods study). Then, we integrated these components via the assimilation of quantitative and qualitative results [58].

In line with the literature on patient-reported outcome measures, that is, measurement from patients’ perspectives rather than from medical or biological ones [60], we focused on parent-reported information outcomes.

The quantitative and qualitative components were integrated in two ways. First, they were integrated using the conceptual framework [57]. Quantitative observations provided results on low-SES mothers’ reported outcomes compared with other mothers. Qualitative observations produced complementary results on low-SES mothers’ information outcome–related contextual factors, information needs, and information-seeking behavior.

Second, quantitative and qualitative results were assimilated as follows [58]: the principal investigator transformed quantitative results into a narrative, which was combined with an interpretation of qualitative results for building a comprehensive ideal type of low-SES maternal perception of outcomes of web-based parenting information, including outcomes-related factors. All coauthors reviewed the ideal type. According to Weber [72], better understanding can emerge from the scientific construction of an ideal type; here, ideal refers to ideas, not to an ideal to achieve. An ideal type does not exist in the real world under this pure, absolutely ideal form [72]. It integrates common aspects of a phenomenon in a comprehensive and coherent manner. This integration adds value to qualitative and quantitative results alone.

Over the 3-year study period, 1889 participants completed 2447 IAM questionnaires (on average 1.3 questionnaires per participant; range 1-12) pertaining to 683 distinct N&G webpages (on average 3.6 questionnaires per page; range 1-26). Participants were geographically dispersed across 352 of the 420 Quebec postal codes (352/420, 83.8%). Participants’ demographic information, IAM ratings, and comments are detailed below.

Among participants 89.31% (1687/1889) were aged between 25 and 44 years, 94.12% (1778/1889) were living with a partner, 85.23% (1610/1889) were living full time with their child; 11.54% (218/1889) had an annual family income less than Can $40,000 (US $30,070), and 4.71% (89/1889) did not complete high school or had only a high school diploma (grade 12). Compared with the 25- to 64-year-old Quebec population, participants had a higher level of education (Table 1). Compared with couples with children in Quebec, participants had a lower level of income (Table 2). Among participants, 77.77% (1469/1889) provided a valid postal code, and according to the Material and Social Deprivation Index [73], 70.30% (1328/1889) of participants had low levels of education and income and 35.2% (665/1889) of other participants lived in deprived areas.

The types of information outcomes are detailed in Table 3. Specifically, all ratings (2447/2447, 100.00%) reported the participants’ intention to use N&G information for themselves or for their child, at least in a conceptual manner. In 97.79% (2393/2447) of ratings, participants expected that information use would lead to health and well-being benefits for themselves or their child.

Figure 3 displays the estimated group differences in proportions for these outcomes, along with 95% CI. The comparison of the different types of intention to use and expected health and well-being benefits of N&G information between the IAM ratings of the participants with low levels of education and income (n=50) and those of the other participants (n=2397), suggested that 5 of 11 outcomes were not inferior for the former group (items a, d, e, g, and k in Figure 3), including 3 being superior (items a, g, and k in Figure 3). The results do not provide conclusive evidence for differences (or the absence of such differences) regarding the 6 other types of outcomes (items b, c, f, h, i, and j in Figure 3). In other words, N&G information was more likely to help participants with low levels of education and income to better understand something (mean difference 0.144; 95% CI 0.007-0.253), decrease worries (mean difference 0.171; 95% CI 0.036-0.302), and increase self-confidence in making a decision with someone else (mean difference 0.115; 95% CI 0.015-0.249).

The ideal type, Kate, is a 30-year-old single mother with 2 children who are 2 and 6 years old. She has completed secondary school and is interested in acquiring professional education. She accesses the internet multiple times daily using her cellular phone. Her annual income is Can $20,000 (US $15,234), and thus she cannot afford additional internet fees to use on a tablet or computer. However, she has easy access to the internet when needed by using her neighbors’ and relatives’ Wi-Fi connection, with their permission, and by using free Wi-Fi services in the mall, a local community organization, and the public library. She says, “I take my phone everywhere and can be on the internet almost any time. It’s mine and is my link to friends, parents, and the world.” When she wants to know something about anything, her first reflex is to look at her phone.

When she needs information regarding the education, health, and well-being of her children, she starts by browsing the top-listed websites on an internet search engine. She compares webpages on a topic from multiple websites, seeking coherence and corroborating information, and confirms her findings using trustworthy governmental and noncommercial websites such as N&G. Usually, she finds relevant and understandable information on N&G that supports what she is planning to do or teaches her something new. Using N&G informational content for her children often decreases her worries and leads her to expect improvement regarding her children’s health or well-being and in her interaction with health or social care professionals. Specifically, she finds that N&G provides valuable information about child behavior, development, and education.

In addition, she seeks complementary information to her Mom-like-me social media group, specifically by a group member tagged as a mother and a nurse. Concomitantly, she calls her mother and her best friend, who has children of the same age as hers. In this way, she receives valuable reassurance and emotional support from relatives who know her and her children well. Finally, she considers all of the information she has obtained and filters it through her values, preferences, and financial capabilities to make child-related decisions. She calls the HealthInfo 24/7 phone line or consults with health professionals in the case of a perceived emergency, when she has doubts regarding the information she has obtained, or when she faces contradictory information and uncertainty.

Results suggest that in families of young children, there are few differences between low-SES mothers compared with higher-SES mothers with respect to the perceived benefits of equity-oriented, that is, trustworthy and easy-to-read, listen to, and watch N&G web-based parenting information (situational relevance, cognitive impact, use, and subsequent health and well-being outcomes of information). According to our qualitative results, this hypothesis is limited to nonurgent problems because participants primarily turn to professionals for emergency and risk-related information needs. This hypothesis can be tested in future research with other websites using the web version of IAM, or the mobile version for smartphones released in January 2019.

Specifically, these results do not support our literature-based initial hypothesis that participants’ level of education and income would be associated with information benefits. Our results suggest the following hypotheses for future research. Compared with other mothers, low-SES mothers might be more likely to report that web-based parenting information that is trustworthy and easy to read, listen to, and watch helps them to better understand child-related issues and to be less worried and more confident in making decisions with someone else, for example, a professional. They may also report that web-based information helps them to discuss issues with someone else, for example, a relative, and do things differently, better manage or prevent a child behavior problem, improve their health or the health of their child, and be satisfied with professional services.

The results from this evaluation of specific N&G information webpages advance the knowledge on information outcomes. Although mixed methods in research on poverty are expanding rapidly [59], our mixed methods study is the first to compare the outcomes of web-based parenting information from the perspective of low-SES mothers seeking specific child-related information versus other mothers. In this study, the advantages of mixed methods are illustrated by the synergy between our quantitative results on how often information outcomes are reported by low-SES mothers compared with mothers of higher SES and our qualitative results on why and how low SES mothers retrieve and corroborate information.

The low-SES mothers we interviewed frequently access the internet. Moreover, the results of a recent survey of a representative sample of the New York State population (n=1350) suggested that the level of health literacy is not associated with the volume of internet utilization [74]. The low-SES mothers we interviewed considered websites and web-based forums as information sources that are helpful in everyday life. Some relied on their social network, for example, relatives, to obtain information and emotional support before seeking information on websites and web-based forums and from care services. This is congruent with the literature [75-79].

Furthermore, our participants might have been atypical for two reasons. First, they volunteered and, thus, may be biased in support of N&G. Second, although McCloud et al [48] suggested that people of low SES in the United States may experience difficulties accessing the internet despite technical support (provision of computers, home internet connection, and technician help), all of our participants reported that they were able to access the internet easily and none reported technical barriers.

Our study has three main limitations. First, for the quantitative component, we used a convenience sampling strategy (self-selected volunteer participants), and participants did not complete an IAM questionnaire each time they visited the N&G website. This limitation probably led to an overestimation of positive outcomes (social desirability bias). However, assuming this bias influenced all participants in the same manner, regardless of their levels of education and income, this limitation did not affect the statistical analysis. Second, for the qualitative component, about half of the participants were just above the poverty line or the low level of education threshold, which certainly influenced the results. Third, parental efficacy was not examined in our interviews, but this might have affected our data, for example, information-seeking behavior. Numerous studies have shown that poverty has a detrimental effect on parenting, including perceived parental efficacy [80-82]. The concept of perceived parental efficacy is defined as “beliefs or judgments a parent holds on their capabilities to organize and execute a set of tasks related to parenting a child” [83].

In contrast to general health information, our results show that participants face difficulties in finding information on child development and education on websites (other than N&G) and web-based forums and navigating health and social care services when they have nonmedical care needs. In line with patient-centered care [84], future research may address several questions regarding child behavioral, developmental, and educational issues. To what extent can parents retrieve easy-to-read, listen to, and watch information in this field when needed? In what ways do websites and forums provide sufficient information to guide parents in managing these issues on their own when they desire and are able to do so? How do parents know who to ask for help when this information is insufficient? How can better understand what social care services are available and how they can be accessed?

Our results support approaches for improving information exchange between low-SES mothers and health or social care professionals, including (1) prescriptions of information that is considered trustworthy, easy to read, listen to, and watch; (2) information seeking by trusted relatives who have experience and higher level of eHealth literacy; and (3) referral to another professional, third party, when needed [84,85]. In other words, professionals can help low-SES mothers find, understand, evaluate, and use information because the provision of informational content alone does not usually satisfy the constraints of low education and empowerment among people of low SES [86].

Finally, the “Kate” narrative can be quickly read by decision policy makers, for example, web editors, and might encourage them to reach the underserved. In line with Reichwein et al [87], an ideal type can raise awareness and allow information providers, for example, web editors, to (1) tailor information content for the information needs and seeking behavior of specific target audience, such as low-SES mothers; (2) promote facilitators and overcome barriers to optimize information outcomes in the targeted low-SES audience; and (3) avoid stigmatization of people of low SES by openly reaching everybody (universalism).

Considering that narrative results of evaluations of interventions that show an impact on inequalities are especially useful for policy making [1], the “Kate” ideal type suggests the following main messages and policy recommendations. Web editors and experts can be encouraged to assess how valuable the informational content they share is from the perspective of low-SES people, providing thereby more equity-oriented content (easier-to-read, listen to, and watch information). In line with Luhmann theory of communication-based autopoietic self-referential interrelated social systems [88-91], the Low Literacy Universal Precaution Principle must be applied to enable trust and satisfactory communication (information exchanges) between Kate and a variety of social systems, such as education, health, and social services. People rely on website information content when they understand and trust it [92,93], whereas they do not use the content when they do not trust it, and often stop accessing it altogether [94]. In addition, O’Neill [95] specified that misplaced mistrust can be harmful. This applies to our work as follows: information providers and decision policy makers should promote trust in trustworthy information and should furnish resistance to mistrust in trustworthy information and trust in untrustworthy information.

In other words, information providers, such as web editors, must be encouraged, trained, and supported to provide trustworthy information using a plain language standard (readability grade 3-5), audio-assistance, and visuals [96]. This is important as a systematic review of 157 cross-sectional studies that assessed 7891 websites with consumer health information showed that the mean readability grade level ranged from grades 10 to 15 [96]. Thus, people of low SES and a low level of literacy are disproportionately disadvantaged regarding web-based information sources (readability grade superior to 5). Specifically, such inequalities remain with most governmental websites [97].

This study suggests a main message, a feasible intervention, and policy recommendations that can be implemented across health care systems.