Characteristics of Study-Participating Care Recipients and Informal Caregivers
Of 17 studies, 7 were conducted in the United States. In the selected studies, children experienced a wide range of medical needs including (1) hearing loss [17]; congenital disease or developmental problems [18]; asthma [19,20]; hydrocephalus [21]; rare genetic diseases [22]; ear, nose, and throat surgeries [23]; and type 1 diabetes [24]. Study topics also included nonspecific diseases requiring primary health care [25-30], emergency care [30-32], and disability [33]. The majority of the 17 studies (12/17, 71%) used a wide range of age criteria even within a single study. Only 5 studies focused on specific age groups such as those aged 2-6 years [19,23,25,29] or preschoolers to 8th grade [19,23,24,29].
The selected studies had limitations in representing diverse populations including relationships to children, gender, race and ethnicity, insurance status, employment, education level, and the regions where informal caregivers live. The majority of studies were limited to parents or legal guardians (14/17, 82%); the remaining studies were of relatives as caregivers (3/17, 18%). The majority of participants were female, usually mothers [17,23,24,27-29,33]. A high proportion of racial and ethnic minorities were found in only 4 studies; African Americans were the largest group in these studies, with proportions ranging from 32% [21] to 83% [32]. Although the selected studies did not aim to recruit low-income families, 6 studies (35%) included low-income study participants (determined based on insurance and employment status). A high percentage of participants receiving Medicaid or government-provided insurance were included in 4 studies (56%-92%). More than 50% of study participants in the DeMartini and colleagues study [26] lived in a high-poverty area. In addition, 3 studies [17,29,33] reported a moderate to high proportion of those who were unemployed or with unsecured jobs. A summary of the study setting, study participants, and their characteristics is shown in Multimedia Appendix 1.
Health-Related Internet Use
The definition, prevalence, purpose, and detailed types of general and health-related Internet use are summarized in Multimedia Appendix 2.
Prevalence of General Versus Health-Related Internet Use
General Internet use was defined based on access to the Internet via computer, cell phone, or other mobile handheld device [34]. The prevalence of general Internet use among informal caregivers was reported with a wide range, 62% to 99%. Half or more were daily users (49%-70%). In 2009, the generic search engines most frequently used were Google (79%), Yahoo (3%), and others (18%) [23]. Among general users, the most common places to access the Internet were at home (45%-87%) followed by anywhere using a smartphone (28%-71%), worksite (33%) and other places (3%-15%) including the library, community agencies, schools, and Internet cafés.
The prevalence of health-related Internet use varied (11%-90%) depending upon how it was operationalized and measured. Only one study used a comprehensive definition of health-related Internet use based strictly on Eysenbach's framework [1]—using the Internet for health-related information, support, and health care education [17]. The most common definition of health-related Internet users included people using the Internet for seeking health-related information for child caregiving [18,19,23,25,28,29,32]. Using a narrow definition was likely to be associated with a lower prevalence of health-related Internet use: 11% used the definition of those with access to care over the Internet focusing on email use [27], 58% used the definition of those with health-specific uses of digital technology [26], and 82% used a general definition of informal caregivers who used the Internet related to their children’s health [21,22,31,33].
Types of Health-Related Internet Use
Information (Content)
The most prevalent purpose for health-related Internet use was seeking information regarding child health care needs; 15% to 90% of caregivers knew how to find health-related information on behalf of care recipients [17-26,28-33]. In one study, many Internet users (87%) chose a generic search engine; almost half (44%) also visited specialized websites for specific health needs [17]. Only 35% used the Internet at the time of the care recipients’ diagnosis [18]; a small group of informal caregivers (9%) sought Web-based information immediately prior to their onsite clinic visit [31].
Informal caregivers were not confident in their ability to appraise health-related information found on the Internet or distinguish the quality of information and support from health care providers. According to Knapp and colleagues [28], only half of the users felt confident enough to evaluate the quality of Web-based information, although the Internet was the most commonly used source for health information according to Bouche [25]. From 10% to 50% of informal caregivers discussed the information found through Internet searches with their health care providers during onsite clinic visits [17,23]. Half of them stated that their health care providers were interested in the Web-based information [17].
Communication
Informal caregivers used the Internet for communicating with their health care providers or peers [21,27,30,32]. The informal caregivers expressed a strong interest in using the Internet and emails to communicate with primary health care providers (80%-86%) [31,33] and health care providers in the emergency department (93%), including receiving lab results [32]. In addition, informal caregivers thought that electronic communication between primary and emergency department care providers would be helpful (34%) [32]. The informal caregivers also expressed a strong interest in using the Internet and emails to contact organizations related to health concerns and promotion (36%) [31,33].
As a communication method with their health care providers, the informal caregivers wanted to receive information via an electronic newsletter about current disease trends (77%), discharge instructions (66.0%), and educational content about common illnesses (73%) [32]. These findings are similar to the study showing the information that respondents want to receive from their health care providers online includes common infections (77%), age-appropriate activities (73%), healthy eating (71%), required well-child visits and screening tests (65%), and resources in community (62%) [26,30].
Support (Community)
The Internet was also commonly used by informal caregivers for obtaining emotional and social support [17,21,22,24]. Almost 30% used emotional support groups and 35% used the Internet for communication with parents in similar situations [20]. As many care recipients had life-long chronic illnesses, the Internet played an important role in helping informal caregivers cope with their emotions by having more information [22,24]. Internet support groups helped informal caregivers adjust to their children’s condition [22]. Peer communication using email was beneficial for expanding their interaction beyond the membership of a certain online group [21].
Education
The Internet was also commonly used by informal caregivers to educate themselves about obtaining care for themselves and their care recipients simultaneously [17]. To educate themselves regarding their personal health care, 86% of informal caregivers found the Internet helpful in learning about diseases [21], and 78% of participants used YouTube for educational videos related to health. Most informal caregivers wanted guidance and recommendations from their health care providers about which online resources to use [21]. Only 58% of those who sought information regarding their personal health care questions trusted the information received, and then only sometimes or somewhat [21].
eCommerce
None of the studies investigated any purchases of medical products or medications via online shopping.
Associated Factors of Health-Related Internet Use
There was evidence that a higher education level in informal caregivers was associated with more frequent use of the Internet related to health [17,21,28], which is consistent with findings from previous studies [2,35]. Higher education levels seemed to be correlated with adequate health literacy [19,20]. A digital divide existed for racial and ethnic minorities such as African Americans and Hispanics and among non-English-speaking groups [21,27,28,32].
Caregiving-specific factors of health-related Internet use included (1) a strong intention to understand children's health information [29], (2) unmanageable situations beyond the capacity of parental adjustment [22,24], and (3) specific treatment requirements of the children [18,24]. However, no relationship with health-related Internet use was found regarding geographic location, age of parents, status of disease, or number of consultations with primary care providers [17,21,25].
Perceived Benefits
Informal caregivers stated that it was easy to find helpful information regardless of the time and their location. Information helped informal caregivers understand a child’s medical condition [18,31], understand specific treatment [19,23], and make decisions about treatment [23,26]. In terms of Internet use as a support system, they were highly satisfied with Internet-based parental support groups, citing obtaining usable ideas, improved informal caregiver relationships with their children, finding people to trust, and seeking stress-coping strategies as specific benefits [22].
Perceived Barriers
There were several barriers that informal caregivers encountered using the Internet for health-related purposes. The quality of websites was a main barrier [17]. Only half felt confident assessing the quality of Web-based information [28]. This may explain why 94% of participants responded that they were not able to find the information they wanted on the Internet [20,31]. Most participants did not remember the specific health-related websites they used [31]. In addition, they hesitated to discuss the Web-based information they found with their health care providers [17,20,23]. This may be based on warnings from their health care providers not to trust Internet-based health information. Additional reasons caregivers do not use Internet health-related information may include personal logistical barriers, fear, and mistrust of information on websites [26]. Other barriers included cost, limited access, lack of knowledge, lack of time, medical disabilities, vision problems, concern about the negative effects of computer use, lack of transportation, and a lack of child care [31,33].
Conceptual and Methodological Evaluations
Issues of Conceptualization
Most studies did not clearly define health-related Internet use. Using or accessing the Internet to find health-related information was the common operational definition. However, researchers did not provide the rationales for why they defined health-related Internet use based on the access to use Internet [34]. Only 2 studies (2/17, 12%) used theoretical frameworks to explain why health-related Internet use was important during the caregiving trajectory. The frameworks used were the theory of planned behavior [29,36] and Antonovsky’s concepts of sense of coping and coherence [22,37]. Other frameworks were used to understand the parental factors and their decisions to use online health information regarding diagnosis and treatment [29,36]. These frameworks considered the Internet a resource for helping informal caregivers reestablish a sense of coherence after they experienced stressful events due to their child's illness [22,37].
Study Design
All 17 studies were cross-sectional. The most frequently used study designs were quantitative (13/17, 77%) and prospective (16/17, 94%). Descriptive (8/17, 47%), correlational (9/17, 53%), qualitative (2/17, 12%) [22,24] and mixed-methods design (1/17, 6%) were also used [17]. Quantitative studies tended to report prevalence and predictors of health-related Internet use, and qualitative or mixed-method studies investigated informal caregiver perceptions of helpful and harmful health-related Internet use. There was no longitudinal study found to imply causality. Walsh and colleagues used multiple observation time points [29]. All of them were 1-group studies without any comparison group. Almost half of the 17 studies used a single recruitment site. Others used multiple sites including caregiver databases [25,33], online recruitment [29], multiple clinical sites [17,21,26], and large-scale clinical trials [19].
Study Sample
Convenience sampling was the most common. Only 4 studies used more rigorous systematic sampling methods based on probability such as random selection [20,28,33] and stratified sampling [25]. Sample sizes ranged widely from 10 to 2371. Based on the selected study design, data analyses, and justification of sample sizes, 8 studies (8/14, 57%) measured quantitative data and had adequate sample sizes, while 6 studies (5/13, 43%) had excessively large sample sizes. Only 2 studies [22,24] used a qualitative study design, but they had very small sample sizes (n=10 and n=27), although saturation was achieved. Most of the study participants spoke English, limiting generalizability to non-English-speaking populations. Nonresponder bias due to low response rates was identified in 5 studies; 4 of them collected data once [18,25,27,28] with response rates between 49% and 76%. Walsh and colleagues [29] collected data at 2 observation times and reported a 48% response rate at follow-up after 2 months from baseline.
Data Collection and Analysis
Almost all studies used surveys; one conducted unstructured interviews individually or as part of a focus group [22,24]. Onsite surveys seemed to be preferred (10/17, 59%) followed by postal (3/17, 18%), telephone (2/17, 12%), and online modality (1/17, 6%) surveys. Porter and Edirippulige conducted an online survey [17], and Walsh and colleagues used online recruitment [29], which decreased generalizability of this study to non-Internet users [38].
The gold standard instruments regarding health-related Internet use were the Health Information National Trends Survey questionnaire [39] and the Pew Internet and American Life study of consumers’ use of the Internet for health care information questionnaire [34]. However, only 2 studies used or modified these questionnaires [19,32]; most studies used their own. These survey questionnaires were developed with a lack of or poorly described psychometrics. Thus, the validity and reliability of these investigator-developed instruments were not well established.
Most of the types of data analyses were descriptive: univariate analyses (chi-square, student t, Pearson r, or other nonparametric tests), descriptive frequency statistics (numbers, percentages, means, and standardized deviations), or multivariate analyses (multivariate analyses of variance, linear regressions, or logistic regressions). The types of analysis statistics were appropriately chosen based on levels of data and measurement types. However, there was very limited information about reporting statistical assumptions checked, handling missing data, reporting pre-analysis, or including significant covariates in the analysis. The summary of methodological evaluation is shown in Multimedia Appendix 3, and the evaluation criteria are explained in Multimedia Appendix 4.