In the United States, more than 1.6 million new cases of cancer are estimated to be diagnosed each year [1]; however, the burden of cancer among the US population is not shared equally. Medically underserved populations are defined as groups with economic, cultural, or linguistic barriers to medical care services [2]. These groups include racial and ethnic minorities and individuals of lower socioeconomic status [3] who have a higher cancer burden compared with their counterparts, which can be partially attributed to differences in the access to, and quality of, care they receive [4-6]. A wide range of technologies is available to patients, which have the potential to improve access to care and empower individuals to participate more actively in their care [7,8]. These technologies include personal health records (PHRs) [9], internet-based (eHealth) technologies [10,11], mobile (mHealth) apps [12], and telemedicine [13]. For example, there is evidence that patient-centered technologies (also commonly referred to as consumer health information technologies) provide patient-centered care by increasing patients’ quality of health care [14], improving communication with providers [13,15-17], providing tailored education and lifestyle messages [14,18], and promoting self-management of health care [19]. Among other cancer health promotion activities, these technologies educate individuals on the benefits of cancer screening, enable individuals to receive reminders for cancer screening and follow-up, and provide tailored decision aids for cancer care. These health information technologies have also been proposed as a means to reduce health care disparities [8,20-22]. The Institute of Medicine identified the internet and computers as critical vehicles to deliver health information to reach diverse populations of cancer patients and survivors [23,24], including low-literacy and low-income African Americans [25-28]. Research is crucial to understand the use and impact of these technologies among underserved populations for the purposes of cancer health promotion, but the medical literature has not been systematically reviewed to understand these patterns or outcomes.

To date, reviews on the use of patient-centered technologies have largely focused on the general population. For example, a recent review by Kim and Nahm [29] found several benefits to the use of patient-accessible PHRs, including consumer empowerment, improved patient-provider communication, increased access to data during times of emergency, improved chronic disease management, and increased likelihood of behavior change. Several concerns were also raised regarding the broader dissemination of PHRs, including data privacy and security, data accuracy, health literacy, and the digital divide. With regard to mobile technologies, Krishna et al [30] conducted a review and found significant improvements in medication adherence, smoking quit rates, self-efficacy, and other health outcomes (eg, asthma symptoms, blood sugar control, and stress levels). Limited attention has also been given to the potential of patient-accessible PHRs among specific disease classes [31,32]; however, Price et al [33] found PHR interventions targeting asthma, diabetes, fertility, glaucoma, HIV, hyperlipidemia, and hypertension (but not cancer) to have beneficial effects such as better quality of care, improved access to care, and increased productivity. In another review, Bennet et al [34] found that racial or ethnic minority populations have been targeted with interventions to facilitate weight loss. Overall, internet-based technologies (eHealth) were only able to affect short-term weight loss, whereas mobile technologies (mHealth) provided no benefit [34]. Montague et al [35] also found that technologies can positively affect the health of the underserved if they are effectively tailored, but little is known about how to effectively tailor cancer-specific technologies to this population. Although eHealth and mHealth studies have also explored the use and adoption of these technologies [36-40] and have been shown to increase screening for cancer [41] and knowledge of cancer and cancer screening [42], these studies have not been systematically reviewed and synthesized to date.

The purpose of this study is to systematically review current evidence on the use of cancer-specific patient-centered technologies among underserved populations. This review contributes to both the informatics and cancer health disparities literature by seeking to address the following issues: (1) to understand the effect or impact of patient-centered technologies on the health or health care outcomes of underserved populations, (2) to understand the use, usability, and acceptance of patient-centered technologies and efforts to tailor their design to improve cancer care among underserved populations, (3) to understand the barriers and facilitators to patient-centered technology use for different populations, and (4) to propose directions for future research based on the current literature.

Patient-centered technology use by underserved populations is influenced by multiple factors. For purposes of this review, we have adapted an existing health services research framework to organize the factors that influence the use and acceptance of information technology among individuals. Originally developed with the organization in mind, the unified theory of acceptance and use of technology (UTAUT) [43] sought to understand the critical factors related to the prediction of behavioral intention to use technologies within the organizational context.

According to the UTAUT, there are 3 constructs (ie, performance expectancy, effort expectancy, and social influence) that are considered direct determinants of the intention to use technology (see Table 1 and Figure 1). Performance expectancy refers to the degree to which using a technology will provide benefits to consumers in performing certain activities. In the health context, benefit examples may include managing chronic conditions or receiving health information to facilitate behavior change. Effort expectancy refers to the degree of ease associated with the consumers’ use of technology (ie, usability). Social influence refers to the extent to which consumers perceive important others, such as their family and friends, to believe they should use a particular technology. Separately, there are 2 constructs within this framework (ie, intention and facilitating conditions) that are considered direct determinants of technology usage behavior, with facilitating conditions referring to the perception of resources and support available to perform a behavior. In addition, there were 4 moderators embedded in this original framework, which contributed to understanding the acceptance of technology by individuals (ie, age, gender, experience, and voluntariness of use).

To tailor this theory to the consumer use of technology, Venkatesh et al developed UTAUT2 [44]. Given the focus of this study, it seems especially apt to use this tailored theory to understand minorities’ use of patient-centered technologies. Under this remodeled framework, 4 key constructs on the general and consumer adoption and use of technologies have been identified and incorporated into UTAUT. These additional constructs are (1) hedonic motivation (the fun or pleasure derived from using a technology), (2) price value (the monetary cost of use on the individual), (3) experience (the passage of time from initial use of the technology), and (4) habit (the extent to which an individual believes the behavior to be automatic).

Recommendations of the statements on enhancing transparency in reporting the synthesis of qualitative research [45] and the preferred reporting items for systematic reviews and meta-analyses [46] were followed. Web-based search was conducted in the following academic databases: (1) PubMed (cancer subset), (2) MEDLINE, (3) PsycINFO, and (4) CINAHL. To optimize search results, we used various combinations of keywords taken from the existing literature and Medical Subject Headings terms. A complete list of search terms is provided in Table 2. Finally, we identified additional studies using a snowball searching technique whereby the reference lists of studies that met our inclusion criteria were examined.

We identified papers that appeared in peer-reviewed journals and were published in the English language up to October of 2016. We included both qualitative and quantitative studies and excluded nonempirical studies such as commentaries as well as international studies. Similar to other reviews [35], we limited eligible studies to those conducted in the United States because sociocultural differences in the United States may be unique from other countries. Studies were included if they assessed patient-centered technologies among underserved populations. The specific topics of interest for this review included (1) the effect of these technologies on the health or health care outcome studied; (2) the use, usability, and acceptance of these technologies and efforts to tailor their design to populations of interest; (3) facilitators and barriers to the use of patient-centered technologies among underserved populations; and (4) implementation lessons learned from studies assessing these technologies among underserved populations. Studies were included if they focused exclusively on underserved populations or underserved groups represented at least 40% of their sample size. To categorize the health information technology (HIT) apps of interest in this study, the following definitions were used.

Each study was individually assessed for relevance. Any disagreements between reviewers were reconciled by consensus. We used a 3-step inclusion process, which is illustrated in Figure 2. In step 1, we examined paper titles and excluded papers that clearly did not have a focus on either patient-centered technologies or cancer care. In step 2, the remaining citation abstracts were retrieved. Similarly, we then excluded paper abstracts that clearly did not have a focus on either patient-centered technologies or cancer care. Finally, the full-text papers of the remaining citations were obtained for independent assessment of all inclusion criteria.

Information systematically extracted from the papers included the following: study design, including the targeted cancer and/or stage of the cancer care continuum [50]. The continuum of cancer care refers to prevention, detection, diagnosis, treatment, survivorship, and end-of-life care [51]. In addition, we extracted information on the underserved population of interest; whether the patient-centered technology focused on healthy individuals, cancer patients or survivors, caregivers, or health care providers; sample size; the type of patient-centered technology used; the study outcome of interest; and whether there was any evidence of tailoring when it came to the technology intervention. Tailoring is defined as “Any combination of information or change strategies intended to reach one specific person, based on characteristics that are unique to that person, related to the outcome of interest, and have been derived from an individual assessment” [52]. Moreover, when available we described the technology’s use (whether patients adopt the tool), usability (the patient’s experience using the tool), and usefulness (the extent to which it meets the patient’s needs). Finally, we described barriers and facilitators to the use of the technologies reported in the paper.

Our keyword search identified an initial yield of 1276 nonduplicative studies (Figure 2). The primary reasons for exclusion are also identified in Figure 2. After applying the exclusion criteria in review of the titles and abstracts, 71 studies were included in the systematic review (marked with an asterisk in the reference list), published between 1995 and 2016.

The characteristics of the studies are summarized in Table 3. Studies varied with respect to the underserved population targeted, the technology used, and the cancer type of interest. A large proportion of studies included in our review target blacks or African Americans (31/71, 44%) followed by rural populations (14/71, 20%). More than half of the included studies assessed eHealth technologies (41/71, 58%). Moreover, the largest proportion of studies focused on breast cancer (26/71, 37%). In addition, the largest proportion of technological outcomes assessed was use of technology (20/71, 2%), whereas knowledge (15/71, 21%) was the largest proportion of health outcomes assessed. Observational studies represented the largest proportion of studies included in our review (32/71, 45%). Overall, 15 studies followed an experimental design (15/71, 21%), whereas the remaining 24 studies were either qualitative (13/71, 18%) or mixed-methods (11/71, 16%).

To provide a consistent structure, the remainder of the Results section is organized as follows. Study summaries are stratified by the type of patient-centered technology. Within each patient-centered technology section, we then further stratify by study design: experimental (Multimedia Appendix 1), observational (Multimedia Appendix 2), and qualitative studies (Multimedia Appendix 3).

The purpose of this study was to review the current evidence on the use of cancer-specific patient-centered technologies among the underserved. Although the reviewed studies targeted various underserved populations including racial and ethnic minorities (eg, blacks or African Americans and Hispanics), low-income, and rural populations, we identified 2 cross-cutting issues that the literature suggests should be taken into account when implementing patient-centered technology interventions: (1) training in the use of patient-centered technologies and (2) tailoring patient-centered technologies to target populations.

The landscape of technology in our digital age is rapidly changing. This growth has led to several advances in health promotion from accessing health information digitally to using technology to track health and fitness [118]. In addition, the internet and mobile devices have become a prominent vehicle to reach diverse minority populations and deliver health information [23-28]. Use of the internet within the home is lower in individuals who are older, belong to a racial or ethnic minority group, are less educated, and have lower incomes [119,120]; however, the internet has become more accessible in many ways because of the proliferation of mobile devices. For example, blacks are more likely to access the internet with their mobile phone than their non-Hispanic white counterparts [119].

Much of the evidence related to mobile devices was devoted to the use of text messaging as a means to provide health information and facilitating behavior change. These results are promising given the consistent findings that underserved populations are receptive to the use of these technologies for cancer prevention and care purposes. Contributing factors to this growth, and the intention to use these technologies, may be traced back to the constructs of social influence, price value, and habit. According to the Pew Research Center, text messaging is being used by more than 90% of the population within each age group (100% of 18-29 years, 98% of 30-49 years, and 92% of >50 years) [121]. Due to this widespread use and already established habit of communicating with others, individuals are likely to adopt this form of technology to communicate with their social group. Due to the way in which this method of communication is ingrained into the day-to-day lives of individuals, the time cost of adopting these technologies is minimal because of their pervasiveness. In addition, using smartphones may be considered a low-cost alternative to accessing the internet compared with home internet.

Mobile devices provide a means to reach minority populations and offer the potential to reduce access issues with respect to health care and health information. However, barriers still exist that prevent the effective use of these technologies. In addition to creating opportunities to advance health promotion, the rapid growth in technology also presents several challenges. Of the most prominent challenges facing users, 1 is the pressure to remain updated with new technologies, their increasing effort expectancy, or the degree of ease associated with technology use. Clearly, realizing the full potential benefit of these technologies is dependent on their effective use. Although studies found that underserved populations are receptive to the use of patient-centered technologies, a recurrent challenge found in the literature was a lack of knowledge as to how to use new apps of the technology as well as the technology itself. These challenges should not be overlooked and range from receiving health information via text messages to using interactive iPads.

Some of these difficulties with use could be remedied by facilitating conditions, for example, providing a short training session at the same time the technology is introduced. Public libraries have been successful in improving decision making in accessing high-quality health information, reducing computer anxiety, and increasing computer interest and self-efficacy among older adults [122,123], and health care providers could leverage or learn from these community institutions. However, other technologies may require substantial modifications to the intervention to remove obstacles and barriers individuals may experience to facilitate their use. Interventions should incorporate usability and feasibility testing with target populations into their development process to identify unanticipated issues as well as appropriate training of target populations in the use of the technologies. Although such methods need to be applied efficiently to minimize their time and resource burden, up-front investment in such approaches can be the difference between a successful or failed implementation. In some cases, new technologies may not be a good fit with underserved populations, for example, low-income individuals may have insurmountable barriers to obtaining expensive new devices. Patients with disabilities, whether mental or physical, may not have the capacity to adapt to new technologies that require significant cognitive load or fine motor skills. In these cases, alternative communication channels may be necessary to deliver a desired behavioral or clinical intervention; we want to be careful to construct patient-centered technologies versus technology-centered patients.

When using interactive technologies (ie, computer-based media that enable users to access information and services of interest, control how the information is presented, and respond to information and messages in the mediated environment [124]), an important feature is the ability to tailor information to the recipients’ needs and interests [124,125]. It is necessary to consider the unique cultural norms and/or challenges of underserved populations when tailoring communication strategies. Robust methods to account for these differences in the design and implementation of technology interventions targeting specific groups is a key area in need of development. Hispanics may be better reached with technologies framed with health education content tailored to this population to improve both content comprehension and acceptance. For example, the use of telenovelas and soap operas is a novel approach that appeals to underserved Spanish-speaking women’s cultural norms and has been found to increase breast cancer screening knowledge and beliefs [57]. Similarly, a culturally tailored educational video, including a soap opera and physician recommendation segment made in Chinese was found to increase Chinese women’s intention to get screened for breast cancer, in addition to increasing their knowledge, perceived risk, and perceived benefits of screening [116].

Cultural competence is another strategy to reduce health and health care disparities that may be applied to the tailoring of patient-centered technologies. Cultural competence is defined as “...understanding the importance of social and cultural influences on patients’ health beliefs and behaviors; considering how these factors interact at multiple levels of the health care delivery system (eg, at the level of structural processes of care or clinical decision making); and, finally, devising interventions that take these issues into account to assure quality health care delivery to diverse patient populations” [126]. Tailoring patient-centered technologies to patients may help overcome sociocultural barriers to providing health care, one being a lack of culturally or linguistically appropriate health education materials [126]. By understanding unique differences among underserved groups, we can better understand how to reach each population, how they spend their time and use technology, and how different forms of technology may be used in different home and community settings. In addition, this approach allows researchers to tailor the technology based on who an individual is and how their identity is constructed before the technology is implemented.

Our study has several limitations. Of the potential limitations, 1 is that our search strategy may not have captured all potential papers meeting our inclusion criteria. To minimize this limitation, we implemented a snowball search method in which we reviewed the references of all included studies for additional citations. Another limitation of our study is that because of the heterogeneity in study design and types of outcomes evaluated, we were unable to aggregate findings in the manner of a meta-analysis. Finally, the included papers may be subject to publication bias as studies that report negative findings are less likely to be published.

There is a growing body of literature exploring patient-centered technology and its influence on the care of underserved populations. Despite the potential of patient-centered technologies and their acceptance among underserved populations, challenges still exist with respect to their effective use and usability. With technology changing at an exceedingly rapid pace, more training needs to be provided to ensure these underserved groups are able to effectively use new and emerging technologies. In addition, tailoring these technologies to unique cultural norms will be critical to facilitating their effective use.