Applied Ergonomics 54 (2016) 218e242

Contents lists available at ScienceDirect

Applied Ergonomics journal homepage: www.elsevier.com/locate/apergo

Review article

Work system barriers to patient, provider, and caregiver use of personal health records: A systematic review Morgan J. Thompson a, Jeremiah D. Reilly b, Rupa S. Valdez b, * a b

Psychology Department, The College of William and Mary, Williamsburg, VA 23187, USA Department of Public Health Sciences, University of Virginia, P.O. Box 800717, Hospital West Complex, Charlottesville, VA 22908, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 May 2015 Received in revised form 16 October 2015 Accepted 18 October 2015 Available online xxx

Objectives: This review applied a human factors/ergonomics (HF/E) paradigm to assess individual, work system/unit, organization, and external environment factors generating barriers to patient, provider, and informal caregiver personal health record (PHR) use. Methods: The literature search was conducted using five electronic databases for the timeframe January 2000 to October 2013, resulting in 4865 citations. Two authors independently coded included articles (n ¼ 60). Results: Fifty-five, ten and five articles reported barriers to patient, provider and caregiver PHR use, respectively. Barriers centered around 20 subfactors. The most frequently noted were needs, biases, beliefs, and mood (n ¼ 35) and technology functions and features (n ¼ 32). Conclusions: The HF/E paradigm was effective in framing the assessment of factors creating barriers to PHR use. Design efforts should address literacy, interoperability, access to health information, and secure messaging. A deeper understanding of the interactions between work systems and the role of organization and external environment factors is required. © 2015 Elsevier Ltd and The Ergonomics Society. All rights reserved.

Keywords: Systematic review Personal health records Human factors and ergonomics

Contents 1. 2.

3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 2.1. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 2.2. Literature search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 2.3. Title screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 2.4. Abstract screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 2.5. Full-text screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 2.6. Data extraction and synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 3.1. Search results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 3.2. Study characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 3.3. HF/E paradigm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 3.4. Patient work system barriers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 3.5. Provider work system barriers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 3.6. Caregiver work system barriers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 4.1. Application of HF/E paradigm to barrier classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 4.2. Application of the HF/E paradigm to caregivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 4.3. Representation of individuals in the work system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

* Corresponding author. E-mail addresses: [email protected] (M.J. Thompson), jdr5bd@ virginia.edu (J.D. Reilly), [email protected] (R.S. Valdez). http://dx.doi.org/10.1016/j.apergo.2015.10.010 0003-6870/© 2015 Elsevier Ltd and The Ergonomics Society. All rights reserved.

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

5.

219

4.4. Implications for PHR design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 4.5. Designing interventions: attending to context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 4.6. Limitations and future research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240

1. Introduction As clinician and patient roles evolve, responsibility for health management is shifting from clinician-governed to patientcontrolled (Wasson et al., 2012). The continuing development of consumer health information technology (IT), health IT designed for use by lay people, supports this change in roles by enabling patients to actively engage in health management both alone and in partnership with formal healthcare providers (i.e., providers) (Center for Advancing Health, 2010). This engagement through consumer health IT has been accelerated by market forces such as the proliferation of consumer-facing mobile health applications (PricewaterhouseCoopers LLP, 2012; Greenspun and Coughlin, 2012) and has been promoted by the Meaningful Use initiative within the United States (The Office of the National Coordinator for Health Information Technology (ONC), 2013a; Center for Medicare and Medicaid Services, 2015): Stage 2 of this initiative (The Office of the National Coordinator for Health Information Technology (ONC), 2013a) incentivizes increased health information exchange, patient-controlled data, requirements for e-prescribing and incorporating lab results, and electronic transmission of patient care summaries across multiple settings. The recently proposed Stage 3 guidelines (Center for Medicare and Medicaid Services, 2015) extend the Stage 2 requirements for patient engagement by increasing the requirements regarding the proportion of a provider's patients who have interacted with their health records (i.e., viewed, downloaded, or transmitted their health records) and used secure messaging through the electronic health record (EHR) (Center for Medicare and Medicaid Services, 2015). Patient engagement in health management, however, does not solely rely on actions taken by patients and providers. Rather, health management in home and community settings often depends upon the active participation of other individuals (Valdez and Brennan, 2015; Skeels, 2010), including informal caregivers (i.e., caregivers). Caregiver as well as patient and provider interest in accessing and exchanging health information has been documented (Friction and Davies, 2008). The Patient Engagement Framework (National eHealth Collaborative and Healthcare Information and Management Systems Society Foundation, 2014), championed by the Healthcare Information Management Systems Society (HIMSS) and the National eHealth Collaborative (NeHC), outlines a vision for Meaningful Use Stages 3 and 4 to further support, through consumer health IT, the full range of individuals (e.g., caregivers, family, friends, clergy) involved in patients' health management. The need to attend to caregivers' perspectives has also been recently identified within the HF/E literature (Holden et al., 2013, 2015). Despite widespread development and promotion of consumer health IT, usage rates have remained relatively low (Valdez et al., 2015b). Proposed explanations for these low patient usage rates include low numeracy and limited technology experience (Taha et al., 2013), difficult login procedures (McInnes et al., 2013), limited family support for using advanced technologies (Mayberry et al., 2011), and fears concerning the misuse of stored information (Tjora et al., 2005). Reasons for low provider usage rates include no prior email communication with patients (Crotty et al., 2013), medical practice

location (Wynia et al., 2011), and concerns regarding appropriate reimbursement and liability (Wynia et al., 2011). Low caregiver usage rates are associated with caregiving responsibilities for a child who has a nonsevere illness (Byczkowski et al., 2014), slow Internet connection (Tom et al., 2012), insurance type (Tom et al., 2012), and fear of discrimination from insurance companies (Weitzman et al., 2012). Given the multi-dimensionality of these rationales, HF/E frameworks should be systematically applied to understand barriers to patient, provider, and caregiver consumer health IT use. Previous research in HF/E has focused on the impact of work system factors on healthcare providers' clinical health IT use, such as EHR and computerized provider order entry (CPOE) (Holden et al., 2013; Carayon and 2006; Fuji et al., 2008; Nazi, 2013; Urowitz et al., 2012). These studies demonstrate that health IT use in professional settings is shaped not only by individual characteristics, but also by a wider range of work system factors. Multiple models of the work system (Carayon and 2006) emphasize that system outcomes are influenced by interactions between users, tasks performed, tools used to accomplish these tasks, and the physical, social, and organizational environments in which users are embedded (Holden et al., 2013). While work system models traditionally characterize the sociotechnical systems of professional work, more recent models translate this concept for patients and caregivers (Holden et al., 2013; National Research Council, 2011; Carayon et al., 2006). These models enable the concept of the work system to be applied to healthcare work conducted by both healthcare providers and lay people, facilitating the study of work system factors in the context of consumer health IT use. In this review, barriers are categorized using the HF/E paradigm developed by Karsh and colleagues (Karsh et al., 2006). Although this paradigm targets patients and providers, we extend and apply it to caregivers. This review focuses on a specific type of consumer health IT, personal health records (PHRs), because Meaningful Use Stages 2 and 3 as well as the Patient Engagement Framework promote increased patient, provider, and caregiver PHR use within the United States. Though varyingly defined (Archer et al., 2011; Connecting for Health Personal Health Working Group, 2003; The Office of the National Coordinator for Health Information Technology (ONC), 2013b), PHRs in the present study are “Internet-based set[s] of tools that allow people to access and coordinate their lifelong health information and make appropriate parts of it available to those who need it” (Connecting for Health Personal Health Working Group, 2003), (p3). Common PHR functionalities include the ability to access lab results, engage in secure messaging with providers, request prescription renewals, schedule appointments, authorize referrals, and view and update medication and allergy lists (Archer et al., 2011). Studies included here report on PHRs that vary in terms of available features and of organizations through which patients were enrolled (e.g., employer, government agency, primary care physician). In contrast to previous reviews of PHRs (Archer et al., 2011; Jabour and Jones, 2013; Amante et al., 2014), this review is systematic, restricted to empirical peer-reviewed articles, focused on work system barriers to use by patients, providers, and caregivers, and based on an HF/E paradigm.

220

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

2. Methods 2.1. Scope Our review was restricted to studies in which users were asked to engage with PHRs in the context of their everyday lives. Studies focused solely on lab-based usability testing or system demonstrations were excluded. This review categorized all factors identified as barriers in quantitative, mixed methods, and qualitative studies. Traditionally, the term “barrier” is defined as “a law, rule, problem, etc., that makes something difficult or impossible” (MerriamWebster). However, for the purposes of this review the term “barrier” has been extended to allow factors associated with decreased PHR use such as individual characteristics (e.g., race, age) to be conceptualized as barriers. The scope did not include conducting a meta-analysis to assess the degree to which each factor affected PHR use. Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) (Moher et al., 2009) criteria were adopted to the extent possible; however, we could not adopt the standards in full. For example, due to the range of methodologies among included articles, we were unable to assess risk of bias in a consistent manner.

or telemedicine systems (unless this had been integrated into a PHR and the article examined both components), (d) were introductions or prefaces to a special issue, conference proceedings, or book, (e) were review articles, and/or (f) were not written in English. Two authors (MJT & RSV) independently coded 100 (2.5%) of the abstracts for eligibility. The first set of 50 abstracts was used to finalize inclusion and exclusion criteria. The second set was used for calculating inter-rater reliability. The resulting value for Cohen's kappa statistic (k¼.90) was interpreted as strong (McHugh, 2012). When discrepancies occurred, decisions were discussed and resolved through consensus-building. Decisions were documented and subsequently used by the first author to include or exclude the remaining abstracts. 2.5. Full-text screening

A health science librarian and a computer science librarian were consulted in developing the search strategy, including the search terms. A range synonymous with personal health record and barrier (see Table 1) was used to ensure that all related articles were included in the initial screening. The search was performed using five electronic databases during October 2013: PubMed, CINAHL (excluding MEDLINE records), Engineering Village (Compendex and INSPEC), IEEE Xplor, and ACM Digital Library. To keep search strategies consistent across databases, controlled vocabulary was not used. All searches were limited to articles published between January 2000 and October 2013.

Articles excluded during the initial full-text screening (a) met abstract exclusion criteria, (b) were opinion articles, (c) were formative papers, (d) were not full papers, (e) described paperbased systems, (f) focused on system architecture, and/or (g) were unavailable after exhausting the University of Virginia's library resources. Two authors (MJT & RSV) independently screened 10% of the full-text articles (n ¼ 21) before engaging in decision making similar to that described for abstract screening. The first author then completed the first round of full-text screening. A second round of full-text screening ensured that included articles focused on PHRs implemented in participants' everyday lives. During this round, articles were excluded if the PHR was (a) not used, (b) used only during a demonstration, and/or (c) not implemented in participants' everyday lives. Two authors (MJT & JDR) individually reviewed all articles during this round. Inter-rater reliability was moderate, according to Cohen's kappa statistic (McHugh, 2012) (k¼.67). Discrepancies were resolved through consensus-building and unresolved differences were referred to the senior author (RSV).

2.3. Title screening

2.6. Data extraction and synthesis

2.2. Literature search

Title screening consisted of reviewing search results for duplicates and publication in peer-reviewed journals or conference proceedings. Duplicates were screened for manually by the first author and the RefWorks duplicate finder tool (ProQuest LLC, 2010). Ulrichsweb was used to exclude non-peer-reviewed journals and conference proceedings. If a conference proceeding was not listed on Ulrichsweb, we determined whether it was peer-reviewed based on the call for participation. 2.4. Abstract screening Abstracts considered eligible and included in full-text screening (a) mentioned PHRs and (b) related to patients', providers', and/or caregivers' experiences with PHRs. Excluded abstracts (a) focused solely on provider-facing electronic systems, (b) focused on technology with functionality limited to symptom tracking without integration into other PHR functions, (c) focused on self-management

Two authors (MJT&JDR) independently extracted the following:           

sample (patient, provider, or caregiver) sample size type of PHR (tethered or untethered) instrumentation (survey, interview, focus group, advisory panel, PHR system logs, and EHR patient data) study design (quantitative, qualitative, or mixed methods) study location (country) journal domain (as classified in Ulrichsweb) patient characteristics (health condition, race, age) provider characteristics (medical practice, age) caregiver characteristics (relationship to patient, race, age) barriers to PHR use associated with each work system (patients, providers, and caregivers) Because the studies reviewed implemented a range of

Table 1 Search terms. Keywords for PHR Patient facing, personal health record, personal health records, patient internet portal, patient internet portals, patient portal, patient portals, patient held record, patient held records, patient accessible electronic medical record, personal medical record, personal medical records, electronic patient portal, electronic patient portals, electronic portal, personally controlled health record, personally controlled health records, electronic patient record, electronic patient records, web-based personal health record, web-based personal health records, personal electronic health record, personal electronic health records, web-based patient portal, web-based patient portals, patient web portal, patient web portals, computerized patient record, computerized patient records, patient accessible electronic health record, and patient accessible electronic health records Keywords for barrier*, adopt*, satisfaction, implement*, facilitat*, challenge*, accept*, dissem*, use, usefulness, useful, attitude*, belief*, advantage*, and barrier disadvantage*

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

qualitative, descriptive, and mixed method approaches, statistical significance could not be reported for all barriers. However, for studies reporting statistical significance, only statistically significant barriers were extracted for analysis. Karsh and colleagues' HF/E paradigm specifies four work system levels (see Table 2) (Karsh et al., 2006). The original model refers to individual factors as patient and provider factors. However, because we extended the model to caregivers, we refer to this work system level as individual factors. Individual factors are characteristics of the person (e.g., weight, personality). Work system/unit factors are tasks performed (e.g., searching for health information) and the tools and technology used to accomplish these tasks (e.g., search engines). Organization factors are characteristics of the home and community (patient- and caregiver-focused) or healthcare institution (provider focused). External environment factors are characteristics of an individual's outside environment (e.g., clinics [caregiver- and patient-focused] or external clinics [providerfocused]). As noted, organization and external environment factors depend on the population being assessed. Data were entered into Microsoft Excel (Microsoft Office Professional, 2010) and analyzed using qualitative content analysis methods (Graneheim and Lundman, 2004; Sandelowski, 2010; Sandelowski, 2000; Hsieh and Shannon, 2005). Using the HF/E paradigm as an overarching framework, barriers were coded both inductively and deductively (see Fig. 1). Deductive coding matched barriers to existing elements of the framework; however, when a quote did not fit an existing category, we created a new code (i.e., inductive coding). Simultaneous coding was used when content extracted from an article reflected more than one factor and/or subfactor. Two authors (MJT & JDR) independently coded all extracted data. Decisions were discussed and discrepancies resolved through consensus-building. Unresolved differences were referred to the senior author (RSV). 3. Results 3.1. Search results The search, using PubMed, CINAHL, Engineering Village (Compendex and INSPEC), IEEE Xplor, and ACM Digital Library, produced a total of 4865 records. After title, abstract, and full-text screening, 60 articles remained for data extraction and synthesis (see Fig. 2).

3.2. Study characteristics Included studies employed a range of approaches for understanding barriers to PHR use (see Appendix A). Forty-five articles focused exclusively on patient work systems (McInnes et al., 2013; Tjora et al., 2005; Ancker et al., 2011; Burke et al., 2010; Day and Gu, 2012; Goel et al., 2011; Gu and Day, 2013; Guy et al., 2012; Hess et al., 2007; Kahn et al., 2010; Kim et al., 2009; Krist et al., 2011; Lau et al., 2013a, 2013b, 2013c; Lober et al., 2006; Mayberry et al., 2011; Miller et al., 2007; Nagykaldi et al., 2012; Nielsen et al., 2012; Osborn et al., 2013; Sarkar et al., 2010, 2011; Schnipper et al., 2008; Tsai et al., 2012; Tuil et al., 2006; Vodicka et al., 2013; Wade-Vuturo et al., 2013; Wagner et al., 2012, 2010; Wang et al., 2004; Weingart et al., 2006; Wen et al., 2010; Wiljer et al., 2010; Zickmund et al., 2008; Zulman et al., 2011; Emani et al., 2012; Nazi, 2010; Nazi et al., 2013; Wald et al., 2009; Tenforde et al., 2012; Denton, 2001; Goldner et al., 2013; Gordon et al., 2012; Lin et al., 2005), three exclusively on provider work systems (Crotty et al., 2013; Wynia et al., 2011; Fuji et al., 2008), two exclusively on caregiver work systems (Byczkowski et al., 2014; Britto et al., 2013), seven on patient and provider work systems (Nazi, 2013; Urowitz et al., 2012; Wald et al., 2010; Earnest et al., 2004; Jung et al., 2011; Do et al., 2011; Poon et al., 2007), two on patient and caregiver work systems (Tom et al., 2012; Weitzman et al., 2012), and one on patient, provider, and caregiver work systems (Woods et al., 2013). Sample sizes ranged from 10 to 100,617. A majority reported barriers to using tethered systems (n ¼ 47), were based in the United States (n ¼ 50), and were classified in the medical sciences journal domain (n ¼ 54). Over half (n ¼ 37) used quantitative designs; the remaining studies used mixed methods (n ¼ 15) or qualitative designs (n ¼ 8). Data collection included a variety of instruments, with most employing surveys (n ¼ 41). Participant demographic characteristics varied across the studies (see Appendix A). Of 28 reporting patient health conditions, diabetes was the most frequent (n ¼ 14). Twenty-nine studies reported patients' average age, with only two reporting an average age less than 40. Thirty-one reported patients' race, most frequently Caucasian (n ¼ 31) or African American (n ¼ 16). Thirteen reported patient participants of Hispanic/Latino ethnicity. Five studies focusing on provider work systems reported the type of medical practice, with two relating to primary care. Three reported provider participants' average age (ranging from 41 to 51 years). Three studies focusing on

Table 2 HF/E paradigm factors and subfactors (Karsh et al., 2006). Factor

Subfactor

Individual factors

Skills, knowledge, training, education Size, weight, reach, strength Age, gender, ethnicity, language Needs, biases, beliefs, mood Task demands, complexity, difficulty Time and sequence demands Availability of usable technology Technology functions/features Noise, temperature, lighting Physical layout and geography Organizational policy/priorities Organizational structure Financial resources Rewards structure Management structure Training provided Staffing levels Social norms and pressures Social climate/culture Extra-organizational rules, standards, legislation enforcement Industry social influence Industry workforce characteristics

Work system/unit factors

Organization factors

External environment

221

222

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

Fig. 1. Coding example.

caregiver work systems reported the relationship between the caregiver and patient, all of which were parents or guardians. Only one study reported caregivers' race (primarily Caucasian) and only one reported caregivers' average age (36 years). 3.3. HF/E paradigm In total, work system barriers were classified under 19 of the 22 subfactors described by Karsh and colleagues and were noted at all levels of the work system. Barriers to both patient and provider use were categorized in 16 of the 22 subfactors, while barriers to caregiver use were categorized in eight. In addition to the subfactors identified from the HF/E paradigm, one new subfactor, behavior, was identified for patient, provider, and caregiver work system barriers. Thus, the total number of subfactors identified for each population was 17, 17, and 9, respectively. Fig. 3 lists the subfactors identified for each population, illustrating subfactors unique to specific populations and those shared by multiple populations. None were identified as unique to caregivers. 3.4. Patient work system barriers A majority of articles (n ¼ 55) reported patient work system

barriers to using PHRs in at least one of the four HF/E paradigm levels: individual (n ¼ 46), work system/unit (n ¼ 39), organization (n ¼ 20), and external environment factors (n ¼ 25). All examples provided in the Results section are either direct quotes from the authors of an article or direct quotes from participants in qualitative studies. Barriers within patient work systems were identified for four subfactors related to individual factors (see Table 3 and Appendix A). Patients' skills, employment, knowledge, training, and education, such as awareness of system features and computer literacy were recognized as barriers to PHR use. Patients' demographics such as age, gender, ethnicity, language, and marital status were described as barriers. Barriers related to patients' needs, biases, beliefs, and mood focused on perceived PHR value, degree of need or existence of alternative technologies, and communication preferences. Patients' behavior was acknowledged as a barrier, including health service utilization, practice of a healthy lifestyle, and norms for technology use and communication. Barriers within patient work systems were identified for five subfactors related to work system/unit factors (see Table 4 and Appendix A). Barriers related to system task demands, complexity and difficulty related to issues encountered when using the PHR, such as unintuitive navigation and password recovery. Barriers to patient use related to time and sequence demands focused on

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

223

Fig. 2. Flow diagram for search results.

provider and system response times and the PHRs' fit with patients' everyday routines. Availability of usable technology including Internet, computer, and smartphone access, was described as a barrier to patients' use. Technology functions and features was the most frequently reported work system/unit subfactor and posed barriers to patients because of challenging content (e.g., difficult terminology), inadequate privacy and security settings, missing functionality, and limited technology platform options. Barriers pertaining to physical layout and geography focused on the location

where the patient lived and whether the patient used the system while hospitalized or after discharge. Barriers within patient work systems were identified for five subfactors related to organization factors (see Table 5 and Appendix A). Barriers to patient use related to organizational policy/priorities focused on whether members of their home and community (e.g., family, friends) would have access to patients' PHRs. Organizational structure, or home and community members' presence at doctors' appointments and in the household, was recognized as a barrier.

Fig. 3. Subfactors identified for each population.

224

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

Table 3 Barriers to patient use related to individual factors. Subfactor

Definition

Example

Patient: skills, employment, knowledge, training, and education (n ¼ 22) Patient: age, gender, ethnicity, language, and marital status (n ¼ 20)

Patients' PHR use reflected individuals' understanding. Patients' PHR use reflected individuals' demographic groups.

Compared to those with a college degree, those with a lower educational attainment were more likely to never have logged on (Sarkar et al., 2011). There were marked race/ethnic differences in use, with African American (31%), Latino (34%), and Filipino (32%) participants least likely, and Asian (53%) and White (51%) participants most likely to both request a password for the internet-based patient portal (a marker for internet access and intent to use) and log on to the portal after requesting a password was completed (an early marker for navigability once access is obtained) (Sarkar et al., 2011). There was a significant difference (p < 0.0001) in the number of medications prescribed by multiple sclerosis staff between PatientSite users (M ¼ 4.61, SD ¼ 2.57) and PatientSite nonusers (M ¼ 2.49, SD ¼ 2.10) (Nielsen et al., 2012). ‘When you use such Net-based systems that have nothing to do with your email account, you have to access it separately. And I read so much email for the rest of the day or do so many other things, that to log on to check if I have had a response todaydI don't bother. Then, it is much better to use an email account that I use on a daily basis.’ These problems have led many patients to use ordinary email instead of MedAxess (Tjora et al., 2005).

Patient: needs, biases, beliefs, and mood Patients' PHR use reflected (n ¼ 33) individuals' mindset. Patient: behavior (n ¼ 24)

Patients' PHR use reflected individuals' actions.

Available financial resources including household income, insurance status, and willingness to pay for PHR service, were identified as barriers. Staffing levels, or the availability of home and community members to assist patients with using PHRs, was categorized as a barrier. Barriers related to social climate/culture focused on home and community members' support, health status, and education. Barriers within patient work systems were identified for three subfactors related to external environment factors (see Table 6 and Appendix A). Extra-organizational rules, standards, and legislation enforcement, such as third party access to PHRs, PHRs' interoperability with EHRs, and the ability/inability to contact multiple providers, were recognized as barriers. Patient barriers regarding industry social influences focused on patienteprovider relationships and the value of the system to providers. Industry workforce characteristics, such as the degree of involvement from the healthcare team and issues related to provider work (e.g., difficulty responding to patient inquiries, inconsistencies in provider notes, and unavailability of clinical support), also presented barriers to patients.

3.5. Provider work system barriers Ten articles reported provider-related barriers to the use of PHRs on at least one of the four HF/E paradigm levels: individual (n ¼ 7), work system/unit (n ¼ 9), organization (n ¼ 6), and external environment factors (n ¼ 4) (see Table 7 and Appendix A). Barriers within provider work systems were identified for four subfactors

related to individual factors. Providers' skills, employment, knowledge, training, and education, including awareness of PHRs and provider specialty, were described as barriers to providers' PHR use. Within the subfactor of age, gender, ethnicity, language, and marital status only gender was identified as a barrier to PHR use. Providers' needs, biases, beliefs, and mood, such as perceived PHR value, degree to which an alternative technology solution was perceived as feasible or appropriate, and beliefs regarding patient self-care, were described as barriers. Behavior, including use of EHRs, degree of engagement with PHRs, and prior use of alternative forms of technology, was also recognized as a barrier. Barriers within provider work systems were identified for five subfactors related to work system/unit factors (see Table 8 and Appendix A). Barriers related to system task demands, complexity and difficulty focused on concerns regarding patients' expectations and issues encountered while using the PHR (e.g., cumbersome system interface and navigation). Time and sequence demands, such as time allotted to patient care, workload, timing of information access, and time required to use and learn the system, were identified as barriers. The availability of usable technology, specifically the lack of patientaccessible computers in waiting rooms, posed barriers to providers. Technology functions and features including accuracy of and patient access to sensitive information, missing functions, privacy, awkward security procedures, and general technical issues, were described as barriers. Barriers related to physical layout and geography focused on the clinic setting (e.g., rural or urban settings).

Table 4 Barriers to patient use related to work system/unit factors. Subfactor

Definition

Example

Patients' PHR use reflected individuals' ability Patients thought the terminology was sometimes difficult to understand and would require to complete related work. help from the healthcare team. ‘[You] have to really read it closely and if you don't know much about terminology … it could be real confusing’ (Wagner et al., 2010). Patients' PHR use reflected the temporal Another participant expressed dissatisfaction after he sent a message to his provider about a length and order of completing related work. medication side effect, and he did not get a response within a reasonable time frame. The consequences of this were threefold: (1) the participant adjusted his medication without provider input, (2) the participant now relies on more traditional forms of communication (e.g., a phone call or an office visit); and (3) the participant has been unsatisfied with his care (Nazi, 2013). Patient: availability of usable Patients' PHR use reflected individuals' access Some access concerns were identified through patient comments, such as: ‘Internet technology (n ¼ 9) to electronic resources. Explorer as the only option for access is very restrictive;’ and ‘Not sure I'd be able to do [the journal] at home [if I didn't have] a job with access to the internet’ (Wald et al., 2009). Patient: technology functions Patients' PHR use reflected characteristics of More than one half of the respondents (62%) wanted to grant PHR access to their spouse or and features (n ¼ 28) the electronic resources. partner [but are unable to do so], and a smaller percentage wanted to grant access to a child (23%), other family member (15%), unrelated caregiver (7%), or friend or neighbor (2%) (Zulman et al., 2011). Patient: physical layout and Patients' PHR use reflected individuals' Utilisation patterns showed that users accessed their patient-accessible electronic health geography (n ¼ 2) location. records significantly more often while the patients were in hospital (median of 6 logins) than after discharge (median of 4 logins, p < 0.001) (Burke et al., 2010).

Patient: task demands, complexity and difficulty (n ¼ 22) Patient: time and sequence demands (n ¼ 14)

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

225

Table 5 Barriers to patient use related to organization factors. Subfactor

Definitions

Patient: organizational policy/priorities (n ¼ 1)

Patients' PHR use reflected established practices for For example, respondents were more interested in sharing access to medication lists, home and community members' PHR use. appointment information, and laboratory and test results with their designee than patiententered health information or communications with providers. Respondents were similarly more interested in delegating prescription refill requests and appointment scheduling than in having the designee communicate with their health care provider … Although respondents tended to be most interested in sharing information with family members (especially a spouse or partner), they expressed high levels of interest in allowing unrelated caregivers to conduct activities in their PHR, such as requesting prescription refills or scheduling appointments (Zulman et al., 2011). Patients' PHR use reflected home and community A common way participants learned about the medical center's PWP [patient web portal] members' structure and participation. was through a knowledgeable family member: ‘My daughter showed [the PWP] to me in my doctor's office, on the computer in the waiting room. No one in the doctor's office ever approached me about it. If it wasn't for my daughter, I wouldn't be a [PWP] user’ (Mayberry et al., 2011). Patients' PHR use reflected available monetary MHAV users were more likely than nonusers to be Caucasian/white, have higher incomes, resources within the home and community. and be privately insured (Osborn et al., 2013). Patients' PHR use reflected the availability of home The system was most frequently used on Thursdays (67%, 5387/8008), followed by Fridays and community members. (14%, 1098/8008), which coincided with the onsite availability of graduate nursing students. Most (77%, 6174/8008) of the system use happened while assistance from graduate nursing students or housing staff was available to the residents. On the other hand, 8% (677/8008) of user activities occurred during off hours when the students or staff were not available (from 5:00 pm to 8:00 am weekdays and weekends) (Kim et al., 2009). Patients' PHR use reflected home and community Personal health record users were younger, more likely to have commercial insurance, members' values, experiences, and practices. identify as Caucasian, have higher household incomes, and live in a region with higher rates of high school completion compared to MyChart non-users (Tenforde et al., 2012).

Patient: organizational structure (n ¼ 2)

Patient: financial resources (n ¼ 13) Patient: staffing levels (n ¼ 2)

Patient: social climate/ culture (n ¼ 5)

Example

Table 6 Barriers to patient use related to external environment factors. Subfactor

Definition

Patient: extra-organizational rules, standards, legislation enforcement (n ¼ 15)

Patients' PHR use reflected policies surrounding the healthcare system.

Patient: industry social influences (n ¼ 10) Patient: industry workforce characteristics (n ¼ 11)

Example

However, some patients were uncertain about a potential misuse of information transmitted through MedAxess. Also, the fact that communication is logged and stored in a database made the situation quite different from that of, for instance, telephone conversations. If such written communication is stored for a very long time, it is difficult to foresee who will have access to the information in years to come (Tjora et al., 2005). Fear of losing relationships: Some participants valued their choices of how to Patients' PHR use reflected communicate, and were concerned that the portal might cause them to lose those choices individuals' interactions with (Zickmund et al., 2008). providers. Patients' PHR use reflected aspects of Although communication seemed to occur primarily with a nurse, dietitian, or other AHP providers' work. via the portal, patients often wished their physician had taken more of an interest in the program and had reviewed the information they had entered on the portal during their clinic visits (this was also largely done by AHPs). Responses revealed that it would have been beneficial if a health care provider had referred them to information in the Health Library. The responses typically reflected a widespread notion that physicians were often busy and may be unable to fulfill this role as much as they would have liked (Urowitz et al., 2012).

Table 7 Barriers to provider use related to individual factors. Subfactor

Definition Providers' PHR use reflected individuals' understanding.

Example

Many health care professionals reported general awareness of My HealtheVet but limited familiarity with its features, with the exception of secure messaging. Health care professionals note that this lack of knowledge limits their ability to endorse patient use, or to integrate use of My HealtheVet features within the clinical practice setting (Nazi, 2013). Provider: age, gender, ethnicity, Providers' PHR use reflected A greater number of male physicians (32.9 percent of males vs. 20.4 percent of females) perceive language, marital status (n ¼ 1) individuals' demographic groups. that none of their patients use PHRs (c2 ¼ 13.846, p ¼ 0.000). Yet, more male physicians than female physicians reported using the patient's PHR information (6 percent vs. 2.8 percent, c2 ¼ 3.780, p ¼ 0.052), having a member of their staff work with the patients and their PHRs (5.2 percent vs. 1.2 percent, c2 ¼ 7.537, p ¼ 0.006), and being capable of electronically integrating PHR information into their own EHR (3.5 percent vs. 2.8 percent) (Fuji et al., 2008). Provider: needs, biases, beliefs, mood Providers' PHR use reflected Providers commonly viewed patients' interactions with the portal positively and their own (n ¼ 3) individuals' mindset. interaction negatively. Negative comments typically concerned time constraints and technical barriers. There were instances where providers indicated that they believed the portal may be more beneficial for patient self-education than for significant provider usage (Urowitz et al., 2012). Provider: behavior (n ¼ 5) Providers' PHR use reflected In general, health care professionals reported limited experiences with patient use (and their individuals' actions. own use) of My HealtheVet health education resources, tools to support medication reconciliation, and tools to support patient self-reported data (with some exceptions), often using alternative tools and resources instead (Nazi, 2013). Provider: skills, employment, knowledge, training, education (n ¼ 4)

226

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

Table 8 Barriers to provider use related to work system/unit factors. Subfactor

Definition

Provider: task demands, complexity difficulty (n ¼ 2) Provider: time and sequence demands (n ¼ 5) Provider: availability of usable technology (n ¼ 2)

Providers' PHR use reflected individuals' ability ‘The system is cumbersome and needs an interface that addresses the needs of patients and to complete related work. data entry requirements’ (Urowitz et al., 2012).

Example

Providers' PHR use reflected the temporal length and order of completing related work. Providers' PHR use reflected individuals' access to electronic resources.

Providers believed that accessing patient information was time consuming and sometimes redundant (e.g., due to manual data entry) (Urowitz et al., 2012). Several factors have inhibited the My HealtheVet PHR adoption, use, and endorsement of patient use: Lack of alignment with structures (e.g., lack of patient-accessible computers in the clinic setting) (Urowitz et al., 2012). Provider: technology Providers' PHR use reflected characteristics of Additional topics suggested [functionality to be included in the PHR] were social history functions/features (n ¼ 8) the electronic resources. update (79%), information about recent appointments with other providers (59%), reason for visit (34%), office questionnaires (31%), and additional screening questions (24%) (Wald et al., 2010). Provider: physical layout Providers' PHR use reflected individuals' The characteristics of users largely tracked those of physicians using electronic records in and geography (n ¼ 2) location. general; for example, they were more often urban, in group practices or hospital employed, and in noneprimary care specialties (Wynia et al., 2011).

Barriers within provider work systems were identified for six subfactors related to organization factors (see Table 9 and Appendix A). Organizational policy/priorities, such as sensitive information permissions and lack of alignment with provider workflow, were recognized as barriers to providers' use. Barriers related to financial resources focused on the cost-benefit of using the system. Staffing levels, or the availability of healthcare institution staff to work with patients on PHRs, was described as a barrier. Providers had concerns regarding the rewards structure and whether or not they would be appropriately reimbursed for time spent reviewing PHRs. Inadequate training provided to healthcare institution staff and patients posed barriers to providers. Barriers associated with social norms and pressures included the healthcare institution staffs' technology use norms. Barriers within provider work systems were identified for two subfactors related to external environment factors (see Table 10 and Appendix A). Extra-organizational rules, standards, and legislation enforcement, such as patient access to provider entered data and provider liability for all information within a patient's PHR, posed barriers to providers' use. Barriers related to industry social influences focused on the amount of time spent with patients and patients' expectations and assumptions (e.g., providers actively monitoring patients' health status on the portal).

3.6. Caregiver work system barriers Five articles reported caregiver-related barriers to the use of PHRs on at least one of the four HF/E paradigm levels: individual (n ¼ 4), work system/unit (n ¼ 4), organization (n ¼ 1), and external environment factors (n ¼ 1) (see Table 11 and Appendix A). Barriers within caregiver work systems were identified for three subfactors related to individual factors. Caregivers' skills, employment, knowledge, training, and education, including system awareness and experience using the Internet, were recognized as barriers to caregivers' use. Caregivers' needs, biases, beliefs, and mood, including the perceived system value, degree of need, comfort sharing medical information on the Internet, and communication preferences, posed barriers. Caregivers' behavior included the likelihood of receiving information from other sources and was described as a barrier. Barriers within caregiver work systems were identified for four subfactors related to work system/unit factors (see Table 12 and Appendix A). Task demands, complexity and difficulty, such as the ease with which caregivers could access and use the system, were recognized as barriers. Barriers related to time and sequence demands focused on provider and system response times, Internet

Table 9 Barriers to provider use related to organization factors. Subfactor

Definition

Example

Provider: organizational Providers' PHR use reflected established practices Several factors have inhibited the My HealtheVet PHR adoption, use, and endorsement of policy/priorities (n ¼ 2) for healthcare institutions' PHR use. patient use: Lack of alignment with workflow (e.g., lack of integration with the primary clinical information system), Lack of alignment with processes (e.g., barriers to information flow) (Nazi, 2013). Provider: financial Providers' PHR use reflected available monetary Each [participant] questioned whether the merits of the intervention would warrant the resources (n ¼ 1) resources within the healthcare institution. resources spent on it (Earnest et al., 2004). Provider: rewards Providers' PHR use reflected healthcare There was particular concern (greater than 70 percent) about both unintentional and structure (n ¼ 2) institutions' incentives for PHR use. intentional data inaccuracies in these records. There were also concerns about privacy, lack of reimbursement for time spent reviewing them, and liability for knowing all of the information in a patient's personal health record (Wynia et al., 2011). Provider: training Providers' PHR use reflected healthcare Provider responses revealed that neither they nor the majority of their patients were able to provided (n ¼ 1) institutions' instruction in PHR use. use the portal easily. More training and improved portal usability testing were said to be needed for the portal to be used more effectively. Issues with specific features such as the display of health indicators and with reading weight and exercise values were mentioned less by respondents (Urowitz et al., 2012). Provider: staffing levels Providers' PHR use reflected the availability of 6.5 percent of users reported that a member of their staff works with patient PHRs. These (n ¼ 1) healthcare institution staff. numbers were even lower for planners (3.9 percent) and nonplanners (1.2 percent) (Fuji et al., 2008). Provider: social norms Providers' PHR use reflected the healthcare Health care professionals often reported using alternative tools and resources. For example, and pressures (n ¼ 1) institutions' technology use norms. although My HealtheVet provides a significant library of health education resources, health care professionals already use alternative resources, such as subscription-based software that is linked from within the primary clinical workflow system or resources retrieved from the Internet, with little incentive to change. Health care professionals reported that they increasingly use Internet resources easily found by search engines, and speculated that patients do as well. As one health care provider said: ‘Why not just Google?’ (Nazi, 2013).

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

227

Table 10 Barriers to provider use related to external environment factors. Subfactor

Definition

Example

Provider: extra-organizational rules, standards, legislation enforcement (n ¼ 2)

Providers' PHR use reflected the policies surrounding the larger (beyond own) healthcare system.

Provider: industry social influences (n ¼ 2)

Providers' PHR use reflected individuals' interactions with patients.

There was particular concern (greater than 70 percent) about both unintentional and intentional data inaccuracies in these records. There were also concerns about privacy, lack of reimbursement for time spent reviewing them, and liability for knowing all of the information in a patient's personal health record (Wynia et al., 2011). The messaging system is great yet can be utilized negatively by patients increasing workload on mydoctor.ca and decreasing time for other patient interactions in office. The messaging system has also increased expectations from patients for immediate response (Urowitz et al., 2012).

speed, and general time constraints. The availability of usable technology, including computer and Internet access, posed barriers to caregivers. Technology functions and features were described as barriers due to the lack of specific functions or features and concerns such as accessing frightening information. Barriers within caregiver work systems were identified for one subfactor related to organization factors (see Table 13 and Appendix A). Financial resources, or the type of insurance caregivers have for their child, were recognized as barriers to caregivers. Barriers within caregiver work systems were identified for one subfactor related to external environment factors (see Table 14 and Appendix A). Extra-organizational rules, standards, and legislation enforcement, or the sharing of data with external parties (e.g., government agencies, outside providers), posed barriers to caregivers. 4. Discussion 4.1. Application of HF/E paradigm to barrier classification Overall, the HF/E paradigm was comprehensive of the barriers

identified in this review. Three subfactors included in that paradigm were not relevant and one additional subfactor was inductively derived. Size, weight, reach, and strength (individual factor), noise, temperature, and lighting (work system/unit factor), and management structure (organization factor) were included in the HF/E paradigm but not recognized as barriers in this review. The salience of these work system factors has also been identified in previous literature. For example, Smith and Sainfort (Smith and Sainfort,1989) described factors similar to size, weight, reach and strength, noise, temperature and lighting, and management structure as influential in work system models focusing on stress reduction. These factors were similarly specified within the SEIPS 2.0 model (Holden et al., 2013), which focuses on provider, patient, and collaborative work systems. Additional work is needed to empirically determine if these theoretically identified work system factors serve as barriers to PHR use. Additionally, despite the applicability of the HF/E paradigm, one new subfactor was identified, behavior, which has been previously described as a germane work system factor in other domains, including stress reduction (Smith and Sainfort, 1989), computer security (Carayon, 2006), and self-care (Holden et al., 2015).

Table 11 Barriers to caregiver use related to individual factors. Subfactor

Definition

Example

Caregiver: skills, knowledge, training, education (n ¼ 2) Caregiver: needs, biases, beliefs, mood (n ¼ 4)

Caregivers' PHR use reflected individuals' understanding. Caregivers' PHR use reflected individuals' mindset.

Caregiver: behavior (n ¼ 1)

Caregivers' PHR use reflected individuals' actions.

Users (versus nonusers) were more likely to have commercial insurance for their child (97% vs. 73%, P,.001) and have at least a 4-year college degree (74% vs. 50%, P < .001) (Tom et al., 2012). Only 2 percent of the parents agreed that they sometimes saw information they wish they had not seen. On the contrary, 12 percent agreed that they sometimes saw information in the portal that frightened them, and 11 percent reported that they sometimes see information that they would have preferred to get directly from their provider (Byczkowski et al., 2014). Parents receive disease-related information through other avenues (n ¼ 3 of 15 nonusers) (Byczkowski et al., 2014).

Table 12 Barriers to caregiver use related to work system/unit factors. Subfactor

Definition

Example

Caregiver: task demands, complexity difficulty (n ¼ 2) Caregiver: time and sequence demands (n ¼ 2) Caregiver: availability of usable technology (n ¼ 2) Caregiver: technology functions/features (n ¼ 3)

Caregivers' PHR use reflected individuals' ability to complete related work.

Make it easier to access the web site and log-in (n ¼ 18 of 126 respondents) (Tom et al., 2012).

Caregivers' PHR use reflected the temporal length and order of completing related work. Caregivers' PHR use reflected individuals' access to electronic resources. Caregivers' PHR use reflected characteristics of the electronic resources.

Faster information (e.g., quick email responses, updates from clinic visits) (n ¼ 9 of 126 respondents) (Tom et al., 2012). ‘I don't have access to the Internet,’ ‘No high-speed Internet,’ ‘I don't have a computer’ (Tom et al., 2012). Statements illustrated challenges stemming from viewing newly revealed information that had not previously been disclosed to patients. One participant, a wife of a patient, expressed stress upon seeing an operative report; when asked if reading such notes was harmful, she denied harm had ensued: ‘I would rather not have known. There was a lot of little things they wrote, you know, step-by-step what had happened in his operation’ (Woods et al., 2013).

Table 13 Barriers to caregiver use related to organization factors. Subfactor

Definition

Caregiver: financial Caregivers' PHR use reflected available monetary resources (n ¼ 1) resources within the home and community.

Example Users (versus nonusers) were more likely to have commercial insurance for their child (97% vs. 73%, P < .001) and have at least a 4-year college degree (74% vs. 50%, P < .001) (Tom et al., 2012).

228

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

Table 14 Barriers to caregiver use related to external environment factors. Subfactor

Definition

Caregivers' PHR use reflected policies Extra-organizational rules, standards, legislation enforcement surrounding the healthcare system. (n ¼ 1)

Example The most common objections to sharing data with an outside provider were relevance to patient care and the potential for discrimination by insurance companies (Weitzman et al., 2012).

Overall, relatively limited attention to higher work system levels was demonstrated. A majority of the articles found barriers related to individual factors (n ¼ 51) and work system/unit factors (n ¼ 45); however, less than half found barriers related to organization factors (n ¼ 26) and external environment factors (n ¼ 27). More specifically needs, biases, beliefs, and mood (individual factor) (n ¼ 35) and technology functions and features (work system/unit factor) (n ¼ 32) were the most frequently recognized subfactors. This may be partly due to the familiarity and widespread use of models such as the Technology Acceptance Model (TAM) (Holden and Karsh, 2010; Legris et al., 2003) and its derivatives, which emphasize the personetechnology interaction and focus less on broader contexts. There is a clear need to apply HF/E models and approaches that attend to higher levels of the work system. Identifying and understanding barriers at higher levels will require complementing traditional lab-based and task-oriented HF/E approaches with fieldbased approaches focused on the context of use. This will likely require HF/E professionals to include qualitative and mixed methods approaches (Desnoyers, 2004; Hancock and Scalma, 2004; Hignett and Wilson, 2004; Moray, 2000; Carayon et al., 2015). Using participatory design methods (Holden et al., 2013; Valdez et al., 2015b) that engage designers in patients', caregivers', and providers' naturalistic environments of use and reveal their perceptions of roles played by environmental factors (e.g., through interviews or focus groups) (Holden et al., 2015; Thompson and Valdez, 2013; Ancker et al., 2015) will be imperative. 4.2. Application of the HF/E paradigm to caregivers Although Karsh and colleagues' HF/E paradigm was originally formulated for patients and providers, we found it feasible to apply it to caregivers. As with patients and providers, it was possible to map all identified barriers, except individual behaviors, to elements of the paradigm. However, few studies examined caregiver barriers, despite growing recognition within the HF/E and medical informatics communities of the critical role these individuals play (Tjora et al., 2005; Skeels, 2010; Chappell and Reid, 2002). As a likely consequence, barriers relevant to patients and providers centered on almost twice as many subfactors as those relevant to caregivers. As previously mentioned, the review found limited research focusing on organization and external environment factors. This general trend also held for caregivers; only two subfactors related to these phenomena were identified (see Fig. 3), despite evidence of the organizational and environmental challenges faced by caregivers (Werner et al., 2012; Chen et al., 2015; Streid et al., 2014). As encouraged by the Patient Engagement Framework (National eHealth Collaborative and Healthcare Information and Management Systems Society Foundation, 2014), future research should include a focus on caregivers, specifically challenges they face at the organization and environmental levels. 4.3. Representation of individuals in the work system Sociotechnical systems, including those specific to healthcare, comprise multiple individuals (Holden et al., 2013; Vincent, 2003). This review presents three distinct groups of individuals involved in patient care (i.e., patients, providers, caregivers). Although all may benefit from PHR use, this review demonstrates that each

group has unique roles and is embedded in a different context. Consequently, work system models that explicitly represent the roles, contexts, and interactions of all actors jointly involved in the processes and outcomes of patient care are needed. This recommendation contrasts with existing work system models which, while theoretically enabling simultaneous representation of multiple individuals, do not distinguish between nor represent their overlapping vs. unique individual, work system/unit, organization, and external environment characteristics (Valdez et al. 2015). The phenomenon of overlapping work system factors existing at different levels is clearly illustrated here. For example, the degree of value providers find in PHRs is an individual factor for provider work systems (i.e., needs, biases, beliefs, and mood), but an external environment factor for patient work systems (i.e., industry social influences). Thus, sophisticated representations of work systems that account for intersections between what may, with less fidelity, be represented as multiple, separate work systems, are needed. 4.4. Implications for PHR design Although we did not conduct a meta-analysis to evaluate the strength of the evidence, we did observe several repeatedly reported barriers. For example, level of education, degree of health and computer literacy, and access to technologies were recurring themes among barriers to patient and caregiver PHR use. According to an Institute of Medicine (IOM) report (2009) (Institute of Medicine, 2009), greater disparities exist among underserved populations, many of which experience these barriers. The National Library of Medicine (NLM), the National Institute on Aging (NIA), and the Office of the National Coordinator for Health Information Technologies (ONC) have prioritized development of consumer health IT aligned with laypersons' health management skills and health literacy (Alper et al., 2015; National Institute on Aging, 2013e2016; National Institute on Aging, 2010e2013a, 2013b). To prevent the growth of health and healthcare disparities caused by barriers to use, HF/E professionals must engage directly with underserved and elderly populations, ensuring that the next generation of consumer health IT is accessible and usable across demographics. Innovative solutions may be needed that focus not only on access through individually owned devices and text-based content, but also through community owned devices and voicebased content (Wells et al., 2015). Another recurring theme was fear of accessing unwanted or frightening information (e.g., test results). To our knowledge, no policies exist that protect patients and caregivers from seeing such information. The absence of such policies may be partly due to a focus on management and control of health information access by others including providers, caregivers, and care partners (Sarkar and Bates, 2014). However, although the intent of objective 2C of the Federal Health IT Strategic Plan 2015e2020 (The Office of the National Coordinator for Health Information Technology (ONC), 2014) is to protect the privacy and security of health information, the strategies it contains may be relevant to ensuring individual control over information access. For example, strategy 3 of objective 2C states, “Support the development of policies, standards, technology, guidance, and solutions to facilitate individuals' ability to manage, control, and authorize the disclosure of specific electronic health information.” This strategy could be extended to

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

guidelines enabling individuals to manage, control, and authorize their own access to specific electronic health information. The ability to tailor personal access preferences will be necessary given that other patients have expressed the desire for faster release (Do et al., 2011) and greater accessibility (Krist et al., 2011) of test results. In other words, effort should be placed on allowing individuals to determine what information is presented in their PHRs and when it is made available, although a minimum waiting period under certain circumstances may be advisable (Do et al., 2011). Moreover, because elements of the medical record may be difficult for laypersons to understand, information access requires tools facilitating understanding and decision-making (Wagner et al., 2010; Do et al., 2011; Wiljer et al., 2010). Patients, providers, and caregivers all commented on concerns regarding secure messaging. Both patients and caregivers worried that secure messaging could replace more personal modes of communication (i.e., calling). Similarly, providers expressed concerns that secure messaging would lead to increased workload and lost face-to-face interaction with patients. Despite all three groups' concerns, Meaningful Use Stage 2 (The Office of the National Coordinator for Health Information Technology (ONC), 2013a) measure 17 calls for patients' ability to electronically message their providers. In design terms, this requires creating a solution to two conflicting viewpoints. Patients' and caregivers' fear of losing personal modes of communication may be partly due to not understanding how or when to use secure messaging. Thus, providing tutorials (Urowitz et al., 2012; Do et al., 2011) through the PHR either via web link or an embedded video may help patients feel more comfortable using another means to contact providers. However, as the use of secure messaging increases (Cronin et al., 2015), design solutions must also focus on lessening provider workload. One potential solution may be to implement a triagebased system that would route secure messages to appropriate staff members (Katz et al., 2003). Another option is to limit the number of characters allowed in messages, thus reducing the amount of text that must be read (Ye et al., 2010) and guiding the patient to use other forms of communication for in-depth inquiries. Some barriers encountered in this review are well recognized within the HF/E and medical informatics communities. One frequently reported under work system/unit factors was that data entry was a time consuming and difficult task with which some patients needed assistance. This barrier may be addressed by traditional human factors methods (e.g., simplified user interface, capacity for multiple forms of data). These methods may be supplemented by training, such as through tutorials on data entry and retrieval. At a higher level of the work system, barriers related to system boundaries and lack of interoperability were also recurring themes. The need to address these barriers not only for PHRs, but also for EHRs, has been widely recognized within the medical informatics community and prioritized as a topic of research and intervention (The Office of the National Coordinator for Health Information Technology (ONC), 2014; Otte-Trojel et al., 2015; Brennan et al., 2015). 4.5. Designing interventions: attending to context Identifying barriers should serve as a foundation for devising design solutions, a key component of HF/E practice (Dul et al., 2012). This review demonstrates that constructing solutions to improve the experience of PHR use requires targeting barriers at all levels of the work system. This means not only targeting the individual, the task performed, and the technology used, but also the larger environment in which these interactions are embedded (Holden et al., 2013; National Research Council, 2011; Dul et al., 2012). Traditionally, when attending to contextual factors, HF/E professionals have focused on implications of these work system levels for technology

229

design (i.e., determining how to design technology to align with contextual factors) (Valdez et al., 2015a; Marquard and Zayas-Caban, 2012; Moen and Brennan, 2005). Yet, as discussed during the recent HFES annual meeting, we acknowledge that the scope of our profession should be broadened to determine how to design and redesign these contextual environments. Current interventions to increase provider PHR use within the United States include the Centers for Medicare & Medicaid Services' (CMS) EHR Incentive Program, which rewards healthcare providers who demonstrate meaningful use (i.e., improved patient care outcomes through meeting core objectives of Meaningful Use Stages 1, 2 and 3) with incentive payments (Center for Medicare and Medicaid Services, 2015). HF/E professionals should collaborate on designing other interventions at higher work system levels to address barriers to PHR use for providers, patients, and caregivers. For example, insurance companies enact patient wellness programs, which incentivize patients' participation in health-related activities such as exercising or attending wellness checkups (Department of Health and Human Services, 2013). Similar programs could address patient and caregiver barriers to PHR use. 4.6. Limitations and future research The limitations of this review yield four directions for future research. First, included articles were limited to studies conducted in six countries, highlighting the need for work that explicitly addresses organization and external environment barriers arising in other national contexts. Second, the coding strategy used in this review focused on assigning barriers to specific work system factors. Future research should build on this by studying interactions between barriers within and across works systems. Third, this review is systematic in that specific criteria were used to determine article inclusion; however, the authors did not assess the level of evidence for the contribution of each work system factor to PHR use. This latter should be the focus of future inquiry. Lastly, although we consulted with one health science librarian and one computer science librarian to assist in our literature search, we acknowledge that developing an exhaustive search strategy is challenging. In retrospect, additional terms such as “impediment” and “obstacle” should be incorporated into future systematic reviews on this topic. 5. Conclusion This review focused on patient, provider, and caregiver work system barriers to PHR use and is timely given developments intended to accelerate PHR use such as the recent proposal of Meaningful Use Stage 3 guidelines within the United States. The results highlight the need to address how organization and external environment factors hinder PHR use, as well as how such barriers can be overcome. Moreover, simultaneous attention to patients, providers, and caregivers lays the groundwork for developing new work system models that explicitly account for the unique and overlapping roles and contexts of distinct individuals within a sociotechnical system. Future design directions should address low health literacy, access to unwanted health information, secure messaging impacts on provider workflow, secure messaging appeal to patients and caregivers, ease of data entry, and interoperability. A key next step is to synthesize the level of evidence for each type of PHR use barrier for each work system to guide intervention design. Acknowledgments We would like to thank Kelly Near and Andrea Denton from the University of Virginia's Library System for guiding and assisting with our literature search.

230

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

Appendix A

Study characteristics of included articles. Author, year Study name (reference)

Sample

Sample Type of size PHR

Instrumentation Methodology Study location

Journal domain

Ancker et al., 2011 Britto et al., 2013 Burke et al., 2010

Patients

74,368

EHR patient data; PHR use logs Interview

Quantitative United States

Medical sciences e Internal medicine

Qualitative

Medical sciences e Pediatrics

Use of an electronic patient portal among disadvantaged populations

Parents' perceptions of a patient portal for Caregivers 24 managing their child's chronic illness Patients 252 Transforming patient and family access to medical information: Utilization patterns of a patient-accessible electronic health record Byczkowski Family perceptions of the usability and value Caregivers 530 of chronic disease web-based patient portals et al, 2014 Crotty et al., Preparing residents for future practice: Report Providers 108 2013 of a curriculum for electronic patient-doctor communication providers Day and Gu, Influencing factors for adopting personal Patients 10 2012 health record (PHR) Denton, Will patients use electronic personal health Patients 136 2001 records? Responses from a real-life experience Do et al., The military health system's personal health Patients; 250; 10 2011 record pilot with Microsoft HealthVault and providers google health Use of a patient-accessible electronic medical Patients; 107; 11 Earnest providers record in a practice for congestive heart et al., failure: Patient and physician experiences 2004 Patients 760 Emani et al., Patient perceptions of a personal health 2012 record: A test of the diffusion of innovation model Providers 955 Fuji et al., Personal health record use by patients as 2008 perceived by ambulatory care physicians in Nebraska and South Dakota: A cross-sectional study Goel et al., Disparities in enrollment and use of an Patients 7088 2011 electronic patient portal The intersection of gender and place in online Patients 7674 Goldner health activities et al., 2013 Patients 7080 Processes and outcomes of developing a Gordon continuity of care document for use as a et al., personal health record by people living with 2012 HIV/AIDS in New York City Gu and Day, Propensity of people with long-term Patients 10 2013 conditions to use personal health records Guy et al., Evaluation of a web-based patient portal for Patients 99 2012 chronic disease management Hess et al., Exploring challenges and potentials of Patients 39 2007 personal health records in diabetes selfmanagement: Implementation and initial assessment Jung et al., Who are portal users vs. early E-Visit Patients, 10,532 2011 adopters? A preliminary analysis providers 136 Kahn et al., Personal health records in a public hospital: Patients 2010 experience at the HIV/AIDS clinic at San Francisco General Hospital Patient 70 Kim et al., Challenges to using an electronic personal 2009 health record by a low-income elderly population Krist et al., Designing a patient-centered personal health Patients 7235 2011 record to promote preventive care

Lau et al., 2013a Lau et al., 2013b

Consumers' online social Network topologies Patients and health behaviors Social and self-reflective use of a web-based Patients personally controlled health management system Patients

709 709

709

Tethered

Tethered Tethered

Tethered

Tethered

EHR patient data; PHR system logs Interview; survey

United States Quantitative United States Mixed methods

United States

Survey, PHR use Quantitative United logs States

Tethered

Interview, observation Untethered Survey

New Zealand Quantitative United States Untethered Survey, advisory Mixed United panels methods States

Tethered

Tethered

Qualitative

Mixed United Survey, States interview, focus methods group Survey Quantitative United States

Medical sciences e Cardiovascular diseases Medical sciences e Computer applications/Nurses and nursing Medical sciences

Medical sciences e Computer applications Health facilities and administration Medical sciences e Computer applications Medical sciences e Computer applications Medical sciences e Computer applicable

Both

Survey

Quantitative United States

Medical sciences e Business and economics e management

Tethered

PHR use logs

Not Specified

Survey

Quantitative United States Quantitative United States

Medical sciences e Internal medicine Medical sciences; Communications

Tethered

Mixed Focus group, survey, PHR use methods logs

United States

Biology e Computer applications; Medical sciences e computer applications

Tethered

Interview

Qualitative

Tethered

Survey, focus group Focus group, PHR use logs

Mixed methods Mixed methods

New Zealand Canada

Medical sciences e computer applications; Medical sciences n/a

United States

Medical sciences e Computer applications

Tethered

PHR use logs

Tethered

PHR use logs, survey

Quantitative United States Quantitative United States

Medical sciences e Computer applications Medical sciences e Computer applications

Untethered Survey, PHR use Quantitative United logs States

Medical sciences e Computer applications

Tethered

Both

EHR patient data, focus group, survey, interview Untethered Survey, PHR use logs Untethered Survey, PHR use logs Tethered

Mixed method

United States

Quantitative Australia Quantitative Australia

Quantitative Australia

Medical sciences e Computer applications

Medical sciences e computer applications; Medical sciences Medical sciences e Computer applications

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

231

(continued ) Author, year Study name (reference)

Sample

Which bundles of features in a web-based personally controlled health management system are associated with consumer helpseeking behaviors for physical and emotional well-being? Patients Lin et al., An internet-based patient-provider 2005 communication system: Randomized controlled trial Patients Lober et al., An internet-based patient-provider 2006 communication system: Randomized controlled trial Bridging the digital divide in diabetes: Family Patients Mayberry support and implications for health literacy et al., 2011 Patients Development and evaluation of an internet McInnes and personal health record training program et al., for low-income patients with HIV or hepatitis 2013 C Miller et al., Determinants of personal health record use Patients 2007 Impact of a wellness portal on the delivery of Patients Nagykaldi patient-centered preventive care et al., 2012 Patients Nazi, 2010 Veterans' voices: use of the American Customer Satisfaction Index (ACSI) Survey to identify My HealtheVet personal health record users' characteristics, needs, and preferences Patients, Nazi, 2013 The personal health record paradox: Healthcare professionals' perspectives and the providers information ecology of personal health record systems in organizational and clinical settings Nazi et al., Evaluating patient access to electronic health Patients 2013 records results from a survey of veterans Patients Internet portal use in an academic multiple Nielsen sclerosis center et al., 2012 Understanding patient portal use: Patients Osborn Implications for medication management et al., 2013 Poon et al., Empowering patients to improve the quality Patients 2007 of their care: design and implementation of a shared health maintenance module in a US integrated healthcare delivery network Sarkar et al., The literacy divide: Health literacy and the use Patients 2010 of an internet- based patient portal in an integrated health systemdresults from the Diabetes Study of Northern California (DISTANCE) Sarkar et al., Social disparities in internet patient portal use Patients 2011 in diabetes: evidence that the digital divide extends beyond access Design and implementation of a web-based Patients Schnipper patient portal linked to an electronic health et al., record designed to improve medication safety: 2008 the Patient Gateway medications module The association between personal health Patients Tenforde record use and diabetes quality measures et al., 2012 Tjora et al., Privacy vs. usability: A qualitative exploration Patients 2005 of patients' experiences with secure internet communication with their general practitioner Patients, Tom et al., Integrated personal health records use: caregivers 2012 association with parent-reported care experiences Patients Tsai et al., Use of the internet and an online personal 2012 health record system by US veterans: comparison of Veterans Affairs mental health service users and other veterans nationally Tuil et al., Patient-centered care: using online personal Patients 2006 medical records in IVF practice

Sample Type of size PHR

Lau et al., 2013c

606; 14 Tethered

Instrumentation Methodology Study location

Journal domain

Survey, PHR use logs

Medical sciences e Computer applications

Survey, PHR use Mixed logs methods

United States

Medical sciences e Computer applications

38

Untethered PHR use logs, observation

Quantitative United States

Medical sciences e Computer applications

75

Tethered

Mixed methods

United States

Medical sciences e Endocrinology

14

Tethered

Mixed methods

United States

Medical sciences; Public health and safety

63,295

Tethered

538

Tethered

Focus group, survey, EHR data Survey; interview; focus group PHR use logs

Quantitative United States Survey, PHR use Quantitative United logs States

Health facilities and administration Medical sciences

100,617 Tethered

Interview, survey

Mixed methods

United States

Medical sciences e Computer application

30

Tethered

Survey, Interview

Mixed methods

United States

Medical sciences e Computer application

668

Tethered

Survey

240

Tethered

Medical sciences; Public health and safety Medical sciences e Computer applications

75

Tethered

Mixed methods

United States

Medical sciences e Computer applications

2779

Tethered

EHR patient data, PHR use logs Focus group, survey, EHR patient data Survey

Quantitative United States Quantitative United States

Quantitative United States

Medical sciences e Computer applications, Medical sciences

14,102

Tethered

Survey, PHR use Quantitative United logs States

Medical sciences, Communications

14,102

Tethered

Survey, PHR use Quantitative United logs States

Medical sciences e Computer applications

5298

Tethered

PHR use logs, survey

Quantitative United States

Medical sciences e Computer applications

10,746

Tethered

Quantitative United States

Medical sciences e Internal medicine

15

Tethered

EHR patient data, PHR use logs Interview

Qualitative

Medical sciences e Computer applications

256

Tethered

Survey

Quantitative United States

Medical sciences e Pediatrics

7215

Tethered

Survey

Quantitative United States

Medical sciences e Computer applications

102

Tethered

Questionnaire, PHR use logs, interview

Mixed methods

Norway

The Medical sciences e Netherlands Endocrinology; Medical sciences e Obstetrics and (continued on next page)

232

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

(continued ) Author, year Study name (reference)

Urowitz et al., 2012 Vodicka et al., 2013 WadeVuturo et al., 2013 Wagner et al., 2012 Wagner et al., 2010 Wald et al., 2010

Sample

Sample Type of size PHR

Patients; 17; 64 Improving diabetes management with a patient portal: a qualitative study of diabetes providers self-management portal Online access to doctors' notes: patient Patients 3874 concerns about privacy

Instrumentation Methodology Study location

Qualitative

Canada

gynecology; Pharmacy and pharmacology Medical sciences e Computer applications

Tethered

Survey; interview

Tethered

Survey, PHR use Quantitative United logs States

Medical sciences e Computer applications

Secure messaging and diabetes management: Patients experiences and perspectives of patient portal use

54

Tethered

Focus group, survey

Mixed methods

United States

Medical sciences e Computer applications

Personal health records and hypertension control: A randomized trial

Patients

443

Tethered

EHR patient data, survey

Quantitative United States

Medical sciences e Computer applications

Patients

16

Untethered Interview

Qualitative

Patients; 3979; providers 272

Tethered

Survey, EHR patient data

Quantitative United States

Medical sciences; Business and Economics e Management Medical sciences e Computer applications

Patients

126

Tethered

Survey, PHR use Quantitative United logs States

Medical sciences e Computer applications

Patients

61

Untethered PHR use logs

Quantitative United States

Patients

980

Tethered

Quantitative United States

Biology e Biotechnology; Medical sciences e Computer applications Medical sciences e Computer applications

Incorporating patient perspectives into the personal health record: implications for care and caring Implementing practice-linked pre-visit electronic journals in primary care: Patient and physician use and satisfaction Wald et al., Survey analysis of patient experience using a 2009 practice-linked PHR for type 2 diabetes mellitus Wang et al., Personal health information management 2004 system and its application in referral management Who uses the patient internet portal? The Weingart PatientSite experience et al., 2006 Weitzman Willingness to share personal health record data for care improvement and public health: et al., a survey of experienced personal health 2012 record users Wen et al., Consumers' perceptions about and use of the 2010 internet for personal health records and health information exchange: Analysis of the 2007 Health Information National Trends Survey Wiljer et al., The anxious wait: assessing the impact of 2010 patient accessible EHRs for breast cancer patients Patient experiences with full electronic access Woods to health records and clinical notes through et al., the My HealtheVet personal health record 2013 pilot: Qualitative study Wynia et al., Many physicians are willing to use patients' 2011 electronic personal health records, but doctors differ by location, gender, and practice Interest in the use of computerized patient Zickmund portals: Role of the providerepatient et al., relationship 2008 Patient interest in sharing personal health Zulman record information, a web-based survey et al., 2011

PHR use logs

United States

Patients, 261 caregivers

Untethered Survey

Quantitative United States

Medical sciences e Computer applications

Patients

7674

Not Specified

Survey

Quantitative United States

Medical sciences e Computer applications

Patients

311

Tethered

Survey

Quantitative Canada

Medical sciences e Computer applications

Patients, 30; 6 providers, caregivers

Tethered

Focus group

Qualitative

Medical sciences e Computer applications

Providers

856

Not Specified

Survey

Quantitative United States

Insurance; Public health and safety

Patients

39

Tethered

Focus Group

Qualitative

United States

Medical sciences e Internal medicine

Patients

18,471

Tethered

Survey

Quantitative United States

Medical sciences e Internal medicine

United States

Participant demographics for included studies focusing on patient work systems. Citation

Health condition

Ancker et al., 2011

M ¼ 40 Hypertension e 16%, Hyperlipidemia 13%, Asthma e 16% Black, 19% Hispanic, 44% White, 6% other, 15% missing/unknown 9%, Diabetes e 7%, Depression e 6%, Drug Abuse/ Dependence e 2.3%, Chronic hepatitis (B, C, or other) 1.8%, Alcoholism e 1.5%, HIV/AIDS e 1.3%, Congenial cardiac disease 37.5% Hispanic, 35.2% White, 10.1% Black, 0.7% Asian, n/a 16.5% Other

Burke et al., 2010 Day and Gu, 2012

Journal domain

Race

Age

One or more long term health condition

n/a

Range: 35e79

Spinal disorders

n/a

Range: 35e85

M.J. Thompson et al. / Applied Ergonomics 54 (2016) 218e242

233

(continued ) Citation Denton, 2001 Do et al., 2011 Earnest et al., 2004 Emani et al., 2012 Goel et al., 2011 Goldner et al., 2013

Health condition

Race

Age

n/a

n/a

M ¼ 53.14

Congestive heart failure

90% White, non-Hispanic

M ¼ 56

Asthma, congestive heart failure (CHF), hypertension, Innovators: 94% Caucasian, Other users: 90% or diabetes Caucasian, Laggards: 94% Caucasian, Rejecters: 86% Caucasian, Non-adopters: 76% Caucasian n/a 49% White, 15% Black. 4% Latino, 2% Asian, 12% other, 19% missing n/a Health seekers: 25.2% Non-White, 74.8% White; Internet users: 24.7% Non-White, 75.3% White

38.1% African American/Black, 0.3% Asian, ¼55 Nonusers: 61.7% < 55, 38.3%>¼55 M¼45.4

62.7% White, 33.3% African American/black

M ¼ 56.9

n/a

n/a

M ¼ 47.4

Diabetes

21% AfricaneAmerican, 28% Non-Hispanic White, 14% Latino, 9% Asian American, 12% Filipino, 17% Other/ mixed 21% AfricaneAmerican, 28% Non-Hispanic White, 14% Latino, 9% Asian American, 12% Filipino, 17% Other/ mixed White 83%

4% 30e39, 15% 40e49, 33% 50e59, 32% 60e69, 17% 70þ

Osborn et al., 2013 Poon et al., 2007 Sarkar et al., 2010 Sarkar et al., 2011 Schnipper et al., 2008 Tenforde et al., 2012

Diabetes

n/a

Diabetes Mellitus

Tjora et al., n/a 2005 Tom et al., Asthma 24%, congenital musculoskeletal 2012 abnormalities 20%, congenital heart disease 18%, inborn errors of metabolism 9%, cystic fibrosis 7%, and hereditary and acquired helytic anemias 5% Tsai et al., n/a 2012

Tuil et al., 2006 Urowitz et al., 2012 Vodicka et al., 2013 WadeVuturo et al., 2013 Wagner et al., 2012 Wagner et al., 2010 Wald et al., 2010 Wald et al., 2009 Wang et al., 2004 Weingart et al., 2006 Weitzman et al., 2012 Wen et al., 2010

4% 30e39, 15% 40e49, 33% 50e59, 32% 60e69, 17% 70þ M ¼ 48

Users: 84.0% Caucasian, 11.0% African American, 3.0% Users: M ¼ 59 Non-Users: M ¼ 62 Other, 2% unknown, 60 years

n/a

n/a

Range: 49e82

Diabetes

28% Non-White

M ¼ 54

n/a

n/a

Work system barriers to patient, provider, and caregiver use of personal health records: A systematic review.

This review applied a human factors/ergonomics (HF/E) paradigm to assess individual, work system/unit, organization, and external environment factors ...
2MB Sizes 0 Downloads 7 Views