Using telehealth consultations for healthcare provision to patients from racial/ethnic minorities: A systematic review

Information sources and search strategy

The protocol for this systematic review was registered in the international register of systematic reviews PROSPERO (CRD 42020221017) and was conducted in accordance with the PRISMA guidelines.[25] Studies were identified from a search of the following 5 electronic databases: Ovid Medline, Ovid PsycINFO, EMBASE, and CINAHL via EbscoHOST and Scopus platforms. The search was limited to studies published in English between 1 January 2005 and 9 October 2020. Additionally, a search for existing systematic reviews on the same topic was conducted in: all EBM reviews via Ovid, Cochrane Database of Systematic Reviews, Joanna Briggs Institute, and PROSPERO. Reference lists of included studies were hand-searched for relevant studies.

A list of terms was formulated for the concepts: “telehealth” and “racial or ethnic minority”. These terms were mapped to subject headings (e.g., Medical Subject Headings or equivalent) and keywords. Boolean operators were used to group subject headings and keywords to create a search strategy (Supplementary file X).

Eligibility Criteria

To be included in the review the study had to meet the following inclusion criteria:

1. Sample population included patients from racial/ethnic minorities, their carers or healthcare staff who provided care to racial/ethnic minorities, with study findings disaggregated by racial/ethnic minority groups.

2. For this review, we describe the intervention as ‘telehealth consultation’. Telehealth consultations occurred instead of a face-to-face consultation to provide a health service such as a clinical assessment or diagnosis and/or management of a health condition (physical or mental). Support services provided as standard practice (e.g., phone interpreting, telemonitoring without direct interaction with a healthcare provider) were excluded. Studies that primarily focussed on delivery of health promotion programs or health-related education via telehealth were also excluded.

3. Included telehealth consultations in healthcare settings (e.g., hospitals, primary care clinics)

4. Reported individual health outcomes (e.g., physical health, mental health). Individual level outcomes could be related to patients/health consumers and care-givers.

5. Reported health service outcomes (e.g., readmission rates, barriers to implementation). Health service level outcomes may be related to access to care, service delivery, satisfaction with care and cost-effectiveness

6. Used quantitative, qualitative or mixed methods study designs

7. Published in a peer-reviewed journal

Reviews of telehealth consultations and healthcare for Indigenous peoples have been published.[26–30] We therefore did not include Indigenous peoples in the current review. Publications were excluded if they were opinion pieces, systematic reviews and meta-analyses, conference abstracts or proceedings or duplicate publications using the same sample and reporting the same outcomes.

Data screening and extraction

Search results were exported from the electronic databases into the systematic review management platform Covidence (Veritas Health Innovation, 2021). Screening was undertaken in two phases. Following the removal of duplicates, studies were screened independently in duplicate by two reviewers on title and abstract. Conflicts were resolved by a third reviewer. In the second screening phase, full text articles were screened independently in duplicate by two reviewers and conflicts were resolved by a third reviewer.

Data extraction was performed by two reviewers using Covidence, with differences resolved by a third reviewer. Extracted data comprised the characteristics of each study: country of study, health setting, study aim, study design, sample size and participants’ characteristics (e.g., gender, age, race/ethnicity/cultural background), health conditions, and the details of the telehealth intervention. Data were also extracted pertaining to study findings: health outcomes (e.g., blood pressure, depression symptoms) and health access and delivery outcomes (e.g., satisfaction with care, implementation barriers and facilitators).

Methodological quality assessment of studies

Risk of bias and critical appraisal were conducted independently in duplicate by two reviewers using the Joanna Briggs Institute (JBI) Critical Appraisal tools.[31] The JBI Critical Appraisal Tools assess the methodological quality of studies of various design and the extent to which the risk of bias is addressed by authors. For each criterion assessed as ‘met’ on the appropriate tool, a score of ‘1’ was applied. Where criteria were assessed as ‘not met’ or if it was ‘unclear’ to the reviewer, a score of ‘0’ was applied. The scores were then calculated for each study and converted to a final quality rating of ‘low’, ‘moderate’ or ‘high’. The following JBI Critical Appraisal Tools and scoring parameters were implemented in this review: Checklist for randomized-controlled trials (RCTs) (score out of 10; Low 0-3, Moderate 4-7, High 8-10); the Checklist for quasi-experimental studies (score out of 9; Low 0-3, Moderate 4-6, High 7-9); Checklist for Analytical Cross-Sectional Studies (score out of 8; Low 0-3, Moderate 4-6, High 7-8); the Checklist for Qualitative Research (score out of 10; Low 0-3, Moderate 4-7, High 8-10); the Checklist for Cohort Studies (score out of 11; Low 0-4, Moderate 5-8, High 9-11); and, the Checklist for Case Series Studies (score out of 10; Low 0-3, Moderate 4-7, High 8-10). (See Supplementary file for critical appraisal criteria for each study design.)

For mixed methods studies, we used two questions screening questions and five questions from the ‘mixed methods studies’ section of the Mixed Methods Appraisal Tool version 2018 (MMAT),[32] (score out of 7; Low 0-2, Moderate 3-5, High 6-7) which are related to the rationale for using a mixed methods study design and level of interpretation and integration of the qualitative and quantitative components of the study.[32]

Synthesis of studies

Extracted data were exported from Covidence into a Microsoft Excel spreadsheet for analysis. A convergent integrated approach to data synthesis was taken in accordance with the JBI methodology for mixed methods systematic reviews.[31] Due to the heterogeneity of study methodology, design, sample characteristics, and outcome measures, there was no opportunity to conduct a meta-analysis of quantitative studies nor a meta-synthesis of qualitative studies. A narrative synthesis was thus conducted where both qualitative and quantitative data were tabulated together and categorized according to the four outcomes of interest: health outcomes, health access, implementation barriers and facilitators to telehealth consultations, and satisfaction with care. A comparison between studies reporting on the same outcome was performed, with similar and divergent findings reported. Study characteristics such as country and setting, study methodology and design, sample sizes and participant characteristics, were examined by calculating frequencies and proportions.

Study selection

The combined searches from the four databases identified 3,314 unique articles which were screened using the inclusion and exclusion criteria. A total of 28 articles met the inclusion criteria for the systematic review (Figure 1).

• Download figure
• Open in new tab

Figure 1:

Flow diagram of study selection. Adapted from PRISMA flow-chart [25]

Study characteristics

Studies were published between 2005 and 2019 and comprised 4,301 patients and 12 caregivers from racial/ethnic minorities, and 34 healthcare practitioners. The majority of studies (n=23, 82.1%) examined groups located in the United States and utilized quantitative study designs (n=19, 67.9%). The most common health conditions examined were mental health conditions (n=17, 60.7%), of which nine focused on depression. The most frequent study setting was community health centers (n=8, 28.6%), followed by primary health clinics (n=5, 17.9%), psychiatric or trauma treatment centers (n=4, 14.3%), the community (n=3, 10.7%) and HIV clinics (n=2, 7.1%). The racial/ethnic minority populations included in the studies were predominantly Latino/a or Hispanic (n=13, 46.4%) and African American (n=7, 25.0%). Seven studies included Asian participants from Korean and Chinese communities, three studies included participants from refugee and immigrant backgrounds and one study described their minority population as ‘non-white ethnic group’. Characteristics of included studies are presented in Table 1. (Supplementary file Table 1 presents detailed information related to study characteristics.)

Quality assessment

The quality of the included studies was mixed (see Supplementary file). All 11 RCTs reported random sampling from patients presenting for clinical care. There was variation in the reporting of allocation concealment and differences in baseline characteristics between groups in some RCTs, however most reported completion of follow-up assessments and an intent to treat approach to data analysis. As all RCT studies investigated a telehealth consultation between a practitioner and a patient, blinding of participant and practitioner was not possible. Accordingly, no RCT studies reported blinding of participants to intervention group or blinding of practitioners delivering the intervention to group allocation. The number of quality indicators for RCT studies reported ranged from 3/10 to 7/10. Each of the two quasi-experimental studies employed a pre/post design and there were mixed results regarding study quality. The quality of reporting for the one study utilizing a cohort design and for the one study using a case series design was high. Overall, the quality of reporting for studies using cross-sectional, mixed methods and qualitative designs was low to medium.

Telehealth consultations

Data pertaining to telehealth consultations are presented in Table 2. Across the included studies, telehealth consultations were primarily mental health consultations (n=17, 61%) such as cognitive behavioral therapy (CBT) or psychiatry (telepsychiatry). Physical health consultations were related to chronic disease management (n=9, 32.1%) and clinical monitoring (n=2, 7.1%). Physical health diseases that were managed by telehealth consultations included diabetes (n=2, 7%), hypertension (n=3, 11%) and infectious diseases (n=2, 7%). Telehealth consultations were used to monitor patients’ clinical status in two studies. One study used a telephone triage system and computer-supported decision-making software.[33] The other study monitored patient’s vital signs and other health related information ‘remotely’; and this was then used to inform virtual ‘wellness’ visits by a nurse practitioner who discussed the patient’s medications and recent health data.[34]

The telehealth consultations and interventions were delivered by a range of healthcare professionals including: nurses (n=10, 35.7%),[33–42] physicians (n=13, 46.4%)[33 ,37 ,39 ,40 ,42–51] including psychiatrists, primary care physicians and specialists, allied health professionals (n=9, 32.0%)[46 ,52–59] such as social workers, counsellors and psychologists, and paraprofessional outreach workers in one study.[39] In eighteen studies (64%), the healthcare professionals delivering the telehealth intervention were bilingual. The mode used to deliver the telemedicine intervention was video conferencing in 16 studies (57.1%) and telephone in 11 studies (39.3%); one study used both internet video and telephone communication.[39]

Health outcomes

In total, 14 studies measured health outcomes with evidence of effectiveness of telemedicine on measures of mental health, cardiovascular health and diabetes related outcomes. Three RCTs focused on health outcomes related to treatment of depression. These three studies assessed the effect of a telehealth CBT intervention on depression severity compared with face-to-face CBT[55 ,60] or compared with anti-depressant therapy.[53] Among Latino patients, worry symptoms were reduced[60] and depression symptoms were improved[55] following the use of culturally adapted telephone-based CBT compared to face-to-face CBT. In the study by Dwight-Johnson[53], Latino patients receiving culturally adapted telephone-based CBT showed greater improvement in depression symptoms over 6 months compared to patients who received usual care with anti-depressant medications.

The effects of telepsychiatry consultations or remote management on improving depression outcomes were reported in three RCTs.[43 ,46 ,49>] In a study of Hispanic patients with depression, monthly telepsychiatry consultations and usual care were equally effective in reducing depression scores after 6 months.[43] In another pilot RCT [46], telephone-based depression care management plus usual care was compared with the usual care only group. The findings of this study showed a trend for lower levels of depression in the intervention group over time compared with the control group.[46] In another study, the effectiveness of culturally sensitive telepsychiatry treatment and care management was compared to usual care for Chinese Americans with depression.[49] This study found greater improvement in depressive symptoms following telepsychiatry compared with usual care.[49] In a single group pre-test post-test design with Korean immigrants, weekly sessions of telepsychiatry for four weeks significantly reduced depression scores at immediate follow-up and at three-months follow-up.[56] In a pilot study using a pre-post two group design focusing on African American caregivers of older adults with dementia, telephone-based and face-to-face CBT interventions were equally effective in reducing depression scores.[54]

Three studies assessed the effects of telemedicine on cardiovascular and diabetes health outcomes.[37–39] In one RCT study of patients with hypertension from Korea, bi-weekly and monthly hypertension-related telephone counselling, and 12-month home monitoring of BP were effective in improving long-term BP outcomes[38]. The second study employing a RCT design assessed the effectiveness of a telemedicine intervention emphasizing medication management and home BP monitoring with African American patients living with hypertension [37]. Jackson et al.[37] found significant improvement in mean systolic blood pressure at 12-months and at 18-months following the intervention compared with usual care. A study by Mayes [39] assessed the impact of telemedicine on blood glucose regulation among Hispanic patients living with type 2 diabetes and found that telemedicine monitoring significantly improved participants’ blood glucose regulation.

Barriers and facilitators

Three studies identified technology as a barrier to the implementation of telehealth, which was not unique to a health condition or study design. A mixed method study involving African Americans and Hispanics with mental health conditions identified initial issues with logging into the videoconferencing software and technical problems with the equipment.[59] A cross sectional study involving refugees with infectious diseases experienced technical difficulties in a quarter of the first telehealth consultations, although this significantly improved with experience.[42] A qualitative study involving African Americans and Hispanics with heart failure found that there were initial concerns with using the equipment, but these dissipated with use.[41] This study also reported issues with internet connection, using the equipment, patients managing the intervention alone at home, and concerns with patient literacy. The combined use of audio and text in the telemonitoring program assisted patients’ understanding.[41]

Six studies identified facilitators to telehealth implementation. The facilitators were not unique to a health condition or race/ethnic background. Telehealth consultations using video conferencing supported health professionals to provide education to patients with mental health conditions by sharing PowerPoint presentations and other written material.[58 ,59] Telehealth permitted caregiver engagement and could be delivered in a language of the patient and caregiver’s preference.[58 ,59] Patients from ethnic minorities reported ease of access to medical help through the GP telephone triage system.[33] Telehealth allowed patients with HIV quick access to their providers and increased their likelihood of attending appointments.[57] Patients with diabetes and hypertension gave favorable ratings to the accessibility of telehealth.[34] Strategies to assist with telehealth consultations included: face-to-face meetings prior to establishing telehealth to build relationships[59] and use of a

Satisfaction with telehealth consultations

Overall, 16 studies (57.1%) explored satisfaction with telehealth consultations. Of these, nine were quantitative studies[33 ,40 ,43 ,45 ,50 ,53 ,55 ,56 ,60], five studies were mixed methods [34 ,46 ,48 ,51 ,59] and two studies used qualitative methods.[52 ,57] Results across the sixteen studies were mixed, with the majority of studies (n=11, 68.8%) reporting high levels of satisfaction with telehealth among patients, carers and health professionals.[34 ,40 ,43 ,45 ,50–53 ,56 ,57 ,59] Reasons for satisfaction with telehealth consultations included increased efficiency and convenience, enhanced privacy and reduced need for travel compared to in person face-to-face consultations. Some studies found no discernible differences in levels of satisfaction between telehealth groups and comparison groups (i.e. usual care or face-to-face interventions)[46 ,55 ,60] and participants in one study gave mixed feedback about telepsychiatry.[48] In a study examining telephone triage, patients from ethnic minorities reported higher satisfaction in the GP triage arm compared to usual care but lower satisfaction compared with white patients overall.[33]

Limitations of the current evidence-base and areas for further research

This review identified key limitations of existing research in telehealth for racial/ethnic minorities. Given the increasing and more widespread adoption of telehealth consultations across the health system due to the COVID-19 pandemic, it is important to understand its impact for patients from racial/ethnic minority groups to ensure equity of healthcare access and utilization. A recent study of telemedicine use during COVID-19 has already identified disparities in telemedicine access for African American patients compared to white patients.[6]

Further research is required in other areas such as acute and chronic disease management and in other care settings for racial/ethnic minorities such as cancer care, kidney disease, emergency department admissions, critical care, pediatrics and obstetrics/midwifery. Healthcare services and institutions should monitor use of telehealth consultations across patient demographics, including language and racial/ethnic background, to measure the impact on health equity. Although our search strategy identified studies conducted in non-English speaking countries, most studies were based in the United States of America. Additional studies are required in other countries with underserved populations who may have less access to information technologies.

Included studies were also limited by a lack of methodological rigor across most study designs, resulting in a significant gap in the current evidence base. Longitudinal studies are needed to determine whether, and to what extent, telehealth consultations affect health-related outcomes over time for racial/ethnic minorities. In particular, the effects of telehealth on early identification of disease and on the long-term management of chronic health conditions are required to maximize long-term health outcomes for racial/ethnic minorities.

Patients from marginalized racial/ethnic minority groups such as refugees were only specifically studied in three of the 28 studies.[42 ,45 ,51] Issues related to caregiver support and the optimal mode of telehealth consultation delivery (i.e., telephone vs videoconferencing) should also be explored in future research.

Limitations of the review

Our review has several limitations. Only articles published in English were included therefore there is risk of publication bias. Additionally, the search period of 1 January 2005 to 9 October 2020, may have omitted earlier relevant studies. However, information technology enabling telehealth consultation has evolved considerably since the early 21^st century. We anticipate that studies published prior to 2005 would have limited relevance.