This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Dermatology Research, is properly cited. The complete bibliographic information, a link to the original publication on http://derma.jmir.org, as well as this copyright and license information must be included.
In emerging market countries in sub-Saharan Africa, access to specialty services such as dermatology is limited. Teledermatology is an innovative solution to address this issue; however, many initiatives have been tried without sustained success. Recently, WhatsApp has been used as a store-and-forward telemedicine communication platform for consultation and education in Botswana.
This study aims to describe the utilization of WhatsApp for teledermatology and the satisfaction levels of participating providers.
A 2-part pilot study was conducted. First, a retrospective review was performed of WhatsApp communications received by participating dermatologists in Gaborone, Botswana, from January 2016 to December 2019. Sender information, patient demographics and history, response time, diagnoses made, and follow-up recommendations were collected. Second, a 12-question cross-sectional survey was distributed to health care providers who utilized WhatsApp for teledermatology during this period. Descriptive statistics were then performed.
There were 811 communication threads over the study period. The majority (503/811, 62%) of communications were consultations from providers inquiring about a specific patient, followed by multidisciplinary care coordination communications (90/811, 11%). Our in-depth analysis focused on the former. In 323 (64%) provider consultations, dermatologists responded within 1 hour. A diagnosis was made in 274 (55%) consultations. Dermatologists gave treatment recommendations remotely in 281 (56%) consultations and recommended an in-person dermatology visit in 163 (32%). Of the 150 health care providers surveyed, 23 (15%) responded. All respondents (100%) felt that there was a need for teledermatology and improved teledermatology education in Botswana. Moreover, 17 (74%) respondents strongly felt that the guidance received via WhatsApp was high quality, and 22 (96%) were satisfied with WhatsApp as a platform for teledermatology.
This retrospective review and survey demonstrated that WhatsApp is a quick, well-received, and sustainable method of communication between dermatologists and providers across Botswana. The app may offer a solution to the challenges providers face in accessing specialty referral systems, point-of-care education, and medical decision-making support for complex dermatologic cases in Botswana. The information gained from this pilot study can serve as the basis for future telemedicine studies to improve the implementation of teledermatology in Botswana and other resource-limited countries.
Despite a high burden of dermatologic diseases, access to dermatologic specialty care is scarce in sub-Saharan Africa [
Teledermatology is a potential solution to address these challenges. Formal telemedicine platforms have been specifically designed to securely communicate predetermined sets of information between providers. Several have been developed and trialed in Botswana. In 2007, the Africa Teledermatology Project began providing health care providers in many sub-Saharan African countries, including Botswana, free access to a web-based platform for consultations, forum discussions, and educational materials [
Although formal teledermatology platforms can offer security and standardization, in low resource settings, the associated logistical and cost burdens frequently render them unfeasible or unsustainable, as previously seen in Botswana. Informal platforms are an alternative that allow the transmission of information via flexible, secure methods that can function on personal mobile phones and with lower bandwidth. Teledermatology through mobile health (mHealth) has demonstrated technical feasibility and reliability in providing care to underserved and remote populations around the world where smartphones are common, but the key to utilization is the ability to send consults within an app that is familiar to the user on their own mobile device [
Mobile phone subscriptions have been increasing in resource-limited countries [
Because WhatsApp is a relatively new platform for teledermatology, it is important to understand how physicians in Botswana are currently using it and gain user feedback to determine its feasibility, effectiveness, and potential to scale for use in other specialties. This pilot study aims to describe how the WhatsApp application is being utilized in Botswana to connect providers to dermatology expertise for patient care and education, as well as to elucidate current provider satisfaction with the platform.
This study was approved by the University of Botswana and the Botswana Health Research and Development Committee institutional review boards (HPDME 13/18/1) and was granted a review exemption by the University of Pennsylvania IRB (835003).
In the first part of the study, we conducted a retrospective review of teledermatology communications sent through WhatsApp in Botswana from January 2017 to December 2019. Messages from health care providers and patients seen within the public health care system of Botswana were sent to 2 dermatology consult mobile phones. Messages were reviewed by a full-time local dermatologist and rotating North American dermatology residents and faculty at Princess Marina Hospital (PMH) in Gaborone, Botswana. WhatsApp messages were downloaded from the mobile phones, and conversation threads found on both were identified as duplicates and removed. There was no specific record kept of which dermatologist reviewed and responded to individual messages.
An overview of the data extraction and categorization methods used to analyze conversation threads is summarized in
Further data collection was only performed on the consultations received from a provider. We extracted data into categories that were considered to be most important to patient care in Botswana, including patient demographics, history provided, photograph parameters, response time, diagnosis, and outcome of consultation. Patient age, sex, and HIV status were recorded if provided. The extent of a history of present illness (HPI) shared by the nondermatologist provider was based on a point system, with 1 point given for each of the following: description of the lesion, location on the body, symptoms reported, timing of onset, change in appearance over time, aggravating or alleviating factors, prior treatments performed, and pertinent lab or imaging results.
Photographs received with consultations were also reviewed. Two authors (TW and AF) with experience reviewing teledermatology consultation photos in Botswana developed a subjective grading system that rated photos as low, medium, or high quality based on criteria of image resolution, lighting, and content (whether photos captured relevant areas of the body). To standardize grading, photos were reviewed by 2 authors (VW and AF) until a consistent agreement on grading was achieved. After standardization was achieved, each photo was graded by 1 author. The file size (kB) of photographs was also recorded.
Response times from dermatology consultants was stratified (0-60 minutes, 1-6 hours, 6-12 hours, 12-24 hours, 24-48 hours, and >48 hours) based on both the time from initial message sent to initial response and initial message sent to final diagnosis or recommendation. Dermatologists provided no diagnosis, a single diagnosis, multiple diagnoses, and/or differential diagnoses in response to consultations from providers. All diagnoses, including those that were differentials, were included in the overall analysis. Diagnoses were classified into the major categories of inflammatory disorders, infection, neoplasm, diseases of vasculature, and other diagnoses. Consultation outcomes were based on the dermatologist’s recommendation and divided into the following categories: advice for local management (when treatment recommendations were provided remotely), referral to see a dermatologist, referral to see a different specialist, or other recommendation. We recorded whether dermatologists provided education to providers (clinical information in addition to a diagnosis and treatment plan).
Descriptive statistics were used to broadly categorize the conversation threads and demographic and clinical data provided in the consultations.
In the second part of the study, we conducted a cross-sectional survey of health care providers in Botswana who used WhatsApp for teledermatology from January 2016 to December 2019.
A research electronic data capture (REDCap) survey was distributed via WhatsApp, and responses were kept anonymous. The target population was a convenience sample of providers that used the platform for consultations. Informed consent was obtained from all participants. This was a voluntary, open survey that consisted of 12 questions aimed at evaluating users' satisfaction and experience with the platform in terms of technical quality, perceived effectiveness and usefulness, privacy and security practices, and suggestions for improvements (
The frequency of responses to survey questions were recorded and reported, and common themes in areas for improvement were identified.
From January 2017 to December 2019, there were a total of 811 conversation threads, with 102 threads in 2017, 350 in 2018, and 324 in 2019. There were 35 threads with a missing date stamp that were also included in the analysis. Approximately 150 senders were identified based on unique phone numbers and names in the phone contact list. An exact number of senders could not be confirmed due to inconsistencies in the way contact information was saved in each mobile phone.
The most common (503/811, 62%) purpose of communication was a consultation from a provider, as seen in the conversation threads, followed by multidisciplinary care coordination (23/811, 11%) (
Categories of communication between dermatologists and nondermatologist providers according to WhatsApp communication threads (N=811).
Purpose of communication | Communication threads, n (%) |
Consult from provider | 503 (62) |
Consult from patient | 23 (3) |
Remote patient management | 7 (1) |
Patient follow-up | 55 (7) |
Teletriage | 44 (5) |
Multidisciplinary care coordination | 90 (11) |
Provider question | 23 (3) |
Incomplete consult | 66 (8) |
Locations of providers that utilized WhatsApp for teledermatology in Botswana, with number of providers in each location.
Our in-depth analysis focused on the 503 WhatsApp consultations from nondermatologist providers (
Format of consultations sent by nondermatologist providers (N=503).
Measures | Consultations | ||
Age provided, n (%) | 380 (76) | ||
Sex provided, n (%) | 383 (76) | ||
HPIa provided, mean (SD)b | 3.1 (1.6) | ||
HIV status provided, n (%) | 234 (47) | ||
Photo provided, n (%) | 477 (95) | ||
|
|||
|
Low | 101 (20) | |
|
Medium | 198 (39) | |
|
High | 178 (35) | |
|
|||
|
<50 Kb | 173 (34) | |
|
50-100 kB | 214 (42) | |
|
100-150 kB | 52 (10) | |
|
>150 kB | 37 (7) |
aHPI: History of present illness.
bHPI provided in the consultation was graded on a point system, with 1 point given for each of the following: description of the lesion, location on the body, timing of onset, change in appearance over time, aggravating and alleviating factors, prior treatments performed, symptoms reported, and pertinent lab or imaging results.
cSubjective photograph quality was determined based on a grading system in which the criteria were image resolution, lighting, and whether relevant areas of the body were captured. Photos were graded on a scale of low, medium, or high quality based on the number of quality criteria met (
Example of a consultation sent from a nondermatologist provider to a dermatologist.
Dermatologists responded to the provider within 1 hour in 64% (323/503) of consultations and provided the final diagnosis or recommendation in 54% (272/503) (
Dermatologists provided 224 unique diagnoses out of a total of 704 diagnoses made. The most common were eczema, contact dermatitis, and warts (
Outcomes of consultations sent by nondermatologist providers (N=503).
Measures | Consultations, n (%) | ||
|
|||
|
0-59 minutes | 323 (64) | |
|
1-6 hours | 120 (24) | |
|
6-12 hours | 35 (7) | |
|
12-24 hours | 8 (2) | |
|
24-48 hours | 6 (1) | |
|
>48 hours | 11 (2) | |
|
|||
|
0-59 minutes | 272 (54) | |
|
1-6 hours | 144 (29) | |
|
6-12 hours | 35 (7) | |
|
12-24 hours | 13 (3) | |
|
24-48 hours | 10 (2) | |
|
>48 hours | 14 (3) | |
|
No final diagnosis or recommendation | 15 (3) | |
|
|||
|
Single diagnosis | 259 (52) | |
|
Multiple diagnoses | 15 (3) | |
No diagnosis made | 229 (46) | ||
Differential diagnosis provideda | 159 (32) | ||
|
|||
|
Local management | 281 (56) | |
|
Dermatology referral | 163 (32) | |
|
Referral to other specialist | 15 (3) | |
|
Other | 44 (9) | |
Education provided | 140 (28) |
aEach patient could have single or multiple conditions presented by the consulting provider. Each of these conditions was considered separately by the evaluating dermatologist. The evaluating dermatologist could provide a single diagnosis, multiple diagnoses (at least 2), no diagnosis, and/or a differential diagnosis for any condition they determined was present.
A survey was sent out to approximately 150 health care providers, of which 15% (23/150) completed the survey (
In terms of privacy and security, only two-thirds (15/23, 65%) of respondents reported always obtaining consent from patients for photos to be sent via teledermatology. Of those who obtained consent, all obtained verbal instead of written consent. Nearly all respondents used a personal phone (21/23, 91%) or camera (1/23, 4%). A majority (14/23, 61%) kept these photos on a password protected device, but nearly one-third (9/23, 39%) did not or only occasionally did. Most respondents (19/23, 83%) were not concerned about privacy or security issues while using WhatsApp for teledermatology. Concerns reported included the possibility of hacking, forwarding photos, and inappropriate access by third parties (
When asked about areas of improvement, respondents shared issues regarding the timing of responses, availability of consultants, and difficulty keeping case discussions organized when multiple separate patient consults were sent within 1 text thread. Another provider expressed concern about patients being able to obtain an in-person follow-up by a dermatologist when needed.
Responses to the Provider Satisfaction Survey questions assessing the overall utility of WhatsApp as a teledermatology platform.
Responses to the Provider Satisfaction Survey questions assessing specific features of WhatsApp as a teledermatology platform and patient data safety practices (N=23).
Questions | Responses, n (%) | |
|
||
|
Easy to send consults | 21 (91) |
|
I already have and know how to use the application | 20 (87) |
|
Easy to ask follow-up questions | 19 (83) |
|
Fast response times | 14 (61) |
|
Doesn’t require a computer with Internet | 10 (44) |
|
Easy to get patients urgently scheduled with dermatology clinic | 11 (48) |
|
||
|
Always | 15 (65) |
|
Almost always | 6 (26) |
|
Often | 1 (4) |
|
Never | 1 (4) |
|
||
|
Verbal | 23 (100) |
|
Written | 0 (0) |
|
||
|
Personal phone | 21 (91) |
|
Work phone | 3 (13) |
|
Personal camera | 1 (4) |
|
Work camera | 0 (0) |
|
Other | 0 (0) |
|
||
|
Yes | 14 (61) |
|
No | 7 (30) |
|
Sometimes | 2 (9) |
|
||
|
No | 19 (83) |
|
Yes | 4 (17) |
This pilot study describes the use of WhatsApp, a popular text messaging app, as an informal teledermatology platform for consultation and education in Botswana and demonstrates that WhatsApp provides a rapid and well-received method of communication between dermatologists and other health care providers.
In our study, the most common use of WhatsApp by health care providers was to consult a dermatologist about a specific patient. Our results indicate that WhatsApp facilitates rapid discussion of dermatology cases, as dermatologists responded within 1 hour for the majority of consults. In addition, group messaging was utilized to provide a platform for simultaneous communication among a team of care providers to facilitate multidisciplinary care coordination, which has been a particular challenge in Botswana [
The sustainability of teledermatology platforms has been a historical challenge in resource-limited countries [
One potential drawback of informal platforms such as WhatsApp is the lack of standardized consultation format, which allows for free-text submission of consults that may be incomplete or contain an inadequate amount of information. The ability to have real-time conversational exchanges can help overcome the lack of structured consults, though this can cause inefficiency. In this study, dermatologists asked clarifying questions in nearly 40% of the consultations. Regarding photo quality, only 20% of consults included photos that were considered low quality in our subjective assessment, primarily due to user errors such as poor lighting or blurriness rather than low resolution. Notably, the subjective rating of photographs did not always correlate with file size, suggesting that high quality photos could be obtained with low-tech mobile cameras. Future studies using validated methods to assess photo quality are needed to further explore this issue.
In this study, dermatologists were able to make a wide variety of skin diagnoses in over half of the consultations, indicating that the history and quality of photos in WhatsApp consultations could adequately support remote evaluation and diagnosis. Knowing whether telemedicine diagnoses are accurate is essential when considering the utility of providing or upscaling such services; however, we were unable to assess diagnostic accuracy in this small pilot study. Some studies have indicated that diagnoses made by teledermatology can be reliable and accurate [
WhatsApp facilitated remote management in over half of the consultations in our study, reducing the need for an in-person consultation and potentially reducing the travel and cost burdens to patients and the health care system. Patients and providers were distributed widely across Botswana, and WhatsApp was able to successfully connect patients and providers across large distances, reaching urban and rural areas. Prior research has also shown that teledermatology can help decrease unnecessary health care spending and improve allocation of resources by reducing unnecessary referrals and outpatient visits [
WhatsApp also has the potential to be used for provider education. In about one-third of consultations, the dermatologist provided education to complement management recommendations. Education is particularly valuable in resource-limited settings, where providers often lack access to clinical educational resources to assist in point-of-care decisions [
Our survey results showed that WhatsApp is a well-received and valuable resource for nondermatology providers. All but 1 respondent were satisfied with WhatsApp as a teledermatology platform, and many reported that it improved the quality of care they delivered. Respondents liked the familiarity of WhatsApp, which is consistent with WhatsApp being the predominant form of mobile communication in Botswana [
When considering telemedicine, the privacy and security of shared patient information is extremely important. In teledermatology, many consultations include protected health information and photos of patients’ faces or sensitive body areas [
Our study has several limitations. Data collection was a manual process with only 1 author reading each conversation thread, increasing the risk for errors and subjectivity, particularly in terms of grading photos. Due to the retrospective nature of the study, the heterogeneity of information provided, and the nature of downloading WhatsApp messages, we were unable to accurately calculate the number of users and all patient demographics. In addition, the number of patients electronically visited was not able to be assessed given the lack of a medical record number or chart linked to each informal teledermatology consult. As previously discussed, this study did not measure accuracy of diagnoses made via WhatsApp, which would be required to measure the overall effectiveness of the platform for teledermatology. The survey was a subjective measurement of the perceived value of teledermatology, not based on a previously validated or reliable survey instrument. A limited number of questions were used to avoid participant burden and survey fatigue. Due to low response rate, survey results may not represent the opinions of all providers using WhatsApp for teledermatology. Reasons for the low response rate are unknown, but they may include the distribution of surveys by cellular messaging, the lack of incentive for participating, and that some providers messaged were no longer participating in WhatsApp teledermatology. Additionally, study findings may not be generalizable to other resource-limited settings due to various regional differences. Despite these limitations, this pilot study serves as an important baseline to inform future investigations of WhatsApp to include diagnostic accuracy, patient acceptability, health outcomes, and the development of standardized guidelines for provider exchange.
Access to dermatology expertise remains a critically limited resource in Botswana. This study shows that there has been consistent and well-received use of WhatsApp for teledermatology in Botswana without dedicated funding, training, or equipment. The platform demonstrates a potential to support a variety of clinical purposes, such as patient consultations, triage and referral, multidisciplinary care coordination and point-of-care education. High satisfaction levels and an improvement in the ability to diagnose and manage a range of dermatologic conditions were evidenced by WhatsApp user feedback. Drawbacks identified include a lack of structured consultation format, potential security risks for patient information, and the inability to integrate consult information into a patient’s record. Despite these drawbacks, convenient, informal teledermatology platforms such as WhatsApp show promise in overcoming the logistical and sustainability challenges that have hampered teledermatology efforts in resource-limited settings. Further studies are needed to assess the effectiveness of WhatsApp and evaluate patient acceptability. The information gained from this study can serve as a baseline for future telemedicine studies and to inform the design and implementation of teledermatology in Botswana and other resource-limited countries.
Summary of data extraction and categorization methods used for analysis of WhatsApp communication threads.
Provider Satisfaction Survey.
Top ten most common conditions diagnosed by dermatologists via WhatsApp. Evaluating dermatologists made a total of 704 diagnoses and differential diagnoses which included 224 unique conditions. The numerical value and percentages in the table represent a portion of the 704 diagnoses and differential diagnoses.
Demographics of survey respondents.
Botswana-UPenn Partnership
history of present illness
institutional review board
mobile health
Princess Marina Hospital
research electronic data capture
The authors are grateful for the support of the American Academy of Dermatology Resident International Grant and the Kramer Family Development Fund, which have supported the development of dermatology care in Botswana for over 10 years.
None declared.