Article Text
PDF
Abstract
Aims Development of non-invasive and minimally invasive glucose monitoring devices (NI-MI-GMDs) generally takes place in high-income countries (HICs), with HIC’s attributes guiding product characteristics. However, people living with diabetes (PLWD) in low-income and middle-income countries (LMICs) encounter different challenges to those in HICs. This study aimed to define requirements for NI-MI-GMDs in LMICs to inform a target product profile to guide development and selection of suitable devices.
Methods This was a multiple-methods, exploratory, qualitative study conducted in Kyrgyzstan, Mali, Peru and Tanzania. Interviews and group discussions/activities were conducted with healthcare workers (HCWs), adults living with type 1 (PLWD1) or type 2 diabetes (PLWD2), adolescents living with diabetes and caregivers.
Results Among 383 informants (90 HCW, 100 PLWD1, 92 PLWD2, 24 adolescents, 77 caregivers), a range of differing user requirements were reported, including preferences for area of glucose measurement, device attachment, data display, alert type and temperature sensitivity. Willingness to pay varied across countries; common requirements included ease of use, a range of guiding functions, the possibility to attach to a body part of choice and a cost lower than or equal to current glucose self-monitoring.
Conclusions Ease-of-use and affordability were consistently prioritised, with broad functionality required for alarms, measurements and attachment possibilities. Perspectives of PLWD are crucial in developing a target product profile to inform characteristics of NI-MI-GMDs in LMICs. Stakeholders must consider these requirements to guide development and selection of NI-MI-GMDs at country level, so that devices are fit for purpose and encourage frequent glucose monitoring among PLWD in these settings.
- general diabetes
- health equity
- health services accessibility
- patient-centered care
- qualitative research
Data availability statement
Data are available on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
The recruitment strategies employed were successful in engaging a diverse sample that included people living with diabetes, both those with experience of non-invasive and minimally invasive glucose monitoring devices and those without, as well as healthcare workers and caregivers.
Multiple data collection techniques were used, which facilitated triangulation of insights across individual, couple and group data collection encounters, and data collection procedures were administered in line with a common protocol across all study sites, supporting comparability of between-site data.
Study informants generally lived in urban areas and spoke the dominant language in these areas, leading to under-representation of socially marginalised people.
As the research took place in each participating country simultaneously, it was not possible to carry out iterative data collection and analysis. Hence, some emerging topics of interest for the study informants could have remained unexplored in-depth.
Observer bias potentially contributed to leading questions and remarks during some data collection encounters.
Introduction
Most development of medical devices for the diagnosis and management of non-communicable diseases takes place in high-income countries, with the specific attributes of high-income markets guiding product characteristics.1 2 Consequently, the suitability of devices for use in low-income and middle-income countries (LMICs) is often overlooked.2 Different environmental and operational conditions, a lack of trained healthcare professionals, limited access to and availability of healthcare facilities, and a shortage of funding in LMICs bring unique challenges in the design, distribution and effective use of medical devices.1–4 Acknowledging and overcoming these challenges is necessary to ensure availability, affordability and usability of medical devices among people with non-communicable diseases in LMICs in order to improve disease diagnosis and management and reduce the occurrence of non-communicable disease attributable severe morbidity.2 5
For people living with type 1 diabetes (PLWD1), frequent measurement of blood glucose levels is essential in tailoring insulin dosage and adjusting behaviour to avoid hypoglycaemia and achieve blood glucose control in order to prevent long-term complications.6 7 Self-monitoring of blood glucose is also important for some people living with type 2 diabetes (PLWD2), mainly those on insulin.7 For all PLWD needing tools to monitor their blood glucose, the lack of access to affordable equipment in LMICs gives rise to low levels of monitoring, contributing to poor glycaemic control and, as a result, poor disease outcomes.3 7–9
Non-invasive and minimally invasive glucose monitoring devices (NI-MI-GMDs) developed in recent years10 11 measure glucose sporadically, frequently or continuously and have the potential to improve glucose monitoring by overcoming some of the barriers that PLWD experience with conventional monitoring tools.12–14 However, poor access to diabetes care and a need for more initiatives to empower and educate PLWD in LMICs lead to a lack of disease awareness, poor health seeking behaviour and low levels of treatment adherence.15 As context-adapted devices may improve adherence and promote effective self-management, there is a need to define minimal and optimal LMIC user requirements for NI-MI-GMDs. The outcomes of this study will be used to inform a target product profile for NI-MI-GMDs, to guide further product development and support selection of the most suitable devices at country level.
Informants, materials and methods
Study design
This was a qualitative, narrative research study conducted in the urban cities of Bishkek (Kyrgyzstan), Bamako (Mali), Lima (Peru) and Dar es Salaam (Tanzania) between February and July 2022. Principal investigators within each country led protocol development, pilot of data collection instruments and adaptation to local needs and ethics approval processes, and coordinated teams of field researchers (trained in culturally sensitive qualitative data collection) to conduct semistructured interviews, couple interviews, focus group discussions (FGDs) and participatory action research (PAR) sessions. The questions in the interview/discussion guides were designed to align with development of the target product profile, developed by 12 international experts. All interviews, FGDs and PAR sessions were conducted in Q2 2022.
Field researchers and informant identification
Informants were identified from a variety of sources (eg, training attendance lists, organisation reports, websites of local associations of PLWD, healthcare worker (HCW) councils). Study recruiters from partner organisations approached potential informants in person or by telephone. Study recruiters were known to some informants where purposive sampling was used, and not known where snowball techniques were introduced. In all cases, a full explanation of the study aims and procedures was provided by field researchers to enable potential informants to make an informed choice regarding study participation. All countries recruited and trained a team of interviewers and PAR session leaders of all genders who were not known to the informants.
Informants
Informants included five target groups: HCW, PLWD1 (adults), PLWD2 (adults), adolescents living with type 1 diabetes (ALWD1) and caregivers. Eligible HCW were nurses or physicians working in a diabetes clinic or healthcare institution providing diabetes care. Eligible PLWD1 and PLWD2 were ≥18 years of age, had been diagnosed with type 1/2 diabetes for ≥2 years and remained in diabetes care. PLWD2 were on insulin at the time of recruitment. Eligible ALWD1 were aged 14–17 years at recruitment irrespective of when they received their diabetes diagnosis, were being cared for by a caregiver approached as a potential study informant and remained in diabetes care. Eligible caregivers were legal guardians, parents or grandparents of minors between 1 and 13 years of age (for FGD informants) or 14–17 years (for couple interview informants) who remained in diabetes care. PLWD1, PLWD2, ALWD1 and caregivers also had to have experience of using glucose metres. Additional inclusion criteria for all groups were willingness to provide written informed consent and fluency with local vernaculars or the countries’ dominant languages.
Individual and couple interviews
Individual interviews were conducted with HCW, PLWD1, PLWD2 and caregivers, and couple interviews were conducted with ALWD1 and caregivers.
Across all participating countries, trained interviewers posed questions from a semistructured interview guide (see online supplemental materials), consisting of open-ended questions grouped in themes of interest to inform the target product profile. Themes included diabetes, control of diabetes, the role of technology, glucose monitoring devices, glucose monitoring device requirements, societal, clinical and technological factors and cost.
Supplemental material
Group discussion sessions
FGDs and PAR sessions were conducted with HCW, PLWD1, PLWD2 and caregivers. For FGDs, trained moderators, assisted by note-takers, used a thematic guide which included the same themes explored during the individual/couple interviews but did not include a fixed set of questions per theme. Each FGD (ie, up to two FGD per study population per country) counted on the participation of 5–7 informants.
For PAR sessions, informants remained in their FGD group.
The PAR sessions included a three-part ‘piling up’ exercise, informed by the handbook ‘Diagnóstico Rural Participativo’.16 During the exercise, informants identified their preferences regarding the dimensions of design, the characteristics and requirements per characteristic that should be prioritised during NI-MI-GMD development. In part 1 of the ‘piling up’ exercise, informants discussed and ordered five high-level dimensions (figure 1). In part 2, 16 further characteristics (subcategories of the 5 high-level dimensions) were discussed, and 10 top priorities were chosen (figure 1). In part 3, requirements within the 10 prioritised characteristics were provided and informants selected their optimal and minimum priority in each characteristic category. As requirements may differ depending on self-monitoring frequency, two different use cases were considered: frequent testing defined as measuring glucose at least twice per day, every day, and sporadic testing defined as measuring glucose once per day or less.
Five high-level dimensions and 16 subcategories discussed during the ‘piling up’ exercise in the PAR sessions. aBefore any hardware needs to be replaced. PAR, participatory action research.
Data analysis
In pursuit of data saturation, it was proposed that in each country at least 42 semistructured interviews (n=42 informants), 6 couple interviews (n=12), 7 FGDs and 1 PAR session (n=35–49) be conducted, with at least 89 informants providing sufficient ‘information power’.17
Data analysis was conducted by field researchers, including one main analyst and several country coanalysts. Computer-assisted qualitative data analysis software Quirkos was used to code transcripts and to prepare analytical memos for reflexive purposes. A reflexive journal was also kept by all analysts. A matrix created in Microsoft Excel was used to extract and tally key findings of utility for the intended target product profile. Content analysis was carried out on the encounters data using labels that mirrored the structured dimensions>characteristics>requirements of the NI-MI-GMD target product profile. Content analysis was deductive, although inductive coding was also allowed, with any new concepts emerging from the data incorporated into the analytical matrix.
Patient and public involvement
A target product profile expert group which included PLWD and diabetes caregivers was instrumental in the conceptualisation and delivery of this study, with some members also coauthoring this manuscript. PLWD were involved in study recruitment through our partner organisations in the four participating countries, which are either patient organisations themselves or work extensively with patient organisations. In addition to publication of the results, partner organisations have reached out to study participants to disseminate the study results.
Results
Informant characteristics
Overall, there were 221 encounters with a total of 383 informants (251 women): 168 semistructured interviews, 24 couple interviews and 29 FGDs. PAR sessions were conducted with informants from the 29 FGDs: 8 with HCWs, 8 with PLWD1, 8 with PLWD2 and 5 with caregivers.
Characteristics of the participating countries are provided in online supplemental table 1 and informant characteristics are shown in table 1. Informants included 90 HCW, 100 PLWD1, 92 PLWD2, 77 caregivers and 24 ALWD1. The majority of PLWD1, PLWD2 and ALWD1 were receiving human insulin (120/216) or analogue insulin (48/216). There was a range of education profiles, with ~50% of informants having no, primary level or secondary level education.
Informant characteristics
Unless otherwise specified, results presented below were similar across the four study sites.
NI-MI-GMD characteristic priorities
During part 1 of the ‘piling up’ exercise, 12/29 PAR groups chose ‘performance and results’ as their number one priority, while 10 chose ‘purchasing considerations’ (table 2).
Prioritisation of dimensions and characteristics for NI-MI-GMD development (findings from the PAR sessions)
During part 2 of the exercise, 27 of the 29 groups chose ‘price and business model’ as a top ten priority and 26 chose ‘alerts/user guidance’. ‘Measurement accuracy’, ‘physical area of measurement’, ‘sample type’, ‘duration of use’, ‘measurement cycle’ and ‘data display’ were each chosen as a top ten priority by 22 groups, while ‘power source’ (14 groups), ‘attachment to the body’ (11 groups), ‘operating temperature’ (11 groups), ‘calibration’ (10 groups) and ‘infrastructure level’ (9 groups) were chosen by fewer than half of the groups.
Outcomes from the third and final part of the ’piling up’ exercise are presented alongside the interview and FGD outcomes in the following sections.
Scope
Intended use
In most of the interviewing encounters (81.4%), informants reported that NI-MI-GMDs would be used to stop finger pricking (table 3). Caregivers and HCW were particularly concerned about the effect of finger pricking in children. Additionally, it was perceived that the real-time glucose measurements taken by NI-MI-GMDs would be important for anticipating hypoglycaemia or hyperglycaemia (50.7%), as the device would help to act on an acute emergency (49.3%). ‘Checking glucose range’ (38.9%) and ‘regulating insulin or oral medication’ (37.6%) were discussed as reasons to use NI-MI-GMDs, as was helping to regulate diet (20.8%).
Perceptions regarding intended benefits and expected functionalities of NI-MI-GMDs and purchasing considerations (findings from the individual interviews, couple interviews and FGDs)
Target population
Expected target users and operators of NI-MI-GMDs were children as users (supported by informants in 60.6% of encounters), PLWD1 (45.2%) and PLWD2 (44.3%). While HCW and PLWD1 were unsure how useful NI-MI-GMD would be for PLWD2, most PLWD2 (all insulin users) self-identified as potential users with reduced risk of long-term complications stated as a potential benefit. This was of particular importance to older informants from Kyrgyzstan and Peru. In Kyrgyzstan and Tanzania, informants believed the NI-MI-GMDs could be operated by children independently allowing confidence that glucose levels are controlled when away from caregivers.
Infrastructure level
Among 9/29 PAR groups who chose ‘infrastructure level’ as a priority, access to electricity, water and the internet were noted as elements that, as a minimum, would be required for frequent or sporadic NI-MI-GMDs use.
Device and utility requirements
NI-MI-GMDs device and utility requirements captured during individual, couple and group interviews (n=221) are shown in figure 2 and described below. Overall, findings from the PAR sessions were consistent with those from the interviewing encounters.
Opinions on device and utility requirements captured during semistructured interviews, couple interviews and FGDs. n numbers and percentages represent the number and percentage of encounters (semistructured interviews, couple interviews or FGDs) in which each requirement category was supported. aOther categories were abdomen (11.3%), legs/thighs/ankle (8.6%), eyes (7.7%), back (4.5%), shoulder (4.5%), neck (3.2%), breast/chest (3.2%), toes (1.4%), armpit (1.4%) and brain (0.5%). bOther categories were must be small, flat, easy to wear and hide (13.1%), like a strip that is placed in the tongue or eyes (9.5%), like an ear clip or earring (9.0%), like a sensor applied with a microneedle (7.2%), like a ring for finger and toes (2.3%), like a necklace or bangles (1.8%), like contact lenses for the eyes (1.4%), like a chip that is introduced in the bloodstream (0.9%), like a hearing device or earphones (0.5%), like a thermometer for the armpit (0.5%), like an anklet (0.5%). cOther categories were every 1–59 min (5.0%) and throughout the night (2.3%). dOther categories were on a smartwatch (4.1%), on other devices (eg, lamp, Alexa, Miao Miao) (1.4%) and on a conventional glucose metre (0.5%). eOther categories were measurements need be traceable by means of allowing the users to run queries in the system (4.5%), measurements could be stored, processed and displayed in social media/messaging applications (0.9%), measurements should not be modifiable once stored (0.5%), measurements should be sent to ‘global diabetes monitoring system’ (0.5%), measurements should be processed in the language of choice of the user (0.5%), measurements should be shared with the public/private health insurance (0.5%), measurements should be processed with a time and geo stamp (0.5%). fOther categories were voice warning the user about the situation (6.3%), gives the user an electric shock (0.9%) and brings coldness to the user (0.5%). FGD, focus group discussion; NI-MI-GMDs, non-invasive and minimally invasive glucose monitoring devices.
Physical area of measurement
In all countries, a preference to measure glucose levels on the wrist/hand (70.1%) or arm (43.4%) was expressed. There was some discrepancy between HCW, who were concerned about stigma and discrimination for PLWD, and PLWD1 and PLWD2, who tended to prefer visible and comfortable areas to avoid harming, misplacing or forgetting the device.
While preferences for body parts other than the wrist/hand or arm were also mentioned in the interview encounters, these were highly dependent on cultural positions and environmental circumstances, that is, some informants expressed that for men, particularly in Kyrgyzstan and Mali, wearing an ear-clip would not be acceptable as it is perceived as women’s jewellery. Some informants in Kyrgyzstan also mentioned wearing a hat and scarf during winter as incompatible with using an ear-clip.
Application/attachment to the body and degree of invasiveness
Informants highlighted the importance of comfort, therefore, methods of attachment to the body such as a band, belt, bracelet or cloth or a watch wrapped around the wrist were generally favoured. Complaints about how unstable minimally invasive sensors were on the body were common. Some thought the adhesives may lead to allergic reactions or itching, while some informants in Kyrgyzstan, where swimming is a common pastime, thought the devices would fall off in water. Non-invasive devices were considered as potentially preferable over minimally invasive devices, particularly for people with diabetes who practise sports such as swimming and for those with a preference for stylish garments. However, some informants in Kyrgyzstan expressed a preference for minimally invasive sensors as they believed these to be more accurate based on their knowledge of the Freestyle Libre glucose monitoring system.
Measurement cycle
Informants in the PLWD1, PLWD2 and caregiver groups divulged that glucose was not being measured frequently due to the cost associated with glucose metres, lancets and strips. The view that the NI-MI-GMD itself should be sufficient without additional consumables was commonly shared. A belief that glucose would be measured more frequently if NI-MI-GMDs were available and affordable was commonly expressed. Device portability was also reported to be likely to facilitate frequent glucose measurement.
Duration of use and interchangeability
There was little feedback on the preferred duration of device use. An NI-MI-GMD resistant to water and extreme weathers and impact was reported to be important in all countries.
Narratives on interchangeability were very scarce.
Data display and management
A preference for access to glucose measurements through mobile phones (69.2% of encounters) or in the NI-MI-GMD (61.5%) itself was expressed, and accessing measurements through downloaded smartphone applications was also favoured in some encounters (22.2%). In Kyrgyzstan, where informants had experience of Freestyle Libre, use of a reader, in addition to the sensor or to the phone, to read measurements was supported in over half of encounters. However, a common narrative was that less equipment was favourable, with some noting that mobile phone use would complicate glucose monitoring, particularly for elderly people.
Methods of data transmission from the NI-MI-GMD to other equipment were not commonly discussed during interviewing encounters; however, PAR groups highlighted the need for wireless data transfer. In most encounters, informants thought that the NI-MI-GMD should have enough internal memory to store several weeks of measurements data (81.4%) and the functionality to transmit these data to relatives, friends or healthcare providers (74.3%).
Power source
Power source was not commonly discussed. While operating and storage temperature was not a focus, informants in Kyrgyzstan suggested that devices should work during harsh winters and in humid environments such as saunas, while those in Tanzania and Mali mentioned the need to withstand extreme heat. In the 11 PAR groups that chose temperature as a priority, informants noted that, as a minimum and optimally, NI-MI-GMD should be able to operate in environments with temperatures of <0°C to 45°C and relative humidity of 10%–90%.
Alerts
Informants in most encounters were interested in having an NI-MI-GMD that could alert them to a drop or increase in glucose levels outside normal glycaemic ranges (62.9% and 60.6%, respectively), with a myriad of preferences for the alert format.
Performance and results
According to opinions voiced during the PAR groups, to avoid mistakes in the adjustment of insulin, an NI-MI-GMD should display, optimally, the precise measurements and, as a minimum, range indications.
There were few suggestions regarding NI-MI-GMD calibration methods. Trust in the NI-MI-GMD was generally high, however, some informants in Kyrgyzstan noted that reliability was difficult to assess when devices were imported from abroad while in Peru and Tanzania, some informants expressed total trust in the reliability of available devices. Some informants said that they would use glucose metres initially or periodically to compare measurements between the glucose metre and the NI-MI-GMD. Among 10/29 PAR groups that chose calibration as a priority, a device needing no calibration was thought to be optimal.
Purchasing considerations
There was no trend regarding how much informants would be willing or able to pay for the NI-MI-GMD (table 3). Informants expressed that equipment required for glucose monitoring, even with conventional methods, is perceived to be expensive. The proportion of encounters in which a price of more than US$100 was proposed was higher in Kyrgyzstan (52.7%) than in any other country (0%–29.1%), although the most favoured purchasing characteristic in Kyrgyzstan was a one-off payment (16.4%).
Importance was placed on fully or partially subsidised devices for financially vulnerable households, with the options of a single payment without subsequent expenditure (8.6%), or with instalments if the cost was high (1.4%) being suggested as convenient for patients and caregivers. There was also concern that local vendors may have freedom to increase prices if the introduction of NI-MI-GMD in the country were not regulated, reducing affordability. Purchasing characteristics was chosen as a priority by almost all PAR groups. As a minimum, the groups thought the NI-MI-GMD should cost the same as their current glucose monitoring device, and, optimally, they should cost less.
Requirements for frequent versus sporadic use are shown in online supplemental materials.
Adolescent voices
Adolescent voices were generally strong during the couple interviews, with adolescents explaining how they self-monitor their glucose levels and what they would like to see in an NI-MI-GMD. For example, direct linkage of glucose monitoring data to healthcare professionals was deemed to be important. The keen participation of adolescents highlights the success of efforts at diabetes care sites in empowering adolescents to manage their own condition.
Summary of important NI-MI-GMD characteristics for LMIC users
Overall, user requirements varied across countries. However, several common themes emerged that can be considered important for NI-MI-GMDs intended for use in LMICs. These included ease-of-use and the possibility to attach to a body part of choice. Additionally, devices should be robust enough to withstand different environmental conditions including long periods of extreme temperatures or high humidity. Notably, though, the ideal device characteristics differ substantially across countries depending on culture and context.
Discussion
This study highlighted a range of differing user requirements for NI-MI-GMDs in LMICs. Preferences for glucose measurement from particular body parts and attachment methods differed across countries based on cultural perceptions and device position convenience with regard to common daily activities or clothing. Moreover, differing data display, alert type and temperature sensitivity requirements were reported. Differences across countries and informant groups are likely driven by varying socioeconomic status, clinical needs, health system access and levels of diabetes knowledge.3 Indeed, knowledge of some of the currently available NI-MI-GMDs differed across countries, with informants from Kyrgyzstan and Peru providing more detailed user requirement narratives due to greater knowledge of devices accessible in their contexts than their counterparts in Mali and Tanzania. This difference highlights the challenge of designing NI-MI-GMDs for use in LMICs, which vary according to current level of NI-MI-GMD use as well as readiness to adopt or expand NI-MI-GMD use due to differences in infrastructure, technology and expertise.3
Device requirements highlighted by informants in this study reflect the additional challenges in diabetes management faced by PLWD in LMICs.3 Devices designed to give users optimal information and guidance on their diabetes management, as well as independence from specific infrastructure requirements, robustness and low cost contribute towards lowering the barriers to achieving glycaemic control through frequent self-monitoring. Willingness to pay is an important consideration in this regard, and this was among the most challenging concepts to discuss during the encounters. Willingness to pay varied across countries, potentially impacted by non-homogeneous questioning across sites and likely impacted by differences in individual countries’ healthcare systems. For example, free insulin provision, as in Kyrgyzstan, is likely to influence willingness to pay for devices. Additionally, some informants in Mali and Tanzania reported that buying basic necessities such as food must be prioritised over investing in new technology such as NI-MI-GMDs. Nonetheless, based on the priorities established during the PAR, it can be recommended that device cost should be lower than or equal to the current cost of glucose self-monitoring. This recommendation is consistent with several studies that have found cost to be a critical barrier to adequate diabetes care in LMICs.3 18 Costing studies or socioeconomic evaluations to determine how manufacturers could be encouraged to reduce costs in order to improve accessibility are warranted.
To the best of our knowledge, this is the first study assessing LMIC user requirements for NI-MI-GMDs in enough detail to allow developers to take these requirements into consideration when designing their devices. The findings will also allow informed decisions regarding purchase of existing devices in individual healthcare systems, which may become relevant in the future, as increasing numbers of developers look towards LMICs for market expansion.19 Some of the stated requirements, including reduced need for finger pricks, low chance of skin irritation and personalised alerts and alarms, are consistent with those identified in several qualitative studies of PLWD in Europe and the USA,20–22 suggesting that some of the characteristics likely already considered during development of NI-MI-GMDs for use in high-income countries will also be of use for users in low-resource settings. However, the informants also highlighted several important characteristics that differ from those that are important in high-income countries. For example, the possibility for a device to attach to a body part of choice would be useful in accommodating the varying cultural practices across different LMICs. In addition, devices suitable for use in LMICs will likely need to be more robust than those used in high-income countries in terms of their sensitivity to environmental conditions such as extreme temperatures and high humidity. Given that these devices are intended for use in contexts where there is often no universal health coverage, limited HCWs with diabetes knowledge and no laboratories or emergency hospitals nearby, they should at least be intended for self-management of hypoglycaemia and hyperglycaemia. Notably, devices for use in LMICs will also need to have characteristics aligned with the REASSURED criteria, which describe the attributes of an ideal diagnostic test to meet the needs of people in resource-constrained settings (real-time connectivity, ease of specimen collection, affordable, sensitive, specific, user-friendly, rapid and robust, equipment-free, deliverable to end users).23 Many of the needs identified in the current study are consistent with the REASSURED criteria, providing real-world assurance that the criteria accurately define tests that are appropriate to meet the needs of those in LMICs. As currently available glucose self-monitoring devices do not fulfil all of the REASSURED criteria, and some fulfil only two or three of the nine,24 the need for further product development is clear. Acknowledging and applying the input of PLWD during product development allows the product to be tailored to users and settings accordingly, and empowers PLWD to be involved in the development process. This is particularly relevant for NI-MI-GMD as devices that are well suited to end-users are most likely to be properly and regularly used, which in the context of PLWD is likely to improve glycaemic control and associated disease outcomes. As such, our study findings can be used alongside the REASSURED criteria to guide development of NI-MI-GMD targeted for LMIC users.
While it is not standard practice to include structured user feedback in a target product profile, such an approach has been recently used in other disease areas,25 26 helping to ensure that suggested device characteristics are relevant to users based on their lived experience. While tailored NI-MI-GMDs are paramount to pave the way to improved diabetes self-care in LMICs, technological innovations can only overcome some of the barriers to glucose self-monitoring and any new device must always be introduced in conjunction with diabetes management education and training on integration into clinical practice. Given the access gaps for NI-MI-GMDs, insulin and diabetes care in LMICs, novel approaches involving end-users in the design of solutions is needed and this study provides an example of how such an approach can be developed.
This qualitative study has several strengths. The recruitment strategies employed were successful in engaging a diverse sample that included both NI-MI-GMD-naïve and NI-MI-GMD-experienced PLWD as well as HCW and caregivers. Multiple data collection techniques were used, which facilitated triangulation of insights across individual, couple and group data collection encounters. Moreover, data collection procedures were administered in line with a common protocol across all study sites, supporting comparability of between-site data. We also believe that saturation was reached as the characteristics of explored themes were thought to be fully understood, meaning there was no need to recruit informants beyond the predefined sample size.
Some limitations must also be noted. Study informants generally lived in urban areas and spoke the dominant language in these areas, leading to under-representation of socially marginalised people. Furthermore, all informants had experience of using a glucose metre and remained in diabetes care, as per the study inclusion criteria. Therefore, PLWD who may not fully understand the relevance of glucose monitoring who may not be accessing specialist diabetes care are under-represented. Further studies to capture the views of these individuals would provide a broader view of user requirements. Additionally, as the research took place in each country simultaneously, iterative data collection and analysis could not be carried out. Hence, some emerging topics of interest for the study informants could have remained unexplored in-depth. Importantly, knowledge and past experience of NI-MI-GMD were identified as motivators for discussion of the minimum and optimal requirements for users in LMICs and these were more frequently reported by PLWD1 and caregivers than other groups, with HCW reporting that they lacked such knowledge. This suggests a need for further evaluations targeting HCWs to gain an understanding of how well equipped they are to support their patients. Observer bias potentially contributed to leading questions and remarks during some data collection encounters. Finally, further socioeconomic evaluations and market analysis may be useful in determining precisely how resource-constrained populations would access, use and benefit from NI-MI-GMD designed with LMIC users in mind.
Conclusions
This multicountry, qualitative study highlights the importance of the perspectives of PLWD in development of a target product profile to inform the design of NI-MI-GMDs for use in LMICs. Informants across all four LMICs prioritised NI-MI-GMD ease-of-use and affordability, while other priorities varied across countries demonstrating a need for broad functionality to include a range of alarms, measurement graphs and attachment possibilities as well as robustness in the face of a range of environmental conditions. These requirements must be considered by all NI-MI-GMD stakeholders in order to bridge the gap between high-income countries and LMICs with regard to access to adequate diabetes care.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and the study protocol was approved by the Institutional Review Board from the Universidad Peruana Cayetano Heredia (SIDISI 207529, Peru), the Comité d’éthique de l’université des sciences, des techniques et des technologies de Bamako (No 2022/36/CE/USTTB, Mali), the Ethical Committee at Muhimbili University of Health and Allied Sciences (MUHAS-REC-05-2022-1153, Tanzania) and the Ethics Committee at the Research and Production Association 'Preventive Medicine' of the Kyrgyz Republic Ministry of Health (extract from the protocol No9, Kyrgyzstan). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
The authors would like to thank all interview and discussion group informants. We also thank all field researchers from the participating sites for their invaluable roles in this study. Thank you also to Miseli who collected data in Mali. Medical writing assistance was provided by Claire Gibbons, PhD, funded by FIND, in accordance with Good Publication Practice guidelines.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors Conceptualisation: BV, ES, DB, ML and GZM-P. Data curation: AA, AD, JPM, SP-L, CM, ES and GZM-P. Formal analysis: MMand GZM-P. Funding acquisition: BV, DB and ML. Investigation: BV, DB, SB, ML-P, AA and KR. Methodology: BV, DB, GZM-P. Project administration: ES. Resources: GZM-P. Software: GZM-P. Supervision: ES, AD, KR, SB, ML-P and HN. Validation: GZM-P. Writing–original draft: ES and BV. Writing–review and editing: ES, DB, BV, ML, AA, AD, SB, ML-P, JPM, SP-L, MM, KR, CM and GZM-P. Guarantor: ES.
Funding German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung) through the Kreditanstalt für Wiederaufbau (KfW); grant number: NA. The Leona M. and Harry B. Helmsley Charitable Trust; grant number: NA.
Competing interests None declared.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.