Impact statements
Mobile phones are increasingly proposed as a means of delivering mental health and psychosocial support in humanitarian settings, but expressed interest in a mobile-delivered intervention does not necessarily translate into actual access, uptake, or sustained use. In this study, most internally displaced participants in two Abuja camps said they would be interested in receiving education on coping with life challenges via mobile phone. Interest varied by education, social media use, and camp setting. Despite high expressed interest, participants reported limited social media use and frequent technological barriers. These findings raise concerns about digital equity. For people with limited literacy, limited familiarity with digital platforms, or limited access to mobile phones, mobile-delivered interventions may be acceptable in principle, but barriers to access and use may persist. Humanitarian organizations should therefore not interpret high stated interest as evidence of implementation readiness or assume that a mobile-based intervention would be equally accessible and usable across population groups. Mobile-delivered interventions ought to incorporate low-literacy and low-connectivity options, as well as in-person or blended approaches, and evaluate these strategies prospectively. Although the study did not assess observed uptake, sustained engagement, or intervention effectiveness, future studies should determine whether expressed interest translates into enrollment, participation, and continued use.Mobile phones are increasingly proposed as a means of delivering mental health and psychosocial support in humanitarian settings, but expressed interest in a mobile-delivered intervention does not necessarily translate into actual access, uptake, or sustained use. In this study, most internally displaced participants in two Abuja camps said they would be interested in receiving education on coping with life challenges via mobile phone. Interest varied by education, social media use, and camp setting. Despite high expressed interest, participants reported limited social media use and frequent technological barriers. These findings raise concerns about digital equity. For people with limited literacy, limited familiarity with digital platforms, or limited access to mobile phones, mobile-delivered interventions may be acceptable in principle, but barriers to access and use may persist. Humanitarian organizations should therefore not interpret high stated interest as evidence of implementation readiness or assume that a mobile-based intervention would be equally accessible and usable across population groups. Mobile-delivered interventions ought to incorporate low-literacy and low-connectivity options, as well as in-person or blended approaches, and evaluate these strategies prospectively. Although the study did not assess observed uptake, sustained engagement, or intervention effectiveness, future studies should determine whether expressed interest translates into enrollment, participation, and continued use.
Introduction
Internally displaced persons (IDPs) are individuals who have been forced to flee their homes because of conflict, violence, disasters or other crises but remain within their country’s borders (Deng, Reference Deng1999). Conflict-related internal displacement remains a major humanitarian challenge (Lischer, Reference Lischer2007; Cuadrado et al., Reference Cuadrado, Libuy and Moreno-Serra2023; Eboreime et al., Reference Eboreime, Anjorin, Obi-Jeff, Ojo and Hertelendy2025). The Internal Displacement Monitoring Centre estimated that at least 75.3 million people were living in internal displacement due to conflict and violence worldwide in 2025 (Internal Displacement Monitoring Centre. 2025) Displacement settings can exacerbate psychological distress and contribute to common mental disorders through multiple pathways, including exposure to traumatic events, loss of livelihoods, disruption of social networks, insecurity and ongoing material deprivation (Hou et al., Reference Hou, Liu, Liang, Ho, Kim, Seong, Bonanno, Hobfoll and Hall2020; Grasser, Reference Grasser2022; León-Giraldo et al., Reference León-Giraldo, Casas, Cuervo-Sánchez, García, González-Uribe, Moreno-Serra and Bernal2023).
In Nigeria, these challenges occur alongside a myriad of predicaments in the mental health system, including limited workforce capacity and weak primary care and referral pathways (Ojagbemi and Gureje, Reference Ojagbemi and Gureje2021; Adepoju, Reference Adepoju2023; Akanni and Edozien, Reference Akanni and Edozien2024). These constraints can complicate the timely and continuous delivery of psychosocial support.
Within this context, life skills education (LSE) may be a practical component of broader mental health and psychosocial support efforts, as it can strengthen transferable social, emotional and coping skills and be delivered by non-specialist facilitators (Barry et al., Reference Barry, Clarke, Jenkins and Patel2013; Eboreime et al., Reference Eboreime, Obi-Jeff, Orji, Ojo, Iyamu, Harri, Said, Oguntimehin, Ibrahim, Anjorin, Duke, Musami, Liebenberg, Crider, Wagami, Dahiru, Uneke, Yaya and Agyapong2024). In humanitarian settings, however, one key question is whether such content is both useful and deliverable amid insecurity, population mobility, suboptimal infrastructure, and limited service capacity. Mobile delivery may offer one way to extend reach when in-person services are difficult to sustain, but its feasibility depends on factors such as device access, digital literacy, connectivity, affordability, language, privacy and trust (Burchert et al., Reference Burchert, Alkneme, Bird, Carswell, Cuijpers, Hansen, Heim, Harper Shehadeh, Sijbrandij, Van’T Hof and Knaevelsrud2019; Matlin et al., Reference Matlin, Hanefeld, Corte-Real, Da Cunha, De Gruchy, Manji, Netto, Nunes, Şanlıer, Takian, Zaman and Saso2025).
These issues are directly relevant in Nigeria. The Rebuilding Emotional Stability and Strength Through Therapeutic and Life-Skills Education for Internally Displaced Persons (RESETTLE-IDPs) trial compares in-person and mobile-facilitated delivery of an adapted LSE program for IDPs and includes implementation outcomes alongside effectiveness outcomes (Eboreime et al., Reference Eboreime, Obi-Jeff, Orji, Ojo, Iyamu, Harri, Said, Oguntimehin, Ibrahim, Anjorin, Duke, Musami, Liebenberg, Crider, Wagami, Dahiru, Uneke, Yaya and Agyapong2024). Within this broader programmatic context, baseline expressed interest was examined as a context-specific indicator of stated receptivity to mLSE.
Although this study was not designed to assess the Technology Acceptance Model (TAM) directly, TAM provided a broad conceptual basis for considering how characteristics external to the proposed mobile intervention might relate to expressed interest in its use. TAM identifies perceived usefulness and perceived ease of use as central to technology acceptance and recognizes that external characteristics may shape these perceptions and subsequent use (Davis, Reference Davis1989; Marangunić and Granić, Reference Marangunić and Granić2015). The present survey did not directly measure perceived usefulness, perceived ease of use, behavioral intention or actual use. Instead, the analysis examined demographic, contextual and technology-related characteristics relevant to the proposed delivery setting. Educational attainment was considered relevant to engagement with structured learning content, social media use reflected prior engagement with mobile digital platforms, reported barriers captured constraints on mobile phone access or use and camp location represented the two study settings. These characteristics were treated as implementation-relevant external factors rather than as measures or proxies of TAM constructs. TAM therefore served as an orienting framework for variable selection and interpretation rather than as a model to be tested, modified or extended. By examining baseline expressed interest in using a mobile format, the analysis aimed to inform the design and delivery of contextually appropriate mobile psychosocial interventions in humanitarian settings.
Methods
Study design and setting
This cross-sectional study analyzed baseline pre-intervention survey data collected between September 2024 and January 2025 as part of the RESETTLE-IDPs trial, a hybrid type II effectiveness-implementation cluster-randomized trial evaluating a contextually adapted LSE program for IDPs in Nigeria (Eboreime et al., Reference Eboreime, Obi-Jeff, Orji, Ojo, Iyamu, Harri, Said, Oguntimehin, Ibrahim, Anjorin, Duke, Musami, Liebenberg, Crider, Wagami, Dahiru, Uneke, Yaya and Agyapong2024). The study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cross-sectional studies (Vandenbroucke et al., Reference Vandenbroucke, Elm, Altman, Gøtzsche, Mulrow, Pocock, Poole, Schlesselman and Egger2007).
The study was conducted in the Durumi and Wassa IDP settlements in Abuja, Federal Capital Territory. Site identification followed a mapping exercise initiated in June 2024. Of 19 IDP camps and host-community settlements identified through the National Commission for Refugees, Migrants and Internally Displaced Persons, 15 were visited and assessed. Four were not assessed because they had closed or relocated, presented security concerns or were outside the Federal Capital Territory. Durumi and Wassa were selected because both were accessible to the research team, had no identified security constraints at the time of assessment, and contained sufficient households to support the planned recruitment. Wassa subsequently served as the site for in-person delivery , whereas Durumi served as the site for WhatsApp-based delivery.
Participant recruitment and survey administration
Following site selection, household enumeration identified 743 households in Wassa and 387 in Durumi and created site-specific household sampling frames containing information on sociodemographic characteristics and patterns of mobile phone access and use. Using computer-generated random numbers, 150 households were selected from each settlement, yielding 300 sampled households. Residents were eligible if they were aged 13 years or older; personal mobile phone ownership was not required. Within each sampled household, field staff listed all eligible adults and used computer-generated random numbers to select one adult for invitation. If no eligible adult resided in the household, eligible adolescents aged 13 to 17 years were listed and one was selected using the same procedure. Selected adolescents participated with assent and parental or guardian consent. No more than one participant was selected from each household. Details of the sampling and recruitment procedures have been reported previously (Duke et al., Reference Duke, Harri, Iyamu, Obi-Jeff, Dahiru, Ojo, Musami, Said, Agyapong, Orji, Yaya and Eboreime2025). Of the 300 sampled households, 264 resulted in an enrolled participant, corresponding to a recruitment yield of 88.0%.
Trained research assistants administered the survey in Hausa by reading each item aloud and recording participants’ responses. Participation was voluntary, and participants could decline any question or discontinue the survey without penalty. Of the 264 participants surveyed, 231 had complete data on the outcome and all variables included in the regression models and constituted the complete-case analytic sample for both the primary and fuller exploratory analyses. The remaining 33 participants were excluded because they had missing data on one or more of these variables.
Outcome and covariates
The primary outcome was expressed interest in using mLSE, assessed using the question, “Would you be interested in using your mobile phone to receive education/training on how to cope with life challenges?” Responses were coded as a binary indicator. Personal mobile phone ownership was not required. This context-specific item assessed stated interest in the proposed delivery format; it was not a validated measure of perceived usefulness, perceived ease of use, behavioral intention, technology acceptance, implementation readiness, uptake or sustained engagement. Because it referred to both coping-related education and mobile delivery, responses could reflect interest in the content, the delivery format or both.
Candidate covariates selected before regression modeling were age, gender, educational attainment, employment status, duration in the camp or host community, psychological counseling since displacement, camp location, English and Hausa language use, usual mobile phone use, social media use during the preceding 3 months and barriers to mobile phone access or use since displacement. English and Hausa language use were assessed separately because participants could report using both languages. These variables reflected participants’ broader linguistic context rather than formally assessed language proficiency and were examined only in the fuller exploratory model.
Educational attainment was categorized as no formal education, Islamic education, primary education or secondary education or higher. Psychological counseling history was assessed by asking, “Have you received any psychological counseling since you were displaced?” Social media use was classified as use of at least one listed platform during the preceding 3 months vs. no platform use.
Reported barriers were grouped into three mutually exclusive categories because some individual responses were absent or infrequent. Technological barriers comprised poor network connectivity, lack of electricity for charging, difficulty using smartphone features and privacy or safety concerns. Accessibility or affordability barriers comprised lack of a personal phone, limited access to a shared phone and inability to afford phone credit or mobile data. Participants reporting any technological barrier were assigned to that category; all others were classified as reporting accessibility or affordability barriers or no barriers.
Because this analysis used baseline data collected within the parent trial, no separate sample size calculation was conducted for the multivariable analysis.
Statistical analysis
Categorical variables were summarized using frequencies and percentages. Participant characteristics were presented descriptively by expressed interest in using mLSE.
Before regression modeling, multicollinearity was assessed using generalized variance inflation factors (Fox and Monette, Reference Fox and Monette1992). Modified Poisson regression with a log link and robust sandwich standard errors was used to estimate prevalence ratios and 95% confidence intervals. Robust Wald p-values were also reported. This approach permits direct estimation of prevalence ratios for binary outcomes rather than odds ratios, thereby facilitating interpretation when the outcome is common (Barros and Hirakata, Reference Barros and Hirakata2003; Zou, Reference Zou2004; Petersen and Deddens, Reference Petersen and Deddens2008).
Because only 15 participants reported no interest in using mLSE, the primary adjusted analysis was restricted to a parsimonious model including educational attainment, camp location, social media use and barriers to mobile phone access or use. These variables were selected on conceptual and contextual grounds rather than through bivariable significance screening: educational attainment represented educational background, camp location represented study setting, social media use represented prior engagement with mobile digital platforms and reported barriers represented constraints on mobile phone access or use.
A fuller model containing the broader set of prespecified covariates was examined as a secondary, exploratory analysis. The small number of participants reporting no interest and the presence of zero or sparse cells in some predictor categories limited the information available for multivariable estimation. Estimates from the fuller model were therefore considered potentially susceptible to sparse-data bias and instability and were interpreted as hypothesis-generating. Greater inferential weight was given to the primary parsimonious model.
Where appropriate, original response categories were combined before modeling to address sparse cell counts and improve interpretability while preserving differences meaningful in the study setting. Given the cross-sectional design, estimates were interpreted as associations rather than causal effects. Analyses were conducted using R version 4.5.2.
Results
Sample characteristics
Among the 231 participants, 216 (93.5%) expressed interest in using mLSE and 15 (6.5%) did not. Table 1 presents participant characteristics by expressed interest in using mLSE. Interest was high across most subgroups, although the proportion expressing interest was lower among participants with no formal education, those reporting technological barriers and those reporting no social media platform use (Table 1).
Participant characteristics by expressed interest in using mLSE (N = 231)
Table 1. Long description
The table presents data for a total sample of N equals 231. It is organized into five columns: Variable, Category, Total sample n (%), Interested in using m L S E n (%), and Not interested in using m L S E n (%).
* Age: The largest group is 25 to 29 years (28.1%), with 96.9% interested in m L S E. The 35 plus group has the lowest interest rate at 89.4%.
* Gender: Females make up 74.9% of the sample (93.1% interested) and males 25.1% (94.8% interested).
* Education: 35.5% have Secondary education or Higher, showing 100% interest. Those with no formal education (13.9%) show the lowest interest at 78.1%.
* Employment: 77.1% are unemployed (92.7% interested) and 22.9% are employed (96.2% interested).
* Barrier: Technological barriers are reported by 51.1% of the sample, with 89.8% of that group interested in m L S E.
* I D P camp location: Durumi I D P Camp (52.8% of sample, 91.0% interested) and Wassa I D P Camp (47.2% of sample, 96.3% interested).
* Language: 78.4% do not speak English (93.4% interested); 81.8% speak Hausa (92.6% interested).
* Mobile phone use: 62.8% use phones for calls (91.7% interested); 37.2% for miscellaneous uses (96.5% interested).
* Social media: 20.3% use one or more platforms (100% interested); 79.7% use none (91.8% interested).
* Psychological counseling: 50.6% have no history (95.7% interested); 49.4% have a history (91.2% interested).
* Duration in camp: 64.5% have been in camp for 6 plus years (93.3% interested).
Multicollinearity diagnostics
Generalized variance inflation factors did not indicate substantial collinearity among the included predictors, with adjusted values ranging from 1.089 to 1.383 (Table 2).
Multicollinearity diagnostics
Table 2. Long description
The table consists of four columns: Variable, G V I F, D f, and Adjusted G V I F. The data for each variable is as follows:
* Age: G V I F 1.975, D f 4, Adjusted G V I F 1.089.
* Gender: G V I F 1.297, D f 1, Adjusted G V I F 1.139.
* Education: G V I F 1.818, D f 3, Adjusted G V I F 1.105.
* Employment: G V I F 1.540, D f 1, Adjusted G V I F 1.241.
* Barrier: G V I F 1.755, D f 2, Adjusted G V I F 1.151.
* I D P camp location: G V I F 1.895, D f 1, Adjusted G V I F 1.377.
* English: G V I F 1.438, D f 1, Adjusted G V I F 1.199.
* Hausa: G V I F 1.282, D f 1, Adjusted G V I F 1.132.
* Mobile phone use: G V I F 1.914, D f 1, Adjusted G V I F 1.383.
* Social media platform use: G V I F 1.389, D f 1, Adjusted G V I F 1.179.
* Psychological counseling history: G V I F 1.448, D f 1, Adjusted G V I F 1.203.
* Duration in camp: G V I F 1.189, D f 1, Adjusted G V I F 1.090.
Primary adjusted analysis
Table 3 presents the primary adjusted modified robust Poisson regression results.
Primary adjusted analysis: reduced modified robust Poisson regression
Table 3. Long description
The table consists of four columns: Category, Robust S E, Adjusted P R (95 percent C I), and p-value.
* Islamic education vs. no formal education: Robust S E 0.087, Adjusted P R 1.16 (0.98, 1.38), p-value 0.079.
* Primary education vs. no formal education: Robust S E 0.094, Adjusted P R 1.18 (0.98, 1.42), p-value 0.080.
* Secondary education or higher vs. no formal education: Robust S E 0.089, Adjusted P R 1.29 (1.08, 1.54), p-value 0.004.
* Wassa I D P Camp vs. Durumi I D P Camp: Robust S E 0.042, Adjusted P R 1.12 (1.03, 1.21), p-value 0.009.
* No social media platform use vs. uses one or more platforms: Robust S E 0.025, Adjusted P R 0.92 (0.88, 0.97), p-value less than 0.001.
* Accessibility/affordability barrier vs. none: Robust S E 0.042, Adjusted P R 1.03 (0.95, 1.12), p-value 0.492.
* Technological barrier vs. none: Robust S E 0.050, Adjusted P R 0.91 (0.83, 1.01), p-value 0.073.
Note: P R stands for Prevalence Ratio, S E for Standard Error, C I for Confidence Interval, and I D P for Internally Displaced Persons.
Note: Primary adjusted model including education, camp location, social media platform use and barriers to mobile phone access or use. Modified Poisson regression with log link, robust sandwich standard errors, robust Wald p-values and 95% confidence intervals. CI, confidence interval; PR, prevalence ratio; SE = standard error.
In the primary adjusted model, the prevalence of expressed interest was 29% higher among participants with secondary education or higher than among those with no formal education (adjusted prevalence ratio [APR] 1.29, 95% CI 1.08–1.54; p = 0.004). There was no clear evidence of differences for participants with Islamic education (APR 1.16, 95% CI 0.98–1.38; p = 0.079) or primary education (APR 1.18, 95% CI 0.98–1.42; p = 0.080), compared with those with no formal education.
The prevalence of expressed interest was 12% higher among Wassa residents than among Durumi residents (APR 1.12, 95% CI 1.03–1.21; p = 0.009) and 8% lower among participants reporting no social media platform use than among those reporting use of one or more platforms (APR 0.92, 95% CI 0.88–0.97; p < 0.001). Neither accessibility or affordability barriers (APR 1.03, 95% CI 0.95–1.12; p = 0.492) nor technological barriers (APR 0.91, 95% CI 0.83–1.01; p = 0.073) were clearly associated with expressed interest, compared with no reported barriers.
Figure 1 visually summarizes the adjusted associations with expressed interest in mLSE.
Primary adjusted prevalence ratios for expressed interest in using mLSE. * Points represent adjusted prevalence ratios from the primary modified robust Poisson regression model, and horizontal lines indicate 95% confidence intervals. Estimates to the right of 1 indicate higher adjusted prevalence of expressed interest relative to the reference category, whereas estimates to the left of 1 indicate lower adjusted prevalence. Filled points denote estimates whose 95% confidence intervals exclude 1, and hollow points denote estimates whose 95% confidence intervals include 1. Reference categories were no formal education for education; Durumi IDP Camp for camp location; use of one or more social media platforms for social media platform use; and none for barriers to mobile phone access or use.
Figure 1. Long description
A forest plot with a vertical reference line at 1.0 on the x-axis, which is labeled Adjusted prevalence ratio with a scale from 0.8 to 1.5. The top of the plot is divided into Lower interest in using m L S E on the left and Higher interest in using m L S E on the right.
From top to bottom, the data points are:
* Islamic education: A hollow blue circle at approximately 1.17 with a wide horizontal confidence interval extending from 0.98 to 1.38.
* Primary education: A hollow blue circle at approximately 1.18 with a confidence interval from 0.98 to 1.42.
* Secondary/higher education: A filled blue circle at approximately 1.29 with a confidence interval from 1.08 to 1.54.
* Wassa I D P Camp: A filled blue circle at approximately 1.11 with a confidence interval from 1.03 to 1.21.
* No social media platform use: A filled red circle at approximately 0.92 with a narrow confidence interval from 0.87 to 0.97.
* Accessibility/affordability barrier: A hollow blue circle at approximately 1.03 with a confidence interval from 0.94 to 1.12.
* Technological barrier: A hollow red circle at approximately 0.91 with a confidence interval from 0.82 to 1.01.
Secondary exploratory analysis
A fuller exploratory model incorporating the broader set of prespecified covariates yielded results generally similar to those of the primary analysis (Table 4). Because only 15 participants reported no interest in using mLSE, estimates from this model should be interpreted as exploratory and hypothesis-generating.
Secondary exploratory analysis: fuller adjusted model
Table 4. Long description
The table is divided into two main sections: Unadjusted analysis and Adjusted exploratory analysis. Each section includes columns for Robust S E, P R or A P R with 95 percent C I, and p-value.
Key findings include:
* Age: All groups (20 to 24, 25 to 29, 30 to 34, and 35 plus years) show no significant difference compared to the 13 to 19 years reference group in either model.
* Gender: Male participants show no significant difference compared to females (A P R 0.97, p-value 0.478).
* Education: Compared to no formal education, secondary education or higher is significantly associated with the outcome in both unadjusted (P R 1.28, p-value 0.008) and adjusted models (A P R 1.31, p-value 0.002). Primary education is significant in the adjusted model (A P R 1.19, p-value 0.049).
* Social media platform use: No social media use is significantly associated with a lower prevalence ratio compared to using one or more platforms (A P R 0.92, p-value 0.013).
* I D P camp location: Wassa I D P Camp shows a significant difference in the adjusted model compared to Durumi I D P Camp (A P R 1.12, p-value 0.032).
* Other variables including employment, barriers, language use (English and Hausa), mobile phone use, psychological counseling history, and duration in camp did not show statistically significant associations in the adjusted model.
Note: Fuller exploratory adjusted model including age, gender, educational attainment, employment status, barriers, camp location, English language use, Hausa language use, usual mobile phone use, social media platform use, history of psychological counseling and duration in camp. Modified Poisson regression with log link, robust sandwich standard errors, robust Wald p-values and 95% confidence intervals.
Discussion
This cross-sectional study examined factors associated with expressed interest in using mobile-delivered life skills education (mLSE) among IDPs living in Durumi and Wassa camps in Abuja, Nigeria. Overall, expressed interest was high , with 216 of 231 participants reporting interest in using mLSE. However, this finding should be interpreted cautiously. The outcome reflects stated interest at one point in time rather than observed technology acceptance, enrollment, uptake, adherence, sustained engagement, or intervention effectiveness.
The high prevalence of affirmative responses may indicate substantial openness to receiving coping-related education via mobile phone. However, these responses may also have been influenced by general openness to services, politeness toward research staff, social desirability, or participants’ perceptions of what was expected of them in the context of a humanitarian research program.. These response processes were not measured directly. In addition, because the survey item combined interest in coping-related education with interest in receiving it via a mobile phone, it is not entirely clear whether participants were responding primarily to the intervention content, the delivery format or both. The prevalence should therefore be understood as the proportion reporting expressed interest, not as evidence of implementation readiness or probable engagement.
Educational attainment was associated with expressed interest in the primary adjusted model. Participants with secondary education or higher had a higher adjusted prevalence of expressed interest than those with no formal education. In contrast, the estimates for Islamic and primary education were in the same direction, but their confidence intervals included the null. This pattern may reflect differences in comfort with instructional content, literacy demands or familiarity with formal learning environments. However, because educational attainment may not accurately represent digital literacy (Eurostat, 2024), the present analysis cannot determine the mechanism underlying the association. Previous studies have similarly reported relationships between education or eHealth literacy and willingness to engage with mobile health interventions (James and Harville, Reference James and Harville2016; Endehabtu et al., Reference Endehabtu, Weldeab, Were, Lester, Worku and Tilahun2018).
Social media use was also associated with expressed interest. Participants who reported no social media platform use had a lower adjusted prevalence of expressed interest than those who reported using one or more platforms. Social media use may indicate prior familiarity with mobile interfaces, messaging applications or digitally delivered content. Still, it is not a direct measure of perceived usefulness, perceived ease of use or technology acceptance. The association is nevertheless relevant to implementation because mobile interventions may be easier to introduce among participants who already use digital platforms (James and Harville, Reference James and Harville2016; Potdar et al., Reference Potdar, Thomas, DiMeglio, Mohiuddin, Djibo, Laudanski, Dourado, Leighton and Ford2020).
The findings on education and social media raise important equity considerations. A mobile intervention that relies heavily on written content, smartphones, messaging applications or independent navigation may be more accessible to participants who already possess greater educational and digital resources. Without appropriate adaptations, mobile delivery may worsen existing inequalities in access to psychosocial support. Low-literacy audio and visual content, simple navigation, assisted onboarding, low-data delivery, non-smartphone alternatives where feasible and in-person or hybrid options should therefore be considered and evaluated from the outset (Till et al., Reference Till, Mkhize, Farao, Shandu, Muthelo, Coleman, Mbombi, Bopape, Klingberg, Van Heerden, Mothiba, Densmore and Verdezoto Dias2023; Sylla et al., Reference Sylla, Ismaila and Diallo2025).
High expressed interest occurred alongside limited prior digital engagement and substantial reported barriers. Only 20.3% of participants reported using one or more social media platforms, while 51.1% were classified as reporting technological barriers. This tension is important because it indicates that interest in principle should not be conflated with the practical ability to access or use a mobile intervention. Participants may endorse the idea of mobile-delivered support while still facing limitations related to phone access, data costs, connectivity, electricity, familiarity with smartphone features or privacy (Burchert et al., Reference Burchert, Alkneme, Bird, Carswell, Cuijpers, Hansen, Heim, Harper Shehadeh, Sijbrandij, Van’T Hof and Knaevelsrud2019; Matlin et al., Reference Matlin, Hanefeld, Corte-Real, Da Cunha, De Gruchy, Manji, Netto, Nunes, Şanlıer, Takian, Zaman and Saso2025; Mwansa et al., Reference Mwansa, Ngandu and Mkwambi2025).
Barrier-related estimates were not statistically significant in the primary adjusted model, although the estimate for technological barriers was lower than that for no reported barriers. These findings should not be interpreted as evidence that barriers are unimportant. The barrier variable was created from a multiple-response item. Because some original categories had no responses or too few responses for separate analysis, they were combined into broader categories. The resulting categories described general barrier profiles rather than the specific barriers each participant reported.
Camp location remained associated with expressed interest, with a higher adjusted prevalence among participants in Wassa than among those in Durumi. The association may reflect unmeasured differences in infrastructure, connectivity, prior service exposure, camp governance, local leadership, population composition, recruitment procedures, interviewer effects or other site-specific conditions. The analysis cannot determine which, if any, of these factors accounts for the observed association. The finding should therefore be interpreted as a difference between Durumi and Wassa rather than as evidence that any specific camp characteristic influenced expressed interest. Because only two camps were included, the finding should not be generalized to other IDP settings.
TAM was applied as a broad orienting framework rather than tested as an explanatory model. Perceived usefulness, perceived ease of use, behavioral intention and actual use were not directly measured. Education, social media use, reported barriers and camp location were examined as implementation-relevant external characteristics rather than as validated measures of TAM constructs. The analysis therefore cannot determine whether the observed associations operated through perceived usefulness, perceived ease of use, prior digital familiarity, access constraints or other mechanisms (Davis, Reference Davis1989; Marangunić and Granić, Reference Marangunić and Granić2015).
The fuller exploratory model showed a similar overall pattern to the primary adjusted model. Because only a small minority of participants expressed no interest in using mLSE, findings from the fuller model were treated as exploratory and hypothesis-generating. With only 15 negative outcomes and zero cells in some predictor categories, the information available for multivariable estimation was limited, particularly in the fuller exploratory model. Estimates from this model may therefore be susceptible to sparse-data bias. Greater inferential weight was given to the primary parsimonious model.
Strengths, limitations and future research directions
This study has several strengths. To begin with, it addresses a practical and policy-relevant implementation question in an underrepresented displaced population where access to mental health and psychosocial support is limited. The analysis used baseline data collected before the intervention was delivered, mitigating the possibility that prior exposure to the program influenced responses. Furthermore, the household sampling process was probability-based within the two selected camps, using site-specific household sampling frames created through enumeration and random selection of one eligible participant per sampled household. The use of modified Poisson regression with robust standard errors also allowed direct estimation of prevalence ratios for a common binary outcome.
Another strength is that the study incorporated demographic, contextual and technology-related characteristics relevant to planning mobile-delivered support. The use of a parsimonious primary model and a fuller exploratory model was appropriate, given the limited variation in outcomes. In addition, the interpretation remained focused on expressed interest rather than extending claims to uptake, adherence or effectiveness.
Several limitations should be considered. A key limitation is that the primary outcome was based on a single survey item. The item combined interest in coping-related education with interest in receiving it via a mobile phone and therefore did not separate interest in the intervention content from interest in the delivery modality. The primary outcome was also not a validated measure of perceived usefulness, perceived ease of use, behavioral intention, technology acceptance, implementation readiness, uptake, adherence or sustained engagement. In addition, the survey was administered only in Hausa, and English and Hausa language use was self-reported rather than formally assessed. Differences in comprehension therefore cannot be excluded, particularly among participants who did not routinely use Hausa.
The very high prevalence of expressed interest may partly reflect acquiescence, politeness toward interviewers, social desirability or perceived expectations associated with research participation in a humanitarian setting. These response processes were not assessed directly. Although interviewer administration enabled participation by respondents with varying literacy levels, it may have increased susceptibility to such responses.
Importantly, only 15 participants reported no interest in using mLSE. Some predictor categories contained no participants who reported no interest. This sparse outcome distribution limited subgroup discrimination and the reliability of estimates, particularly in the fuller exploratory model. The exploratory findings should therefore be regarded as hypothesis-generating rather than confirmatory.
The cross-sectional design also limits causal attribution. Although educational attainment and camp residence likely influencedthe expression of interest, the analysis cannot establish that these characteristics caused differences in interest or exclude confounding by unmeasured factors.
Although the questionnaire assessed several determinants of mobile access and use, these measures were not all included separately in the adjusted models. The survey assessed phone ownership and type, shared-phone access, internet access, affordability of phone credit and data, network connectivity, electricity for charging, familiarity with smartphone features and privacy or safety concerns. However, several response categories had no observations or insufficient frequencies for separate analysis, and the limited number of participants reporting no interest limited the number of variables that could be included in the regression models. .
The survey also did not fully characterize the practical conditions of phone sharing, household or gendered control over technology, privacy while receiving sensitive content, and autonomy over phone use. . Thus, the analysis cannot determine whether participants who expressed interest had reliable, private, affordable and autonomous access to the technology required for participation.
Camp location may capture multiple unmeasured structural and operational differences between Durumi and Wassa. Because the study included only two camps, which were not randomly selected from all IDP settlements, the findings may not be generalizable beyond these sites. The complete-case analysis presents an additional limitation. If the 33 excluded participants differed systematically from those included, the estimated prevalence of expressed interest and the observed associations may not fully represent the enrolled sample.
Future research should use multi-item measures that directly assess perceived usefulness, perceived ease of use, behavioral intention and related technology acceptance constructs. Studies should characterize device ownership and control, shared-phone access and use, digital literacy, data affordability, connectivity, charging access, privacy and gender-related restrictions on technology use in sufficient detail for more robust analyses. Qualitative research could clarify how displaced persons interpret questions about mobile-delivered support and whether affirmative responses reflect interest in the content, the delivery format or access to services more generally. Prospective longitudinal and implementation-focused studies should assess whether expressed interest translates into enrollment, uptake, sustained engagement and intervention effectiveness, while also evaluating the feasibility of assisted and hybrid delivery approaches for people with limited literacy or digital access.
Conclusions
Most participants in Durumi and Wassa expressed interest in the proposed mLSE format, assessed by asking whether they would be interested in using a mobile phone to receive education or training on how to cope with life challenges. In the primary adjusted analysis, expressed interest was more prevalent among participants with secondary education or higher and among Wassa residents, and less prevalent among those reporting no social media platform use. These findings indicate high stated interest but do not establish technology acceptance, implementation feasibility, uptake or sustained engagement. Planning for mLSE should consider varying literacy needs, prior digital engagement, constraints on mobile phone access or use and site-specific context. Prospective studies should assess whether stated interest translates into enrollment, uptake, sustained engagement and intervention outcomes.
Open peer review
To view the open peer review materials for this article, please visit http://doi.org/10.1017/gmh.2026.10305.
Data availability statement
The data that support the findings of this study are available on request from the corresponding author.
Author contribution
A.E.E.D. and E.E.: study conceptualization and design; A.E.E.D. and E.E.: data acquisition; A.E.E.D.: data interpretation; A.E.E.D.: data analysis and visualization; A.E.E.D. and E.E.: raw data verification; A.E.E.D.: drafting of the original and final drafts of the manuscript; A.E.E.D.: methodology; R.O. and E.E.: supervision; E.E.: resources; A.E.E.D. and E.E.: full access to the data in the study. All authors (A.E.E.D., B.I.H., I.I., A.I., A.A., C.O.J., A.J.J.H., S.Y., V.I.O.A., R.O. and E.E.) provided critical feedback on revising the manuscript for intellectual content. All authors (A.E.E.D., B.I.H., I.I., A.I., A.A., C.O.J., A.J.J.H., S.Y., V.I.O.A., R.O. and E.E.) were ultimately responsible for the decision to submit for publication.
Ethics and approvals
The study was approved by Nigeria’s National Health Research Ethics Committee (NHREC/01/01/2007-18/01/2024) and the Health Sciences Research Ethics Board at Dalhousie University (REB #2024-7085), and was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from adult participants, while assent and parent or guardian consent were obtained for participants younger than 18 years.
Funding statement
This work is supported by Creating Hope in Conflict: a Humanitarian Grand Challenge; a partnership of the United States Agency for International Development (USAID), the Foreign, Commonwealth & Development Office, United Kingdom of Great Britain and Northern Ireland (FCDO), the Stabilization and Humanitarian Aid Department, Ministry of Foreign Affairs of the Netherlands (NL MFA) and His Majesty the King in right of Canada (“His Majesty”) represented by the Minister for International Development of Global Affairs Canada acting through The Department of Foreign Affairs, Trade and Development of Global Affairs Canada (“DFATD”) with support from Grand Challenges Canada [grant number R-HGC-POC-2408 67370]; and the Government of Canada, Canadian Institutes of Health Research, Institute of Population and Public Health [grant number PAA-192178]. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
Competing interests
The authors declare no competing interests.
Open access
Comments
Dear Editor,
We are submitting our original research manuscript, “Expressed Interest in Mobile-Delivered Life Skills Education for Mental Health and Psychosocial Support Among Internally Displaced Persons,” for consideration for publication as an Original Research Article in Cambridge Prisms: Global Mental Health.
In this study, we examine factors associated with expressed interest in mobile-delivered life skills education (mLSE) among 231 internally displaced persons in two camps in Abuja, Nigeria. Using modified Poisson regression, we identify demographic, contextual, and technology-related factors associated with interest in mLSE. Our findings show the potential of mobile-delivered support in humanitarian settings while also highlighting important barriers related to literacy, affordability, and digital access.
We believe this manuscript is a strong fit for the journal because it addresses global mental health in a displaced population and offers evidence with clear relevance for the design and equitable implementation of scalable digital mental health and psychosocial support in low-resource humanitarian settings.
This manuscript is original, has not been published previously, and is not under consideration elsewhere. All authors have reviewed and approved the submitted version.
Sincerely,
Ejemai Eboreime, MD, PhD