Lack of ownership of mobile phones could hinder the rollout of mHealth interventions in Africa

  1. Justin T Okano
  2. Joan Ponce
  3. Matthias Krönke
  4. Sally Blower  Is a corresponding author
  1. Center for Biomedical Modeling, Department of Psychiatry and Biobehavioral Sciences, Semel Institute for Neuroscience and Human Behavior, David Geffen School of Medicine, University of California, Los Angeles, United States
  2. Afrobarometer / Institute for Democracy, Citizenship and Public Policy in Africa, University of Cape Town, South Africa
4 figures, 4 tables and 1 additional file

Figures

Figure 1 with 3 supplements
Basic mobile phone and smartphone ownership in 33 African countries.

(A) Boxplots show the probabilities of not owning a mobile phone (cream), owning a basic mobile phone (BP; orange), or owning a smartphone (SP; red). Country-level probabilities (dots) are overlaid and jittered to reduce overlap. (B) Barplot shows the country-level probabilities of BP ownership (orange) and SP ownership (red) ordered by SP ownership. Geographic distribution showing probabilities of (C) BP ownership and (D) SP ownership in 33 Afrobarometer countries at the sub-national level.

Figure 1—figure supplement 1
Flow diagram of participant sample sizes.

Shown for 33 countries that collect phone ownership data as part of the Afrobarometer Round 7 (R7) survey.

Figure 1—figure supplement 2
Internet usage by type of phone owned.

Afrobarometer Round 7 (R7) surveyed individuals on how often they used the internet. Individuals were not asked to specify how they used the internet (i.e. they did not necessarily have to use or own a phone to access the internet).

Figure 1—figure supplement 3
Phone ownership in the poorest individuals.

(A) Country-level basic mobile phone (BP) ownership in individuals with Lived Poverty Index (LPI) = 3. (B) Country-level smartphone (SP) ownership in individuals with LPI = 3.

Proximity to health clinics and mobile phone ownership in 33 African countries.

(A) Boxplots show the proportion of individuals who live in close proximity to a health clinic (HC) based on whether they do not own a mobile phone (mean 0.49), own a basic mobile phone (BP; mean 0.56), or own a smartphone (SP; mean 0.66). Country-level probabilities (dots) are overlaid and jittered to reduce overlap. (B) Scatterplot shows the country-specific proportions of individuals who live in close proximity to an HC amongst individuals who: do not own a mobile phone (white dots), own a BP (red dots), or own an SP (black dots).

Phone ownership by age and gender in 33 African countries.

Population pyramid displays the distribution of the population stratified by gender and 5 year age groupings (with the exception of the 18–20 age class) by ownership of a mobile phone: no mobile phone (cream), a basic mobile phone (BP; orange), or a smartphone (SP; red).

Figure 4 with 8 supplements
Country-level gender and urban-rural effects.

(A) Posterior distributions of the country-level effect on mobile phone ownership of being male (compared to female), sorted by median. (B) Posterior distributions of the country-level effect of living in an urban area (compared to living in a rural area), in the same country order as (A). Both (A) and (B) are on the logit-scale – and should be viewed respectively as country-specific adjustments to the population-level effect of (A) being male or (B) living in an urban area.

Figure 4—figure supplement 1
Model 1 posteriors – (country-level) intercept.

Posterior distributions (medians and 95% highest posterior density [HPD] regions) of u0j , the country-level intercept, for country j.

Figure 4—figure supplement 2
Model 2 posteriors – (country-level) intercept.

Posterior distributions (medians and 95% highest posterior density [HPD] regions) of u0j , the country-level intercept, for country j.

Figure 4—figure supplement 3
Model 2 posteriors – (country-level) urban/rural.

Posterior distributions (medians and 95% highest posterior density [HPD] regions) of u1j , the country-level effect of living in an urban area, for country j.

Figure 4—figure supplement 4
Model 2 posteriors – (country-level) gender and proximity I.

Posterior distributions (medians and 95% highest posterior density [HPD] regions) of u2j , the country-level effect of being female in close proximity to a health clinic (HC), for country j.

Figure 4—figure supplement 5
Model 2 posteriors – (country-level) gender and proximity II.

Posterior distributions (medians and 95% highest posterior density [HPD] regions) of u3j , the country-level effect of being male not in close proximity to a health clinic (HC), for country j.

Figure 4—figure supplement 6
Model 2 posteriors – (country-level) gender and proximity III.

Posterior distributions (medians and 95% highest posterior density [HPD] regions) of u4j , the country-level effect of being male in close proximity to a health clinic (HC), for country j.

Figure 4—figure supplement 7
Model 1 Markov Chain Monte Carlo (MCMC) diagnostics and ROC curve.

(A) Violin plots of the log posterior (top) and No-U-Turn Sampling (NUTS) acceptance statistic (bottom). There are no divergent transitions. To further assess chain convergence, we used R^ , the potential scale reduction factor (Gelman and Rubin, 1992). All values of R^ are nearly equal to 1 (|R^|<1.005), indicating all chains have converged. (B) NUTS energy plot (Betancourt, 2017) for all four chains. The transition distribution (darker histogram in each plot) and target distribution (lighter histogram in each plot) are moderately well aligned, indicating efficient sampling. (C) The effective sample size Neff from the posterior distribution of each parameter was used to gauge the level of autocorrelation in each sample. Neff/N>0.1 for all model parameters indicating independent draws within each sample. (D) ROC curve used to assess the diagnostic capability of fitted Bayesian multilevel logistic regression (BMLR) model. The AUC was 0.79 indicating good predictive accuracy.

Figure 4—figure supplement 8
Model 2 Markov Chain Monte Carlo (MCMC) diagnostics and ROC curve.

(A) Violin plots of the log posterior (top) and No-U-Turn Sampling (NUTS) acceptance statistic (bottom). There are no divergent transitions. Additionally, all values of R^ are nearly equal to 1 (|R^|<1.005), indicating all chains have converged. (B) NUTS energy plot for all four chains. The transition distribution (darker histogram in each plot) and target distribution (lighter histogram in each plot) are moderately well aligned, indicating efficient sampling. (C) Neff/N>0.1 for all model parameters indicating independent draws within each sample. (D) ROC curve used to assess the diagnostic capability of fitted Bayesian multilevel logistic regression (BMLR) model. The AUC was 0.78 indicating good predictive accuracy.

Tables

Table 1
Probability of mobile phone ownership by country and gender.
CountryAny mobile phoneSmartphone
FemaleMaleORFemaleMaleOR
Benin0.660.904.78*0.190.412.84*
Mali0.600.884.77*0.150.261.93*
Burkina Faso0.670.904.31*0.180.271.64*
Senegal0.830.953.90*0.440.521.40*
Guinea0.790.933.46*0.220.231.06
Niger0.470.743.23*0.140.231.92*
Cote d`Ivoire0.850.953.10*0.320.491.98*
Nigeria0.780.913.02*0.480.541.29*
Gambia0.850.942.82*0.600.651.23
Togo0.750.892.80*0.330.451.61*
Uganda0.590.802.76*0.330.371.18
Tunisia0.920.972.45*0.610.600.97
Sierra Leone0.620.802.42*0.440.441.01
Tanzania0.740.872.41*0.230.311.50*
Malawi0.410.612.29*0.340.431.49*
Morocco0.900.952.28*0.600.641.22
Ghana0.820.912.15*0.350.471.70*
Mauritius0.930.972.11*0.630.641.02
Cameroon0.850.922.10*0.550.540.96
Mozambique0.610.751.87*0.450.451.02
Sao Tome and Principe0.750.851.83*0.550.581.14
Liberia0.640.761.79*0.380.531.90*
Sudan0.890.941.77*0.650.671.11
Eswatini0.920.951.74*0.570.601.14
Zambia0.630.731.61*0.400.411.06
Zimbabwe0.820.871.46*0.420.451.15
Madagascar0.440.501.29*0.360.361.00
Gabon0.930.941.260.630.661.13
Namibia0.880.901.240.540.621.40*
Botswana0.890.911.230.550.621.33*
Cabo Verde0.850.881.200.710.771.36*
Lesotho0.840.861.180.390.391.02
South Africa0.910.910.970.690.680.93
  1. *

    Significant at α=0.05.

  2. Odds ratio of male phone ownership to female phone ownership.

Table 2
Probability of mobile phone ownership by country and urban/rural status.
CountryAny mobile phoneSmartphone
RuralUrbanORRuralUrbanOR
Gabon0.820.977.17*0.310.725.66*
Zimbabwe0.780.955.94*0.280.644.58*
Uganda0.630.905.58*0.290.492.42*
Burkina Faso0.740.945.25*0.120.527.57*
Ghana0.770.945.09*0.240.533.60*
Senegal0.820.964.78*0.340.602.90*
Morocco0.860.974.70*0.420.713.51*
Zambia0.560.844.23*0.290.502.54*
Madagascar0.400.734.17*0.270.543.25*
Togo0.760.934.14*0.250.573.92*
Tanzania0.750.924.00*0.170.423.45*
Malawi0.450.763.87*0.330.542.37*
Guinea0.820.953.84*0.120.425.11*
Liberia0.580.843.69*0.350.552.33*
Botswana0.840.953.65*0.430.733.56*
Mali0.700.883.24*0.130.476.09*
Niger0.570.803.10*0.140.373.63*
Sudan0.890.963.09*0.590.762.16*
Nigeria0.790.923.05*0.420.612.20*
Lesotho0.820.932.91*0.270.573.62*
Mozambique0.600.812.86*0.330.603.02*
Cameroon0.830.932.74*0.430.642.42*
Cote d`Ivoire0.860.942.45*0.260.553.35*
Tunisia0.910.962.32*0.520.641.62*
Sierra Leone0.640.802.28*0.320.552.61*
Namibia0.850.932.22*0.440.682.72*
Benin0.720.852.18*0.230.402.21*
Eswatini0.930.972.170.560.732.17*
Cabo Verde0.820.891.69*0.630.792.26*
South Africa0.890.921.360.570.742.07*
Sao Tome and Principe0.770.821.34*0.520.591.30
Mauritius0.950.950.980.610.671.29*
Gambia0.920.890.710.610.641.15
  1. *

    Significant at α=0.05.

  2. Odds ratio of urban phone ownership to rural phone ownership.

Table 3
Fitted Bayesian models for determinants of mobile phone ownership in Africa.
Population-level effectsBivariable (Bayesian logistic regression [BLR] models)Multivariable (Bayesian multilevel logistic regression [BMLR] model 1)
Crude OR95% highest posterior density (HPD) regionaOR95% highest posterior density (HPD) region
Femaleref-ref-
Male2.04(1.94, 2.16)*2.37(1.96, 2.84)*
Ruralref-ref-
Urban3.47(3.27, 3.68)*2.66(2.22, 3.18)*
Age (61+)ref-ref-
18–201.36(1.21, 1.53)*1.73(1.51, 1.96)*
21–251.75(1.59, 1.93)*2.29(2.05, 2.56)*
26–401.98(1.82, 2.15)*2.68(2.44, 2.94)*
41–501.85(1.68, 2.03)*2.44(2.19, 2.72)*
51–601.54(1.39, 1.71)*1.78(1.59, 2.00)*
Lived Poverty Index (LPI) = 0 (wealthiest)ref-ref-
LPI = 10.50(0.44, 0.56)*0.69(0.61, 0.78)*
LPI = 20.29(0.26, 0.32)*0.46(0.41, 0.52)*
LPI = 3 (poorest)0.24(0.21, 0.26)*0.35(0.31, 0.40)*
No health clinic (HC) in close proximityref-ref-
HC in close proximity1.56(1.49, 1.65)*1.31(1.24, 1.39)*
Country-level effectsEstimate95% highest posterior density (HPD) region
σu00.87(0.67, 1.11)
σu10.48(0.33, 0.62)
σu20.49(0.36, 0.64)
ρ01–0.43(–0.73, –0.13)
ρ02–0.32(–0.63, –0.02)
ρ120.21(–0.16, 0.55)
  1. *

    Significant at α=0.05.

  2. Bayesian 95% HPD regions.

Table 4
Fitted Bayesian models for determinants of smartphone ownership among mobile phone owners in Africa.
Population-level effectsBivariable (Bayesian logistic regression [BLR] models)Multivariable (Bayesian multilevel logistic regression [BMLR] model 2)
Crude OR95% highest posterior density (HPD) regionaOR95% highest posterior density (HPD) region
Female, no health clinic (HC) in close proximityref-ref-
Female, HC in close proximity1.23(1.15, 1.32)*1.15(1.03, 1.30)*
Male, no HC in close proximity1.10(1.03, 1.18)*1.50(1.30, 1.72)*
Male, HC in close proximity1.54(1.44, 1.65)*1.92(1.63, 2.26)*
Ruralref-ref-
Urban2.95(2.81, 3.09)*2.67(2.33, 3.10)*
Age (61+)ref-ref-
18–305.59(5.08, 6.16)*7.85(7.04, 8.77)*
31–403.53(3.19, 3.90)*4.73(4.24, 5.30)*
41–502.25(2.02, 2.50)*2.75(2.44, 3.09)*
51–601.58(1.41, 1.77)*1.72(1.51, 1.96)*
Lived Poverty Index (LPI) = 0 (wealthiest)ref-ref-
LPI = 10.61(0.57, 0.66)*0.67(0.62, 0.73)*
LPI = 20.38(0.35, 0.40)*0.45(0.41, 0.49)*
LPI = 3 (poorest)0.27(0.25, 0.29)*0.35(0.32, 0.39)*
Country-level effectsEstimate95% highest posterior density (HPD) region
σu00.85(0.67, 1.05)
σu10.39(0.28, 0.50)
σu20.21(0.08, 0.34)
σu30.31(0.18, 0.43)
σu40.41(0.28, 0.53)
ρ01–0.61(–0.81, –0.37)
ρ02–0.38(–0.77, 0.01)
ρ03–0.51(–0.80, –0.19)
ρ04–0.78(–0.93, –0.60)
ρ120.10(–0.38, 0.54)
ρ130.16(–0.24, 0.55)
ρ140.29(–0.06, 0.60)
ρ23–0.07(–0.59, 0.41)
ρ240.47(0.04, 0.82)
ρ340.61(0.30, 0.88)
  1. *

    Significant at α=0.05.

  2. Bayesian 95% HPD regions.

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Justin T Okano
  2. Joan Ponce
  3. Matthias Krönke
  4. Sally Blower
(2022)
Lack of ownership of mobile phones could hinder the rollout of mHealth interventions in Africa
eLife 11:e79615.
https://doi.org/10.7554/eLife.79615