Informing mobile health strategies: Study of mobile phone ownership among rural women for breast cancer control in Ghana

Abstract

Background

Breast cancer (BC) screening uptake remains low among women in rural Ghana because of structural and informational barriers. Understanding mobile phone ownership, use, and willingness to receive short message service (SMS) messages is essential for designing SMS-based mobile health interventions to improve BC awareness and uptake of clinical breast examinations.

Objective

To examine mobile phone ownership, usage patterns, and willingness to receive SMS-based BC information among women in rural Ghana to inform the development of a future SMS-based mobile health intervention.

Methods

We analyzed data from 708 women aged ≥25 years who participated in a community-based cross-sectional survey conducted in 14 rural communities in the Ashanti Region of Ghana. Descriptive statistics, chi-square tests, and Firth’s penalized logistic regression were performed.

Results

Mobile phone ownership was high (98.2%), and most women stated the willing to receive SMS messages about BC (93.9%) and clinical breast examination appointments (93.8%). Respondents who anticipated an early (same-day) response to messages had 13.91 times higher odds of being willing to receive text messages about breast cancer and screening (AOR = 13.91; 95%CI: 6.11, 31.65; p < 0.001) and 9.58 times higher odds of being willing to receive reminders for setting up CBE appointments (AOR = 9.58; 95%CI: 4.44, 20.66; p < 0.001) compared to late responders. Additionally, being comfortable with reading and receiving SMS was significantly associated with a willingness to receive general breast cancer texts (AOR = 10.93; 95%CI: 1.01, 117.70; p = 0.049).

Conclusion

High mobile phone ownership and willingness to receive SMS messages suggest that SMS-based mobile health interventions may be a feasible strategy for promoting BC awareness and clinical breast examination in rural Ghana. The cross-sectional design assessed stated willingness rather than actual use of an SMS-based intervention and does not permit causal inference.

Keywords

mobile phone breast cancer awareness clinical breast examination short message service women Ghana

Background

Ghana, like lower-and middle-income countries (LMICs), is experiencing an increasing burden of non-communicable diseases (NCDs), including cancer.¹ The burden of cancer is becoming a growing global public health concern. It is estimated that the worldwide incidence of cancer among women will rise to 13.3 million (a 44% increase) by 2040, resulting in approximately 7.1 million deaths.¹ In sub-Saharan Africa and Ghana, breast cancer is the most commonly diagnosed cancer among women.^2,3 In 2022, around 5,026 new cases of breast cancer were diagnosed in Ghana, accounting for 31.4% of all cancers in the country.⁴ Although there are no clear prevention pathways for breast cancer like those for cervical cancer (such as HPV vaccination),⁵ most empirical evidence highlights the importance of breast cancer screening for early detection and diagnosis,^6–9 However, in Ghana, the uptake of breast screening practices remains low, with reports showing that 4.5% to 42.6% of Ghanaian women have ever participated in breast examinations, and fewer than 3.0% have had a mammogram.^10,11 In rural communities, the lack of screening facilities; long travel distances to cities with limited screening options; priorities for work; cultural narratives around the perceived causes of breast cancer; low perceived susceptibility; lower household income; lack of insurance coverage; lower educational levels; and cultural norms such as shyness have been identified as significant barriers to the uptake of breast screening practices.^12–14 Besides these barriers, informational obstacles include a lack of awareness about breast cancer screening and limited knowledge of how to perform self-examinations and where to seek clinical breast examination (CBE).^11,15,16

In response to these structural and informational barriers to breast cancer awareness and screening, there is a growing interest in exploring how technological tools and applications can enhance screening. For example, Eden et al.¹⁷ have shown that using a mobile health (m-health) app can significantly reduce women’s decision conflict regarding mammography and boost their confidence in screening. Likewise, another study reports that a smartphone app increased breast screening uptake by approximately 8.9% among women in Korea.¹⁸ However, in Ghana, there are limited studies on the use of mobile phone applications to advance breast cancer awareness and screening uptake among women. Further, the current state of mobile phone ownership and use is unknown. The only available Ghanaian study¹⁹ focused on identifying barriers and facilitators, as well as women’s preferences, regarding the modes of delivering short message service (SMS) messages about cervical cancer screening in urban areas. Bonful et al.’s study¹⁹ focused on the urban population and did not clarify the women’s willingness to accept the intervention. Understanding mobile phone ownership, usage patterns, and populations’ willingness to engage with mobile health interventions in specific contexts is critical to developing effective, accessible digital health strategies. However, evidence on these factors among women in rural Ghana remains limited. As a result, the following research questions remain important: (a) What are the patterns of mobile phone ownership and usage among women in rural Ghana? and (b) Are women living in rural areas of Ghana willing to receive breast cancer awareness and screening information through a mobile health intervention?

To address these questions and inform the development of a future SMS-based mobile health intervention, we examined mobile phone ownership and use, as well as women’s willingness to receive breast cancer-related information through mobile phones in rural Ghana. This study was conducted as part of a larger research project aimed at developing an SMS-based mobile health intervention to enhance breast cancer awareness and promote the uptake of CBE among women in Ghana. Importantly, the present formative study did not evaluate a breast cancer awareness or screening intervention, nor did it assess the acceptability or effectiveness of a specific mobile application. Rather, its purpose was to generate baseline evidence to inform the design, development, and content of a future mobile health intervention. The findings indicate that mobile phone access and use are widespread among women in the study setting and suggest that SMS-based mobile health approaches may be a feasible platform for delivering breast cancer awareness and screening information. These findings provide an important foundation for the development and future evaluation of mobile health interventions to improve breast cancer awareness and screening uptake among women in Ghana.

Methods

Design and setting

This study employed a cross-sectional design. It was part of a larger community-based project that gathered information from women living in rural areas to help develop a mobile phone-based app for breast cancer control in Ghana. The study took place in the Onwe, within the Ejisu Municipality of the Ashanti Region, Ghana. Onwe is mainly rural, comprising 14 communities and four healthcare facilities, two hospitals and two Community-Based Health Planning and Services (CHPS) centers. The demographic profile of Onwe comprises approximately 100,384 households, with around 37,745 women aged 25 to 79 years in 2021.²⁰

Inclusion and exclusion criteria

Women aged 25 years and older who had no history of breast cancer, were residents in the Onwe communities, and were willing to provide informed consent were eligible to participate in the study. The age threshold of 25 years was based on current breast cancer screening recommendations in Ghana, which encourage women aged 25 years and above to perform monthly breast self-examination (BSE) and undergo annual CBE.²¹ This age criterion was approved by the ethics committee and applied throughout data collection. Additionally, participants needed to speak either Twi or English as their primary language. Individuals who did not meet these criteria were not eligible to participate in the study.

Sampling and sample size

We used the existing Performance Monitoring for Action (PMA) household sampling frame to identify and recruit participants.^22,23 PMA is a comprehensive nationwide household survey tracking the adoption of family planning methods and other health indicators across health facilities.^22,23 The PMA sampling frame is based on a multistage cluster sampling design in which enumeration areas were originally selected by the Ghana Statistical Service using probability proportional to size (PPS) sampling with urban–rural stratification.^22,23 The PMA household listing available to the research team included all households in the 14 study communities. Accordingly, the sampling frame provided complete household coverage of the study communities from which eligible households were identified and selected.

The study was conducted in the Onwe of the Ejisu Municipal Assembly, comprising 14 communities and an estimated 100,384 households. All 14 communities were included in the study; therefore, no additional cluster selection was undertaken at the study level. The PMA sampling frame served as the basis for identifying and recruiting potential participants from households within these communities. Within each community, households were randomly selected from the PMA sampling frame using systematic random sampling. A sampling interval was calculated by dividing the total number of listed households in the PMA frame for each community by the target number of households to be approached. A random start was selected between 1 and the sampling interval, and every kth household was approached for participation. Where a selected household was unavailable or declined to participate, the next consecutive household on the sampling frame was approached according to a predetermined replacement protocol. Within each selected household, all women aged 25 years or older were enumerated and screened for study eligibility. To ensure broad community representation while minimizing overrepresentation of larger households, a fixed number of participants was randomly selected from each eligible household. Where fewer than five eligible women resided in a household, one participant was randomly selected and invited to participate. In households with five or more eligible women, two participants were randomly selected and invited to participate. Because one participant was selected from households with fewer than five eligible women, whereas two participants were selected from households with five or more eligible women, women had unequal probabilities of selection. Participant-level sampling weights were not available and could not be reconstructed because household-specific selection probabilities and the information required to derive individual sampling weights were not retained for the analytic dataset. Consequently, all analyses were conducted without sampling weights. On average, approximately 65 women were enrolled from each community. The sample size was estimated using the single-population proportion formula, adjusted for the study design: n = (z^2* p* q) /e^2 deff.:

• z = 1.96, corresponding to a 95% confidence level.

• p = 0.50, a conservative estimate used because reliable local estimates of mobile phone ownership and willingness to receive SMS-based breast cancer information among rural women in Onwe were unavailable, and because p = 0.50 yields the maximum sample size for a given precision.

• q = 1 - p = 0.50

• e = 0.05, the margin of error.

• deff = 2.5, the design effect.

Based on these assumptions, the minimum required sample size was 961 women. After allowing for an anticipated 10% nonresponse rate, the target sample was 1,068 women (961/0.90 ≈ 1,068). A total of 1,068 women were approached, of whom 914 agreed to participate and were enrolled, corresponding to an 85.6% participation rate. Among the enrolled participants, 206 had incomplete data on key exposure variables (mobile phone ownership, comfort with SMS use, and anticipated response time) and were excluded from the complete-case analysis. Because these variables were completely missing rather than partially observed, multiple imputation was considered inappropriate. The final analytic sample therefore comprised 708 women, representing 77.5% of enrolled participants and 66.3% of all women approached. Figure 1.

It should be noted that the PMA household listing served as the sampling frame for identifying eligible households within the study communities. Although the original PMA platform employed a multistage probability sampling design with sampling weights for nationally representative surveys, the present study was conducted only within the Onwe catchment area and used the PMA household listing as a practical sampling frame rather than for population estimation. Consequently, sampling weights specific to this study were not available. Likewise, because all 14 study communities within the Onwe catchment area were included rather than a sample of communities, no additional clustering or weighting adjustments were incorporated into the regression analyses. Because the PMA household listing provided complete household coverage of the 14 study communities, the findings are applicable to women recruited from those communities. However, because sampling weights were unavailable and the analyses did not account for unequal participant-selection probabilities or household clustering, the results should not be interpreted as statistically representative of all women residing in Onwe, the Ashanti Region, or Ghana.

Study variables

Outcome variable

The outcome variable was the willingness to use an SMS-based mobile health intervention. This was assessed by participants stated willingness to receive text messages on a mobile phone about scheduling appointments for CBE and to receive information about breast cancer and its screening practices. These two indicators were analyzed independently of each other rather than as a composite outcome. Respondents who answered ‘yes’ were considered willing to use a mobile health intervention for breast cancer control.

Explanatory variables

Guided by insights from previous studies,^9,12,15,16 we included the following variables as explanatory factors: age (25-29, 30-34, 35-39, 40-44, 45-49, and 50+ years), educational level (no formal education, primary, secondary, and tertiary), occupation (agriculture/farming, self-employed, housewife, civil/government/private, and other), and marital status (single, married/cohabiting, separated/divorced, and widowed). Other variables considered included comfort with receiving and reading text messages (yes, no), comfort with sending and replying to text messages (yes, no), frequency of buying internet bundles (daily, weekly, monthly, and none), and the urgency of responding to messages delivered to mobile phones (same day and later). To optimize statistical power and ensure distinct, interpretable categories for analysis, original ordinal questionnaire items were dichotomized based on conceptual thresholds. For the SMS comfort variables (originally measured on 5- and 6-point scales), responses were split at the midpoint: scores of 3 or lower were classified as “Uncomfortable”, representing neutral-to-negative sentiment, while scores of 4 or higher were classified as “Comfortable”, representing positive sentiment. For anticipated message-reading time, responses were dichotomized to isolate immediate action from delayed action; the highest-urgency option (Level 1) was categorized as “Early”, while all subsequent levels indicating a delay (Level 2 or higher) were categorized as “Late”.

Data collection process

A questionnaire was used as the data collection tool. It was adapted from a published study that assessed mobile phone ownership and use among 3299 women undergoing cervical cancer screening in Kenya²⁴ and 259 patients with diabetes in Nigeria.²⁵ The adapted questionnaire was initially drafted in English, translated into Twi by a bilingual language expert, and then back-translated into English by a bilingual independent language expert to check for potential errors. The revised items were reviewed by public health researchers (n=2) and breast cancer specialists (n=2) in Ghana and pretested with 30 rural women from a nearby community not included in the main study to assess cultural relevance, clarity, and translation accuracy before final administration.

The construct measuring comfort with SMS use (reading/receiving and writing/sending SMS) demonstrated excellent internal consistency (Cronbach’s α = 0.91). The construct assessing health-related content searched on mobile phones showed acceptable reliability (Cronbach’s α = 0.74). The acceptability construct, comprising willingness to receive breast cancer information messages and appointment-related messages, demonstrated excellent reliability (Cronbach’s α = 0.97). The questionnaire (Supplementary file 1) was programmed in REDCap (Research Electronic Data Capture) to facilitate real-time data collection and upload, and to support quality control and efficient data processing.²⁶

Bilingual graduate research assistants (RAs) from the School of Nursing and Midwifery at the Kwame Nkrumah University of Science and Technology received training in informed consent, data collection, and quality assurance. On the first day, RAs were introduced to the research purpose, methods, ethical considerations, and operational plan. The second day focused on the data collection tool and REDCap software, while the third day involved rehearsing the data collection process, pretesting the questionnaires, and discussing procedures. After obtaining informed consent, the study staff interviewed women face‐to‐face at their homes in either the local language (Twi) or English, depending on their preference, and recorded participant responses electronically on tablets using REDCap. Each questionnaire took approximately 35 minutes to complete. The study started on 1^st March 2023 and ended on 30^th June 2023.

Statistical analyses

All statistical analyses were conducted using STATA 18 (StataCorp, College Station, TX, USA). There was no imputation. Missingness was concentrated in the principal mobile phone variables (phone ownership, comfort using SMS, and anticipated response time), which constituted key study exposures. Because these variables were unavailable rather than partially observed, complete-case analysis was considered the most appropriate analytical approach. Consequently, missing responses (n=206) were excluded from the analysis. The underlying missing-data mechanism could not be formally evaluated; therefore, the possibility that data were not missing completely at random cannot be excluded. Because missingness occurred primarily in the principal exposure variables, multiple imputation was considered unlikely to provide reliable estimates and was therefore not undertaken. Similarly, a meaningful sensitivity analysis based on imputed exposure data was deemed inappropriate. Accordingly, all reported analyses are complete-case analyses and should be interpreted with appropriate caution regarding potential selection bias. Because participant-level sampling weights were unavailable, all regression analyses were conducted as unweighted models, and no survey-design adjustments for unequal selection probabilities or household clustering were applied. A descriptive analysis was initially performed to determine the distribution of the proportion of women willing to receive text messages about scheduling CBE appointments and information about breast cancer and its screening methods. We then performed cross-tabulations and used the Chi-square test and Firth’s Penalized logistic regression to assess whether willingness to receive text messages about scheduling screening appointments and information about breast cancer differed significantly across participant characteristics. Firth’s penalized logistic regression was used because the outcome variables were highly imbalanced, with relatively few ‘no’ responses, which could introduce small-sample bias and quasi-complete separation in standard logistic regression models. Moreover, standard logistic regression models were initially examined; however, evidence of model instability due to sparse data and separation issues prompted the use of Firth’s approach, which provides bias-reduced, more reliable parameter estimates under such conditions. A p-value less than 0.05 was considered statistically significant. The results of the multivariable logistic regression model were reported as adjusted odds ratios (AOR) with 95% confidence intervals (CI). The final multivariable models included age, educational level, marital status, and occupation as adjustment variables based on prior evidence and their potential role as confounders.

Ethical considerations

The study was conducted in accordance with the Declaration of Helsinki. Ethical approval was sought and obtained from the Ghana Health Service Ethics Review Committee (GHS-ERC: 016/01/23).

Results

Sample characteristics

Table 1 shows the distribution of the sample characteristics. Most participants were aged 25-29 years (25.7%) and had secondary or higher education (71.9%). Marital status was predominantly married or cohabiting (60.6%). The most common occupation was self-employment (52.5%). Additionally, 65.8% of respondents reported being comfortable reading and receiving SMS messages, whereas 60.9% reported being comfortable writing and sending SMS messages. The majority of participants (82.6%) responded to messages on their mobile phones the same day.

Table 1.

Distribution of sample characteristics.

Variables	Frequency (n)	Percentage (%)
Age categories
25-29 years	182	25.7
30-34 years	127	17.9
35-39 years	100	14.1
40-44 years	78	11.0
45-49 years	65	9.2
50 years+	156	22.0
Education
No formal education	92	13.0
Primary	107	15.1
Secondary/higher	509	71.9
Marital status
Single	143	20.2
Married/cohabiting	429	60.6
Separated/divorced	79	11.2
Widowed	57	8.0
Occupation
Agriculture	152	21.5
Self-employed	372	52.5
Housewife	30	4.3
Civil/government/private	81	11.4
Other	73	10.3
Comfortable: reading and receiving SMS
No	242	34.2
Yes	466	65.8
Comfortable: writing and sending SMS texts
No	277	39.1
Yes	431	60.9
Anticipated response to messages delivered to mobile phone
Same day	585	82.6
Later	123	17.4

Mobile phone ownership and stated willingness to receive text messages for breast examination

Most women owned a mobile phone (98.2%). The results also reveal that about 12.7% of those who owned a mobile phone regularly shared it with one or more other people (such as a partner, family member, or friend). Almost all of the participants expressed willingness to receive text messages about breast cancer and its screening (93.9%) and set up appointments for CBE (93.8%) (see Table 2).

Table 2.

Mobile phone ownership and stated willingness to receive text messages for breast examination.

Variables	Frequency	Percentage
Owns a mobile phone
No	13	1.8
Yes	695	98.2
If you own the mobile phone, do you regularly share your phone with one or more other people (such as a partner, family members, or friend)?
No	607	87.3
Yes	88	12.7
Willingness to receive text messages on your mobile phone about setting up appointment for CBE
No	44	6.2
Yes	664	93.8
Willingness to receive text messages on your mobile phone about breast cancer and its screening
No	43	6.1
Yes	665	93.9

Types of mobile devices owned

The study revealed that cell phones (55.9%) and smartphones (51.2%) were the most commonly used mobile devices among participants (Figure 2).

Mobile phone usage dynamics

Regarding mobile phone use, 95.5% of respondents use their phones every day. For web search content, 9.3% searched for information on healthy living; 3.7% searched for breast cancer signs and symptoms, 2.5% looked up breast screening information, 2.0% searched for cancer prevention methods, and 1.1% sought information on breast cancer treatment or medication (Table 3).

Table 3.

Mobile phone usage dynamics.

Indicators	Frequency (n)	Percentage (%)
Frequency of mobile phone use
Never	14	1.9
Less than once a week	13	1.8
At least once a week	5	0.7
Every day	676	95.5
Contents searched on website
Breast cancer and its signs and symptoms
No	682	96.3
Yes	26	3.7
Breast screening
No	690	97.5
Yes	18	2.5
How to prevent cancer
No	694	98.0
Yes	14	2.0
Breast cancer treatment or medication
No	700	98.9
Yes	8	1.1
Healthy living
No	642	90.7
Yes	66	9.3

Figure 3 shows the reasons for accessing websites. The findings indicate that only a few participants accessed health-related information via a smartphone website. Specifically, fewer than 1 in 10 participants reported accessing websites for information on breast screening (9.7%), tracking health behavior (7.6%), or obtaining information on breast cancer and other cancers (3.1%) (Figure 3).

Distribution of women’s stated willingness to receive an SMS-based mobile health intervention

Women aged 25-29 years show the highest willingness for both receiving messages about breast cancer and its screening (97.3%) and setting up appointments (96.7%) with p-values <0.001 (see Table 4). Education-wise, those with secondary/higher education are most willing (96.1% and 96.3%, respectively; p-values <0.001). Among marital statuses, singles show the highest willingness to receive screening messages (95.1%) and to schedule appointments (96.5%), although these differences are not statistically significant (p-values = 0.860 and 0.266). By occupation, the civil/government/private sector shows the highest willingness to both messages (98.8% for both; p-values 0.010 and 0.001). Those comfortable with reading and receiving SMS (99.1% and 98.7%; p-values <0.001) and writing and sending SMS (99.1% and 98.8%; p-values <0.001) are most willing. Women who shared their mobile phones with others had a higher willingness to receive screening messages (96.8%) and to schedule appointments (95.7%), although these differences were not statistically significant.

Table 4.

Bivariate distribution of women’s stated willingness to receive an SMS-based mobile health intervention.

Variables	Yes, willing to receive text messages on your mobile phone about breast cancer and its screening. n (%)	p-values	Yes, willing to receive text messages on your mobile phone about setting up appointment for clinical breast examination. n (%)	p-values
Age categories		<0.001		<0.001
25-29 years	177 (97.3)		176 (96.7)
30-34 years	123 (96.8)		124 (97.6)
35-39 years	95 (95.0)		94 (94.0)
40-44 years	75 (96.1)		75 (96.2)
45-49 years	63 (96.9)		63 (96.9)
50 years+	132 (84.6)		132 (84.6)
Education		<0.001		<0.001
No formal education	78 (84.8)		77 (83.7)
Primary	98 (91.6)		97 (90.6)
Secondary/higher	489 (96.1)		490 (96.3)
Marital status		0.860		0.266
Single	136 (95.1)		138 (96.5)
Married/cohabiting	403 (93.9)		402 (93.7)
Separated/divorced	73 (92.4)		71 (89.9)
Widowed	53 (93.0)		53 (92.9)
Occupation		0.010		0.001
Agriculture	134 (88.2)		132 (86.8)
Self-employed	353 (94.9)		353 (94.9)
Housewife	29 (96.7)		29 (96.7)
Civil/government/private	80 (98.8)		80 (98.8)
Other	69 (94.5%)		70 (95.9)
Comfortable: reading and receiving SMS		<0.001		<0.001
No	203 (83.9)		204 (84.3)
Yes	462 (99.1)		460 (98.7)
Comfortable: writing and sending SMS texts		<0.001		<0.001
No	238 (85.9)		238 (85.9)
Yes	427 (99.1)		426 (98.8)
Anticipated response to messages delivered to mobile phone		<0.001		<0.001
Same day	575 (98.3)		572 (97.8)
Later	90 (73.2)		92 (74.8)
Shared mobile phone with others		0.217		0.412
No	575 (93.5)		575 (93.5)
Yes	90 (96.8)		89 (95.7)

Factors associated with stated willingness to receive an SMS-based mobile health intervention

Respondents who anticipated an early (same-day) response to messages had 13.91 times higher odds of being willing to receive text messages about breast cancer and screening (AOR = 13.91; 95%CI: 6.11, 31.65; p < 0.001) and 9.58 times higher odds of being willing to receive reminders for setting up CBE appointments (AOR = 9.58; 95%CI: 4.44, 20.66; p < 0.001) compared to late responders. Additionally, being comfortable with reading and receiving SMS was significantly associated with a willingness to receive general breast cancer texts (AOR = 10.93; 95%CI: 1.01, 117.70; p = 0.049), though its association with willingness to receive CBE appointment texts did not reach statistical significance (AOR = 3.92; 95%CI: 0.79, 19.53; p = 0.095) (Table 5).

Table 5.

Firth regression results on stated willingness to receive an SMS-based health intervention.

Variable	Willing to receive text messages on your mobile phone about breast cancer and its screening		Willing to receive text messages on your mobile phone about setting up appointment for CBE
Variable	Adjusted odds ratio (AOR [95%CI])	p-value	Adjusted odds ratio (AOR [95%CI])	p-value
Comfortable: reading and receiving SMS
Uncomfortable	Ref		Ref.
Comfortable	10.93 [1.01, 117.70]	0.049	3.92 [0.79, 19.53]	0.095
Anticipated response to messages delivered to mobile phone
Late	Ref		Ref.
Early (same day)	13.91 [6.11, 31.65]	< 0.001	9.58 [4.44, 20.66]	<0.001
Model statistics
Wald chi2(17)	75.45		73.24
Prob > chi2	<0.001		<0.001
Number of observations	708		708

Adjusted predicted probabilities of willingness to receive breast cancer text message interventions (N = 708)
Variables	Receive text messages about breast cancer and screening adjusted predicted probability (95% CI)	Receive breast cancer appointment reminders adjusted predicted probability (95% CI)
Comfort reading/receiving messages
Uncomfortable	88.8% (80.1, 97.5)***	91.2% (86.3, 96.0)***
Comfortable	98.7% (97.0, 100.0)***	97.5% (94.9, 100.0)***
Anticipated response to messages delivered to mobile phone
Late	81.4% (75.3, 87.4)***	82.6% (76.7, 88.6)***
Early	98.1% (96.9, 99.2)***	97.5% (96.2, 98.8)***

NB: Check supplementary file for complete table.

***p<0.001.

Predicted probabilities and marginal effects

When evaluating the absolute practical magnitude via adjusted predicted probabilities, respondents who anticipated an early response to messages had a 98.1% predicted probability of being willing to receive text messages about breast cancer screening, compared to 81.4% among late responders. Also, women who were comfortable reading and receiving SMS had a significantly higher predicted probability of willingness to receive information about breast cancer screening (98.7%) compared to those who reported being uncomfortable (88.8%). Participants who were comfortable reading or receiving messages had a predicted probability of 97.5% (95%CI: 94.9%, 100.0%) of being willing to receive breast cancer appointment notifications, compared with 91.2% (95%CI: 86.3%, 96.0%) among those who were uncomfortable (see Table 5).

Discussion

Mobile health interventions have the potential to promote breast cancer awareness, early detection and diagnosis of breast cancer and improve health outcomes.²⁷ This study represents an initial formative phase of a larger research program aimed at developing a mobile phone-based intervention to improve breast cancer awareness and promote uptake of CBE among women in Ghana. The study aimed to investigate women’s ownership and patterns of mobile phone use as well as their willingness to accept an SMS-based mobile health intervention to promote breast cancer awareness and increase CBE uptake in rural Ghana. We did not evaluate a breast cancer awareness or screening intervention; rather, the findings, including participants’ language preferences, literacy levels, and communication needs, will inform the design, cultural adaptation, and linguistic tailoring of SMS content and app-based messages in subsequent phases of the study. This approach will help ensure that future communication materials are appropriate, understandable, and accessible to the target population.

Our findings show high ownership of mobile phones among rural-dwelling women in Ghana, with cellphones being the most prevalent type. The proportion of mobile phone ownership (98.2%) is higher than rates reported in rural Uganda (76.1%).²⁸ We also observed that 9.7% and 7.6% of the participants accessed websites for information on breast screening and for tracking health behaviors, respectively. This suggests a potential to increase the use of mobile health interventions to promote breast cancer awareness and CBE uptake. It also indicates that, although current usage is relatively low, there is a substantial opportunity to expand and enhance mobile-based health education and related initiatives.

Although mobile phone ownership was nearly universal among participants, very few reported using their phones to search for breast cancer-related information. Specifically, only 3.7% had searched for information on breast cancer signs and symptoms, 2.5% for breast screening information, and 1.1% for breast cancer treatment and management. These findings suggest that digital access alone may not translate into active digital health information-seeking behavior. A plausible explanation could be that rural women may have limited awareness of reliable online breast cancer information sources, insufficient digital health literacy to identify and evaluate health information, or a preference for obtaining health information from trusted sources such as healthcare professionals, family members, community networks, and the media.^11,15 Constraints related to internet data costs, limited connectivity in rural areas, concerns regarding information credibility, and competing priorities for mobile phone use may also reduce health-related information-seeking behavior.^12–14 Importantly, these findings indicate that passive access to information through mobile technology is likely insufficient to promote breast cancer awareness and screening behaviors. Rather, proactive delivery of health information through SMS-based app messaging may be needed to overcome barriers to information-seeking and engagement. To maximize effectiveness, future intervention content should be culturally relevant and linguistically tailored to participants’ preferred languages (e.g., English and Twi), use clear and non-technical language appropriate for varying health and digital literacy levels, and incorporate locally meaningful terminology and examples. These considerations are expected to enhance comprehension, engagement, and uptake of breast cancer information and screening services among the target population.

We found that most women in rural Ghana were willing to use an SMS-based mobile health intervention to advance their breast cancer awareness and screening uptake. This result is consistent with James et al.,²⁹ who found that most women (69%) were willing to use a mobile health intervention. Our study further revealed that all women, irrespective of their age, demonstrated a high level of willingness to use an SMS-based mobile health intervention to enhance breast cancer screening. However, as individuals aged, their desire to use an SMS-based mobile health intervention declined. This is similar to an earlier study by James et al.,²⁹ which reported greater willingness among younger women (18-29 years) to use mobile health interventions. A similar pattern of willingness to use mobile health interventions has been reported in a systematic review,³⁰ which found that younger people are more comfortable using mobile technologies. Indeed, we expected younger women to be more willing to use a mobile health intervention due to their familiarity with technological tools and platforms.³¹

Our study further underscores the significant role of comfort level in receiving and reading text messages in influencing women to use SMS-based mobile health interventions for breast cancer awareness and screening uptake. Those who reported being comfortable with text messaging were significantly more likely to be receptive to such SMS-based mobile health intervention and to take proactive steps toward CBE. This finding is consistent with existing literature^32,33 emphasizing the importance of tailoring communication methods to individual preferences and comfort levels to convey health-related information effectively. From a psychological perspective, comfort with text messaging may serve as a proxy for a broader inclination toward digital health literacy. Women who are uncomfortable with mobile technologies may perceive mobile devices as less useful rather than as lacking in functionality. Nevertheless, there is a window of opportunity to change the narrative by implementing interventions such as brief training sessions on mobile phone usage to increase technology adoption and engagement among less tech-savvy populations.

This study has also shown that delayed response to text messages significantly reduces the odds of women being willing to receive mobile health informational messages about breast cancer and messages aimed at setting up CBE appointments. Specifically, women who anticipated an early response had an adjusted predicted probability of 97.5% of being willing to receive breast cancer appointments, whereas those preferring a late response had a predicted probability of 82.6%. This represents an absolute difference of approximately 14.9 percentage points, indicating substantially greater willingness among those favoring earlier communication. Thus, highlighting a critical potential barrier to the use of SMS-based mobile health interventions that must be addressed. This delayed responsiveness can disrupt communication within the mobile health platform,³⁴ leading to missed opportunities for breast cancer screening. Our findings, thus, underscore a need for any plans to develop an SMS-based mobile health intervention for breast cancer awareness and screening to include personalized messaging schedules, reminders, or motivation to enhance the prompt responses and active participation.

The finding that a proportion of participants share mobile phones raises important ethical and operational considerations for the design of breast cancer screening SMS interventions. In shared-device contexts, confidentiality cannot be assumed, and sensitive health information may be exposed unintentionally to partners or family members.³⁴ This creates potential risks related to stigma, misinterpretation of screening messages, or unintended disclosure of health concerns. Although willingness to receive and act on SMS reminders remained high among women who shared phones, the risk of inadvertent disclosure suggests that message content should be carefully designed to avoid explicit or sensitive wording that could reveal the purpose of the communication. Future SMS-based breast cancer programs should incorporate appropriate safeguards for participant privacy and autonomy. Women should provide explicit informed consent before enrolment into messaging programs, retain the option to withdraw at any time without consequence, choose their preferred language and timing of message delivery where feasible, and receive messages designed to minimize unintended disclosure of sensitive health information when phones are shared within households.

Implications for application development

As noted above, this study is part of a larger research project aimed at developing a mobile phone-based application to enhance breast cancer awareness and increase uptake of clinical breast examinations among women in Ghana. The present formative study assessed mobile phone ownership and patterns of mobile phone use among women in rural Ghana. We did not evaluate a breast cancer awareness or screening intervention; rather, the findings are intended to provide baseline and preliminary data to inform the design, development, and content of a mobile application for breast cancer awareness and screening in the subsequent phases of the larger study. An additional consideration is to define the breast cancer control pathway, which was beyond the scope of this study. However, our intended pathway includes (1) SMS-based mobile health education about breast cancer symptoms and risk; (2) prompting or booking a clinical breast examination; (3) a referral pathway for abnormal findings; (4) linkage to diagnostic imaging and biopsy; and (5) follow-up after referral. The findings of the present study reflect women’s readiness for SMS-based mobile health intervention. This suggests that such interventions may be a feasible way to deliver breast cancer awareness information and appointment reminders about CBE to women in rural Ghana. The high prevalence of mobile phone ownership and the strong stated willingness to receive breast cancer-related messages indicate that mobile phones could serve as an accessible communication platform for future breast cancer awareness initiatives. However, intervention developers should consider participants’ digital literacy levels, particularly their comfort with reading, receiving, and responding to SMS messages, as these factors were strongly associated with willingness to engage with mobile health communications. Future studies should move beyond assessing willingness and evaluate the implementation, acceptability, and effectiveness of SMS-based mobile health interventions in rural populations. Randomized controlled trials and implementation studies are needed to determine whether SMS-based education and appointment reminder systems can improve CBE uptake and breast cancer awareness.

Strengths and limitations

A cross-sectional design was appropriate for assessing ownership, usage patterns, and willingness to adopt SMS-based mobile health interventions; however, causal relationships and actual adoption behavior could not be determined. The data were self-reported and may therefore be subject to recall bias and social desirability bias, particularly regarding mobile phone usage patterns and willingness to receive breast cancer-related messages. A further limitation is that this formative study was not guided by a specific theoretical framework for health behavior change or technology adoption. Consequently, the study was primarily descriptive and exploratory in nature, focusing on mobile phone ownership and usage patterns rather than examining the theoretical determinants of breast cancer screening behaviors or mHealth acceptance. This should be considered when interpreting the findings. Future studies should consider incorporating relevant behavioral and technology adoption theories to strengthen intervention design and evaluation. This study used the PMA household listing as a sampling frame to identify eligible households within the 14 study communities in the Onwe catchment area. Although the original PMA platform uses a probability sampling design with sampling weights for nationally representative surveys, weights were not available for this study. Additionally, because all study communities were included, no further clustering or weighting adjustments were applied in the analyses. Therefore, while the findings are applicable to women recruited from the participating communities, they should not be interpreted as representative of all women in Onwe, the Ashanti Region, or Ghana. Given that the analysis was restricted to participants with complete information on key mobile phone variables, the exclusion of participants with missing data may have preferentially retained women who were more engaged with mobile phone technology, potentially leading to modest overestimation of mobile phone ownership and willingness to receive SMS-based health messages. Because the mechanism underlying missingness could not be established, some degree of complete-case selection bias cannot be ruled out. Furthermore, because missingness primarily affected key exposure variables, multiple imputation and formal sensitivity analyses were not considered appropriate, and the findings should therefore be interpreted as complete-case estimates. Nonetheless, this study is arguably the first of its kind in Ghana and provides novel findings to advance the development of mobile health interventions for breast cancer awareness and CBE uptake. These findings may inform mobile health planning in similar rural communities in Ghana, but further studies across diverse regions are needed before national generalization.

Conclusion

The findings suggest a high stated willingness to receive SMS-based breast cancer information and appointment-related messages among women in these rural communities. Our findings highlight the importance of considering individual preferences and literacy levels when designing such interventions. A personalized approach may support future efforts to improve breast cancer screening uptake if demonstrated to be effective in subsequent intervention studies. The higher odds of engagement among those who responded early to messages suggest that timeliness may be an important design consideration. Therefore, any plans to develop a mobile health intervention for breast cancer awareness and screening should focus on creating systems that allow for prompt responses and follow-up, ensuring messages are delivered and acted upon quickly. This might involve automated reminders or interactive messaging platforms that promote immediate responses and actions. Future intervention studies are needed to assess actual engagement, appointment completion, screening uptake, and clinical impact.

Supplemental material

Supplemental material - Informing mobile health strategies: Study of mobile phone ownership among rural women for breast cancer control in Ghana

Supplemental material for Informing mobile health strategies: Study of mobile phone ownership among rural women for breast cancer control in Ghana by Adwoa Bemah Boamah Mensah, Kofi Boamah Mensah, Anita Eseenam Agbeko, Aliyu Mohammed, Joshua Okyere, Collins Dodd, Keith Horvath, Beth Virnig, Joe-Nat Clegg-Lamptey, and Shalini Kulasingam in DIGITAL HEALTH.

Supplemental material

Supplemental material - Informing mobile health strategies: Study of mobile phone ownership among rural women for breast cancer control in Ghana

Footnotes

Acknowledgements

We thank our participants for their willingness to take part in this study. We also acknowledge the funding support from the American Association of Cancer Research (AACR) and the US National Cancer Institute Centre for Global Health (NCI/CGH) through the 2022 Beginning Investigator Grant for Catalytic Research (BIG Cat) mechanism. This initiative is by the African Organization of Research and Training in Cancer (AORTIC).

ORCID iD

Adwoa Bemah Boamah Mensah

Ethical considerations

The study was conducted in accordance with the Declaration of Helsinki. Ethical approval was sought and obtained from the Ghana Health Service Ethics Review Committee (GHS-ERC: 016/01/23).

Consent to participate

All participants provided written consent. The participants were provided with an information sheet detailing the study’s purpose, duration, potential benefits, risks, and discomfort.

Author contributions

ABBM: Conceptualization, methodology, investigation, data curation, formal analysis, funds acquisition, project lead and administration, writing – original draft, writing – review & editing. KBM: Conceptualization, methodology, investigation, data curation, formal analysis, funds acquisition, writing – review & editing. AEA: Methodology, investigation, data curation, formal analysis, funds acquisition, writing – review & editing. AM: Methodology, investigation, data curation, formal analysis, writing – original draft, writing – review & editing. JO: Methodology, investigation, data curation, formal analysis, software; writing – original draft, writing – review & editing. KH: Conceptualization, Methodology, formal analysis, project mentor and supervision, writing – review & editing. BV: Conceptualization, methodology, project mentor and supervision, writing – review & editing. JCL: Conceptualization, methodology, investigation, data curation, formal analysis, project mentor and supervision, writing – review & editing. SK: Conceptualization, methodology, investigation, data curation, formal analysis, project mentor and supervision, writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study is part of a larger research project supported by the “2023 Beginning Investigator Grant for Catalytic Research (BIG Cat), Grant Number 22- 15-75-MENS”.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The datasets supporting the conclusions of this article are publicly available in the Figshare repository [10.6084/m9. figshare.31082404].*

Supplemental material

Supplemental material for this article is available online.

Appendix

References

Sung

Ferlay

Siegel

, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021; 71: 209–249. https://doi.org/10.3322/caac.21660

Anyigba

Awandare

Paemka

. Breast cancer in sub-Saharan Africa: The current state and uncertain future. Experimental Biology and Medicine 2021; 246: 1377–1387. https://doi.org/10.1177/15353702211006047

Ssentongo

Agbese

, et al. Prevalence and determinants of breast cancer screening in four sub-Saharan African countries: a population-based study. BMJ open 2020; 10: e039464. https://doi.org/10.1136/bmjopen-2020-039464

Ferlay

Ervik

Lam

, et al. Global cancer observatory: cancer today. International agency for research on cancer, 2020, p. 20182020.

Lei

Ploner

Elfström

, et al. HPV vaccination and the risk of invasive cervical cancer. New England Journal of Medicine 2020; 383: 1340–1348. https://doi.org/10.1056/NEJMoa1917338

Ginsburg

Yip

Brooks

, et al. Breast cancer early detection: A phased approach to implementation. Cancer 2020; 126: 2379–2393. https://doi.org/10.1002/cncr.32887

Scheel

Giglou

Segel

, et al. Breast cancer early detection and diagnostic capacity in Uganda. Cancer 2020; 126: 2469–2480. https://doi.org/10.1002/cncr.32890

Mensah

ABB

Mensah

Aborigo

, et al. Breast cancer screening pathways in Ghana: applying an exploratory single case study methodology with cross-case analysis. Heliyon 2022; 8: 1-13.

Okyere

Dey

NEY

Owusu Ansah

, et al. Prevalence and determinants of breast self-examination practices among women in their reproductive age in Namibia: an analysis of the 2013 Namibia demographic and Health Survey. BMC Public Health 2023; 23: 30. https://doi.org/10.1186/s12889-023-14985-5

10.

Agyemang

Tei-Muno

Dzomeku

, et al. The prevalence and predictive factors of breast cancer screening among older Ghanaian women. Heliyon 2020; 6: e03838. https://doi.org/10.1016/j.heliyon.2020.e03838

11.

Osei-Afriyie

Addae

Oppong

, et al. Breast cancer awareness, risk factors and screening practices among future health professionals in Ghana: A cross-sectional study. PloS one 2021; 16: e0253373. https://doi.org/10.1371/journal.pone.0253373

12.

Wang

Gregg

Qiu

, et al. Breast cancer screening for patients of rural accountable care organization clinics: a multi-level analysis of barriers and facilitators. Journal of community health 2018; 43: 248–258. https://doi.org/10.1007/s10900-017-0412-x

13.

Kathrikolly

Shetty

Nair

. Opportunities and barriers to breast cancer screening in a rural community in Coastal Karnataka, India: a qualitative analysis. Asian Pacific Journal of Cancer Prevention: APJCP 2020; 21: 2569–2575. https://doi.org/10.31557/APJCP.2020.21.9.2569

14.

Mena

Wiafe Addai

Sauvaget

, et al. Evaluation of the impact of a breast cancer awareness program in rural Ghana: a cross‐sectional survey. International journal of cancer 2014; 134: 913–924. https://doi.org/10.1002/ijc.28412

15.

Wuur

Duodu

Tarkang

. Factors that influence breast cancer screening among women of reproductive age in the Nandom Municipality, Ghana. BMC Women's Health 2022; 22: 359. https://doi.org/10.1186/s12905-022-01946-0

16.

Afaya

Seidu

A-A

Sang

, et al. Mapping evidence on knowledge of breast cancer screening and its uptake among women in Ghana: a scoping review. BMC Health Services Research 2022; 22: 526. https://doi.org/10.1186/s12913-022-07775-z

17.

Eden

Scariati

Klein

, et al. Mammography decision aid reduces decisional conflict for women in their forties considering screening. Journal of Women's Health 2015; 24: 1013–1020. https://doi.org/10.1089/jwh.2015.5256

18.

Heo

Chun

Lee

, et al. Effects of a smartphone application on breast self-examination: a feasibility study. Healthcare informatics research 2013; 19: 250–260. https://doi.org/10.4258/hir.2013.19.4.250

19.

Bonful

Addo-Lartey

Sefenu

, et al. Developing a culturally tailored short message service (SMS) intervention for improving the uptake of cervical cancer screening among Ghanaian women in urban communities. BMC women's health 2022; 22: 154. https://doi.org/10.1186/s12905-022-01719-9

20.

GhanaStatistical Services . Ghana2021 Population and Housing. Preliminary Report. GhanaStatistical Services, 2021.

21.

Ministry of Health Ghana . National Strategy for Cancer Control in Ghana2012-2016. Health Mo, 2012.

22.

Zimmerman

Olson

Group

PPI

, et al. PMA2020: Rapid turn around survey data to monitor family planning service and practice in ten countries. Studies in family planning 2017; 48: 293–303. https://doi.org/10.1111/sifp.12031

23.

Performance Monitoring and Accountability 2020 (PMA2020) Project aKNUoSaTK . Detailed Indicator Report: Ghana 2013. GhanaStatistical Services, 2013, PMA2020.

24.

Stocks

Ibrahim

Park

, et al. Mobile phone ownership and use among women screening for cervical cancer in a community-based setting in western Kenya: Observational study. JMIR public health and surveillance 2022; 8: e28885. https://doi.org/10.2196/28885

25.

Olamoyegun

Emuoyibofarhe

Ala

, et al. Mobile phone use in the management of diabetes in Nigeria: a new potential weapon. West Afr J Med 2020; 37: 201–208.

26.

Harris

Taylor

Minor

, et al. The REDCap consortium: building an international community of software platform partners. Journal of biomedical informatics 2019; 95: 103208. https://doi.org/10.1016/j.jbi.2019.103208

27.

Lee

Ghebre

, et al. Mobile phone multilevel and multimedia messaging intervention for breast cancer screening: pilot randomized controlled trial. JMIR mHealth and uHealth 2017; 5: e7091. https://doi.org/10.2196/mhealth.7091

28.

Leahy

Abomoslim

Booth

, et al. Women’s cellphone access and ownership in rural Uganda: implications for self-care interventions. BMC Global and Public Health 2024; 2: 8. https://doi.org/10.1186/s44263-024-00038-5

29.

James

Harville

Whitehead

, et al. Willingness of African American women to participate in e-Health/m-Health research. Telemedicine and e-Health 2016; 22: 191–197. https://doi.org/10.1089/tmj.2015.0071

30.

Borghouts

Eikey

Mark

, et al. Barriers to and facilitators of user engagement with digital mental health interventions: systematic review. Journal of medical Internet research 2021; 23: e24387. https://doi.org/10.2196/24387

31.

Gallagher

Roach

Sadler

, et al. Mobile technology use across age groups in patients eligible for cardiac rehabilitation: survey study. JMIR mHealth and uHealth 2017; 5: e8352. https://doi.org/10.2196/mhealth.8352

32.

Wang

Wei

, et al. Smartphone interventions for long-term health management of chronic diseases: an integrative review. Telemedicine and e-Health 2014; 20: 570–583. https://doi.org/10.1089/tmj.2013.0243

33.

Fischer

Chadyiwa

Tshuma

, et al. Acceptability of mobile health interventions to increase diabetic risk factor awareness among the commuter population in Johannesburg: descriptive cross-sectional study. JMIR diabetes 2019; 4: e12600. https://doi.org/10.2196/12600

34.

Danquah

Hasham

MacFarlane

, et al. Use of a mobile application for Ebola contact tracing and monitoring in northern Sierra Leone: a proof-of-concept study. BMC infectious diseases 2019; 19: 1–12. https://doi.org/10.1186/s12879-019-4354-z

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.20 MB

0.43 MB