Mothers Intention to Use Mobile Phone Text Message Reminders for Routine Child Vaccination in Northwest, Ethiopia

Background: The immunization program in most developing countries including Ethiopia is challenged by lack of effective methods to track vaccination schedules. With the unprecedented penetration of mobile devices in developing world, mHealth applications are being leveraged for different disease domains. Among the different factors that affect the use of mHealth interventions is the intention of end users to use the system. In this research, we aimed to fill the evidence gap by investigating mother’s intention to use text-message reminders for vaccination in north-west, Ethiopia. Methods: A cross-sectional study was conducted among 460 mothers presenting with their infants for vaccination at a health facility from October 1 to 26, 2018. A systematic random sampling technique was applied to select study participants. Data were collected using validated instrument. Descriptive statistics were computed and binary logistic regression analyses was used to assess factors associated with the outcome variable. The regression results were reported as AORs with their 95% CIs. Results: Of 456 mothers included for analysis, 360 (78.9%) with 95% CI: (74.9% - 82.4%) of mothers have intention to use mobile phone text message reminders for vaccination. Of these, 270 (75%) also wanted to receive the reminders a day before the vaccination due date for the child. The preferred language for receiving reminders was Amharic language (58.1%). Mothers aged 35 years and above [AOR = 0.352; 95%CI: 0.149- 0.833], secondary education and above [AOR= 4.428; 95% CI: 2.046- 9.580], mobile phone use for more than two years [AOR= 3.627; 95%CI: 1.657-7.938], perceived usefulness [AOR: 6.372; 95%CI: 3.128-12.981] and perceived ease of use [AOR: 3.847; 95%CI: 2.061- 7.180] were predictors of intention to use mobile phone text message reminders for vaccination. Conclusion: Majority of mothers have intention to use mobile phone text message reminders for child vaccination. Mother’s age, education, duration of mobile phone use,

3 resource limited setting of Ethiopia.

Background
Timely completion and uptake of the childhood vaccination is key to reducing the high morbidity and mortality of vaccine-preventable diseases among infants globally. However, many children still miss scheduled vaccines in the Extended Program of Immunization (EPI) or are being vaccinated after the recommended ages (1,2). Adherence to childhood vaccination schedules is a function of various factors including the information gaps both from the service supply and demand sides (3).
The immunization program is also challenged with lack of effective methods to track vaccination schedules (4). Immunization programs usually involve the use of the child health card as a tool for reminding caregivers of children of the dates of their next vaccination (5). However, it was observed that majority of the mothers who missed their vaccination appointment were due to forgetfulness and difficulty in tracking vaccination schedules indicating a need to identify more innovative approaches (6)(7)(8).
With the continuous growth of mobile network coverage and unprecedented penetration of mobile devices, leveraging mHealth applications in health sector is becoming popular (9). By the end of 2018, 5.1 billion people around the world subscribed to mobile services, accounting for 67% of the global population (10). In the same year, the mobile subscriber's in Ethiopia reached 44% (11). A study conducted on mobile phone access in Gondar city among pregnant women reported that 76.7% of mothers owned a mobile phone (12).
Mobile health (mHealth) is the use of mobile phone technology to deliver health care (13). mHealth is now extensively used in healthcare and there is a growing global trend in harnessing this technology for behavior change, disease surveillance, prevention and control of various health problems and enhancing attendance for health services. Hence, the field of mHealth, has been proposed as a potential solution to many of the challenges that developing countries face (14)(15)(16)(17)(18)(19)(20)(21)(22). mHealth applications and programs make use of several aspects of mobile technology such as text messaging, voice and video services (13). Short message service (SMS) is widely applicable appointment reminder intervention to improve health care seeking behaviors considering participant characteristics such as forgetfulness and lack of knowledge (23)(24)(25). Text messaging is a potential tool for health care improvement for several reasons; applicability on almost every model of mobile phone, relatively low cost, widespread, does not require technological expertise, and widely applicable to a variety of health behaviors and conditions (26,27).
When introducing mobile phone based health services, it is critical to study the factors that influence user intention to adopt those services (28). Hence, investigating user's intention, preference and engagement is crucial (29)(30)(31)(32)(33)(34) to design more effective mHealth interventions (35,36). Cultural and contextual differences should also be considered when developing text message reminder systems (28,37,38).
Although text messaging has expanded in popularity worldwide, there is limited evidence on user's intention and preference in Ethiopia, particularly in the study area. Therefore, this study aimed to assess the intention of mothers and its associated factors to use of mobile phone text message appointment reminders for routine vaccination in Gondar city, north-west, Ethiopia.

Study design and setting
A health-facility based cross sectional study was conducted from October 1 to October 26, 2018 in Gondar city administration, northwest, Ethiopia. Gondar city administration has a total of 24 Keble's (13 urban and 11 rural). The city administration had an estimated total population of 390, 644. Of which, 12,149 were under one year of age. The city administration has also a total of 23 public health facilities (one comprehensive specialized hospital, eight health centers and fourteen health posts (39).

Source and study populations
The source population consisted of mothers paired with infants attending the vaccination units at health facilities. The study population included those eligible mother-infant pairs who visited the selected health facilities during the study period.

Inclusion and exclusion criteria
Those mothers of infants who visited vaccination units of health facilities and remaining with at least one vaccination appointment were included. Mothers who resided in the study area for at least 6 months prior to the study period and who owned mobile phone were included for this particular study.
Mothers whose infants had already received the last doses of vaccines were excluded from the study.

Sample size and sampling procedures
We could not find any study conducted in Ethiopia to determine the intention of mothers to use text message reminders for routine vaccination. Therefore, we did pilot study to determine the proportion of those mothers who have intention to use the text message reminders and it was found to be 77.6%. Finally, the sample size required for this particular study was calculated by considering the following assumptions; proportion of intention to use text message reminder for child vaccination as 77.6% (from pilot study), 95% CI and 4% margin of error. With this assumptions, the sample size was 418. Taking 10% non-response rate, the final sample size was 460.
All the eight health centers and the comprehensive specialized hospital in Gondar city were included for this study. The sample in each health facility were allocated proportional to the number of clients who attended vaccination schedules in the same period of the previous year. A systematic random sampling technique was applied to select the study participants.

Study variables
The outcome variable was intention to use text message reminder for vaccination. Based on the technology acceptance model; perceived ease of use and perceived usefulness were considered as predictor variables for this study (40). Additionally, the socio-demographic characteristics of mothers were included as predictors for the outcome of interest.
Intention to use mobile text message reminders was defined as the user's likelihood to use mobile phone text message reminders for child vaccination. (28,29,(41)(42)(43). Perceived ease of use was defined as the extent to which a person believes that using a particular system (in this case the text message reminder) would be free from effort (28,29,41,42). Perceived usefulness was defined as the degree to which a person believes that using a particular system (in this case the text message reminder) would enhance his or her task (in this case timely vaccination of children) (28,29,41,42).
All items were measured on a 5-point likert-type scale ranging from "strongly disagree" (score 1) to "strongly agree" (score 5). Item scores for each composite variable were summated and averaged to create variable scales for data analysis (44,45). Finally the composite variables were dichotomized based on the average scores taking 3 as cut of point. Hence, those variables with average score of higher than 3 were categorized as "Yes" while those variables with average score lower than or equal to 3 were categorized as "No" (46,47).
The household socio-economic status was created by principal components analysis (PCA), including variables on asset ownership, housing characteristics and ownership of animals and farming. This was done for urban and rural households separately. Urban and rural households PCA loading scores were merged for household wealth index classification using quintiles. Finally, the merged scores for urban and rural were divided into three quintiles as poor, middle and rich households (48).

Data collection tools and procedures
Data collection instrument for this study was adapted from the scales used in the technology acceptance model (TAM) which has four major variables: Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Behavioral Intention (BI) and actual use. The scales perceived usefulness and perceived ease of use were adapted from Davis's study (42) and the scale intention to use was adapted from Venkatesh et al's study (46) to fit the study context. During adaption, the data collection instrument underwent forward and backward translation. Face and content validity of the data collection instrument was assessed by six experts and the proposed changes from the expert panels were considered for refinement of the data collection instrument.
Then, prior to the actual data collection, pilot study was done out of the study area, in health facilities of Bahir Dar city administration with a sample size of 100. The results of the pilot study were used to assess the validity and reliability of the data collection instrument. The internal consistency for each dimension of the data collection instrument was checked using Cronbach's alpha and scores on perceived usefulness (Cronbach alpha = 0.95), perceived ease of use (Cronbach alpha = 0.91) and intention to use text message reminders (Cronbach alpha = 0.93) were deemed acceptable.
Finally, nine data collectors and three supervisors were recruited for the actual data collection.
Training was given to data collectors and supervisors on the data collection tools and procedures.
Mothers who visited the selected health facilities for their infant's vaccination were approached for interviews. Face to face interview technique was used to collect data from eligible study participants using the validated data collection instrument.

Data processing, analysis and parameter estimation methods
The data were checked for completeness, entered into Epi-data version 3.1 and exported to STATA version 14 software for analysis. Descriptive statistics on frequencies and percentages were computed and have been presented using graphs and tables. Binary logistic regression analysis model was used to identify the predictor variables for intention to use text message reminders for child vaccination. Those independent variables which had p-value of less than 0.2 during the bivariable logistic regression analysis were considered for the multivariable logistic regression analysis. Then, the results were reported as Adjusted Odds Ratio (AOR) with their 95% CIs.
The presence of multicollinearity was checked among independent variables using Variance Inflation

Ethical considerations
Ethical approval was obtained from the University of Gondar ethical review board before the commencement of the study. Study permission was sought at all levels of governmental administration systems including health offices and health facilities. Informed written consent was obtained from each study participant. Moreover, confidentiality assurance was provided to study participants on any information provided by them and their privacy was kept during data collection.

Socio-demographic characteristics
This study considered a sample size of 460 with a response rate of 99.1%. As shown in Table 1, 260 (57%) mothers belonged to an age group of 25-34 years. Majority of the mothers were currently married (90.8%), orthodox by religion (87.5%) and more than half (54.8%) of the mothers attained secondary education and above [ Table 1].
Pertaining to occupation of the mothers, the largest 263 (57.7%) of the total mothers were housewife's followed by merchants 89 (19.5%). The study also indicated that vast majority (93.9%) of the mothers were resided in urban kebele's [ Table 1].

Predictors of intention to use text message reminders for child vaccination
The bivariable analysis indicated that age, mother's education, mother's occupation, marital status, household wealth index, place of residence, duration of mobile phone use, mobile phone type, perceived ease of use and perceived usefulness were significantly associated (p < 0.2) with intention of mothers to use text message reminders for child vaccination. In the final multivariable logistic regression model the variables age, mother education, duration of mobile phone use, perceived ease of use and perceived usefulness were found to have statistically significant association with intention of mothers to use text message reminders for vaccination.
As shown in In the final multivariable model, marital status, occupation, household wealth index and the type of mobile phone mothers are currently using didn't have a statistically significant association with intention of mothers to use mobile phone text message reminders for child vaccination. This study indicated that majority of the mothers have intentions to use text message reminders for child vaccination. This finding corroborated findings from a study in Lagos Nigeria where most mothers indicated a willingness to receive reminder text messages for vaccination appointments through mobile phones (50). A willingness study on pregnant women in Gondar city also reported consistent finding where around three fourth of women were willing to receive text messages (12).
However, this finding was slightly higher than a finding from another study in Nigeria (35). On the other hand, this finding is lower than a study finding from Kenya (9). The difference might be due to the difference in the ICT infrastructure and investment in digitalization across countries that can potentially be reflected among the study participants intention to use mobile technologies for health service.
In this study, older mothers were less likely to use text message reminders for child vaccination in the future that is consistent with studies conducted in Ethiopia among pregnant women (12) and ART patients (51). A study from China also indicated that young individuals are more open to new technologies (52). Evidences also indicated that younger population has higher tendency of interacting to new technological products including mHealth with more control over ICT to adapt to the operating procedures (53,54).
Educational status of mothers was also positively associated with their intention to use mobile phonebased text message reminders for child vaccination. This finding is in accordance with other studies (6,30,50,52,55). This may be explained by the fact that educated women are likely to be aware of incoming text messages and are likely to read and act on the received messages promptly. Evidence also showed that literacy status was shown to be an issue in text message reminder system implementation that has to be addressed when text message reminder system is being planned for implementation (26). A potential drawback to implementing a mobile phone based text messaging intervention is that it requires the recipient to have an adequate level of literacy, marginalizing some population groups who could potentially benefit from the mHealth intervention (22). In our study population this could affect around 14% of women having no formal education.
In this study duration of mobile phone use was found to positively affect intention of mothers in receiving text message reminders for child vaccination. Another study also indicated that offering experienced people an opportunity to get acquainted with an e-Health application will increase their intention to use e-Health in future (55). These days, as clients often use mobile phones in their daily activities for longer duration, it may reflect their intention towards the usage of their mobile phones for additional services including health care (56).
The study also found that perceived usefulness has a positive significant association with intention of mothers to use text message reminders for child vaccination. This finding is consistent with other studies where users who did not believe in the possible advantages of e-Health were less inclined to use e-Health (52,55,57,58). If the users perceived that the new system can provide useful health information and improve performance, then this usefulness will directly lead to the intention to use the proposed system (56,57,59). End users need to perceive the system as being useful or they will not attempt to use it regardless of how easy or difficult it is to use. Therefore, during system development, there is a need to ensure that the system will improve the intended health outcomes (26,56).
The findings also showed that perceived ease of use was positively associated with intention of mothers to use text message reminders for child vaccination. This finding corroborates with findings of other studies (45,52,55,57). When users have no or little previous experience of using a system, they usually pay more attention to the system's ease of use. This implies that users would be unwilling to use a new mHealth service regardless of how useful the system would be, if they perceive it to be difficult to use. Research also showed that users will stop using mHealth interventions that are not user friendly (45). Difficulty in using a new system could be solved if the user thinks that the system will be useful to them. One study reported that training users on the new mobile health technology improves perceived ease of use and, thereby, increases intention to use the actual system (45). Hence, deployment of new mHealth service may require extra guidance on how to operate and use the new system for improved implementation (26,56).
Mobile services are mainly designed for individual users, who may have different expectations and needs in accordance with their preferences. In order to develop an effective text message reminder system for vaccination, parental preferences must be fully understood and taken into consideration (54). In this study, more than half of mothers would like to receive the text message reminders in Amharic (local) language. This finding is consistent with evidence from India (27). In contrary, from studies in Nigeria (8,30,50) and Kenya (60) majority of the mothers preferred English language for reminders on their mobile phones which could be attributed to their high literacy levels.
For successful implementation of mHealth programs, clients should be able to choose when and how frequently they would receive text messages (22). The findings of this study indicated that majority of mothers preferred to receive one text message reminder per each vaccination appointment. The optimal timing most preferred by mothers for receiving the text message reminders is the day preceding the vaccination appointment date which corroborates the findings in other studies (6,35,50). This might be due to the fact that sending text messages to mothers one day before their vaccination appointments will increase the chances of the messages being seen and help them to get prepared for their child vaccination appointments ahead of time.
The time of day at which mothers would receive text messages varied greatly. In this study afternoon is the most preferred time of the day to receive text message reminders which is consistent with another study (54). On the other hand in a study from Nigeria, about two fifths of mothers were willing to receive reminder text messages any time of the day (4).
This study also showed that marital status, mother's occupation, household wealth index and type of current mobile phone were not found to have a statistically significant association with intention of mothers' to use mobile text messages for child vaccination. In another study it was also reported that the type of mobile phone didn't have significant association with intention to use SMS reminders (12).
Thus, the type of mobile phone the mother had and the differences in economic status would not be a major challenge for implementing text message reminder interventions for child vaccination. Though it didn't have significant effect in another study (12) the variable place of residence has been removed from the final model due to its multicollinearity effect.

Implications for practice and research
This study has practical implications in particular for immunization program managers. Given the high proportion of mothers who had intention to use mobile phone-based text message reminders for vaccination, incorporating mobile text messages is a promising avenue to strengthen the routine immunization program in Ethiopia. If designed appropriately by considering users preference in terms of frequency, timing and language; text message based mHealth interventions may be an innovative way for engaging users in care for improved child vaccination outcomes. The study also provides a basis for further interventional studies that can develop and assess effectiveness of mobile text messaging interventions as a tool to improve the routine immunization program in Ethiopia.

Limitations
The findings of this study should be interpreted in light of some limitations. First, we studied intention for text message based appointment reminder for those who already has mobile phone and visiting vaccination units of health facilities in north-west, Ethiopia. So, the findings may not be generalizable to the population of entire country particularly for those residing in rural areas.
As most mHealth programs focus on those with mobile phones, a potential drawback to the use of mobile phone based text-message-reminders is potential marginalization of certain populations, such as those that do not have a mobile phone. In the study area, it has been reported in another study that three-fourth of pregnant women owned a mobile phone (12) who actually are eligible for mHealth implementation. Finally, supplying mobile phones for those mothers who do not have mobile phones would not be a sustainable approach during actual implementation of mHealth interventions.
However, these limitations may be reduced as mobile technology advances and mobile subscriptions grow in developing countries. In addition, further study would be helpful to assess the needs of the population that is not eligible for mHealth interventions and explore available methods to reach them.
Secondly, the responses might have been affected by bias introduced by the interviewers. To reduce this bias, we trained the interviewers and standardized the interviewing procedures.

Conclusion
In this study, we found that majority of mothers have intention to use mobile phone text message reminders for child vaccination. Most of the mothers also would like to receive the text message reminders in Amharic (local) language one day before the vaccination due date. Predictors of mothers' intention to use mobile phone text message reminders include mother's age, mother's education, duration of mobile phone use, perceived ease of use and perceived usefulness of the proposed system.
Considering these predictors and user's preference could inform the implementation strategy for use of the mHealth supported text messaging interventions in the resource limited setting of Ethiopia.
Program planners should also consider utilizing mobile text message reminders as a strategy to increase adherence to vaccination services. Taking in to account the findings of this study, development of automated mobile phone based text message reminder system and testing its effectiveness is recommended.  Figure 1 Intention, perceived usefulness and perceived ease of use in using text message reminders for vaccination, Gondar city administration, northwest, Ethiopia [N=456]