Exploring the feasibility of eHealth solutions to decrease delays in maternal healthcare in remote communities of Ghana

Skilled care attendance and MMR

The MMR in Ghana was reduced from 590 in 1995 to an estimated 358 in 2015. Hence, the MDG target to reduce MMR to 185 in 2015 has not been achieved [9]. Moreover, the MMR of non-institutional deaths remains unknown. It is estimated that more than 80% of maternal deaths occur in the communities or within 24 h of admission to a health facility [10]. Moreover, 80% of maternal deaths can be attributed to five treatable and preventable complications: hemorrhage, sepsis, eclampsia, obstructed labor, and complications of abortion [11]. The MMR in Kpando was 623 in 2010, reduced to 161 in 2013. However, it increased to 205 in 2014. The main causes were: cardiac failure, hemorrhagic shock, respiratory failure, and acute chest syndrome. Causes of mortality in Kpando and Ghana are different but might be related, e.g. cardiac failure can be caused by eclampsia [12], and obstructed labor and complications of abortion can lead to hemorrhagic shock [13]. It is difficult to compare MMR in Kpando with the rest of Ghana because of the lack of patient data available in Kpando: unlike health data elsewhere, the underlying conditions leading to maternal deaths in Kpando are unknown because only the last symptoms were reported [12, 13].

Deliveries supervised by SBAs are increasing in Ghana but there is a big disparity between urban and rural areas. For instance, the Greater Accra Region around the capital of Ghana recorded 92% of deliveries supervised by a SBA in 2014, in contrast, the Volta Region recorded one of the lowest of the country with less than 50% [9, 11]. On the other hand, the contraceptive prevalent rate (CPR) defined as “the proportion of women of reproductive age who are using (or whose partner is using) a contraceptive method at a given point in time” is the highest in the country, with 32% in the Volta Region ([9], pp.-44). In Kpando, the proportion of births attended by SBAs was increasing steadily from 70% in 2010, reaching a peak value of 87% in 2013. However, in 2014 there has been a decrease of 13% and an increase in deliveries attended by TBAs. The health directorate attributes this to the lack of midwives in most facilities [7].

The percent of women attending ANC for at least one visit was 90.9% across the country; however, in the Volta Region this number was below average, with 81.9% in 2013. In fact, only 67.7% of registrants attended all four visits (as recommended by WHO), but ANC attendance has actually increased in the region since 2013 [14]. In Kpando, most women, register for ANC services during the 2nd or 3rd trimester of pregnancy, and only 30% of registrants attended the all four visits.

EHealth in MHC

The biggest challenges in MHC are shortage of healthcare professionals, equipment, transportation, and health services at the municipal level. Other challenges are a lack of structure, a lack of regulation and inadequate data to assess maternal mortality issues [9]. Nevertheless, Ghana has high quality of care where this is available, a national health insurance system, and well-recognized facilities for medical education and accreditation [14].

On the other hand, the government struggles to extend these services to rural areas, even with the introduction of CHPS zones two decades ago. The program coverage is affected by logistical problems, supervisory lapses, and resource shortages. The Ministry of Health is trying to bridge the inequity gaps between rural and urban areas and rich and poor, looking for new ways to improve the outcome of investments made in the health sector. New policies such as free MHC and incentives for health workers to work in remote areas have been implemented [15, 16].

In 2003, as part of an attempt to optimize resources, the country started to consider eHealth as a way of improving healthcare [17]. Through this work, the term eHealth is used as defined by WHO in the WHA58.28 Resolution, i.e. “eHealth is the cost-effective use of information and communication technologies (ICT) in support of health and health-related fields, including health-care services, health surveillance, health literature, and health education, knowledge and research” ([18], pp.-1). EHealth has the potential to enhance healthcare systems’ capacity in developing countries [19]. Ehealth solutions are new healthcare delivery models capable of making qualified healthcare accessible to underserved areas—remote areas as well as areas lacking on-site medical expertise [19, 20]. In addition, eHealth solutions can also reduce costs for health service delivery, maintenance, and support. This makes them a viable and affordable way to improve healthcare in rural Ghana [21, 22].

In 2010 The Ghana National eHealth Strategy was published with the goal to: “Harness the potential of Information and Communication Technology to improve the health status of people living in Ghana” ([23], pp.-35). The report identified eHealth as a potential solution to several problems: large inequalities of access to healthcare among regions, weak referral, poor emergency systems, the increase of non-communicable diseases, and the lack of medical personnel [23].

Regarding the implementation of eHealth solutions, the identified challenges were: lack of ICT infrastructure in most facilities, low IT literacy, and lack of funding for ICT infrastructure. Local stakeholders cannot afford the increased cost of eHealth solutions [23].

In recent years, two eHealth solutions to improve maternal health have been piloted in different areas in Ghana. The first one aimed to send reminders by SMS to pregnant women [24]. The second one was a smartphone app to improve the knowledge of CHNs working in remote areas [25]. However, according to local authorities, no previous research has been done on implementing eHealth solutions in the study areas. Ehealth as it relates to our study encompasses telemedicine and mHealth, including a computerized decision-support system.

Objective

This paper assesses the feasibility, in terms of potential of and requirements, of eHealth solutions to improve maternal healthcare in remote areas of Kpando, Ghana. Improvement in maternal care is approached as a reduction of the delay in receiving adequate maternal healthcare.

The geographical scope of this article is the remote peninsulas and islands in Lake Volta whose inhabitants seek healthcare in Kpando Municipality. Factors that influence skilled maternal care attendance and eHealth implementation were identified through a variety of data collection methods.

Use cases

In order to get a deeper insight into the role of each stakeholder in MHC, use cases of typical pregnancies in the area were developed using the results of semi structured interviews, literature review and focus groups where pregnant women and health care workers explained typical cases of pregnancies in the area. Each use case includes the role of each of the stakeholders involved. The causes that could lead to maternal deaths were organized according to the TDM model. This resulted in the list of challenges for MHC in the area. The use cases also led to the identification of work routines and cultural beliefs that are challenges for eHealth implementation. In the use cases, details of events that occurred during pregnancy and childbirth for each stakeholder involved were mapped. Use cases are a suitable approach because they help to: exemplify typical pregnancies in the area, identify the weaknesses and strengths of the healthcare system, identify the reasons why the delays occur, and demonstrate how technology could help overcome the complications.

Literature review

Searches were conducted using Google and Google Scholar search engines, BMC Medical Informatics and Decision Making Process Journal, and ScienceDirect. Official documents from the UN and WHO were gathered from their online repositories. The keywords for the search were: maternal eHealth, eHealth in developing countries, maternal healthcare in developing countries, and impact assessment of eHealth solutions. Local reports and research related to maternal health were provided by the Health Directorate of Kpando and a local NGO at the study site.

Documents that contained impact assessment studies of eHealth in developing countries or presented eHealth solutions to improve MHC in developing countries were included. In addition, reports and recommendations on MHC, case studies of eHealth interventions, and detailed qualitative research on MHC in the study areas were also included. Each abstract was reviewed to determine if the document met the criteria.

Semi-structured interviews

Semi-structured interviews [41] were conducted with healthcare providers at various levels at six facilities in Kpando Municipality to identify the MHC needs, as well as identify work routines in ANC, delivery care, and PNC. Fifteen health workers were each interviewed for about 20 to 30 min. Through these interviews, their perspectives on barriers and facilitators for pregnant women accessing MHC were gathered and work routines were identified. Finally, these health workers discussed the possibility of using eHealth solutions in their jobs and their willingness to do so.

Focus group discussion

A focus group discussion [42] was organized with twelve MHC stakeholders: one representative of Ghana Education Service, one of National Health Insurance, two TBAs, three mothers from the islands, one representative of a private health facility that provide services to the islands, two midwives from GHS, one father from the islands and the Public Health Nurse from Kpando Municipality. The objective was to elaborate on the main pregnancy-related challenges and explore different solutions to the problem of poor access to MHC for the women living on the islands around Kpando. The challenges and their respective solutions were discussed and prioritized.

ICT infrastructure assessment

In order to quantitatively analyze the telecommunication infrastructure, naturalistic observations [42] of the equipment in the healthcare facilities were performed. The focus was on mobile phone penetration among patients and ICT infrastructure in the facilities. Moreover, a network coverage test for the two biggest telecommunication operators in the area was performed, using the android application OpenSignal. Four different measurements were taken at each of 20 geographical locations at three-minute intervals for each operator. The average of the four measurements was used as an evaluation factor.

OpenSignal was used for quantitative analysis of the percentage of time the study areas have network access. OpenSignal automatically measure the time each location has access to network and provides a summary of it. The percentage provided in the results is the summary of 20 geographical locations, in each location the app was on for an average of 20 min, then turned off while moving to a different location. The total measuring time was 400 min.

Theoretical frameworks for analysis

Two theoretical frameworks were used, the TDM and the 5C. The TDM, by Thaddeus and Maine [43], is a leading framework for assessing delays in maternal healthcare. Prompt, adequate treatment after the onset of obstetric complications will most likely result in a satisfactory outcome; however, a delay might lead to maternal death. Delays are not caused by one single factor, but by various interrelated factors. These factors are categorized into three groups (see Fig. 2).

The 5C conceptual framework by Drury [44], shown in Fig. 3, was used to structure research activities, outcomes, and recommendations. While the application of the TDM identifies factors that delay maternal healthcare, this framework analyzes the potential — and requirements — of eHealth solutions. The 5C framework was used to select the delay on which an eHealth solution would have the biggest impact. The five Cs are context, content, connectivity, capacity, and community. Information on the context, connectivity, capacity, and community was collected and analyzed during this research. The fifth C, content, will structure implementation research on the proposed solution. Challenges were placed on each category by researcher consensus.

All the gathered data led to the outline of a future eHealth solution to improve MHC. This strategic proposal should significantly decrease delays in maternal healthcare, improving the health status of mothers living in remote areas by addressing the main challenges of the healthcare system and the difficulties faced by mothers seeking access to healthcare. The prioritization is based on the relative importance that each participant of the study gave to the main challenges.

Factors that influence delay

Factors influencing delay in maternal healthcare according to the TDM model are summarized in Fig. 4. They are divided into three categories: socio-economic/cultural factors, accessibility of care, and quality of care. In addition, all factors influencing care are separated into those on the demand side (the patient) and those on the supply side (the healthcare provider). Following paragraphs further explain challenges identified in: Demand side through literature research, demand side through field research, supply side through literature research, and supply side through field research (i.e. semi-structured interviews and focus group discussion), respectively.

In previous research in the area, financial dependence of women, lack of education, and traditional beliefs were stated as the main of causes of delay in seeking care (first delay) on the demand side [45, 46]. Another key factor influencing the first delay identified in the literature was social stigma [10]. In [8, 47] pregnant women declared that distance to facilities, geographical situation and their economic status makes access to emergency care impossible at nighttime, hence, affecting the time it takes to identify and reach a healthcare facility (second delay). Finally, negative previous experiences came up as an influencing factor for the first and third delays [48]. As a result of these factors, among other problems, only 20% of women in the area go for ANC during the first semester. ANC visits are important to early detection and prevention of pregnancy related complications.

Some of these results were confirmed during the field research. On one hand, during the focus group discussion and in the semi-structured interviews with nurses, different stakeholders explicitly explained how lack of education led to maternal deaths. Furthermore, CHNs and stakeholders, confirmed poor road infrastructure, geographical situation, and the communities’ economic status as factors affecting the second delay during interviews and focus group. The lack of affordable transportation was a new factor that CHNs and mothers in the focus group identified during the field research. On the other hand, social stigma and negative previous experiences as factor influencing the quality of care (third delay) and first delay did not come up frequently during the interviews, but some stakeholders freely talked about it during the focus group.

On the supply side, the literature identifies social stigma, staff attitude towards patients, and competitiveness between facilities as factors influencing the first delay. Lack of equipment and human resourcers, organizational loopholes, and poor road infrastructure were identified as factors affecting the second delay, particularly in emergency cases [47, 48]. Finally, the literature reported that TBAs are a preferred choice of care due to their social recognition [10]—however, their lack of training affects the third delay.

On the field research, lack of equipment and human resources, organizational loopholes, poor road infrastructure and TBAs as prefered choice of care were confirmed during interviews and focus group discussions. A factor that was found only in the interviews was a lack of knowledge of patients previous complications, and how it affects the first delay. Finally, factors that were identified only in the interviews and affect the first and third delays were: inconsistency in communication, a poor referral system, and lack of access to current knowledge.

Ehealth solutions to reduce delay

Figure 5 summarizes how eHealth solutions can help reduce delays in receiving MHC. The solutions are further developed in the following subsections.

Challenges to eHealth implementation

In this section, the different challenges to eHealth implementation are explained, based on the 5C model proposed by Dr. Peter Drury in [44]. Accordingly, they are divided into the five different components: context, content, connectivity, capacity, and community.

Solution outline

This subsection outlines an eHealth solution that might optimize efficacy and feasibility based on the findings this work. Additionally, it might address the some of the main needs in MHC for patients and health workers in the situation considered in this work. Moreover, it is designed to fit under the current work practices of Kpando and areas with similar characteristics. A schematic is presented in Fig. 6. This solution, falling within the field of mHealth, is an application for Android mobile phones or tablets handled by front line health workers (CHNs in CHPS zones). It is cloud-based and uses the mobile communication networks available in the facilities.

The solution should be focused on providing a pregnancy monitoring system and bridging the distance between facilities and communities. It has the potential to make a substantial impact in the third delay and solve direct challenges in content and context as defined by the 5C model. To fulfill this potential, decision support systems and data gathering services need to be included in the application; they would help develop evidence-based birth plans to increase skilled maternal health attendance. The solution should be adapted to fit into the workflow of healthcare facilities in order to be used in facilities and outreach programs. Equipping CHNs with mobile phones or tablets and basic medical equipment to monitor pregnancies would be a requirement.

The application would allow facilities to share information, improving patient tracking and collaboration between professionals. It would provide health administrators with access to data summaries, which could improve reporting and logistics systems — addressing the content and capacity of the 5C model. The system should be designed to work offline, asynchronously sending the information to the cloud server whenever a connection is available. Cloud services are accessible in Ghana and allow remote maintenance of the network, a feasible solution to reduce connectivity challenges.

Strengths and limitations

Among the strengths of the research, are its qualitative design and the inclusion of many communities and stakeholders. The latter adds to the conceptual generalizability of the results to other remote areas in Ghana and neighboring countries. The qualitative setup of the study permitted the research objective to be examined in depth and comprehensively, allowing subtleties and complexities to be discovered.

On the other hand, one limitation is the cultural differences (including language barrier) between participants and researchers. These differences could have created response bias towards socially desirable answers; some participants may have been reserved towards the researcher in some aspects of their answers. Culture differences may also have led to misapprehension of sensitive and subtle topics during analysis. Further, interviews were not recorded since participants did not feel comfortable to be recorded due to personal and professional issues. Consequently, misinterpretation of questions and answers may have altered the true meaning of some discussed subjects.

Triangulation of different research methods and data, with the inclusion of different stakeholders, was used to increase the validity of this study. Moreover, the incorporated member checks and the active seeking of negative cases increased understanding of the stakeholders.

Future work

A mHealth solution has been outlined, based on health workers’ needs as identified during the research. In the next steps, the solution will be prototyped and re-designed following a human-centered approach. The aim is to create a solution that health workers will be willing to use.

In a following publication the mHealth solution will be described including aspects such as networking, security specifications, backup systems, and privacy protocols. These aspects will be assessed and incorporated into the design of the solution.