Stakeholders perspectives on paper-based and electronic clinical decision support systems in Malawi Africa

Abstract

The objective of this paper is to explore the perceptions of key stakeholders involved and/or affected by existing paper-based decision support guidelines (known as Community Case Management (CCM)) and a proposed digitised mobile clinical decision support system (CDSS) of CCM in rural settings of Malawi, Africa. Data was collected using field notes and semi-structured interviews with 17 key stakeholders (i.e. clinical, technical, development aid support (NGO), government and community health workers both in Malawi, Europe and USA). Stakeholders provide a rich insight into the variety of both perceived benefits and challenges of the existing guidelines and the proposed electronic CDSS. It was found that all stakeholders believe that the CDSS will improve adherence to guidelines and subsequently result in better care for children. It is further envisioned that the time needed for administration with the current paper-based approach could be reduced using electronic, as opposed to manual, collation and sending of records. This paper acts to underpin the rationale and motivation for the development and rollout of an electronic CDSS to support community health workers in their assessment, classification and treatment of young children in rural settings in Malawi, Africa.

Keywords:

• Clinical decision support systems
• developing countries
• barriers
• facilitators

Introduction

Decision support interventions have long been utilised within the healthcare domain (Kilsdonk, Peute, Riezebos, Kremer, & Jaspers, 2016), delivered via paper-based or electronic means of communication. Such interventions are considered to create profound opportunities to both healthcare providers and patients (Politi, Adsul, Kuzemchak, Zeuner, & Frosch, 2015). The use of decision support interventions at the point-of-care have for example led to enhanced communication between provider and patients, increased medical knowledge, reduced decision conflict, and generally improved diagnosis and treatment (Kilsdonk et al., 2016; Politi et al., 2015). While many advantages are reported in a wide array of literature (e.g. decision support, clinical practice, information systems) the perspectives are often derived from those in resource-rich environments (Esmaeilzadeh, Sambasivan, Kumar, & Nezakati, 2015). Little evidence exists on the perceptions of decision support interventions in developing countries.

In Malawi, Africa, decision support interventions are predominantly paper-based. One such example is Community Case Management (CCM) aimed at delivering treatment for common childhood illnesses (suspected pneumonia, fever/suspected malaria, and diarrhoea) in the community. Introduced by World Health Organisation(WHO) and UNICEF, CCM is a ‘feasible, effective strategy to complement facility- based management for areas that lack access to facilities’ (Marsh, Gilroy, Van de Weerdt, Wansi, & Qazi, 2008, p. 381). CCM aims to improve child health by enabling trained community health workers (known locally as Health Surveillance Assistants) or volunteers to assess, classify, treat and refer sick children in rural communities, who reside beyond the reach of fixed first-level health facilities.

Based on the poorest child health indicators in Malawi, the Ministry of Health(MoH) selected ten districts to initially implement CCM (Malawi Ministry of Health, 2008). CCM decision support clinical guidelines were customised based on these indicators to ensure that the highest child health risks in rural Malawi would be addressed. Since 2008, CCM has now been implemented throughout Malawi, and the MoH continues to deploy national community-based health worker programmes to improve access to primary health care (Callaghan-Koru et al., 2013). Community Health Workers serve Malawian communities at a ratio of 1 to 1000 population (Nsona et al., 2012).

Research suggests that correct implementation of CCM can lead to improved child survival (Mugeni et al., 2014). Yet research also shows that CCM guidelines are often not correctly implemented (Amouzou, Morris, Moulton, & Mukanga, 2014), potentially compromising efforts to improve child survival. It is therefore important to understand stakeholders’ perceptions of the ‘as-is’ scenario to guide future initiatives within this domain. Moreover, the introduction of electronic clinical decision support systems (CDSS) has been proposed as a way to improve paper-based interventions, such as CCM guidelines (DeRenzi et al., 2008). Yet, a dearth of research has been conducted on perceptions of CDSS from a multi-stakeholder perspective.

The objective of this paper, therefore, is to explore the perceptions of key stakeholders involved in and/or affected by the introduction of a mobile health application encompassing CDSS (based on WHO and UNICEF paper-basedguidelines) in rural village clinics in Malawi: (1) regarding the current status and use of the paper-based decision support intervention known as CCM and (2) perceived benefits and challenges to future use of a mobile CDSS artefact for clinical management of sick children in these settings.

Methodology

This study employs a qualitative approach to achieve the objective. A number of stakeholders or key informants with varied roles developing and implementing initiatives to enhance health service delivery, both locally (i.e. Malawi) and internationally (i.e. Europe and USA) were identified to provide diverse insights into the current Malawian healthcare service delivery, specifically focusing on CCM guidelines. Stakeholders were selected using the purposive sampling technique (Patton, 1980) where the interviews were selected based on the researchers’ judgement or reputational and positional methods (Knoke, 1993). Table 1 provides an overview of the stakeholders involved in the data collection process.

Table 1. Overview of stakeholders.

Stakeholder type	Description of role	Respondent number
Health Surveillance Assistants (HSAs)	Community health workers deployed in rural settings with an assigned catchment area, responsible for preventive health care management in children as well as numerous other community health and education tasks. These stakeholders currently follow paper-based CCM guidelines thus; they can provide information from a healthcare practitioner perspective	11, 12
Government	Ministry of Health (MoH) officials who understand the governmental policy procedures in operation in Malawi, Africa. Examples include the District Health Officer (DHO) and the Programme Manager for Integrated Management of Childhood Illness	5,6
Aid support(NGOs)	A citizen-based association that operates independently of government, usually to deliver resources or serve a social or political purpose. These stakeholders provide opinions on the health service delivery in Malawi from a neutral perspective	10
Clinical Stakeholders	Individuals who have an understanding of paediatric health services/family medicine. These stakeholders understand children’s health and infectious disease in primary care and can assess likelihood of success/failures from a clinical perspective	2, 4, 7, 8, 9,13, 14, 17
Technical Stakeholders	Individuals who are directly associated with the development of the technological artefact. These stakeholders understand the underlying soft/hardware artefact and can assess likelihood of success/failures from a technical perspective	1, 3, 15, 16

Malawi is a land-locked country in southeast Africa with a population of 16.5 million,¹ and is ranked as one of the ten poorest countries in the world (Callaghan-Koru et al., 2013). The Malawian economy is predominantly agricultural, with a largely rural population. The Government of Malawi decided to reform its telecommunications sector in 1998 (Clarke, Gebreab, & Mgombelo, 2003) and as a result, the Malawian population nowadays has greater access to mobile network coverage. Yet, mobile penetration remains very low when compared to the African average (Buddle, 2015).

Data was collected using field notes and semi-structured interviews with 17 key stakeholders over the course of a twelve month period between September 2013 and September 2014. The interviews were carried out face-to-face, by telephone, or Skype. Interviews were transcribed and data anlysis was analysed using the method proposed by Miles and Huberman (1994); namely, data reduction, data display and drawing conclusions/verification. The Supporting LIFE (Low cost Intervention For disEase control) project developed a smartphone Android application with a CDSS embedded based on the paper-based clinical guidelines of CCM (abbreviated to SL eCCM App), with additional features and functionalities, for example user instruction materials and an electronic diagnostic tool for capturing breathing rate of a sick child. Stakeholders were asked for their perceptions surrounding current use of the paper- based CCM guidelines and future use of the SL eCCM App. The next section presents the findings from the stakeholder analysis.

Findings

Perceptions of ‘as-is’ paper-based decision support intervention

The ‘as-is’ situation of implementing and abiding by the paper-based CCM guidelines is not without its challenges. The stakeholder analysis revealed three key challenges currently faced on the ground in Malawi: namely, (1) Limited Training and Resources, (2) Complying with CCM Guidelines and (3) the Current Healthcare Reporting System. Each challenge is subsequently discussed in the proceeding sections.

Limited training and resources

Limited training on the application of CCM guidelines is the first challenge identified from the stakeholder analysis. The findings revealed that HSAs receive 10–12 weeks of training to cover a variety of tasks, which some stakeholders perceived to be insufficient to fully train the cadre of health worker based to their educational backgrounds. To exacerbate the issue of short term training, it was further revealed that approximately ‘six days’ over this period is dedicated to CCM. Additionally, people-oriented resources were identified as inadequate in Malawi with only a limited amount of HSAs completing the 12 week training programme.

Stakeholders identify that the role of a supervisor for monitoring HSAs implementation of the CCM guidelines is critical in ensuring that the correct policies and procedures are performed. Yet, the analysis revealed that the Malawian MoH suffers from inadequate supervisory officials on the ground. Comments reflecting limited training and resources include:

the HSA is the first port of call and they have no official clinical knowledge, rely on policy,and dependant on reliability of instruments (Stakeholder 8)

The Ministry of Health has 9,971 HSAs in Malawi, but those that have undergone training for this are less than 4000; 3,700 or thereabouts. (Stakeholder 5)

The biggest problem is service provision and especially looking at the kind of HSAs that are providing this service, and looking at our current healthcare system, the challenge is monitoring implementation- it’s a very important task that we cannot manage for each and every health care provider. (Stakeholder 5).

Unfortunately, limited training and resources may be linked with poor compliance with CCM guidelines identified in the stakeholder analysis.

Complying with paper-based CCM guidelines

The second challenge, complying with CCM guidelines, was highlighted by a number of key personnel associated with CCM implementation on the ground. While the village register (which the paper based summary of each HSA’s clinic activity) is often fully completed, little is actually known about how thoroughly HSAs execute relevant CCM assessment items and the data quality of the register; stakeholders assume that HSAs may not be completing all of the required steps. This is exemplified in the following comments:

When we did our quality of care assessment the results showed the ability of our HSAs todo clinical care assessment in 75–80% of cases. (Stakeholder 5)

The challenge we have is we are still not sure whether our HSAs are doing the right things at all times. (Stakeholder 5)

They fill in the register but don’t follow the guidelines very well. (Stakeholder 9)

Stakeholders perceive that noncompliance or poor adherence to CCM guidelines could have a negative impact on the quality of care children receive in rural clinics, resulting in instances of misdiagnosis and mismanagement of sick children.

Current healthcare reporting system

The third challenge is related to the inadequacies with the current healthcare reporting system in Malawi, Africa. Describing the reporting process, stakeholders (7 and 12) mentioned that outside of recording patient visits on the village register, HSAs are required to use this to generate a regular monthly report for the health management information system (HMIS) usage and further submit to the District Health Office (DHO). Once this report is received the data is fed up the chain from district to national levels but this reporting is criticised as potentially having errors and is slow where the HMIS officers need to wait for paper form reports to reach DHO. At present, the process for reporting the relevant statistics of childhood illnesses in various rural catchment areas is a time consuming process, and does not provide real time data for decision-making at the district, regional or national level for example with disease outbreaks or emerging public health event.

The amount of time it takes to merge health data on a monthly basis and then submit to the MoH- one HSA takes 1 and half days, which takes from treating children. (Stakeholder 4)

Concerns around the quality of the data given the fragmented data processes currently in place. (Stakeholder 1)

There are concerns that the current reporting system in Malawi fails to capture complete, accurate, real-time disease monitoring and surveillance. This could potentially have wider societal implications due to substandard quality of data being used for decision-making purposes at a local and national level.

Perceived benefits of the proposed SL eCCM App

The stakeholders’ analysis revealed a number of perceived benefits associated with the introduction of the SL eCCM App into HSAs work practices.

Improving adherence to guidelines

The findings reveal that the current paper based CCM guidelines encourage a haphazard attitude to filling in the village register form which risks compromising quality of care, and suggesting a need for the app. Therefore, improving adherence to guidelines is one perceived benefit from using the SL eCCM App:

The Supporting LIFE application engenders adherence i.e. it gives accurate guidance to HSAs and removes possibility for human error through the use of the decision support software. (Stakeholder 16)

It will ensure compliance to the step wise approach for CCM of sick children; hence it will ensure quality of care is adhered to. There will be no guesswork for the HSA as the app will provide the information which is necessary for the appropriate treatment for the patient. (Stakeholder 8)

A common positive consensus surrounding the need for the SL eCCM App emerged from the data, with one potential benefit being improved adherence to the clinical algorithm in CCM.

Better quality of care for children via training resources

To ensure that the highest quality of care is received by sick children, it is imperative for HSAs to maintain their healthcare delivery skills. The SL eCCM App provides builtin training material which can be accessed at any time. This could ensure that HSAs maintain and/or enhance their current training skillset, assuming that the training feature is utilised.

This innovation provides the missing gap that ensures that indeed the skill to provide an assessment approach, stepwise, is not missed and then we are sure that the children are getting the quality of care that is needed. That is the most important element. I think the collaboration and the project that is being done in Mzuzu tries to address this for us and that is essential. (Stakeholder 5)

It should also support their educational needs as the training videos and other material will help them keep up to date with the latest assessment and diagnostic techniques. e.g. breathing rate is performed by giving HSAs step by step instructions on how to complete each type of assessment accurately. (Stakeholder 17)

The findings reveal that stakeholders recognise the benefit of the training element of the app for HSAs to enhance their decision-making at the point-of-care.

Enhanced effectiveness in terms of decision making

The SL eCCM App facilitates display of the clinical information required for decision-making at point-of-care. The application uses a clinical guideline decision-support rule engine which encapsulates classification and treatment rules for assessing a sick child. This ultimately reduces the complexity and standardises delivery of healthcare by HSAs for children, as exemplified in the following comment:

I think the application could assist HSAs with their decision making processes when diagnosing and treating sick children. Hopefully, as a result there may be a reduction in the misdiagnosis and mistreatment of sick children. (Stakeholder 15)

The real-time identifying and tracking of disease outbreaks in rural districts of Africa is envisioned to further improve decision-making of HSAs:

From the District Health Officer perspective having real-time data on the health of young children in Malawi will be invaluable as it will enable them to monitor infectious diseases in this population more closely, identify serious patterns of disease quickly and intervene earlier if required. (Stakeholder 6)

Yet, stakeholders (8 & 14) highlighted that the recommendations (in terms of diagnosis and treatment) presented by the SL eCCM App should not be used in isolation and that individuals’ tacit knowledge and clinical experience should be leveraged through each patient assessment process. Overall, stakeholders perceive that the decision-making of the HSA will improve.

Enhanced efficiency in terms of time saving

Stakeholders perceived that the introduction of the SL eCCM App as part of delivering children’s healthcare services could potentially save time. For instance, having the CDSS functionality in addition to relevant patient data, at the point-of-care, reduces the time required to make a decision. Furthermore, stakeholders envision that the workload of the HSA will be reduced, The SL e CCM App should potentially reduce the time spent on diagnosing and treating a sick child, it should also provide a more efficient process for reporting clinical activity data. The following comments reflect these perceptions:

It will reduce the workload as it will be made more efficient through technology. (Stakeholder 10)

It is going to reduce the time spent in terms of assessing the child as well as delivery of information to the authorities is going to reduce, so there will be a big reduction of time that is going to be taken. (Stakeholder 14)

It will also reduce transport needed to get the form A to the district health office. (Stakeholder 11)

Overall, the stakeholders involved in this report perceived that the SL eCCM application can bring about benefits at an individual, regional, national and societal level if implemented in the appropriate way. However, successful implementation of the SL eCCM App can only be achieved if common challenges are overcome. The following section describes the perceived challenges associated with implementing the SL eCCM application in Malawi.

Perceived challenges of the proposed SL eCCM App

While stakeholders acknowledged many proposed benefits of the app a number of perceived challenges were also reported. These challenges stem from a myriad of technical and socio-economic difficulties and are highly contextual in nature. That is, the environment in which the application will be utilised is considered to be a potential impediment to the overall success of the project.

Operationalising the mobile device

The key concerns associated with operationalising the application on the mobile device include: battery performance, device safety (i.e. possibility of being stolen and/or damaged) and network connectivity. Firstly, charging mobile devices in rural areas where there is very limited access to the electrical grid and frequent power outages is a challenge. This is exemplified in the following comments:

Reliable power supply will need to be available to enable the HSA’s to power the devices. (Stakeholder 13)

The difficulty will be battery, when the battery becomes flat, in the community when there is nowhere to charge the phones. (Stakeholder 14)

The second challenge associated with the operationalisation of the mobile device is mobile safety. In resource-poor settings stakeholders highlighted that mobile phones may be stolen as they are highly valued in society. Moreover, phones have a tendency to break in the dusty and humid environment in Africa and human error such as dropping the device on the ground in settings where clinic facilities are rudimentary:

“There is a risk the mobile device will be damaged or stolen so mechanisms need to be in place to minimise this risk where possible.” (Stakeholder 17)

The third challenge is that of mobile network connectivity. In rural areas of Malawi, many people do not have access to mobile network, according to numerous stakeholders. However it is perceive that network availability will only improve over the coming time due to ongoing infrastructural developments in Malawi:

“It will be interesting to determine the performance of telecom data transmission in Malawi and the general availability of the 2.5G networks. Whilst data availability does not affect the capability of the HSA to use the app and perform assessments, it is important for gaining an accurate picture of disease surveillance.” (Stakeholder 16)

Recommendations are proposed by some stakeholders on how such issues can be overcome. These include the use of solar panels for charging the device and awareness of areas with mobile connectivity. Using the device is one challenge but also it is important that the users have some degree of technical competency, which is discussed below.

Technical competencies of Health Surveillance Assistants

Another potential challenge faced by this project is that of the skill-set of HSAs. The findings reveal that many Malawians have not previously interacted with a smartphone and thus, their technical competencies may be relatively low. A number of stakeholders reported this concern:

A final challenge of the project will be technological learning curve required of the HSAs who may not have previously been exposed to smartphone devices. (Stakeholder 16)

Some HSAs are not used to technology unlike me, so it will be a problem for them if they are not trained to use the technology correctly. (Stakeholder 11)

Some HSAs may find the app difficult to use (because they may not have used a smartphone device prior to the trial) leading to resistance towards the application. (Stakeholder 15)

We are bringing an innovation that does not completely confuse or bring an overhaul to what a HSA is currently doing, because if that brings everything new then we need to start all over again, and that will be a problem with the HSA (Stakeholder 5)

It is imperative that the SL eCCM App is used correctly but the stakeholders perceive that this should not be a huge concern since the app is digitised replica of the paper- based approach.

Sustainability and scale up

Stakeholders argued that more funding need to be secured, relationships with the Ministry of Health need to be maintained and communication with other NGOs on the ground is integral to ensure sustainability and scale up of mHealth initiatives. This is exemplified in the following comments:

I wonder about how the project will be sustainable, as I know some projects who had a mobile technology for some years, and when the project was over, the switch back to paper had to happen. (Stakeholder 12)

How sustainable will it be? Will it be kept going. It’s something which should be looked at critically as well. I was thinking to ask the Minister for Health that after we complete the trial, that maybe the government will support the roll out of the application for sustainability. (Stakeholder 14)

A significant challenge facing SL eCCM for a wide scale rollout is in relation to the technological infrastructure in place within Malawi. (Stakeholder 16)

One stakeholder (15) proposed that the cooperation among Malawian academic partners and Malawian MoH can resolve the sustainability issue and build capacity within the country for future project expansion and scale up.

The success of the Supporting LIFE project and other projects implementing mobile Health apps (mHealth) encompassing CDSS is contingent on overcoming the barriers as outlined in order to support the team and those who are affected by the project, i.e. HSAs, DHOs, the caregivers and children, to work together in a meaningful way in Malawi.

Discussion and conclusion

Decision support interventions are important for improving the delivery of child health services in developing countries. The findings of the study reveal that the situation on the ground using the paper-based CCM guidelines faces many barriers such as noncompliance, insufficient resources and workflow inefficiencies in the current reporting and surveillance systems. These barriers act as motivational drivers for the implementation of the SL eCCM App (encompassing CDSS).

According to existing literature (Chanda, Hamainza, Moonga, Chalwe, & Pagnoni, 2011; Graham, English, Hazir, Enarson, & Duke, 2008) there is no agreed consensus on the rate of adherence to CCM guidelines by frontline health workers in developing countries. This stakeholder analysis corroborates this analysis as our findings reveal that the village register entries are perceived to be incomplete, but a dearth of evidence exists which focuses on quantifying the assessment of adherence to the actual guidelines. This study further supports literature (Nsona et al., 2012) which argues that insufficient resources and the reporting structure within the Malawian health system are problematic.

Many stakeholders refer to the SL eCCM App as a method of improving compliance with the guidelines, by helping to reduce the possibility for human error and guesswork through its stepwise decision support software. This coincides with research which argues that introducing technology at the community level health care system can improve the delivery of healthcare services to patients (DeRenzi et al., 2008). The advantages of using the SL eCCM App potentially outweighs the disadvantages of the current system by:

•	Ensuring better adherence of healthcare providers to clinical guidelines by enforcing all data entry into the app compared with skipping critical steps towards diagnosing and treating sick children,
•	Providing precise informed clinical decision-making over receiving 6 days of formal training and manually diagnosing and treating sick children,
•	Offering accurate and timely digitised datasets which can be digitally provided to various reporting bodies compared with manually calculated reports, delivered in person.

Introducing technology in developing countries is not without its challenges. Such an introduction necessities training and support on the ground for HSAs, as they may have never previously interacted with smartphone technology. Although a cost will be associated with training and support, the long term benefit of mHealth technology potentially outweighs the short term investment in training and support if patient outcomes are improved. Without training and sustained support health workers may resist the technology, as observed in similar studies (Amouzou et al., 2014). Unfortunately, Malawi suffers from a ‘brain drain’ and is losing educated, technically savvy individuals (Nsona et al., 2012). Offering training and support will build technical capacity of individuals on the ground, which may be of benefit to other sectors outside of research.

With the electronic data collation, storage and decision support facilitated by the app, healthcare workers’ workload could be improved in terms of effectiveness and efficiency. Automating the current paper-based decision support intervention will reduce misdiagnosis and mistreatment, improve adherence to the guidelines and ultimately improve the quality of care children receive. Moreover, it could reduce the time and associated costs required to collate and disseminate the information for monthly reports. It is therefore imperative that the introduction of mHealth technology benefits users (directly and indirectly) as advocated by Black et al. (2009).

The research also found that many stakeholders were concerned with the ‘bigger picture’ of the electronic CDSS. Interestingly, this represents the passion of those involved in the project in improving the healthcare system of Malawi. The key concern is that the MoH will not be able to fund the scale up to the rest of the country, and the project will remain as just another pilot study. As a result, it is important to prepare an economic evaluation of the costs associated with introducing and diffusing mHealth technologies so that it can be presented to the MoH at the end of the project to provide policy makers with the evidence needed to make decisions on resource allocation forthis and other needs.

Gathering a multi-level stakeholders’ perspective of decision support interventions in Malawi produced many insights into the possibilities for future research within this domain. We propose that future researchers working in developing countries should conduct a stakeholder analysis to gather and analyse information by interested and/or involved parties when developing and/or implementing a policy or programme pertaining to mHealth initiatives. Social scientific studies such as these will endeavour to enable the convergence of critical thinking in the mHealth domain (Chib, van Velthoven, & Car, 2015); thus facilitating the generation of national and international standards, best practice and policy to support the delivery of low cost, high quality healthcare services in these unserved areas of the world.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by the European Union’s Seventh Framework Programme (FP7/2007-2013) [grant number 305292].

Acknowledgement

The research leading to these results has received funding from the European Union’s Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 305292

Notes

1. Population figure: http://countryeconomy.com/demography/population/malawi.