i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

journal homepage: www.ijmijournal.com

Guidelines for maternal and neonatal “point of care”: Needs of and attitudes towards a computerized clinical decision support system in rural Burkina Faso S Alphonse Zakane a,b , Lars L Gustafsson b , Göran Tomson c,d , Svetla Loukanova e , Ali Sié a , Josefine Nasiell f , Pia Bastholm-Rahmner c,g,∗ a

Centre de Recherche en Santé de Nouna, BP 02 Nouna, Burkina Faso Department of Laboratory Medicine, Division of Clinical Pharmacology, Karolinska Institutet at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden c Department of Learning, Informatics, Management and Ethics, Medical Management Centre (MMC), Karolinska Institutet, SE-171 77 Stockholm, Sweden d Department of Public Health Sciences, Health Systems and Policy, Karolinska Institutet, SE-171 77 Stockholm, Sweden e Institute of Public Health, University of Heidelberg, INF 324-69120 Heidelberg, Germany f Department of Clinical Science, Intervention and Technology, Division of Obstetrics and Gynecology, Karolinska Institutet at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden g Public Healthcare Services Committee Administration, Stockholm County Council, SE-118 91 Stockholm, Sweden b

a r t i c l e

i n f o

a b s t r a c t

Article history:

Background: In 2010, 245,000 women died due to pregnancy-related causes in sub-Saharan

Received 10 July 2013

Africa and southern Asia. Our study is nested into the QUALMAT project and seeks to

Received in revised form

improve the quality of maternal care services through the introduction of a computerized

29 January 2014

clinical decision support system (CDSS) to help healthcare workers in rural areas. Health-

Accepted 31 January 2014

care information technology applications in low-income countries may improve healthcare

Keywords:

tance to CDSS and that the fit between the system and the clinical needs does present

Burkina Faso

challenges.

Computerized clinical decision

Aims: To explore and describe perceived needs and attitudes among healthcare workers to

provision but recent studies demonstrate unintended consequences with underuse or resis-

support systems

access WHO guidelines using CDSS in maternal and neonatal care in rural Burkina Faso.

Maternal and neonatal care

Methods: Data were collected with semi-structured interviews in two rural districts in Bur-

Perceived needs and attitudes

kina Faso with 45 informants. Descriptive statistics were used for the analysis of the

Rural healthcare

quantitative part of the interview corresponding to informants’ background. Qualitative data were analyzed using manifest content analysis. Results: Four main findings emerged: (a) an appreciable willingness among healthcare workers for and a great interest to adapt and use modern technologies like computers to learn

Abbreviations: ANC, antenatal care; CDSS, clinical decision support system; CIE, Comité Institutionnel d’Ethique; CRSN, Centre de Recherche en Santé de Nouna; IT, information technology; MDG, Millennium Development Goal; PMTCT, prevention mother to child transmission; PNC, postnatal care; QUALMAT, quality of prenatal and maternal care (EU-funded project); TAM, technology acceptance model or theory acceptance model of technology. ∗ Corresponding author at: Department of Learning, Informatics, Management and Ethics, Medical Management Centre (MMC), Karolinska Institutet, SE-171 77 Stockholm, Sweden. Tel.: +46 8 123 135 76. E-mail address: [email protected] (P. Bastholm-Rahmner). 1386-5056/$ – see front matter © 2014 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijmedinf.2014.01.013

460

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

more in the workplace, (b) a positive attitude to easy access of guidelines and implementation of decision-support using computers in the workplace, (c) a fear that the CDSS would require more working time and lead to double-work, and (d) that the CDSS is complicated and requires substantial computer training and extensive instructions to fully implement. Conclusions: The findings can be divided into aspects of motivators and barriers in relation to how the CDSS is perceived and to be used. These aspects are closely connected to each other as the motivating aspects can easily be turned into barriers if not taken care of properly in the final design, during implementation and maintenance of the CDSS at point of care. © 2014 Elsevier Ireland Ltd. All rights reserved.

1.

Background

Maternal mortality varies widely between rich and poor countries. As many as 287,000 women died due to pregnancyrelated causes worldwide in 2010 of which 245,000 deaths were in low-income countries. This equals a death risk of 1 in 25 women during delivery in the poorest countries [1]. Millennium Development Goal 5 (MDG 5) aiming for improving maternal health was adopted in the late 1990s but is the one that has failed to show any major improvement in recent years [2]. The priority is to increase the proportion of deliveries with skilled attendance, this being a key indicator for measuring a progress of MDG 5 [1,2]. However the medical care in many healthcare settings in low-income countries is of poor quality and the performance of health workers is suboptimal since there is a gap between knowledge and clinical practice [2,3]. Innovative methods and tools to improve skills and performance of rural maternal healthcare workers are needed to decrease the risks of death for the mothers and their newborns [2–5]. One strategy to assist health workers in rural settings in poor countries to more skilled and informed decisions is to provide access to easy-to-use guidelines and information with Information Communication and Technology (ICT) tools at the point of care [5–8]. Among ICT tools, computerized clinical decision support systems (CDSSs) are one of most interest combining action-oriented algorithms and interactivity to help and fulfil the needs of the healthcare providers [5,9–11]. Patient characteristics and results of clinical assessments by the provider are matched to a computerized clinical knowledge base and recommendations are then presented to the clinician for a decision [5,10,11]. The benefits include immediate and targeted delivery of necessary knowledge or evidence required to make informed decisions at the point of care [4,5,11,12]. The QUALMAT (Quality of Maternal and Neonatal Care) project addresses the potential needs of CDSS used in rural care settings in three African countries (Burkina Faso, Ghana and Tanzania) [5]. The system has four parts: (i) the user interface, (ii) the xml data base, (iii) algorithms to screen entered clinical values in the xml database and (iv) a section with training and information documents using WHO guidelines for maternal and neonatal care [4,5]. The idea is that needed knowledge is available at “point-of-care” in a pedagogic manner on a laptop in each resource-strained health facility supplied with electricity using rechargeable batteries connected to solar panels [5]. The CDSS system was developed

in English for use in Ghana and later translated into French for Burkina Faso and into Kiswahili for use in Tanzania [5]. The benefits, implementation challenges and impact of CDSS have, however, received no or limited attention in resource-strained healthcare settings in Africa [4,5,8,13] even though mobile phones have appreciably improved connections, procedures for referrals and distance consultation patterns in rural healthcare in Africa [6]. All studies reveal challenges including power failure, poor knowledge in using computers, shortcomings in adapting the CDSS according to needs and linking it to other health information systems thereby impeding the desired benefits of these services [10,14,15]. Recent studies of unintended consequences with underuse, or resistance to apply CDSS, show that these observations are explained by the poor fit between the system and clinical needs [16]. The better understanding we have regarding facilitators, barriers and issues of non-acceptance, the better we can minimize unexpected adoption behaviour and thereby benefit most from the use of CDSS in rural healthcare settings [16]. In particular, it is important to understand how a CDSS has to be adapted to the local context [14]. The aim of this study was to explore and describe perceived needs and attitudes among healthcare workers to access WHO guidelines using CDSS in maternal and neonatal care [5] in rural Burkina Faso. The study was carried out in maternal healthcare units in two rural districts in north-western Burkina Faso [5]. They are challenged by lack of trained well-paid staff, poor resources, are simple facilities with erratic access to electricity making it mandatory to understand how a CDSS can be contextualized and strengthen performance of staff [5,14,17].

2.

Methods

2.1.

Study design and setting

To understand the perceived needs and attitudes to access WHO guidelines through the use of CDSS in maternal and neonatal rural healthcare in Burkina Faso, we used a qualitative approach with semi-structured interviews. A qualitative approach is a fruitful way to explore people’s needs, experiences, attitudes, thoughts and perceptions of different phenomena [18]. This study on understanding the value and complexities of CDSS in rural maternal care is part of a comprehensive implementation research project called “Quality of prenatal and maternal care: Bridging the know-do gap (www.qualmat.net)”

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

461

[5,19]. The QUALMAT project is funded mainly by the European Union and includes the development and implementation of a system of performance based incentives and a computerassisted clinical decision support system (CDSS) based on WHO guidelines for maternal and neonatal care. The interventions are evaluated in a pre-and post-controlled study design in rural Burkina Faso, Ghana and Tanzania between 2009 and 2014. The project assumes that healthcare workers competence and motivation interact with the work environment to contribute to their work effort, demonstrated in their clinical performance [5,17,20]. In each country, one district with 6 primary healthcare facilities is subject to the intervention package whereas the control district with 6 primary health facilities has no interventions. The healthcare workers in primary healthcare facilities include in accordance to policies in Burkina Faso nurses, nurse assistants, midwives, midwife assistants and a community member trained to sell drugs at facility pharmacy. All primary healthcare facilities have equal type of services: one clinic for general care, one maternity for primary obstetrical care and one pharmacy unit. The staff in maternity services at the intervention facilities in Burkina Faso has a total of 26 healthcare workers and in the control facilities there are 23 healthcare workers. One health facility in the intervention district and one in the control district in Burkina Faso are located in a semi-urban area with electricity from grid connections. All other are rural health facilities with electricity from old solar panels in the control district but the intervention facilities were provided with new solar panel system as a package with the computer (laptop), CDSS and light through three or four bulbs. This specific qualitative study on needs and attitudes to CDSS was performed in twelve primary healthcare facilities, six from Nouna (intervention district) and six from Solenzo (control district), in Burkina Faso. All interviews were carried out before the CDSS intervention started in June 2012. All the healthcare centres involved in this study have maternity facilities equipped to accommodate uncomplicated deliveries including a 24-h observation period after the delivery. In total 49 healthcare workers are working in maternity units at the included facilities, but with no access to physicians. All healthcare centres are maximum 2 h’ drive from the district referral hospitals. The district hospitals, one in Nouna and one in Solenzo, provide full emergency obstetric care including the caesarean sections. Ambulances from the district health authorities are available upon request. Nouna Institutional Review Board, Burkina Faso approved the study design (approval number 2011-004/CIE/CRSN).

of “working in a maternity unit in a local healthcare facility”. All interviews were face to face directly in the informant’s workplace and were conducted from December 2011 to March 2012. In the beginning of the interview, the purpose of the study was explained to the informant for about 30 min. The informants received both written and oral information about the research before signing the consent form. The interviews were performed in French and lasted approximately 40 min. Prior to the interview, the informants from Nouna had participated in a demonstration and training workshop about the concept and handling of the CDSS, including training in computer use for three days. The informants from Solenzo had been informed about the planned CDSS project prior to the interview. The reason for interviewing healthcare staff from that district was to understand if there were differences between the two groups in perceived needs and attitudes to computerized tools and gain baseline knowledge about needs and attitudes also in the control district for future implementation of the CDSS in the non-intervention district. The interview guide included two parts:

2.2.

All closed questions with fixed answering alternatives about the informants’ background characteristics (professional background, experience in healthcare, computer access and use) were summarized as frequencies. For the open-ended questions, the transcribed data were analyzed using qualitative manifest content analysis according to Graneheim and Lundman [21]. This analysis includes multiple steps as summarized in Table 1. First the text was sorted to identify relevant meaning units, i.e., informants’ responses to the aim of the study. In the second step, the meaning units were shortened to condensed meaning units,

Data collection

One interviewer (SAZ) performed the interviews with 45 out of a total of 49 informants working in maternity units in these rural healthcare facilities. Four of the informants were absent for sick leave. The interviewer was familiar with the context, the design and the contents of the CDSS [5] and the method of interviewing. Twenty-three informants were from the intervention district Nouna and 22 from the non-intervention district Solenzo. The informants were selected from register lists provided by the District Medical Officers with the criteria

(1) The first part contained closed questions with fixed answers on informants’ background characteristics (gender, age, educational level, and specialization), experience in current facility and healthcare, computer access, use of guidelines and how they want access to guidelines. For one question “How would you describe yourself as a computer user?” the informants’ were asked to respond on a 5-grade scale from “very inexperienced” to “very experienced”. (2) In the second qualitative part, open-ended questions were asked to explore the informants’ needs of information and equipment in the workplace and attitudes to CDSS in maternal care. The following three broad questions were asked in the interview: • What is your daily or routine work in maternal and neonatal care? • What do you find difficult or tricky with your work? How are you managing these difficulties? • What is your opinion about the value of access to electronic guidelines and CDSS in maternal care in your healthcare unit? The aim of using this type of broad questions was to identify needs that could be related to the workplace of the informants. The second part of the interview was audiorecorded and transcribed verbatim by the first author Zakane.

2.3.

Data analysis

462

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

Table 1 – The structure of the qualitative content analysis process. One example is include to show the structure and the steps of the analytical process from identification of a meaning unit to summarizing findings into a category. Meaning unit We found difficulties only when you are alone, and you ought to follow patient in the delivery room and have to do consultation at the same time. So we can say that the workflow is our major difficulty. We need support with personnel because I’m alone and dedicated for the work in the maternity unit

Condensed meaning unit The workflow and lack of health care workers are a major problem

the condensed meaning unit were then coded as showed with an example in Table 1. The codes were compared for differences and similarities and sorted into sub-categories. Finally, the sub-categories were systematically abstracted to represent the ‘manifest’ aspects of the dataset. This part of the procedure answered the question “What?” about the transcribed data and was classified in a category. The analysis was primarily performed by two of the authors having different backgrounds, a computer scientist with experience in interviewing (SAZ) and a behavioural scientist with extensive qualitative research experience (PBR). An intercoder reliability check was done by the other authors (LLG, GT, JN, SL, AS) to confirm the content of the codes and categories as compared to the transcribed original material. When opinions differed between research members, such as about meaning or origin of statements, we returned to the transcripts and sought evidence to establish consensus. This iterative process was used throughout the whole analysis, i.e. moving from the whole transcripts to the condensed description and back again. The results are comprised of citations, which were selected to illustrate the informants’ attitudes of the different categories.

Code Needs of health care workers

Characteristics

A total of 45 healthcare workers were interviewed: (n = 23) in Nouna and (n = 22) in Solenzo districts in Burkina Faso. There were 16 males and 29 females with an average age of 35 years (Table 2). A total of 22% of healthcare workers in Nouna district reported lack of experience with computers compared to 68% in Solenzo district (in total 44% of the interviewees). In Nouna district, 91% of healthcare workers have access to computers privately compared to 27% of healthcare workers in Solenzo district. Forty-four informants reported that they were using medical guidelines in practice. The absolute majority of the informants (n = 40) wanted to have access to guidelines through computers.

3.2.

Results from the transcribed interviews

Four categories were identified from the interviews with the rural healthcare workers:

Number/average (Range)

Gender Female Male

14 9

15 7

Age (years)

35 (26–42)

35 (27–52)

7 14

8 10

2

3

0

1

5 5 2 11

7 0 1 14

4 years (1 month–10 years) 9 years (2 years–19 years)

3 years (3 months–10 years) 7 years (6 months–27 years)

Education level Primary Secondary ordinary level Secondary advanced level University Specialization Nurse Nurse assistant Midwife Midwife assistant

Results Background characteristics of informants

Needs and problems in work

Nouna (n = 23) Solenzo (n = 22)

Experience in health

3.1.

Human resources

Category

Table 2 – Background characteristics of informants and their reports about accessibility to guidelines and experience of computers in Nouna and Solenzo districts.

Experience in current facility

3.

Sub-category

Do you have access to a computer? 21 Yes 2 No

6 16

How would you describe yourself as a computer user? 14 Very 0 inexperienced Inexperienced 5 1 16 6 Some experience 2 0 Experienced 0 1 Very experienced Do you use any type of guideline? Yes 23 0 No

21 1

In which type of support do you need these guidelines? (1 or 2 choices) 21 19 Computer 2 1 Mobile 7 4 Paper 4 5 Poster

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

Table 3 – Reported daily routine activities and types of decision support used routinely by healthcare staff (Category A). Activities Antenatal care (ANC) Delivery

Decision support Guidelines in paper, books and as posters. Consulting colleagues (in same facility or consulting with experts in the hospital)

Postnatal care (PNC) Prevention of mother to child transmission of HIV programmes (PMCTP) Family planning (PF) Vaccination

A. B. C. D.

Table 4 – Reported needs and experienced problems in the daily work (Category B). B.1 Human resources

Need of health care workers

B.2 Educational training

B.3 Infrastructure of the facility including supplies and access to guidelines

Lack of training and experiences to: • Fill data in the partograph during delivery • Identify the position of the baby • Diagnosing patients

Lack of: • Drugs e.g. for malaria prophylaxis and iron • Lab test e.g. malaria rapid diagnostic test • Scan (ultrasound) • Bed net • Electricity • Ambulances

Daily routine activities Needs and problems at work Expectations and attitudes of using CDSS Barriers to use CDSS

In category A and B, the informants describe contextual factors into which the CDSS should be implemented. In category A, the informants described their daily routine activities in maternal work and in category B they reported what kind of needs and problems they have in relation to the working situation today. Category C and D describes the informants views of the CDSS after they have got information about the system. Subsequently, in category C the focus is on the informants’ expectations and attitudes of using CDSS and category D describes what kind of barriers they expect to use the CDSS. Despite differences in experiences in use of computers between the two districts Nouna and Solenzo (Table 2), we couldn’t find any variability between the two districts in the interview data on needs of information and attitudes to access guidelines through CDSS. A. Daily routine activities

463

Need easy access and time saving guidelines (electronic guideline or posters)

to fulfil their work tasks. The reported needs and problems in their daily work are summarized in Table 4. B.1. Human resources Needs of healthcare workers The informants pointed out the need of more healthcare workers present in the maternal unit since the flow of work is intense. However, it should be desirable with three persons; one for antenatal care (ANC) another for delivery and one for other counselling activities including family planning (FP) and preventing mother-to-child transmission of HIV (PMTCT). One informant explained that: “We need support with personnel because I’m alone and dedicated for the work in the maternity unit (Female midwife-assistant)”. B.2. Educational training

The informants described the work in maternal care as organized according to a planned programme for delivering services and fulfil duties. The routine activities and decision support systems used in practice are summarized in Table 3. Informants stated that in all of their daily activities the patients and their needs are in focus. They clarified the cause of the visit of a patient by questioning her in order to be sure that they understand what she was telling or trying to tell before making a decision. One informant concluded: “To have success in care, you need to understand the patient, understand what she said and what she can’t say. Then you need to summarize the conversation with the patient. . . (Male, nurse)”. B. Needs and problems at work According to the experience of the informants, they face problems in their daily work determining their needs for human and medical resources besides competence to be able

Lack of training and experiences All informants emphasized the need of basic and continued medical training as most of them have little experience in maternal care. When the diagnosis is not sure or the delivery is complicated they refer the patients to the district hospital. The experience of some workers was that they were not aware on how to fill in and use the partograph during the course of delivery or for identifying the position of a baby. “. . . we need training. I need training on how to use the partograph and fill in required information (Female midwife assistant)”. “. . . for example if it is a placental retention we refer, or if we have a situation with a big foetus and the woman can’t give birth we make episiotomy. If it is a too complicated delivery we refer the patient (Male nurse-assistant)”. The informants wanted to have training through exchange of experience between colleagues. They also wanted regular

464

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

training sessions on maternal healthcare as well as information on how to handle ANC, managing delivery, or training in neonatal and post-delivery care. Furthermore, the informants described how they had a desire for re-training as they were aware the science is changing, and they want to be updated with new knowledge and practices in maternal care. Furthermore, they express how they also wanted to learn about how the new techniques like CDSS and mobile phones could help them to gain such knowledge. “. . . we need retraining. You know that science is developing quickly and we ought to follow these changes, so to be aware we ought to have continuous training (Female midwife-assistant)”. B.3. Infrastructure of the facility including supplies and access to guidelines Poor access to laboratory tests, drugs and ultrasound The informants mentioned poor access to laboratory tests, lack of drugs, ultrasound and some equipment. They acknowledged a lack of drugs for malaria prophylaxis or iron or no access to malaria rapid diagnostic test [22] and to bed-nets. One informant describes this as: “To take care of the patients we need equipment, antimalarial drugs for prophylaxis and bed nets (Female, midwife-assistant)”. Need for electricity Seven of the twelve healthcare facilities have electricity. Five healthcare facilities in the intervention district Nouna and one in the non-intervention district Solenzo have electricity from solar panels that seem good and two facilities, one in the Nouna district and one in the Solenzo district, have grid connection. The electricity of the facilities from the Solenzo district was old solar equipment that did not function correctly. Early in the morning, the solar panel system is not functioning and this requires use of pocket lamps during delivery. Without electricity, it is difficult to make a delivery safe but also to document each patient activity. One informant expressed: “. . .we are using solar panels to get electricity, but early in the morning around 4 am there is always power shortage. So if we have a patient in delivery room we need to use a torch (Male nurse)”. Need of ambulance When an ambulance is needed the healthcare workers call the district hospital. In many cases the ambulance is not available due to shortage of cars and poor roads. An informant said: “Sometimes for emergency evacuation of a patient to the district hospital the ambulance is not available (Female, midwifeassistant)”. Needs of easy access to time saving guidelines All informants mentioned the need for quick access to medical guidelines or advice. Today they mainly search for information, especially about diagnosis, in books, which is time consuming and the information is difficult to find, especially during on-going consultations with patients present. So to get quick answers, they preferred calling colleagues using

Table 5 – Expectations and attitudes for use of the CDSS (Category C). C.1 CDSS with updated guidelines

C.2 CDSS as help for good and high quality care

C.3 CDSS for training and gaining access to knowledge

Exciting with CDSS and IT

Help for: • Making prompt and correct decision • Avoid diagnostic failure • Managing delivery

Opportunity for: • Continued medical training • Administrative work

Need update of guideline

mobile phones. Some informants argued that information in posters with treatment algorithms can be easily accessed to be part of advice to act by just a glance. Informants stated: “When we get problems we call the head of maternity (the midwife) or the head of the facility (Female midwife-assistant)”. “I was in one facility where they have a poster on the breast diseases during the breast-feeding on the wall, so it was very easy for us to see treatment recommendations. Here we need to check guidelines in books, which is not easy especially when we are face to face with a patient (Female midwife-assistant)”. C. Expectations and attitudes of using CDSS Most of the informants had an expectation that CDSS should solve several of the problems in work. Therefore, they were keen to use the CDSS and showed a positive attitude and had great expectations on testing the CDSS (Table 5). C.1. CDSS with updated guidelines Excited to use CDSS and IT The informants were excited of the possibilities to use CDSS. They described the CDSS as an opportunity for them to learn how to use computers and be in the frontline of using new technology for healthcare in a rural area in Burkina Faso. An informant described that: “We need this system, Nowadays electronic tools are the solution, because in most areas we do not have access to paper and book. Therefore I think that we should go ahead with this new technology (Female midwife-assistant)”. Needs of updated guidelines Informants explained that new guidelines in maternity care are needed, as the available guidelines are old and not updated according to the new national directives in maternal care. As the science is in continuous changes the existing guideline has to be updated. Informants explained that: “We need electronic access to guidelines, because with computer it will be easy for us to find advice (Male nurse)”. “We need new guidelines, because the guidelines we are using now are very old (Male midwife)”.

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

Table 6 – Reported barriers to use the CDSS (Category D). D.1 Barriers related to the use of CDSS in daily practice

D.2 Barriers related to the users

Administrative “double” work, e.g. document all activities both in paper and in the CDSS. Prolonged consultations visits

Computer illiteracy

No financial incentives to health care workers to use the CDSS

C.2. CDSS as help to provide good and high quality care

D.1. Barriers related to the use of CDSS in daily practice Administrative “double” work and prolonged consultation visits Informants mentioned that they have some doubts about the way to use the CDSS. They noticed that instead of using only the CDSS, they would have to document in parallel all activities performed in a register and in other documents. So they are afraid that these double activities will introduce new time-consuming work. The informants said if the CDSS took time from patient work they would not use it. In maternal care, time to make a decision has to be less today than yesterday. Otherwise a disaster can quickly occur. They explained that the introduction of CDSS in healthcare would need changes in the organization of work. An informant said: “During the test of the CDSS, we saw that the software prolonged the consultation time. I mean it takes a lot of time to fill all the information asked. If we need to document information on paper and enter data on computer, you must motivate us otherwise it will not be easy (Male nurse)”.

The informants view CDSS as a tool for helping them in their work to make prompt and correct decisions during the patient visits, e.g. avoid diagnostic failures. Furthermore, informants mentioned that CDSS will help them with graphically and user-friendly partograph curves to manage delivery. An informant explained that:

“My question now is with this new system shall we continue to fill paper and use computer? My suggestion is that we go through with computer and stop filling paper otherwise it would be very tricky and time consuming to use the CDSS (Female midwifeassistant)”.

“. . . If we are following women during ANC to delivery, it will be easy for us with the computer to have all information about the pregnancy and to conduct delivery easily (Male nurse)”. C.3. The use of CDSS for training and gaining access to knowledge

465

D.2. Barriers related to the users

Continued medical training The informants’ believed that they will use the CDSS as a motivating tool for learning. Many informants from the intervention district mentioned that the training module in the CDSS on maternal care is useful. They were impressed by the possibility of the CDSS to include recommendations on how to pursue documentation on maternal care, delivery and neonatal care. They noticed that the training module would be a good tool to improve their capacity and fill gaps in their knowledge of maternal care. Furthermore, they thought that when they become more diagnostically skilled, the referral cases to the hospital would be reduced as well as the number of calls for distance assistance from colleague. An informant said:

Computer illiteracy The informants emphasized the need of training to overcome their computer illiteracy. They express that the success and the use of the CDSS will depend on their ability to use computers. Two informants said:

“This system will be very welcomed. It will help us to improve our knowledge and will be a good tool for training (Female midwifeassistant)”.

Lack of financial incentives The informants pointed out that there are many challenges in daily work in rural Africa compared to other places. One motivating factor when they start using CDSS would be to have financial incentives. An informant said:

Administrative work Informants hope the CDSS will help them to make the monthly statistical reports of the facilities electronically. They argued the computer would make it easier to produce the report each month and submit it to the district management staff. An informant explained that: “This system will help us to write our monthly report easily (Male nurse-assistant)”. D. Barriers to use CDSS Informants talked about different conceivable barriers when using the CDSS. These barriers could be related to the system but also to the users (Table 6).

“We need training to be able to use computers but also to be able to use this new system (Male nurse)”. “It will be good to have regularly meetings to share experience with all facilities that will use the CDSS. Retraining in computer basic skill and with the CDSS will be good as most of us are inexperienced in using computers” (Female midwife-assistant).

“. . .We need also good condition for work, like . . . financial incentives (Female midwife-assistant)”.

4.

Discussion

There are four main findings in this study: (a) the fascination among healthcare workers for and a great willingness to adapt and use modern technologies like computers and learn more in the workplace, (b) a positive attitude to easy access to guidelines and implement decision-support using computers in the workplace, (c) a fear that the CDSS requires more working time and leads to double-work, and (d) that the CDSS is

466

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

complicated and requires substantial computer training and extensive instructions. These findings can be divided into aspects of motivators (a and b) and barriers (c and d) in relation to how the CDSS usage is perceived. These aspects are closely connected to each other as the motivating aspects easily can be turned into barriers if not properly taken care of in the final design of the CDSS and during the implementation of the system at the point of care [5,8,13]. As an example, the informants conveyed an enthusiastic attitude to CDSS as they see the system as a facilitator to learn themselves. They believed this would be achieved by getting easy access to guidelines and data about the patients. However, this motivating factor can quickly be turned into barrier if the informants’ see that the CDSS requires more working time or if they do not understand how to use it properly. This is in accordance with previous reports [13,23]. There was a surprising interest among our informants to get access to new technologies and tools via CDSS as it was believed this would enable them to keep up with developments and scientific progress. This interest is in line with considerations and findings from other studies and reviews related to uptake of ICT tools including mobile phones among healthcare workers in rural Africa [8,13,17,24–26]. In a questionnaire study in Ghana and Tanzania as part of the QUALMAT project, rural healthcare workers scored as high as a mean value of 4.5 (5-point Likert scale) when rating the value of computers for their daily work [26]. Enthusiasm for and acceptance of new technologies in healthcare to provide immediate help has also been documented for rapid diagnostic tests used by rural village health workers in Tanzania [22]. It is apparent that during development and implementation of new CDSS for rural healthcare settings, it is important to build on the enthusiasm for new technologies and the need to get easy access to information [4,5]. Adherence to clinical guidelines is essential for quality improvement in healthcare. Despite efforts in developing and disseminating these guidelines, healthcare workers commonly lack access to them. They can also ignore them, explained by problems with finding appropriate advice and assistance rapidly when providing maternal and neonatal care [4,5,13,27,28]. A few studies have shown the use of medical records and CDSS can improve healthcare workers performance in developing countries [5]. One success story is the medical record system used in Malawi to provide clinical overviews of treated HIV-patients with embedded decision support tools [13,24]. The success of this system is most likely due to the use of simple solutions, repeated improvements, training of staff in computer use and focusing on easy-touse graphical interfaces [13,24]. Our findings on needs and attitudes to a CDSS for maternal care are in line with how the Lilongwe medical record system has been designed and implemented. In our study, the informants explained that they wanted easy access to up-dated guidelines as those used today, often in books, are old and lack user friendliness. This is reasonable since all the facilities in this study lacked healthcare workers, often without any physicians or obstetricians available for questioning and advice and the need to rely on easy and rapid access to pedagogic recommendations [3,5,28].

It is not surprising that the informants reported that they have limited computer experiences and need substantial training. It is well documented that computer literacy skills are an important determinant in the decision to use the CDSS, or not, in both rich as well as in poor countries [10,13,16,24,26]. This should be of even higher importance in rural settings with lack of computer skills as we report. Eighty percent of healthcare workers were recently reported as either computer illiterates or beginners in Ghana and Tanzania [27], making it mandatory to introduce computer training when implementing computer tools in rural healthcare settings. To overcome poor computer literacy skills among healthcare workers in Burkina Faso, the QUALMAT project is going to use different strategies. Firstly, all users will be trained in basic computer skill and how the CDSS works. Secondly, a training module with instructions of how to use the system is incorporated in the CDSS. In this module, healthcare workers can see and read instruction on the systems different parts. Furthermore, the users will receive a visit of an expert of the system at least twice per month. These visits are to help healthcare workers in case of computer problems but also to train them in the use of the CDSS. The users can also use mobile phones free of charge to call the expert during the initial phase and thereby get advice online. The users’ expectation of what the CDSS can do for them in work is highly important for their subsequent use in practice. In this study, the informants wanted to learn more about how to use the partograph as a tool to monitor the progress of labour and maternal/foetal well-being so they know if immediate or later referrals are needed. Based on our results, we conclude that the value of a CDSS in rural healthcare institutions does require that the system can deliver information rapidly, without interruptions, with no double work. In addition, using strong priorities of access to information graphically and in charts. This is in line with other recent reports from Malawi, Ghana and Tanzania [13,24,27]. We found that the informants considered it critical to have a suitable and sustainable electricity infrastructure like solar panels or rechargeable batteries for the CDSS. From earlier studies in rural Africa, we know that solar panels do work properly but they need good installation and upkeep to be functional over the long term and avoid repeated interruptions in power supply [8]. The Lilongwe system has apparently considered all these technical challenges by using computers operating on low-voltage systems [13,24]. Our results are in accordance with one validated model to predict and explain end-users acceptance and use of information technology (IT) in healthcare called the technology acceptance model (TAM) [16,29–31]. Originally, the TAM was developed in the 1980s to understand how ICT-systems were taken up and used by workers in factories in US [29–31]. The model has two determinants: perceived ease of use and perceived usefulness [29–31]. Perceived ease of use is defined as “an individual’s perception that using an IT-system will be free of effort” and perceived usefulness is defined as “an individual’s perception that using an IT-system will enhance job performance” [29]. According to the model, these two determinants interact with each other as user’s perceived ease of use can impact positively on the perceived usefulness of the IT-system and create a positive attitude to the end-users acceptance of the

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

IT-system [16,29–31], i.e. users who have a perception of ease of use of IT-system can automatically perceive usefulness of a system. In our study, the motivational factors (e.g. positive attitude to learn more to perform better in work) could be related to the dimension of ‘usefulness’ of the CDSS and the factors of barriers (e.g. fear that CDSS take more time from consultation with the patient and create administrative double-work) related to the ‘ease of use’ of the CDSS. Knowing these factors among healthcare workers perception of barriers before implementation of the CDSS would allow the project to address these barriers, enhance healthcare workers acceptance of CDSS and subsequently increase its use in practice. It is thus critical to consider the healthcare workers perception of barriers before implementing any CDSS system in clinical practice. We believe further studies are needed to ascertain the acceptability of CDSS and their subsequent usage in changing behaviours [5,27] to provide future guidance.

6.

467

Conclusions

Healthcare workers in rural maternal care in Burkina Faso are enthusiastic for easy access to guidelines through CDSS and appreciate the learning potentials of such systems. However, there are fears regarding the extra workload and the interviewees foresee a need for training in knowledge and skills of handling computers. These aspects are closely connected to each other as the motivating aspects can easily be turned into barriers if not taken care of properly in the final design of the CDSS and during its implementation at the point of care. To overcome these barriers, the project has a responsibility to continuously follow-up the fit between the CDSS and the user’s needs during implementation, which was concluded in the questionnaire study of computer knowledge among rural healthcare workers [26].

Authors’ contribution 5.

Methodological considerations

In qualitative interview studies, twenty informants are seen as usually enough to capture the potential variation in informants’ ways of experiencing a phenomenon [32]. The method used, involving interviews [18] and content analysis [21], is well established and the researchers are familiar with these methods and should therefore give consistent and reliable results. We choose to interview two groups of healthcare workers from different districts in Burkina Faso. Consequently, the forty-five informants in this study should be sufficient to reflect and describe informants’ view on the CDSS. Despite the fact that healthcare workers in Nouna got a demonstration and training in the use of the CDSS and the informants from Solenzo just got information about the CDSS system, they both expressed and reported the same perception of motivators and barriers to the CDSS. This view of consensus might be explained by the context in rural Burkina Faso where these groups of healthcare workers all have limited medical education and are dependent of a doctor or a gynaecologist in decision making in maternal care. To decide the level of analysis of data there is always a challenge of how the data was collected and by whom. In this study, we interviewed healthcare workers about their thoughts of the CDSS. In all interviews of such kind there is always a risk of recall bias, i.e. that the informant may state what they believe that the interviewer wants to hear. To avoid this, we use three broad questions were the informants had to reflect on own experiences from the facility. To stay close to the informants own experience, we decided to stop the analysis at the manifest level [21]. This focuses on what the informants’ actually say in the interview i.e. the lowest level of interpretation. As a qualitative study, the findings could be transferable to similar settings in other resource low countries, and cannot be generalized to the whole population. However, the results of this study could be used as a basis for constructing questions for a larger questionnaire study to be able to test the TAM in a rural African setting.

All authors contributed to the conception and design of the study and to interpretation of the data. PBR guided the development of the interview guide and the coding framework together with SAZ. SAZ conducted all interviews and analyzed the data guided by PBR and with quality controls and inputs from all authors. PBR, SAZ, SL, JN, AS, LLG and GT outlined the disposition of the manuscript. The first draft was written by PBR and SAZ and completed by PBR, SAZ and LLG with major inputs by GT. The final manuscript was circulated among all authors that approved the final manuscript.

Conflict of interest None declared.

Acknowledgments The QUALMAT research project (Quality of Prenatal Care and Maternal Care: Bridging the Know-do Gap) funded by the 7th Framework Programme of the European Union (Grant agreement 22982) as a collaboration between the Centre de Recherche en Santé de Nouna (Burkina Faso), Ghent University (Belgium), Heidelberg University (Germany), Karolinska Institutet (Sweden), Muhimbili University of Health and Allied Sciences (Tanzania), and Navrongo Health Research Centre (Ghana). The general objective of this research is to improve maternal health through better pre-natal and maternal care services offered by better motivated health workers. The intervention packages include the development and implementation of a system of performance based incentives and a computer-assisted clinical decision support system (CDSS) based on WHO guidelines. The interventions are evaluated in a pre-and post-controlled study design in rural Burkina Faso, Ghana and Tanzania between 2009 and 2014. In part this research was also funded by grants from Swedish Research Council to Lars L Gustafsson (VR-2011-3440) and through the EU-project eI4Africa (e-Infrastructures for Africa, grant agreement 313582). For more information, please refer to the website of the project (www.qualmat.net).

468

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

Summary points What was already known on the topic: • Perceived usefulness and reported ease of use of computerized e-health applications will determine to what extent they are used in high-income countries. • User-friendly guidelines integrated in CDSS have been shown to improve the quality of care in high-income countries. What this study added to our knowledge: • Healthcare workers in rural maternal care in Burkina Faso are enthusiastic for easy access to guidelines through CDSS, availability of a graphical tool such as partograph and appreciate their learning potential. • There are fears regarding the extra workload and interviewees foresee the need for training and skills in handling of computers to enhance the use of any CDSS. • Aspects of users’ adoption behaviour to CDSS in rural healthcare settings can be divided into motivators and barriers and these aspects are closely connected.

references

[1] Trends in maternal mortality: 1990–2010. WHO, UNICEF, UNFPA and the World Bank estimates World Health Organization 2012, Geneva, 2012, Available from http://whqlibdoc.who.int/publications/2012/9789241503631 eng.pdf [2] R. Lozano, H. Wang, K.J. Foreman, J.K. Rajaratnam, M. Naghavi, J.R. Marcus, L. Dwyer-Lindgren, K.T. Lofgren, D. Phillips, C. Atkinson, A.D. Lopez, C.J. Murray, Progress towards Millennium Development Goals 4 and 5 on maternal and child mortality: an updated systematic analysis, Lancet 378 (378) (2011) 1139–1165. [3] A.K. Rowe, D. de Savigny, C.F. Lanata, C.G. Victora, How can we achieve and maintain high-quality performance of health workers in low-resource settings, Lancet 366 (2005) 1026–1035. [4] U. Baker, G. Tomson, M. Somé, B. Kouyaté, J. Williams, R. Mpembeni, S. Massawe, A. Blank, L.L. Gustafsson, J. Eriksen, How to know what you need to do: a cross county comparison of maternal health guidelines in Burkina Faso, Ghana and Tanzania, Implement. Sci. 7 (2012) 7–31. [5] A. Blank, H. Prytherch, J. Kaltschmidt, A. Krings, F. Sukums, N. Mensah, A. Zakane, S. Loukanova, L.L. Gustafsson, R. Sauerborn, W.E. Haefeli, Quality of prenatal and maternal care: bridging the know-do gap (QUALMAT study): an electronic clinical decision support system for rural sub-Saharan Africa, BMC Med. Inform. Decis. Mak. 13 (2013) 44. [6] S. Etzo, G. Collender, The mobile phone ‘revolution’ in Africa: rhetoric or reality? Oxf. J. Soc. Sci. Afr. Aff. 109 (2010) 659–668. [7] J. Barrington, O. Wereko-Brobby, P. Ward, W. Mwafongo, S. Kungulwe, SMS for life: a pilot project to improve anti-malarial drug supply management in rural Tanzania using standard technology, Malar. J. 9 (2010) 298. [8] H. Lucas, Information and communications technology for future health systems in developing countries, Soc. Sci. Med. 66 (2008) 2122–2132.

[9] S.A. Pearson, A. Mosey, J. Robertson, I. Hains, M. Williamsson, J. Reeve, D. Newby, Do computerised clinical decision support systems for prescribing change practice? A systematic review of literature (1990–2007), BMC Health Serv. Res. 9 (2009) 154. [10] B. Eiermann, P. Bastholm-Rahmner, S. Korkmaz, B. Lilja, T. Shemeikka, A. Veg, B. Wettermark, L.L. Gustafsson, Knowledge databases for clinical decision support in drug prescribing – development, quality assurance, management, integration, implementation and evaluation of clinical value, in: C.S. Jao (Ed.), Clinical Decision Support, INTECH, Croatia, 2010, ISBN:978-953-7619-64-00, Available from www.intechopen.com/books/decision-support-systems/ knowledge-bases-for-clinical-decision-support-in-drugprescribing-development-quality-assurance-mana (30.04.13). [11] B. Sjöborg, T. Bäckström, L.B. Arvidsson, E. Andersén-Karlsson, L.B. Blomberg, B. Eiermann, M. Eliasson, K. Henriksson, L. Jacobsson, U. Jacobsson, M. Julander, P.O. Kaiser, C. Landberg, J. Larsson, B. Molin, L.L. Gustafsson, Design and implementation of a point-of-care computerized system for drug therapy in Stockholm metropolitan health region – bridging the gap between knowledge and practice, Int. J. Med. Inform. 76 (2007) 497–506. [12] D.L. Hunt, R.B. Haynes, S.E. Hanna, K. Smith, Effects of computer-based clinical decision support systems on physician performance and patient outcomes: a systematic review, JAMA 280 (1998) 1339–1346. [13] G.P. Douglas, O.J. Gadabu, S. Joukes, S. Mumba, M.V. McKay, A. Ben-Smith, A. Jahn, E.J. Schouten, Z. Landis Lewis, J.J. van Oosterhout, T.J. Allain, R. Zachariah, S.D. Berger, A.D. Harries, F. Chimbwandira, Using touchscreen electronic medical record systems to support and monitor national scale-up of antiretroviral therapy in Malawi, PLoS Med. 7 (2010) 1000019. [14] F. Rabbani, S.N. Lalji, F. Abbas, S.W. Jafri, J.A. Razzak, N. Nabi, F. Jahan, A. Ajmal, M. Petzold, M. Brommels, G. Tomson, Understanding the context of balanced scorecard implementation: a hospital-based case study in Pakistan, Implement. Sci. 6 (2011) 31. [15] J. Lanrewaju, M.A. Pate. L. Lin, K.A. Schulman, A model for the adoption of ICT by health workers in Africa, Int. J. Med. Inform. 81 (2013) 773–781. [16] R.J. Holden, B.T. Karsh, The technology acceptance model: it past and its future in health care, J. Biomed. Inform. 43 (2010) 159–172. [17] T. Oluoch, X. Santas, D. Kwaro, M. Were, P. Biondich, C. Bailey, A. Abu-Hanna, N. deKeizer, The effect of electronic medical record-based clinical decision support on HIV care in resource-constrained settings: a systematic review, Int. J. Med. Inform. 81 (2012) 83–892. [18] C. Pope, N. Mays, Introduction to qualitative methods in health and health services research, BMJ 311 (1995) 42–45. [19] EU-project QUALMAT – Quality in maternal care, 2014, Described at www.qualmat.net (accessed 02.09.13). [20] B. Chaudhry, J. Wang, S. Wu, M. Maglione, W. Mojica, E. Roth, S.C. Morton, P.G. Shekelle, Systematic review: impact of health information technology on quality, efficiency, and costs of medical care, Ann. Intern. Med. 144 (2006) 742–752. [21] U.H. Graneheim, B. Lundman, Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness, Nurse Educ. Today 24 (2004) 105–112. [22] M. Mubi, A. Janson, M. WarsameM., A. Martensson, K. Kallander, B. Ngasala, G. Magana, L.L. Gustafsson, A. Massele, G. Tomson, Z. Premji, A. Bjorkman, Malaria rapid testing by community health workers is effective and safe for targeting malaria treatment: randomised cross-over trial in Tanzania, PLoS ONE 6 (2011) e19753.

i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 3 ( 2 0 1 4 ) 459–469

[23] Q. Wan, M. Makeham, N.A. Zwar, S. Petche, Qualitative evaluation of a diabetes electronic decision support tool: views of users, BMC Med. Inform. Decis. Mak. 12 (2012) 61. [24] G.P. Douglas, R.A. Deula, S.E. Connor, The Lilongwe central hospital patient management information system: a success in computer-based order entry where one might least expect it, AMIA Annu. Symp. Proc. (2003) 833. [25] C.O. Jones, B. Wasunna, R. Sudoi, S. Githinji, R.W. Snow, D. Zurovac, Even if you know everything you can forget: health worker perceptions of mobile phone text-messaging to improve malaria case-management in Kenya, PLoS ONE 7 (2012) e38636. [26] F. Sukums, N. Mensah, R. Mpembemi, A. Blank, J. Kaltschmidts, W.E. Haefeli, Health workers’ knowledge of and attitudes towards computer applications in rural African health facilities, 2013, (under review). [27] T.D. Some, I. Sombie, N. Meda, Women’s perception of homebirths in two rural medical districts in Burkina Faso: a qualitative study, Reprod. Health 8 (2011) 3.

469

[28] G. Elwyn, A. Edwards, P. Kinnersley, R. Grol, Shared decision making and the concept of equipoise: the competences of involving patients in healthcare choices, Br. J. Gen. Pract. 50 (2000) 892–899. [29] V. Venkatesh, M. Morris, G. Davis, User acceptance of information technology: toward a unified view, MIS Quart. 27 (2003) 425–478. [30] W.W. Kushniruk, D.R. Kaufman, V.L. Patel, Y. Lévesque, P. Lottin, Assessment of a computerized patient record system: a cognitive approach to evaluating medical technology, MD Comput. 13 (1996) 406–415. [31] F.D. Davis, Perceived usefulness, perceived ease of use, and user acceptance of information technology, MIS Quart. 13 (1989) 319–340. [32] J. Sandberg, Understanding human competence at work: an interpretive approach, Acad. Mange. J. 43 (2000) 9–25.