Barriers and facilitators to antenatal and delivery care in western Kenya: a qualitative study

Study area

The study was carried out in the rural area of Asembo in Rarieda district, western Kenya. With a midyear population of 64,442, recent estimates of maternal mortality ratio in this area are between 524 and 847 per 100,000 live births for the period 2003–2008 whilst infant mortality ranges from 76 – 132 per 1000 live births over the same period [13]. HIV and malaria rates are high, and anaemia is also a common problem. The population are predominantly members of the Luo tribe who are subsistence farmers or fishermen. Within the study area there are 7 health facilities (3 health centres, 3 dispensaries and 1 mission hospital) all of whom offer antenatal care. Women have to travel a maximum distance of 5 km to reach any facility.

Recruitment and participants

Ethics

The study was approved by the Kenya Medical Research Institute (KEMRI), the U.S. Centers for Disease Control and Prevention (CDC) and the Liverpool School of Tropical Medicine.

Analysis

Thematic analysis was used to identify a narrow range of themes reflecting the textual data [14], using the following steps: The English transcripts were read several times by 2 of the authors (LM and SD) in order to familiarise themselves with the raw data. Neither author had been involved in data collection and considered themselves to be impartial in relation to study expectations. Each transcript was then entered onto Nvivo 9 where an initial skeleton coding frame was set up using the key themes identified by the authors. The coding frame comprised the major issues identified as barriers or facilitators to utilising either ANC or delivery care. Each transcript was then coded separately by the 2 authors according to the framework. On completion, they each revisited the framework and divided further using emergent subthemes. The narratives were then recoded where relevant, under these subthemes. The 2 separate completed coding frames were then compared to check consistency across them. Only minor differences were noted, which were discussed and agreement reached. The data for each theme and subtheme were then pieced together by one of the authors (LM) to provide an overview of the content relating to that specific theme. Quotes were chosen to represent a typical response, unless where stated, to illustrate a deviant opinion.

Facilitators for health facility based antenatal care

Attending a health facility for ANC was viewed very positively by women across all of the FGDs with many reasons put forward as to why. Three elements of care seemed to be most important and raised by all groups: testing for disease, checking the position of the foetus, and receiving injections and/or medications.

Barriers for health facility based antenatal care

Four barriers to attending a health facility for ANC were predominant, mentioned by many of the women across the different groups. These were HIV testing, attitudes of clinic staff, which was linked to another barrier – long clinic waiting times, and cost.

This view was not however held by all; a couple of women when asked specifically, refuted this, reporting that they received care in good time, and did not spend a lengthy time at clinic.

A couple of women mentioned transport costs as an issue, and there was disagreement within groups as to whether the health facilities were far, although any distance was recognised as a problem when the women were sick and had to walk.

Facilitators for health facility based delivery

Just one participant mentioned being HIV positive as a reason for a health facility delivery.

However, not all women agreed on health facility as the best place to deliver – it was suggested that a healthy woman would go to a TBA, ‘although a sick mother would deliver in the hospital’. A minority preferred to give birth at home (or at an older relative’s home) although sometimes this was simply for financial reasons, whilst just occasionally the view was expressed that it did not really matter where delivery took place – it was Gods will as to the outcome.

Barriers to health facility based delivery

Although almost all participants agreed that health facilities were the safest place to deliver, a range of factors influencing the place of delivery and limiting women’s use of SBA were identified. These included: access to health facilities, which was influenced by unpredictable timing of labour and transport, as well as cost of facility-based delivery care, husbands’ preference, health facility staff attitudes and previous experiences and habits.

From all FGDs it was suggested that although TBAs had better interpersonal skills and were more accessible than health facilities, they had serious limitations such as not having adequate equipment for “safe delivery” such as gloves or medicines for PMTCT of HIV and were not able to help in case of obstructed labour when a caesarean section is required.

As with ANC, nurse attitudes were a barrier to health facility delivery, although to a lesser extent with fewer women mentioning this as a problem.

Predicting readiness to act using a theoretical framework

We have used the “Health Belief Model” to assess our study participants ‘readiness to act’ in terms of attending a health facility for ANC and delivering within a health care setting [15]. The model was developed to explain behaviours but is also useful for health care programming in designing change strategies. The health belief model suggests that health-related behaviour depends on the individuals perception of four concepts:1. the severity of a potential illness, 2. the person's susceptibility to that illness, 3. the benefits of taking a preventive action 4. the barriers to taking that action. Positive implications for the future use of facility based antenatal and delivery care in the local area emerged from our study suggesting that change is possible, although strategies are needed to activate ‘readiness’ (cues to action). Women suggested that ANC services offered would be beneficial for their health and that of their baby. Perceived benefits of ANC were healthier pregnancy, healthier infant and a registration card or a bed net. Cost of care, nurses’ attitudes, long waiting times and HIV testing were perceived barriers to attending clinic. Most, though not all perceived susceptibility to complications, and understood the benefits of delivering in a health facility as complications would be managed there, however, as Thaddeus and Maine reported as far back as 1994 [11] cost was a substantial perceived barrier to delivery in a health facility. Importantly, results from our study indicate that many women are receptive, in theory, to using formal ANC and delivering within a healthcare facility. Cues to action are now needed in order to ensure these women are captured within the healthcare system that provides a continuum of care

Failing to meet WHO recommendations

The present study provides explanations as to why most women in this area do not present at an ANC in their first trimester or attend the recommended 4 FANC visits [9]. Our results suggest that women wait until near the end of their pregnancy rather than first or even second trimester so that they have more time to gather the money to pay for any tests, also waiting until the last minute to get the card to present in case they had a health facility delivery. This also means that they do not have the time to attend for the recommended number of visits. Women were mostly positive about receiving ANC, recognising the importance of having a health check and receiving medications. However, once ascertained that they and the baby were fine, and having received a registration card, and/or a bed net cost, waiting time or nurses attitude were a greater barrier, following prior bad experience, or the reluctance to spend additional time away from home duties or additional cost, once the first visit had been made. This finding resonates with a recent systematic review [16] which reported that some women do not return to ANC once they have received their free ITN.

Lack of knowledge negated by trust

We were not able to ascertain whether women were aware of the need for 4 visits but speculate this is unlikely because their knowledge appeared limited. Whilst mentioning the importance of receiving medications or injections none appeared to fully understand what these were for, an issue reported elsewhere [17,18], whilst a recent systematic review [16] found women were unaware of the benefits of intermittent preventive treatment (IPTp) with sulphadoxine-pyrimethamine (SP) for malaria, or why SP was being given. Vague knowledge of the services offered at ANC was also corroborated in a household cross-sectional survey carried out in the KEMRI/CDC HDSS in 2010 by Ouma P, Were V, Hamel M, Desai M. Results from the IPTp Uptake Cross-Sectional Survey [Unpublished]. KEMRI/CDC. 2010 ( [19]unpublished data). Women who were recently pregnant in the last 2 years were asked about information received at ANC. Just 56% reported that nurses providing care explained what they were doing. Despite this, women’s acceptance of medication appeared to demonstrate trust in both the facility based health care system, and in western medicine. Despite not knowing what they were for, participants took them believing they would help their, or the baby’s health. Likewise, Essendi et al. [19] reported women in Nairobi trusted health personnel implicitly, whilst a further study in Mombasa [20] similarly reported that women used and preferred western pharmaceuticals during their pregnancy. These findings contrast with other studies that show a distrust for modern medicine, and a stronger preference for traditional medicines and carers [21,22]. The study area has a high prevalence of malaria, also HIV, anaemia and helminth infections [13,23] necessitating screening during pregnancy. Malaria and HIV were the only infections/diseases mentioned often by pregnant women in our FGDs, and tetanus immunisation was mentioned just once. Increasing knowledge of services offered at ANC and their benefit may help earlier ANC attendance and compliance with the recommended minimum number of visits.

Issues of male support

Women perceived their husbands provided little support, either financially or in terms of help with household duties, during pregnancy and delivery. This finding contrasts with the results from our FGDs among men, which focussed on the perspectives of men on antenatal and delivery care service utilisation in the same region [24]. In this parallel study, men were positive in their views of antenatal and delivery care, and as decision makers they stated they encouraged, some even ‘forced’, their wives to attend health facility for antenatal or delivery care. Further work would be useful to ascertain the true influence that men have on their spouses healthcare.

The influence of healthcare provider attitudes

Our findings on service dissatisfaction contradict the result from the cross-sectional survey by Ouma P, Were V, Hamel M, Desai M. Results from the IPTp Uptake Cross-Sectional Survey [Unpublished]. KEMRI/CDC 2010 [19] (unpublished data), which showed 90% of respondents felt welcomed to ANC by the nursing staff. Furthermore, 71% felt they were encouraged to ask questions and 97% had them answered satisfactorily. In contrast, participants from our FGDs reported that the attitude of the nurses, as well as long clinic delays had the effect of preventing return visits. The difference in findings could be explained by the different methodological approaches used, the individual face to face survey being more prone to courtesy bias [25] compared to FGDs where participants are encouraged to share their feelings freely amongst their peers. Complaints about poor staff attitude has also been reported in studies in other low and middle income countries [26,27] and specifically in Kenya [20,28,29]. Common complaints about health care providers are lack of care, rudeness or harshness of attitude. Whilst nursing is a vocation, some may be attracted by the prestige associated with working in the health sector. Ojwang et al. [28] described nurses as ‘all knowing benefactor’, and the perception that some nurses feel superior or ‘so learned’ was made by some women in our study. A poor attitude to patients by healthcare workers has long been recognised as a problem in the Kenyan setting, with the Kenya Ministry of Health declaring commitment to redress this issue. However, a study carried out in Kisumu, western Kenya [29] highlighted that the reduction in numbers of staff, coupled with increased workload led to burnout, which health workers admitted led to impatience and irritation with clients. There was some suggestion that nurses attitude or work ethic contributed to the long waiting times that were mentioned by many as a barrier to accessing ANC, however we are unable to assess whether this is a major factor, or whether this is primarily due to lack of organisation whereby women can turn up to clinic at any point during their pregnancy. Numbers can’t be predicted, neither can the nature of their individual care meaning health care professionals may have to perform a variety of roles in any session. This may be compounded by a shortage in healthcare workers in the Kenya health system which has been reported elsewhere [30-32]. Clearly however, investment is needed in staff numbers, alongside a shift in attitude of the healthcare workers with a parallel shift in the negative perceptions of women towards health care workers.

HIV testing – barrier and facilitator

Studies [33,34] have demonstrated that HIV testing is a barrier to attendance at ANC, resulting from a fear of being tested because of involuntary disclosure to others, fear of stigma, and fear of finding out that they have the disease. These reasons were similarly important in our study, although participants also reported HIV testing as a reason for attending for care. It is likely that knowledge and use of ARVs and PMCT will mean that HIV testing becomes less of a barrier and more of a facilitator to antenatal services in the future. Indeed, Hardon et al. [35] in a 4 country study conducted reported PMCT services were valued because they protect the health of the unborn child and provide treatment for the mother. Kinuthia et al. [36] reported stigma was less of a barrier to using PMTCT than health service related issues.

Cost – a significant barrier

As reported in numerous studies [16,19,37,38], cost was a barrier to ANC attendance, and to delivery in a formal healthcare setting. However, unlike other Kenyan settings where additional factors such as women’s lack of autonomy [19,27,39], HIV testing [33,37], and baby theft [37] including insecurity at night [19] had a strong negative influence, cost and transport were the predominant barriers arising through our FGDs. Women were aware that in cases where pregnancy complications arose, the healthcare setting was really the only place where they could deliver safely. What prevented them from doing this, was predominantly the cost. However, Ouma’s cross sectional survey which reported on actual behaviours, found the dominant reason for not delivering in a health care facility was that the labour/delivery came too fast. This was similarly reported in a recent Ghanaian study, and although 94% gave birth at home, only one quarter planned to do so [40].The timing of the onset and the unpredictability of labour was also mentioned as a barrier to deliver at a health facility in our FGDs. Individualized birth planning and preparedness is an integral part of FANC. However its implementation in Kenya has been poor, estimated at 68% in 2010 [41]. Full implementation of individualized birth planning could improve rates of delivery with a SBA as shown by Mpebemeni et al. [42]. Subsidizing cost of both transport and delivery fees are also likely to act as facilitators to delivery care. However, whilst these solutions might help create more demand for ANC and delivery with a SBA this also has to be balanced against inadequate supply. Currently, in western Kenya there are inadequate numbers of facilities able to provide quality delivery services, and a shortage of equipment and supplies for obstetric care [43].

Study limitations

Some limitations need to be considered. Whilst our study included a number of focus groups, with women across a large age range our findings can be generalised only to the parent population, and not across other settings where different health care, geographical or cultural factors may affect access and utilisation of ANC and delivery services. It is possible, as with all focus group discussions in general, that the opinions given do not reflect each individual but rather the collective norm of the wider group. So whilst there appeared to be consensus around many issues within groups this may not reflect the true picture. However, similar opinions across groups reassure us that we have captured at least the prevailing views of women in Asembo. Furthermore, whilst it is possible there was an element of social desirability bias in response, with participant awareness that formal healthcare provision is being promoted by the government in preference to informal healthcare provided by the TBA, we uncovered opinion that disagreed the hospital is the best place in which to deliver. This suggests that some of the women felt comfortable in proffering opposing opinion. Constraints on time meant we were unable to explore some issues in depth, which may have added further value to the study. This includes exploring what the women understood by complications and their knowledge and recognition of danger signs, both during pregnancy and during labour. It would be useful to investigate whether they understand the need for planning ahead and what prevents any plans being put into practice successfully. Further understanding of the causes and possible solutions of long delays in ANC could help to address one of the main barriers to repeat ANC care.