‘I stayed with my illness’: a grounded theory study of health seeking behaviour and treatment pathways of patients with obstetric fistula in Kenya

The study participants were recruited from three fistula repair centres in Kenya—namely, Kenyatta National Hospital, Kisii Level 5 Hospital and Gynocare Fistula Centre—in 2013. Kenyatta National Hospital is a national referral hospital located in Nairobi city that holds at least one free fistula surgical camp annually to an estimated 400 patients. Kisii Level 5 Hospital is a public regional referral facility located in the southwestern part of the country and holds a free fistula surgical camp annually to about 50 patients. Gynocare Fistula Centre located in the North Rift holds an estimated 300 fistula surgeries throughout the year. We recruited patients who had undergone corrective surgery for fistula caused by labour and childbirth (obstetric fistula) as reported by the women from three hospitals. Women were approached by the researchers (AMK) and research assistants (ML and MO) during their inpatient stay and offered information about the study. Upon written informed consent, their permission to voluntary participate in the study was obtained. Only participants who voluntarily accepted participated in the study. A number refused participation though they were not obliged to give us their reasons. Those who refused to participate were not recorded.

Ethical clearance for the study was obtained from the University of Nairobi and Kenyatta National Hospital ethics and research committee (ERC) project number P618/11/2012. A standard written consent information sheet was used to explain the scope of the study and their role in participation, and the long term follow up nature of the study. Women gave verbal and written informed consent to participate in the study. They were made to understand that participation was voluntary and those who did not participate in the study would still receive care as planned by the clinical team. Only those participants who consented were recruited into the study.

This is not applicable since no photos, clinical images or videos of participants were collected nor are presented. Participants gave permission to use of a voice recorder in interviews and anonymous and un-identifying publication of the transcribed data resultant from the study. Participants’ anonymity has been guaranteed through labelling of their quotes using code numbers in place of names. In addition, their exact location is concealed and in its place, the name of the larger county. For instance, a participant labelled as (018_1, Siaya) means participant code number 018_1 lives in Siaya county.

The study used grounded theory methodology [22]. Primary data was collected using interview guide. Narrative approach allowed participants to construct the events surrounding the index birth and context of occurring fistula illness, living with fistula and their health seeking behaviour. These interviews were audio recorded. The transcripts were imported and managed within NVIVO 10 software. This paper reports the findings of the first phase of the study that focused on the health seeking behaviour and treatment pathways of fistula patients. Due to space limitations, phase 2 of the study which focused on the topic of reintegration of fistula patients is reported in a separate article. We included women of various ages, level of education, geographical location, type of facility and number of children in keeping with deviant case analysis in grounded theory methodology to describe all possible conditions of fistula patients across three hospitals in Kenya.

AMK a female PhD student and registered nurse midwife conducted the interviews with the help of ML and MO, female research assistants. The research team had training in qualitative health research in conducting qualitative research and hospital ethnography. In hospitals, the data collection research team wore normal clothes, no uniform nor badge, presenting as nursing student researchers. This naïve positionality enabled us to reduce the power positions clinician researchers have over the participant and improved interactions with participants. The researcher assumption was that women did not want to go to hospital hence long decades with illness. Later, we discuss how this assumption was refuted, as women lay claim to an unresponsive system than their unwillingness to seek treatment.

(AMK) and trained research assistants (ML and MO) conducted the interviews in a hospital setting using a narrative guide in Kiswahili. Additional file 1 shows the narrative guide that was used in the study. The first broad question was highly exploratory allowing the participant to start from when they delivered the baby and the subsequent steps they followed. Subsequent probes and follow up questions then steered the conversation further. The interview guide is attached. Interviews generally lasted for forty to ninety minutes depending on the length of the participant and their style of narration. There were no repeat interviews on the topic of health seeking behaviour. Instead, participants were let to tell their narrative and probes used to ascertain any facts they may have left out. The audio recording was translated from Kiswahili to English then transcribed by AMK assisted by YB a trained research assistant. The researchers kept field notes, observing and recording any detail information to further refine the research and help in interpretation and analysis. Concerning participant checking, transcripts were not returned to participants for this phase of the study. The nature of their short hospital stay (maximum 9 days) meant any clarifications would be sought during the second phase of the study. Rather, thorough probes and follow up questions were used on the interview guide to ensure clear and complete information. In addition, during the initial interviews, some participants were asked if they agreed with the perceived meaning of their interview data.

Purposive sampling was used to recruit patients whose vaginal fistula had been caused by childbirth and had undergone vaginal fistula surgery. The systematic nature of grounded theory led to analysing data from the initial 6 interviews at the Gynocare centre where data saturation occurred at 30 interviews. However, an additional 5 participants were recruited to cater for loss to follow up that may occur in a 1 year follow up study (that was focused on reintegration outcomes). There was need for a flexible approach to sampling due to the rapid nature of fistula camps held just for a week annually at Kenyatta and Kisii. In these two centres we took similar cue from the first study site leading to 30 participants at Kisii and 48 participants at Kenyatta with an additional 8 to total 56. Data saturation in the last study site was obtained using a larger sample because the participant characteristics were less homogenous and their narratives had broad variations. The sample size was guided by existing research on number of participants in qualitative studies, variation across the three study sites, the need for ample numbers to cater for loss to follow up and sufficient numbers for theory development [23]. Finally, to further develop certain emerging categories into theoretical construct, theoretical sampling was used to pursue and interview participants, whose narrative contained the chosen phenomena to fully describe it, for instance, ‘staying home with fistula illness’ as a health action in the health seeking pathway of fistula patients.

Grounded theory methodology following the classical Glaserian process of 3 stage coding – namely open, axial and selective axial coding - was applied [22]. One researcher (AMK) coded all data on health seeking behaviour, obtaining the emerging themes. Apriori themes were not set before interview and analysis. Open coding was conducted on initial 6 narratives where as many as 16 emerging themes (categories) and 67 subthemes were identified and their properties defined. Open coding generated first order and in vivo themes such as ‘I stayed with my illness’. Axial coding then linked the various categories with the subcategories. In selective coding, the findings were further integrated and refined into theoretical constructs. For instance, the invivo code ‘I stayed with my illness’ was further examined and fell under themes of living with fistula illness and the moral properties of suffering; as well as a strategy and (in)action in response to the illness. Selective coding of the transcripts led to the core category of health seeking behaviour pathway for fistula patients, that incorporates ‘staying with illness’ as a subcategory, alongside other core phenomenon such as ‘seeking the formal health facilities’. Eventually, more narratives dwelt on their experiences seeking formal health care making it a core phenomenon, whilst less and less instances of staying home arose. Another core phenomenon is the length of time spent living with fistula illness.

Throughout the study, we wrote memos that were case based and reflecting on what we have learnt in a particular interview. In addition, we wrote conceptual memos about an initial code or focused code, describing its meaning, comparing between data and asking questions that would need to be answered in further interviews. For example, a memo on the focused code ‘living with fistula’ stated:

“Themes of having psychological thoughts; social isolation; suffering overlap in the women’s narratives and allude to the moral properties of living with fistula. Although the women rarely refer to suffering directly, their narratives elicit feelings of helplessness, sadness, emotional pain, shame and at times crying during interviews. Occasionally, suicidal thoughts and feelings of (perceived) iminent danger of death from the illness permeate women’s narratives of their lives with obstetric fistula. These are properties that define their experience of suffering. At times, they refer to their illness as a 'problem' shida a term which in local Kiswahili dialect that refers to a situation in one’s life that needs help. Content describing women’s suffering closely follows or links with their isolated lives and the mental anguish they experience. Although counselling exists in varied form across the three sites, how much and at what frequency and style is it needed to undo the ill mental health women with fistula have undergone?”

Participants’ individual narratives were read and a pathway constructed. A composite pathway emerging from the narratives was then generated and the model drawn. The analysis yielded the seven key actors and likely actions in the treatment pathway of fistula patients. Living with fistula and seeking treatment took on a chronic illness career as presented in section 2.0. Many women wove their health seeking narrative with the difficulties, challenges, frustrations and disappointments of living with fistula.

While, to a researcher, an objective health seeking pathway should be devoid of other life events, to the women the context is inseparable from the experience. It is this rich and nuanced data that became the hallmark of our research that is presented to further illuminate these pathways. The narratives with fistula patients led us into a glimpse of their entire life story centred on the fistula illness.

Data saturation is defined as ‘the point in data collection and analysis when new information produces little or no change to the codebook’ [23]. On the other hand, theoretical saturation is assumed ‘once all main variations of the phenomenon had been identified and incorporated into the emerging theory’ [23]. Thus, data saturation may be obtained with fewer interviews than that required for theoretical saturation. For instance, Wilson, et al. used 66 interviews to develop a grounded theory on HIV/AIDS reconciling incompatibilities theory [24], whilst Mason and Harris used a sample of 114 in a study of environmental factors that influenced market orientation [25]. A heterogeneous sample of fistula patients, the broad research question and on a subject of health seeking behaviour where little was known beforehand necessitated a sample of 121 before we could reach theoretical saturation. Thus the composite pathway of fistula patients’ health seeking behaviour incorporated all the possible variations of this phenomenon.

The conceptual framework presented fits with the data and incidents described. Fit was obtained through constant comparative coding where one emerging concepts were checked and compared with other participants narratives in similar and varying contexts to verify whether they exist. Considering the findings presented in quotes and composite health seeking pathway in Fig. 1, the conceptual framework in Fig. 2 is adequate in reflecting the contextual issues surrounding fistula care in Kenya. The qualitative findings strongly suggest, for example, that the conceptual model fits the constructs of fistula as a chronic illness over time; seeking formal care in hospitals alongside alternative therapies; and the interplay of contextual social, economic, health system and individual factors that determine health actions of women with fistula illness.

A total of 121 participants (women with obstetric fistula) gave their narratives of the experience of health seeking behaviour during fistula illness. These narratives were used to construct the composite pathway. The demographic characteristics of 2 women whose narratives are included in the composite pathway were not obtained, and consequently 119 women’s are presented. The participants’ mean age was 33.2 (17 to 62 years). The mean age at fistula development was 23.2 years. Women generally had low level of education, with 12.6% having no formal education and a further 54.7% had obtained primary education only. About a third of the women had no surviving child and half had developed fistula during their first pregnancy delivery. The participants described long episodes of labour at home under unskilled birth attendance mean 48 h (Table 1).

We present the health seeking behaviour of fistula patients using a conceptual framework grounded on the narrative data that sufficiently takes into account the context surrounding fistula care in Kenya. A typical storyline of health seeking behaviour during fistula illness depicted in our study is that of sequential pathways with initial many visits to hospital getting few surgeries. Fistula therefore becomes a chronic illness that women have to live with. Living with fistula entails moral properties of suffering with generally long pathway to care that did not always guarantee access to care. There is a deviation from normal and women strive to regain normalcy in their physical health, social acceptability, marriage and economic independence.

Women’s realization of symptoms of both rectovaginal and vesicovaginal fistula sets off an account of a pathway that did not always guarantee access to care. The women meet a health system unresponsive to needs of women with few fistula experts, missed diagnosis, lack of pertinent information and logistical support. Further, the women themselves may not be motivated to undergo surgery, for fear of death and due to poverty. Lack of social support, divorce and separation add to the other stressors in the process of health seeking. Few women experience the ideal enablers of healing and regaining normalcy, namely, correct information through radio, access and linkages to care through mobile phones, and most vital surgical facility and expertise at a cost they can afford.

A typical narrative of health seeking for a fistula patient entails sequential pathways where she moves from place to place, person to person seeking treatment. Initially, many visits to hospital get few surgeries, owing to lack of facility, lack of medical team with surgical expertise at the hospital or high cost of the surgery. What follows however, for this and many other similar stories is daunting. Fistula therefore becomes a chronic illness that women have to live with, attempting multiple times to regain normalcy.

The composite treatment pathway is characterized by a lengthy time living with illness, multiple visits to different actors and a core narrative of seeking the formal health facilities (denoted as hospital) in the quest for healing. Emerging themes examined in this pathway namely, cost of surgery, unavailability of surgery; missed opportunities to diagnose and refer patients are amongst the key stumbling blocks that women with fistula face in seeking treatment at the hospital. Consequently, other health steps or actions include trial of home remedy and traditional medicine in trying to heal their condition. Granted, patients’ encountering a system that does not cater for their needs is frustrating and leads to long decades of living with fistula illness. Further, not knowing what illness it is and how it is treated (unawareness) further delays women from obtaining surgical treatment. Treatment pathways are impeded by lack of knowledge, transport, money; multiple hospital visits, referrals, and cost of surgery.

Patients’ perception of their experience to screening and treatment may add to a perception that the system is weak and inefficient in treating fistula survivors; and by word of mouth encourage women to refer or not to refer other women who have the illness for treatment. Due to dysfunctional health system, obstetric fistula becomes a chronic stigmatized illness with the hospital as a key disperser of patients to other health care choices as demonstrated.

Ideal cases of obtaining fistula surgery at the first step of seeking treatment in hospital are rare, with only 3 out of 99 women who sought care at hospital at the first step undergoing corrective surgery for either Recto Vaginal Fistula (RVF) or Vesico Vaginal Fistula (VVF). The proportion accessing surgery increases upon multiple visits at the second and subsequent steps at hospital visits. A close scrutiny of the pathways reveals that initially, women stay home, as time advanced, more women procured surgery. The extremely slow cases are represented by women who had five or more visits to the hospital and got treatment having lived with fistula for many years.

There is scarce research concerning the health seeking behaviour patterns of women with obstetric fistula. By the time of this study, no study existed that wholly examined the health behaviour patterns of fistula patients beyond reporting the time spent before surgery [18]. However, existing research has illuminated some of the components of health seeking behaviour similar to our study.

A recent study by Maulet, Berthé [27] reported that obstetric fistula illness may be viewed as a chronic illness owing to the length of care process, limitations of surgery and the persistent physical and moral suffering [27]. Another qualitative study spanning 18 months in Mali and Niger, established the need for long term care, and tertiary prevention measures, but acknowledge that the holistic approach to care, prevention-treatment-reintegration axis may be feasible within standalone fistula repair centres and international NGOs but a challenge to the national referral hospitals that are biomedical oriented and overwhelmed with emergencies and other referrals [28]. In our study, women report multiple visits to hospital for surgery and less than half had repeated surgeries. However, our study participants were not resident in hospitals as they awaited the next surgery, as opposed to the Mali study [27]. Granted, Maulet, Berthé [27] contribution highlights the long care process of fistula patients, suggesting that surgery is not a one-time episode as women undergo multiple surgeries. This care process after proper diagnosis may be long [27]. In our study, about half the women had undergone one previous surgery by the time of the study, confirming the long term nature of fistula care.

Although obstetric fistula is not considered to be amongst chronic illnesses [29] the length of time spent with illness and illness trajectory mimics other chronic illness. This emerging concept has been alluded to by other authors [31]. However, based on the findings of this study we contend that the chronicity of fistula illness is beyond the time spent during the care process as suggested by Maulet., et al. [28]. The total time lived with fistula illness is the first clearest indicator, in the biography of women who suffer this condition, that obstetric fistula is a chronic illness. Secondly, depending on the surgery outcome, irreparable fistula presents the woman with the possibility of living with fistula illness for the rest of her life. Those who encounter other forms of corrective surgery like urinary diversion permanently live with the aftermath of fistula illness through the changed function of their bowel movement from the ‘normal’. Infertility is another fistula illness symptom that lingers after surgery, along with traumatic psychological moral experiences. Obstetric fistula illness is thus in this sense, a chronic illness with dire consequence on the woman and her family. Further, women who have undergone fistula surgery are required subsequent pregnancies through caesarean section.

Regarding time lived with fistula illness, a recent study in a western Kenya hospital indicated that 35% of the women had experienced incontinence for one to five years and a further 19% for over five years [18]. Similarly Yeakey, et al. [30] report a period of up to 20 years living with the condition in a qualitative study of the lived experiences of women with obstetric fistula, suggesting a long term nature of suffering fistula illness. Our study unravelled similar and more dire findings; of women who had lived with fistula illness for up to 43 years. The chronicity of fistula illness is suggested, first in the longevity of the illness and secondly in the way the woman’s life is disrupted and the normal as she knew is halted.

Wall [31] has suggested that obstetric fistula is a neglected tropical condition. Our findings support this author’s position, and even go further to demonstrate that neglect of obstetric fistula patients has led to the chronicity of fistula illness. Through the pathways, we demonstrate a health system which often disperses women seeking treatment to other actors in the field who are ill equipped to treat fistula. Although hospital is a popular choice as steps of action in women’s pathways, it does not always obtain for them the treatment and healing they seek. Further, our findings suggest a systematic failure in the health care system to cater for the needs of women who present with obstetric fistula at the hospital. As such, the hospital is a key disperser of women to other forms of care providers; yet it is the most popular step taken by women when they realize symptoms of bowel and urinary incontinence.

Our study introduces new explanations to what characterizes the chronic nature of fistula illness. As such, the lack of a broad range of other studies on the health seeking behaviour of fistula patients limits chances of comparison. For instance, we report 7 key actors on the pathways chain including the hospital and health clinics, private health care providers, traditional medicine and diviners, religious discourses and prayer, Non-Governmental Organization and home remedies and staying home the participants can take. We report for the first time a broad range of multiplicity of actors in fistula care. Some studies in Tanzania a [32] and Uganda [33] have reported the use of religion as coping mechanism while in Ghana [34] reliance on traditional medicine has been reported. A prolonged lag time between onset of fistula illness and first hospital visit, was related to lack of awareness and few fistula repair centres and unacceptable modern care [11].

The numerous hospital visits without proper referral and surgery is for the first time focused on in our study. However, seeking of other avenues of healing including traditional medicine, prayer and healers are phenomenon reported elsewhere in studies of health behaviour and therapy management for malaria in Kenya [35]. Similarly, Mwini-Nyaledzigbor, Agana [34] reports use of traditional herbs to treat fistula in a study conducted in Ghana.

In our study, the innovative use of mobile phone communication was a good enabler to women’s reaching fistula repair facilities. These findings complement those of Fiander, Ndahani [36] in the innovative usage of mobile phones to link women with fistula to care in providing information and transport costs. The Comprehensive Community Based Rehabilitation program in Tanzania used mobile phones to transfer money to ambassadors who then assist the patients to cover transport costs while they seek treatment. This improved the number of referrals for fistula surgery combined with free treatment was owed to efficient mobile communication, training of the ambassadors and eliminated cost of transport [36]. In our study however, the mobile phone was used to contact the regional representative concerning scheduled dates in one of the dedicated fistula centres; and later during follow up care after surgery. This innovative use of commonplace means of communication to provide accurate information concerning an individual woman’s care is sustainable and offers possible exploration in future.

Enablers of health seeking behaviour should be emphasized on basing on women’s treatment experience. For instance, information about treatment services and access to treatment to patients living with obstetric fistula through radio advertisements, mobile phone contact by regional representative, and contact by health care providers and referral by health care provider are possible avenues of improving the experience and shortening the women’s treatment seeking pathway.

Based on the study findings, we make the following recommendations. Firstly, there is a need to treat women with obstetric fistula promptly; given the immense suffering the illness poses and the pathway through which they seek care can be improved. This treatment should be by qualified fistula surgeons. Secondly, in view of a lot of women living with fistula for decades, the current treatment is a little too late and early referral and treatment ought to be available, to decrease backlog and treat all new cases. In this regard, building the capacity of the health system and personnel through training of midwives and other health personnel on the proper information regarding fistula prevention, recognition of symptoms and management and improving surgical facilities are priority areas to be focused on by policy and resource allocation. The members of general communities should be informed about obstetric fistula prevention, signs and symptoms and where to seek treatment when the illness occurs and how to support patients with obstetric fistula and post- surgery period.

The participants in this study had undergone fistula surgery and no information was obtained from women who didn’t access the formal hospital for surgery.