Unravelling patient pathways in the context of antibacterial resistance in East Africa

doi:10.21203/rs.3.rs-1933132/v1

Download PDF

Research Article

Unravelling patient pathways in the context of antibacterial resistance in East Africa

https://doi.org/10.21203/rs.3.rs-1933132/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 19 Jun, 2023

Read the published version in BMC Infectious Diseases →

You are reading this latest preprint version

Background

A key factor driving the development and maintenance of antibacterial resistance (ABR) is individuals’ use of antibiotics (ABs) to treat illness. It is therefore crucial to understand the dynamics of patient treatment-seeking - where patients go when unwell, what motivates choice, and treatments they obtain. Developing a better understanding of social inequalities of antibiotic use in low resource settings has been identified as a research priority, as has understanding pluralistic healthcare landscapes, but so far robust multi-country mixed methods data has been lacking. This paper aims to investigate patterns and determinants of patient treatment-seeking pathways, and how they intersect with AB use in East Africa, a region where ABR-attributable deaths are exceptionally high.

Methods

The Holistic Approach to Unravelling Antimicrobial Resistance (HATUA) Consortium collected quantitative data from 6,827 adult outpatients presenting with urinary tract infection (UTI) symptoms in Kenya, Tanzania, and Uganda, and conducted qualitative in-depth patient interviews with a subset (n=116). We described patterns of treatment-seeking visually using Sankey plots and elucidated explanations using mixed-methods. Using Bayesian hierarchical regression modelling, we investigated the associations between socio-demographic, economic, healthcare, and attitudinal factors and three outcomes related to ABR: self-treatment as a first step, having a multi-step treatment pathway, and consuming ABs.

Results

Despite patients predominantly seeking help from medical facilities as a first step, many described complex and repetitive treatment-seeking pathways, which increased the likelihood of consuming ABs. Higher socio-economic status patients were more likely to consume ABs and multi-step pathways. Reasons for choosing providers (e.g., cost, location, time) were conditioned by wider structural factors such as healthcare systems and AB availability.

Conclusion

There is likely to be a reinforcing cycle between complex, repetitive treatment pathways, AB consumption and ABR, which could become more common as ABR evolves. A focus on individual behaviours obscures the contextual challenges patients face when treatment seeking. Pluralistic healthcare landscapes may promote more complex treatment seeking and therefore inappropriate AB use. We recommend intervening on aspects of the treatment cycle amenable to change, focussing on medical facilities (e.g., accessible diagnostics, patient-doctor interactions, information flows), and community AB access points (e.g., drug sellers).

Antibacterial resistance (ABR) is a significant global health threat which compromises treatment of infections with antibiotics (ABs). ABR was associated with an estimated 4·95 million deaths globally in 2019.[1] Such deaths are highest in the Sub-Saharan African region,[1] where there is high burden of infectious diseases and fewer resources to tackle ABR. The process of ABR is influenced by the way individuals use ABs, which is, in turn, impacted by social, political, and economic systems operating on a variety of scales.[2–5] It is therefore crucial to understand treatment-seeking behaviours (or ‘patient pathways’) people take when they are unwell, how these relate to AB use, and the potential presence of ABR[6–8]. This study takes as its central focus individual patient pathways, as told by patients in Kenya, Tanzania, and Uganda, and interrogates these behaviours within their social, economic, and political contexts.

Self-treatment with ABs is one example of ‘inappropriate’ individual treatment-seeking behaviour that is posited to contribute to ABR,[9] but which is influenced by a plethora of social and structural factors. A recent mixed-methods paper from six low- and middle-income countries (LMICs) showed that AB self-treatment was common: reported by 55% of respondents surveyed in Vietnam, 46% in Bangladesh, and 36% in Ghana.[10] Propensity for AB self-medication, while influenced by individual characteristics such as socioeconomic status is affected by structural determinants, including AB dispensing regulations, availability of ABs through alternative providers (e.g., drug sellers), and public trust in different types of healthcare providers.[11] In LMICs, some patients choose to self-treat with ABs rather than access them via prescription because of reduced access and perceived deficiencies in the formal healthcare system, such as long queues, short consultation times, and high out-of-pocket expenses.[12, 13] Other influencing factors include recurrent episodes of infections, cultural beliefs and practices, and symptoms stigma.[11]

In this paper, we focus on urinary tract infections (UTIs) as a lens for understanding treatment-seeking and ABR more generally. UTIs are mild bacterial illnesses that are usually treated with ABs,[14] but which can be stigmatizing and reduce quality of life.[15] Globally, over 150 million individuals are diagnosed with UTIs each year, most of whom are women.[16] Empiric use of ABs globally to treat UTIs contributes to ABR of the uropathogens responsible (typically Escherichia coli and other Enterobacteriacae), and presents a growing challenge to effective treatment.[17] In East Africa, high community prevalence of UTIs, combined with high levels of AB self-medication, may further exacerbate ABR,[14] particularly considering self-management for UTI symptoms is extremely common.[18, 19]

Research questions

We aim to assess the socioeconomic, attitudinal, and contextual factors associated with patients’ treatment-seeking pathways for UTI-like symptoms and explore how key pathway points intersect with AB use. We focus on three aspects theoretically related to the ABR: patients’ self-treating rather than seeking help at a medical facility, having multi-step pathways (e.g., serial attendance at different healthcare facilities), and AB consumption during the pathway. To do this, we use quantitative and qualitative patient interview data to explore the lived experience of treatment-seeking and shed light on situational barriers and facilitators of such behaviour.

Theoretical and Methodological Approach

This study is part of a multi-country interdisciplinary consortium “Holistic Approach to Unravel Antibacterial Resistance in East Africa (HATUA)”.[20] Our conceptual framework which sees ABR as a complex assemblage consisting of interconnected, multi-scalar social, political, and biological influences (see Fig. 1 in [20]). At the heart of this complexity is the ‘patient pathway’, nested in a pluralistic healthcare landscape comprising various formal and informal healthcare providers.[21] Pathway analysis has been used to investigate care for tuberculosis,[6] abortion,[22] and cancer[23] but has rarely been used to understand AB use.[8] Incorporating medical syncretism, the patient pathway may contain any number of steps, and include delays in seeking treatment, different choice of provider, self-treatment and self-medication, and differences in regimen adherence. Rather than a linear sequence of actions,[7] the patient pathway can be iterative, cyclical, and repetitive. To unravel this complexity, we analyse qualitative and quantitative data about AB use pathways in parallel. We avoid the assumption that human behaviour is entirely driven by the individual decisions[24] by evaluating patient pathways within their social, economic, and political contexts.

Sample

The sample consists of 6,827 adult outpatients, aged 18 and over (or pregnant and <18) who were recruited from healthcare facilities in Kenya, Tanzania, and Uganda, within three sites per country (Kenya: Nairobi, Nanyuki, and Makueni; Tanzania: Mwanza, Kilimanjaro, and Mbeya; Uganda: Mbarara, Nakapiripirit, and Nakasongola), between February 2019- September 2020. We recruited patients at primary, and secondary facilities (levels 2-5) which were predominantly government-funded (Table S1). Clinicians identified patients with symptomatic and probable UTI for inclusion. In all sites, less than 1% of those approached declined to participate. Patients provided a clean catch mid-stream urine sample and answered a questionnaire on treatment-seeking for UTI symptoms, AB use practices and attitudes, and sociodemographic characteristics. We excluded 219 patients who came to the recruitment clinic for non-UTI symptoms and had not attempted to treat their symptoms, leaving 6,608 patients for analysis: 3,190 (48·3%) from Tanzania, 1,757 (26·6%) from Uganda, and 1,661 (25·1%) from Kenya.

In-depth interviews (IDIs) were conducted 1-2 weeks after the clinic visit with a purposively selected subset of patients (n=116) who had reported longer treatment pathways or were diagnosed with a multi-drug resistant UTI pathogen. IDIs were conducted in person, at the respondents’ homes and in their primary language, using standardised topic guides, which were subsequently translated into English by fieldworkers. IDIs focused on mapping individual pathways based on the patient’s account of recent treatment-seeking action, AB use, knowledge and attitudes, and motivations for behaviours. Patient qualitative and quantitative data were linked using study barcodes. Participants gave written informed consent. Patients were not involved in the design, or conduct, or reporting, or dissemination plans of our research. Ethical approval was obtained from National and Institutional Research Ethics Committees (see protocol).[20]

Study variables

Treatment seeking behaviours

Figure 1 illustrates the structured questionnaire used to collect patient pathway information, which identified the types of providers consulted, treatments taken, and reasons for these choices.

[Fig 1 about here]

We identified three binary outcomes with a theoretical or empirical link to ABR risk identified in previous studies:[11, 25, 26]

Self-treatment as a first step (defined as going to a drug shop/pharmacy, seeking advice from friends or family, or using traditional or home remedies) vs seeking help at a medical facility;
Having two or more steps in the pathway prior to coming to the recruitment centre vs. having fewer steps;
Taking ABs from any source to treat their symptoms vs not taking any ABs (taking other types of medicine or none)prior to coming to the recruitment clinic.

The last variable was derived from patients’ self-reports of the names of medicines taken during the pathway. During the interview, respondents were prompted using a drug bag or drug card developed specifically for each site.[27]

Other variables

We included self-reported variables that might impact treatment-seeking patterns.[11] Sociodemographic factors included gender (male/female), age (categorised into <25; 25-34; 35-44; 45-54; 55-64; 65+ years), marital status (married, never married, and other, which included cohabiting, widowed, divorced), and household size (1-2 people, 3-6 people, 7+ people). Socioeconomic status was measured by education level (no formal, primary, secondary, and tertiary), employment status (formal employment, informal employment, homemaker, not working), self-reported ability to meet healthcare costs (easy, difficult, very difficult), and a within-country asset index grouped into quintiles (Table S2 for details). Healthcare factors included the level of medical facility the patient was recruited from (lower-level community health centres (levels 2-3) vs. high-level clinics or referral hospitals (levels 4+), and whether they had previous experience of UTI (did not have UTI before, had UTI before, or did not know what UTI was). We also included indicators which measured whether patients felt that UTI symptoms were stigmatised (yes/no), which may impact treatment-seeking.

Statistical Methods

We used Sankey plots to visualise quantitative data on patient pathways, showing counts and percentages of types of providers chosen and type of treatment obtained (if any) at each step. We excluded patients without valid data on the steps considered (n=230), leaving a sample of n=6,378. Subsequently, we used Bayesian hierarchical logistic regression models to assess socioeconomic and attitudinal factors associated with three binary outcomes outlined above (full model specification in supplementary section 3). The models were estimated in R using the Nimble package.[28]Approximately 8% of our sample had missing values on the outcomes or covariates, which we account for within a Bayesian modelling framework. Regression models had four levels: patients were nested in 25 clinics, clinics in nine sites, and sites within three countries. Results were reported in terms of odds ratios (ORs) and 95% highest posterior density intervals (HPDI) due to the skewed posterior distribution of all independent variables.

Qualitative data analysis

Translated English-language interview transcripts were linked to quantitative data using patients IDs, and analysed using NVivo software.[29] We used iterative thematic content analysis, beginning with first-round coding based on interview questions, such as how patients sought treatment and obstacles to treatment-seeking. Subsequent rounds of in-depth coding were undertaken to identify differences and similarities in treatment-seeking pathways between patients, as well as potential contributing factors to decision-making around treatment seeking.

Sample characteristics

Most patients were females of reproductive age, and the majority were married (Table S3). In Uganda and Tanzania, most had little or no formal education, whereas in Kenya 86% had secondary or higher education. In the pooled data, most were either in informal employment or homemakers, and lived in households of 3–6 people. Ugandan respondents were least likely to report it was ‘easy’ to meet healthcare costs, compared to Tanzania and Kenya. Around half of respondents from Tanzania and Kenya reported they had been previously diagnosed with UTI, compared with one quarter in Uganda. Possibly related is that symptoms stigma was higher among Ugandan participants (39%) compared to those in Tanzania and Kenya (16% and 23%, respectively). Patients in Tanzania and Uganda were mostly recruited at lower-level facilities (level 2–3), whereas in Kenya most were recruited at higher-level facilities or referral hospitals (levels 4+). Sociodemographic characteristics for the qualitative sample reflected those in the quantitative (Table S4).

Overview of patient pathways

Figure 2 illustrates the choices of provider and whether ABs were obtained for each pathway step pooling data across 3 countries; analogous country-specific plots are shown in supplementary Figure S7. The pathways as shown are composed of a maximum of three steps. At step 1 or 2, patients are classified as going to either a government clinic, private clinic, drug shop, self-treating, or going directly to a recruitment clinic. At the third step, participants reported either trying a third option or attending the recruitment clinic.

[Fig. 2 about here]

First choice of treatment

Across all countries, most patients (86%) went to medical facilities as their first step in treating UTI-like symptoms (Fig. 2). The largest group of patients (45%) went directly to a HATUA recruitment clinic, 26% visited another government-funded facility, 15% visited a private facility, 7% visited drug shops/pharmacies and 7% self-treated either with home remedies, herbs, or drugs provided by friends/relatives. Kenyan patients had the simplest pathways and were most likely to go to a recruitment centre as their first step. Visiting private clinics as a first step was more common in Tanzania and Uganda than in Kenya, whereas visits to drug shops and pharmacies were more common in Kenya than the other countries. A minority of patients attempted self-treatments before attending a medical facility, including consuming herbal remedies and ABs obtained in drug shops without a prescription, as shown in a multi-step pathway from a female patient in Nakapiripirit, Uganda.

Step 0, Delay: “When I started feeling this pain I first ignored [it]”

Step 1, Traditional Medicine: “Later I went to [a traditional healer] where I was told that I was bewitched … The herbalist gave me some leaves for bathing … then for drinking I was given moringa, he told me to take for one month ... I felt fine for that one month, but it didn’t help me.”

Step 2, Drug seller: “I began buying the medicine in the drug shop. I was given amoxicillin capsule and metronidazole. I bought [a] full dose that I took for 3 weeks I was told to take 2*3 a day, which I did. In addition, I stayed for some months.”

Step 3, Drug Seller: “After some time, it started again, and I decided to go and buy more medicine which was not full dose because I didn’t have enough money to buy the medicine.”

Step 4, Recruitment Facility: “Then in March I met sister who… [advised me] to come for check-up and I was given medication for one month from [this] heath centre.”

Here, initial self-treatment behaviour was driven by confusion: “I just don’t know what to do” (female patient, Nakapiripirit, Uganda). After unsuccessful treatment attempts with the traditional medicine and drug sellers, the patient reflected that for future ailments: “I will be coming to get medication at the hospital not somewhere else since they prescribe well and give the right medication.”

The most reported reasons for first choice of treatment are shown in Fig. 3 (country-specific results in Fig S8). Those who chose a government or private facility were most often motivated by location, but cost was also cited as a reason for choosing government clinics, drug sellers, and self-treatment (less so private facilities). Time was a common reason for choosing drug shops or self-treatment. Qualitative data shows that cost was often related to both location (due to transportation expenses) and time (due to loss of work), which can influence both the choice of medical facility and the decision to self-treat at step one:

“[I live] far from the hospital or health centre, if I feel pain the first thing is to go [to the] pharmacy, there they will assist me with a drug that will ease pain at that moment. But if pain persists, I will have to go to the hospital for further treatment and advice from the doctor.” (male patient, Mbeya, Tanzania)

[Fig. 3 about here]

Results from multivariable regression suggest the importance of perceived costs on choice of first treatment. After adjustment for socio-demographic variables, choosing to self-treat from a drug shop or with home remedies at step 1, versus choosing a medical facility, was more likely among patients who find it ‘a little difficult’ (OR = 1·40; 95%HPDI = 1·15, 1·66) or ‘very difficult’ (OR = 1·87; 95%HPDI = 1·47, 2·29) to meet healthcare costs, relative to those who found it ‘easy’ (Fig. 4, panel 1). Self-treatment was most strongly associated with financial constraints in Tanzania and Kenya, and not in Uganda (see country specific plots, Figure S9).

[Fig. 4 about here]

‘Trust’ was the most cited reason among those using self-treatments. Qualitative accounts indicated this can be related to trusting one’s own knowledge to treat illness:

“Normally, I first observe the condition as I take the concoction for ginger, garlic, lemon and honey. That mixture heals everything, I don’t know why it does not heal HIV/AIDS.” (male patient, Nanyuki, Kenya)

In some cases, ‘trust’ in formal medical facilities is eroded when patients’ expectations are not met, and symptoms remain unresolved, prompting patients to favour other treatment sources. Qualitative data also highlighted that drug shops were seen as convenient alternatives to medical facilities, and that drug sellers were sometimes viewed as part of the trusted cadre of healthcare professionals. For example, one patient explained: “I trust only medical help from doctors and nurses, so the health centres and drug shops is where I only go” (female patient, Nakapiripirit, Uganda).

Understanding multi-step pathways

Overall, 56% of patients tried some other form of treatment before arriving at the recruitment facility; we label these as having ‘multi-step pathways’. Of these patients (n = 3,546), 53% took one step before the recruitment clinic (30% of the full sample), 29% had two additional steps (16% of the full sample), and 18% had three or more steps (10% of the full sample) (Fig. 5). Multi-step pathways were more common in Uganda and Tanzania than Kenya.

[Fig. 5 about here]

While some multi-step pathways did include self-treatment or drug seller visits, most began with patients seeking treatment from medical facilities, after which pathways became convoluted, often involving repeated visits to medical facilities, causing great frustration, as evident in the pathway description by a male patient from Mbarara, Uganda:

Step 1, Facility 1: “When I got infected, I went to [facility 1] … He injected me and gave me drugs to take and told me that I had recovered. I … took the medicine but I did not recover. I remained sick.”

Step 2, Facility 1: “I went back and … I asked him, haven’t I healed now, he wrote for me more medicine, I went to the pharmacy and bought the drugs ... I took the medicine and I seemed to recover but pain reduced.”

Step 3, Drug Seller: “When I finished [the prescription], I thought I would recover but I went back to the same state. So, I looked at the place I went to buy the medicine, I went there, and they wrote for me this one.”

Step 4, ‘Leave it’: “[The treatment] has not worked. So, I decided to leave it for a while.”

Step 5, Facility 2: “When the sickness continued, I [visited another] doctor … He tested me, gave me medicine, and injected me six times ... Then he took my urine samples ... He told me that the medicine he was injecting couldn’t cure it. He then brought another type of medicine … [but] when the dose was over … I started feeling like the disease had come back. Then I wondered how I was, if I had gone everywhere and the disease was failing, now where was I to go!”

Step 6, Recruitment facility: “The doctor … had called a specialist who would test my urine … and give me proper treatment … I decided to go there, and they got my urine samples.”

To explore whether absent diagnostics may contribute to treatment failures, we conducted further analysis among a subset who visited clinics in study phase 2 (Jan 2020 onwards) with data on diagnostics. This showed that 73% had had their urine sample tested (although it is unknown what kind of test); and that urine testing was more common at private than government facilities (Fig S12).

Having a multi-step pathway was more common in patients older than 35 years and those with a previously diagnosed UTI (Fig. 4, panel 2). Stigma around UTI symptoms was associated with longer pathways, dependant on context. In Uganda, where 39% of patients felt stigma, it was associated with higher odds of having a multi-step pathway (OR = 1·44; 95%HPDI = 1·03, 1·86) (Figure S10). However, Ugandan patients were more likely to describe their decision-making as driven by financial considerations rather than stigma, reflecting the fact that more Ugandan patients reported that it was ‘very difficult’ to meet healthcare costs compared to those in Kenya (34% vs. 12%, Table S3). By contrast, in Tanzania, where only 16% of patients reported UTI symptoms stigma, this was associated with simpler pathways. IDIs also suggested that in some contexts, stigma drove treatment choices. For example, Kenyan patients discussed choosing private or distant medical facilities to avoid being recognized by members of the community. Private clinics were also favoured because treatment was faster, meaning patients could avoid having to explain absences from work: “If I went to a public hospital, I could have taken a long time, and my friends at the market could have wanted to know what my problem was and this way I tried my level best to hide it” (female patient, Makueni, Kenya).

Antibiotic consumption during the pathway

Increasing the number of pathway steps provided more opportunities for AB consumption. Among patients with multi-step pathways, nearly half (48%) reported taking ABs at step one, and 42% of those with a second step took ABs (Fig. 2). As Fig. 2 shows, most ABs were consumed after visits to medical facilities. The most common ABs taken were amoxicillin and ciprofloxacin (see Table S6), and some ABs taken were not recommended for treating UTI symptoms (e.g., doxycycline). AB consumption was higher among those with UTI symptoms stigma (OR = 1·34; 95%HPDI = 1·09, 1·61), those who had been previously diagnosed with UTI (OR = 1·34; 95%HPDI = 1·06, 1·63), and those who had higher-level educational qualifications (OR = 1·46; 95%HPDI = 1·06, 1·87) (Fig. 4, panel 3); AB consumption was lower among those who found it ‘very difficult’ to meet healthcare costs (OR = 0·83; 95%HPDI = 0·66, 1·02).

Qualitative data enriched the picture with stories of repeat prescriptions for the same drugs procured from multiple visits to medical centres and drug sellers, evident in the pathway of a female patient from Moshi, Tanzania:

Step 1, Private Facility: "They started with amoxicillin. Then they gave me ampiclox. Then they gave me amoxiclav. But symptoms persisted."

Step 2, Government Facility 1: "Then I … went to Kilimanjaro hospital, and there I still had UTI … They gave me amoxiclav."

Step 3, Government Facility 2: “Then I went to [another] hospital, they also gave me amoxiclav… and they said that was a strong drug."

Step 4, Recruitment Facility: "I went there because I didn’t get a cure. And I had two problems. The biggest one was ulcers, and I still had UTI symptoms… [the doctor] suggested the tests I needed to take. I agreed and they tested me … [then] he told me that he is giving me a seven-day dose [of amoxiclav].”

This patient further expressed concern that her insurance coverage impacted the drugs she was given: “Sometimes a doctor may prescribe same drugs that you have used before and when you ask him he says it is the insurance” (female patient, Moshi, Tanzania).

Visits to medical facilities and drug sellers often went hand in hand, due to stockouts:

“You find yourself not having money, [and at the hospital] you are just given a prescription and told to buy drugs. If you don’t have money you won’t buy, until you get money that is when you will buy” (female patient, Mwanza, Tanzania).

In some cases, this contributed to decisions to go directly to drug sellers to save the cost of medical consultation.

Using mixed-methods data, we show that patient pathways in East Africa for a common infection (UTI) are often convoluted, involving reiterative steps and different healthcare providers. Such complexity was not driven by individual choice; patients were struggling to get care that worked in a confusing, hybrid healthcare landscape riddled with structural constraints. Our findings support others which stress the importance of location, cost, and time in treatment decision making,[11, 30] but we show these are contextualised by community factors such as stigma, illness behaviour, and local understandings of illness which are in turn conditioned by wider socioeconomic, geographic, and healthcare structures.

Even though patients predominantly chose what might be regarded as clinically ‘optimal’ paths, i.e., attending medical facilities, this often did not resolve their symptoms. The causes of these treatment failures deserve further investigation to reduce patient suffering and relieve pressure on an already under-resourced health infrastructure. Recurrent UTI-like symptoms (indicating repeat or chronic infections) were associated with multi-step pathways and repeated AB consumption, indicating that longer pathways might be a consequence of ABR making UTI more difficult to treat. Another part of the story could be the paucity of diagnostic capabilities for UTI in LMICs,[31] which, if more widely available, could detect drug resistance and guide appropriate treatments. Yet another issue may be equitable access to appropriate treatments: around two thirds of those attending clinics ended up not taking any medicines, and our data suggest this may be due to stockouts at clinics and prohibitive costs of medicine.[11, 13] This pattern further highlights the importance of medical facilities as a valuable site of intervention in addressing ABR and the pressing need to more fully understand limitations present within these spaces. Future studies should address clinic-related factors in depth, including the patient-doctor consultation, medical records keeping and information flows, and other barriers that prevent patients from following best practice advice around ABs.

Patients’ treatment choices were often motivated by time and financial constraints, which can lead to systematically different pathways among the poorest and the richest patients. More educated and wealthier patients were both more likely to consume ABs and have multi-step pathways. As described elsewhere,[13] wealthier subgroups may be also more likely to choose private facilities and buy medicines from drug sellers to save time. On the other hand, poorer subgroups or those struggling to afford healthcare are more likely to self-treat and struggle to afford appropriate AB treatment.

This study has some limitations. The linked quantitative-qualitative sample is representative of the patient population attending public facility outpatient services for UTI symptoms at study sites (future consortium papers will address community members not attending clinics). There were proportionally more respondents from higher-level referral facilities in Kenya compared to the other countries. Comparisons of our sample to population representative DHS surveys showed some differences in terms of marital status and educational distributions (Table S6), which caution against wider generalisation. Given the self-reported nature of treatment-seeking behaviour and its predictors, we cannot rule out reporting bias.

This study has taken a patient-centric perspective, but our results suggest that treatment-seeking is never an individuated behaviour; actions are influenced by situational constraints and are contextually dependent. Thus, we contribute evidence that AMR should be considered a system rather than a product a set of individual actions[5].

Complex treatment pathways are likely related to various individual and structural factors, but another important driver is likely to be ABR itself. As ABR continues to evolve, the cyclical treatment attempts we observed here for UTI-like illness will become more common, reflecting the vicious socio-biological cycle of ABR. Drug resistance means treatment attempts are more likely to fail, thus fuelling ABR by necessitating further AB treatments. In many LMICs, there are key structural weaknesses which facilitate this vicious cycle: including (but not limited to) under-resourced public healthcare, insufficient diagnostic capacity, and ample opportunities to purchase ABs without prescription. Thus, we advocate that attention be paid towards addressing these upstream factors which drive both ABR and complex patient pathways.

ABR antibacterial resistance

ABs antibiotics

HATUA The Holistic Approach to Unravelling Antimicrobial Resistance Consortium

UTI Urinary tract infection

LMIC Low- and middle-income country

IDIs In-depth interviews

OR Odds ratio

HDPI Highest posterior density intervals

Ethics approval and consent to participate

The study received ethical approval from the University of St Andrews, UK (number MD14548, 10/09/19); National Institute for Medical Research, Tanzania (number 2831, updated 26/07/19); CUHAS/BMC Research Ethics and Review Committee (number CREC /266/2018, updated on 02/2019); Mbeya Medical Research and Ethics Committee (number SZEC-2439/R.A/V.1/303030); Kilimanjaro Christian Medical College, Tanzania (number 2293, updated 14/08/19); Uganda National Council for Science and Technology (number HS2406, 18/06/18); Makerere University, Uganda (number 514, 25/04/18); and Kenya Medical Research Institute (04/06/19, Scientific and Ethics Review Committee (SERU) number KEMRI/SERU/CMR/P00112/3865 V.1.2). For Uganda, administrative letters of support were obtained from the district health officers to allow the research to be conducted in the respective hospitals and health centres. All participants provided written informed consent to participate. All methods were carried out in accordance with relevant guidelines and regulations.

Consent for Publication

Not applicable.

Availability of data and materials

The datasets generated and/or analysed during the current study are not publicly available due ethical and data access agreements with individual country ethical boards but are available from the corresponding author on reasonable request.

Conflict of Interests

None

Funding

The Holistic Approach to Unravel Antibacterial Resistance in East Africa is a Global Context Consortia Award (MR/S004785/1) funded by the National Institute for Health Research, Medical Research Council, and the Department of Health and Social Care. The award is also part of the EDCTP2 programme supported by the European Union. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. This work is supported in part by the Makerere University-Uganda Virus Research Institute Centre of Excellence for Infection and Immunity Research and Training (MUII). MUII is supported through the DELTAS Africa Initiative (grant number 107743). The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences and Alliance for Accelerating Excellence in Science in Africa and is supported by the New Partnership for Africa’s Development Planning and Coordinating Agency with funding from the Wellcome Trust (grant number 107743) and the UK Government. This paper was funded in part by a grant from the National Institutes of Health (grant number U01CA207167), and a Scottish Funding Council GCRF Consolidator Award.

Author contributions

KK co-designed the study, conceptualised the paper, supervised analysis of quantitative data, and wrote the manuscript. KJF conceptualised the paper, analysed qualitative data, and wrote the manuscript. MAAA conceptualised the paper, assisted with data preparation, analysed the quantitative data, and edited the manuscript. SN co-designed the study, supervised data collection, and reviewed the manuscript. JRM co-designed the study, supervised data collection, and reviewed the manuscript. MK co-designed the study, conceptualised the paper, supervised qualitative data analysis, and edited the manuscript. MFM supervised data collection and reviewed the manuscript. AA supervised data collection and reviewed the manuscript. DLG assisted with data preparation and reviewed the manuscript. AGL co-designed the study, supervised quantitative data analysis, and edited the manuscript. SIH analysed qualitative data and edited the manuscript. BTM co-designed the study, supervised data collection, and reviewed the manuscript. HW provided statistical advice and edited the manuscript. CK supervised data collection and reviewed the manuscript. EO assisted with quantitative data analysis and reviewed the manuscript. NEN supervised data collection and reviewed the manuscript. OL reviewed the manuscript. JB supervised data collection and reviewed the manuscript. AM-B reviewed the manuscript. VAS co-designed the study and reviewed the manuscript. AGD reviewed the manuscript. JNM supervised data collection and reviewed the manuscript. AS coordinated the study and reviewed the manuscript. JS co-designed the study and reviewed the manuscript. AE co-designed the study and reviewed the manuscript. DA co-designed the study and reviewed the manuscript. SHG co-designed the study and reviewed the manuscript. GK co-designed the study and reviewed the manuscript.WS co-designed the study and reviewed the manuscript. DJS co-designed the study and reviewed the manuscript. BBA co-designed the study, supervised data collection, and reviewed the manuscript. JK co-designed the study, supervised data collection, and reviewed the manuscript. SEM co-designed the study, supervised data collection, and reviewed the manuscript. MTGH led the design of the project, was the guarantor of the work, and reviewed the manuscript.

Acknowledgements

HATUA Consortium includes everyone listed above, plus: Pendo Ndaki, Eveline T. Konje, Fernando Benitez- Paez, Madeleine Clarkson, Xuejia Ke, Benjamin Sunday.

Murray CJ, Ikuta KS, Sharara F, Swetschinski L, Robles Aguilar G, Gray A, et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399:629–55.
Davis A, Sharp J. Rethinking One Health: Emergent human, animal and environmental assemblages. Soc Sci Med. 2020;258:113093.
Chatterjee A, Modarai M, Naylor NR, Boyd SE, Atun R, Barlow J, et al. Quantifying drivers of antibiotic resistance in humans: a systematic review. Lancet Infect Dis. 2018;18:e368–78.
Ahmad M, Khan AU. Global economic impact of antibiotic resistance: A review. Journal of Global Antimicrobial Resistance. 2019;19:313–6.
Tompson AC, Manderson L, Chandler CIR. Understanding antibiotic use: practices, structures and networks. JAC-Antimicrobial Resist. 2021;3.
Hanson CL, Osberg M, Brown J, Durham G, Chin DP. Conducting Patient-Pathway Analysis to Inform Programming of Tuberculosis Services: Methods. J Infect Dis. 2017;216 suppl_7:S679–85.
Haenssgen MJ, Ariana P. Healthcare access: A sequence-sensitive approach. SSM - Popul Heal. 2017;3:37–47.
Lucas PJ, Uddin MR, Khisa N, Salim Akter SM, Unicomb L, Nahar P, et al. Pathways to antibiotics in Bangladesh: A qualitative study investigating how and when households access medicine including antibiotics for humans or animals when they are ill. PLoS One. 2019;14:e0225270.
Ocan M, Obuku EA, Bwanga F, Akena D, Richard S, Ogwal-Okeng J, et al. Household antimicrobial self-medication: A systematic review and meta-analysis of the burden, risk factors and outcomes in developing countries. BMC Public Health. 2015;15:742.
Do NTT, Vu HTL, Nguyen CTK, Punpuing S, Khan WA, Gyapong M, et al. Community-based antibiotic access and use in six low-income and middle-income countries: a mixed-method approach. Lancet Glob Heal. 2021. https://doi.org/10.1016/S2214-109X(21)00024-3.
Torres NF, Chibi B, Middleton LE, Solomon VP, Mashamba-Thompson TP. Evidence of factors influencing self-medication with antibiotics in low and middle-income countries: a systematic scoping review. Public Health. 2019;168:92–101.
Larson PS. Access to Health Care in Sub-Saharan Africa: Challenges in a Changing Health Landscape in a Context of Development. Heal Ecol Perspect Anthr. 2019;:95–106.
Green DL, Keenan K, Huque S. The Role of Multidimensional Poverty in Antibiotic Misuse: A Study of Self-Medication and Non-Adherence in Kenya, Tanzania, and Uganda. 2021.
Flores-Mireles AL, Walker JN, Caparon M, Hultgren SJ. Urinary tract infections: epidemiology, mechanisms of infection and treatment options. Nat Rev Microbiol 2015 135. 2015;13:269–84.
Izett-Kay M, Barker KL, McNiven A, Toye F. Experiences of urinary tract infection: A systematic review and meta-ethnography. Neurourol Urodyn. 2022. https://doi.org/10.1002/NAU.24884.
Ozturk R, Murt A. Epidemiology of urological infections: a global burden. World J Urol. 2020;38:2669–79.
Mazzariol A, Bazaj A, Cornaglia G. Multi-drug-resistant Gram-negative bacteria causing urinary tract infections: a review. https://doi.org/101080/1120009X20171380395. 2017;29:2–9.
Smith CM, Shallcross LJ, Dutey-Magni P, Conolly A, Fuller C, Hill S, et al. Incidence, healthcare-seeking behaviours, antibiotic use and natural history of common infection syndromes in England: results from the Bug Watch community cohort study. BMC Infect Dis. 2021;21:1–11.
Lelie-van der Zande R, Koster ES, Teichert M, Bouvy ML. Womens’ self-management skills for prevention and treatment of recurring urinary tract infection. Int J Clin Pract. 2021;75:e14289.
Asiimwe BB, Kiiru J, Mshana SE, Neema S, Keenan K, Kesby M, et al. Protocol for an interdisciplinary cross-sectional study investigating the social, biological and community-level drivers of antimicrobial resistance (AMR): Holistic Approach to Unravel Antibacterial Resistance in East Africa (HATUA). BMJ Open. 2021;11:e041418.
Moshabela M, Pronyk P, Williams N, Schneider H, Lurie M. Patterns and implications of medical pluralism among HIV/AIDS patients in rural South Africa. AIDS Behav. 2011;15:842–52.
Coast E, Norris AH, Moore AM, Freeman E. Trajectories of women’s abortion-related care: A conceptual framework. Social Science and Medicine. 2018;200:199–210.
Zøylner IA, Lomborg K, Christiansen PM, Kirkegaard P. Surgical breast cancer patient pathway: Experiences of patients and relatives and their unmet needs. Heal Expect. 2019;22:262–72.
Chandler CIR. Current accounts of antimicrobial resistance: stabilisation, individualisation and antibiotics as infrastructure. Palgrave Commun. 2019;5:15–7.
Ayukekbong JA, Ntemgwa M, Atabe AN. The threat of antimicrobial resistance in developing countries: Causes and control strategies. Antimicrobial Resistance and Infection Control. 2017;6.
Lu J, Sheldenkar A, Lwin MO. A decade of antimicrobial resistance research in social science fields: a scientometric review. Antimicrobial Resistance and Infection Control. 2020;9.
Dixon J, MacPherson E, Manyau S, Nayiga S, Khine Zaw Y, Kayendeke M, et al. The ‘Drug Bag’ method: lessons from anthropological studies of antibiotic use in Africa and South-East Asia. Glob Health Action. 2019;12:1639388.
NIMBLE Development Team. NIMBLE: MCMC, Particle Filtering, and Programmable Hierarchical Modeling (0.12.1). 2021.
Ltd. QIP. New NVivo, release 1.0. 2020.
Dixon J, Macpherson EE, Nayiga S, Manyau S, Nabirye C, Kayendeke M, et al. Antibiotic stories: a mixed-methods, multi-country analysis of household antibiotic use in Malawi, Uganda and Zimbabwe. BMJ Glob Heal. 2021;6:e006920.
Yadav H, Shah D, Sayed S, Horton S, Schroeder LF. Availability of essential diagnostics in ten low-income and middle-income countries: results from national health facility surveys. Lancet Glob Heal. 2021;9:e1553–60.

No competing interests reported.

supplementarymaterialfinal.docx

Download PDF

Journal Publication

published 19 Jun, 2023

Read the published version in BMC Infectious Diseases →

Editorial decision: Major revision
02 May, 2023
Reviews received at journal
15 Apr, 2023
Reviews received at journal
20 Mar, 2023
Reviewers agreed at journal
05 Mar, 2023
Reviewers agreed at journal
01 Mar, 2023
Reviewers agreed at journal
06 Feb, 2023
Reviewers agreed at journal
05 Sep, 2022
Reviewers invited by journal
22 Aug, 2022
Editor assigned by journal
22 Aug, 2022
Editor invited by journal
13 Aug, 2022
Submission checks completed at journal
13 Aug, 2022
First submitted to journal
05 Aug, 2022

You are reading this latest preprint version

Unravelling patient pathways in the context of antibacterial resistance in East Africa

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Research questions

Theoretical and Methodological Approach

Methods

Sample

Study variables

Treatment seeking behaviours

Statistical Methods

Qualitative data analysis

Results

Sample characteristics

Overview of patient pathways

First choice of treatment

Understanding multi-step pathways

Antibiotic consumption during the pathway

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1