Getting to the Emergency Department in time: Interviews with patients and their caregivers on the challenges to emergency care utilization in rural Uganda.

OBJECTIVES: Karoli Lwanga Hospital and Global Emergency Care, a 501(c)(3) nongovernmental organization, operate an Emergency Department (ED) in Uganda's rural Rukungiri District. Despite available emergency care (EC), preventable death and disability persist due to delayed patient presentations. This study seeks to understand the emergency care seeking behavior of community members utilizing the established ED.
METHODS: We purposefully sampled and interviewed patients and caregivers presenting to the ED more than 12 hours after onset of chief complaint in January-March 2017 to include various ages, genders, and complaints. Semistructured interviews addressing actions taken before seeking EC and delays to presentation once the need for EC was recognized were conducted until a diverse sample and theoretical saturation were obtained. An interdisciplinary and multicultural research team conducted thematic analysis based on descriptive phenomenology.
RESULTS: The 50 ED patients for whom care was sought (mean age 33) had approximately even distribution of gender, as well as occupation (none, subsistence farmers and small business owner). Interviews were conducted with 13 ED patients and 37 caregivers, on the behalf of patients when unavailable. The median duration of patients' chief complaint on ED presentation was 5.5 days. On average, participants identified severe symptoms necessitating EC 1 day before presentation. Four themes of treatment delay before and after severity were recognized were identified: 1) Cultural factors and limited knowledge of emergency signs and initial actions to take; 2) Use of local health facilities despite perception of inadequate services; 3) Lack of resources to cover the anticipated cost of obtaining EC; 4) Inadequate transportation options.
CONCLUSIONS: Interventions are warranted to address each of the four major reasons for treatment delay. The next stage of formative research will generate intervention strategies and assess the opportunities and challenges to implementation with community and health system stakeholders.

Introduction

In low- and middle-income countries (LMICs), 24 million lives are lost annually due to emergency conditions that could be addressed through implementation of comprehensive emergency care (EC) systems, with 21% of these deaths occurring in sub-Saharan Africa (SSA) [1]. In 2007, the World Health Assembly resolution on EC systems encouraged health systems research enabling the expansion of effective EC systems globally. More specifically, it urged member states to identify the unmet needs of populations, institute formal EC systems, and develop capacity as a means to reduce catastrophic morbidity and mortality [2]. In 2019, World Health Assembly resolution 72.16 redoubled and expanded these calls to action, including encouraging initiatives to increase awareness of, and expand management capacity for, emergency situations at the community level [3]. Despite the recent establishment of academic EC training and clinical care programs in SSA, there has been limited large-scale implementation of community-based EC services.

Until recently, Uganda, like many LMICs lacked EC services. Efforts are under way to develop a well-coordinated EC system, train appropriate workforce to staff EC facilities, and inform the community about new EC resources available. As part of these efforts, in 2008, Global Emergency Care (GEC), a US-based 501(c) [3] nongovernmental organization, partnered with Karoli Lwanga Hospital, known locally as Nyakibale, to build and staff rural Uganda’s first Emergency Department (ED) with nonrotating, formally trained staff. This ED provides high-quality, timely care, evaluating and stabilizing patients before requiring payment. However, patients often arrive at the ED late in the course of illness or injury, reducing the potential impact of EC interventions.

Delays in utilization of available EC are widely recognized though not well understood, ranking second among 7 priority research questions identified to facilitate the improvement of EC in LMICs through a research priority setting exercise conducted by a multidisciplinary coalition of researchers and stakeholders, the Global Emergency Care Research Network [4]. The global emergency medicine community recognizes the need for qualitative research to better understand the perspectives and experiences of those seeking to utilize EC, as well as assess the socioeconomic, cultural, and structural barriers leading to treatment delay [5]. Our research goal was to better understand these barriers, particularly the challenges rural Ugandans face in accessing and utilizing EC, by interviewing patients and caregivers in the ED.

The Rukungiri District

The Rukungiri District is located in rural Southwest Uganda. The district is mainly farmland with one urban center. Dirt roads traverse steep, rocky hills [6]. There are 2 “rainy seasons,” when heavy rains render dirt roads barely passable for days. Motorcycle taxis are the most prevalent mode of vehicular transportation, as they can navigate footpaths reaching homes and small villages. As of the last published district-level census in 2014, very few households owned automobiles; 7% of households owned motorcycles and 19%, bicycles [7].

The vast majority of inhabitants are subsistence farmers (78%), and most lack regular access to money, especially between harvest seasons [6]. Most homes have no electricity [7]. In 2019, there were 57 mobile phone subscriptions per 100 Ugandans nationwide [8]. While mobile phone ownership is increasing substantially, credit is typically purchased only as needed and may not be available at the time of an emergency.

The Ugandan health care system

Like much of SSA, the Rukungiri District has a pluralistic health care delivery model comprising:

“Traditional” and allopathic practitioners

Private, nonprofit, and government health facilities

Vendors selling medicines in small shops and local markets

Traditional healers are commonly consulted, and one source estimated that 90% of rural Ugandans utilize traditional remedies [9-11]. In many small villages, private clinics and pharmacies are often staffed by laypersons with no formal health care training. In major trading centers, private clinics staffed by nurses, and possibly a physician, may have limited laboratory and/or imaging capacity. The quality of care at private health facilities varies widely, and their patients generally must pay out of pocket before receiving services.

The Ugandan government provides a multitiered health care system with the health service infrastructure following a pattern of the country’s administrative units, with health centers (HCs) of increasing capacity (designated levels II–IV), general hospitals (formerly known as district hospitals), and regional referral hospitals. HC IIs are nurse-run clinics at the village level, while larger facilities (HC III and IV) in trading centers are staffed by midlevel providers, typically Clinical Officers with 2–3 years of medical training, and deliver primary care and maternity services, with capacity to provide limited surgical, inpatient, laboratory, and blood transfusion services [12, 13]. User fees for government health care were abolished in 2001. Following this, an initial decline in services and drug shortages increased the use of private health care [14]. In 2017, when we conducted these interviews, there was no ambulance available in the district.

Karoli Lwanga Hospital ED

Karoli Lwanga Hospital (Nyakibale) is a private not-for-profit district hospital. In 2008, GEC partnered with the hospital to open the first fully functional ED in Uganda and a 2-year Emergency Care Practitioner (ECP) training program for Ugandan nurses. The ECP program was developed to address Uganda’s chronic physician shortage (3 physicians per 100 000 residents in the Rukungiri District as of 2015) and lack of specialty trained EC professionals [6]. The program’s goal was to train ED-based nonphysician clinicians to triage, evaluate, stabilize, and treat patients before admitting them to already-overburdened hospital wards, sending them for surgery, or discharging them home. Unlike other Ugandan health care facilities, patients are clinically stabilized prior to payment, and the ED treats any patient who presents [15]. However, acuity is high, with ED patient mortality of 2.1% and an admission rate of 66.3% (2017, unpublished data). Delayed ED presentation can reasonably be surmised to be among the causative factors.

Methods

To better understand the challenges rural Ugandans face in accessing and utilizing EC we purposefully sampled and interviewed patients presenting to the ED more than 12 hours after onset of chief complaint, or their caregiver when the patient themselves was unavailable. Semistructured interviews addressed actions taken before seeking EC and delays once the need for EC was recognized. We purposefully interviewed a diverse sample of patients by demographics and chief complaints. An interdisciplinary and multicultural research team conducted thematic analysis based on descriptive phenomenology using the Framework Method.

Participant selection

Semistructured interviews with ED patients or their caregivers were conducted at Karoli Lwanga Hospital. A purposeful sample of 50 patients arriving to the ED ≥12 hours after onset of the complaint that they sought care for (chief complaint (CC)) were interviewed between January 19 and March 29, 2017. ≥12 hours was set as a benchmark based on advice from the Ugandan clinicians working in the ED due to their observation that patients arriving outside of this timeframe are subjectively noted to have poor outcomes attributable, at least in part, to delayed presentation. The sample was selected to take into account age, gender, type and severity of CC, as well as “information-rich cases,” which we defined as individuals having significant experience with challenges in EC access and utilization, based on conversations with ED staff [16]. These conversations often focused on challenges with identifying the need for EC, obtaining EC from local facilities, finance, or transport. Further, as weather may influence delay, the interviews spanned dry and rainy seasons. For patients with known chronic disease, we included in our sample only those with acute presentations of the disease, rather than those simply seeking ongoing treatment related to the disease. Patients or their caregiver were approached in person in the ED or hospital wards. Recruitment and interviews did not interfere with patients’ clinical care and were not conducted when the patient was severely ill but rather once they were stabilized, if at all. The “caregiver” accompanying the patient, was interviewed with pediatric patients and patients who could not be interviewed directly because of illness or preference.

Ethics

Research ethics approval was obtained through the University of Arizona Institutional Review Board (Tucson), The AIDS Support Organization Research Ethics Committee (Kampala, Uganda), and the Ugandan National Council of Science and Technology (Kampala). Karoli Lwanga Hospital leadership supported the study. Verbal informed consent was approved by the organizations listed above due to low literacy rates in the community. It was obtained from all participants, witnessed by the lead researcher (AP) and head research assistant (NC) and documented on a verbal consent form with AP’s signature. In the case of caregiver interviews, assent was obtained from patients old enough, or alert enough to provide it. In most instances this applied to the young children, the elderly and those with severe illness.

Interviews

Each interview included 10 core questions and appropriate follow-up queries probing actions taken before seeking EC and delays to presentation once the need for EC was recognized. The interview guide was developed by AP (graduate training in public health, qualitative methods and medical anthropology), BD (emergency medicine global health fellowship and graduate training in public health), MN (Regents Professor of medical anthropology), and HD (graduate training in public health and qualitative research experience) in consultation with Ugandan members of the GEC Research Team. It was based on the guiding theory of the Three Delays Model [17], described in depth in the Discussion, and on prior quantitative data regarding barriers to ED utilization at Karoli Lwanga Hospital ED (unpublished). The interview guide remained constant, with minor changes to wording throughout the 50 interviews. The lead researcher (AP) and Ugandan members of the GEC Research Team translated the core questions into the local dialects (Runyankore/Rukiga). Once consensus was reached in the phrasing of questions, back-translation to English was undertaken. The guide was initially piloted with community members not seeking EC followed by approximately ten ED patients or their caregivers. The English version is provided as S1 File.

After didactic training in qualitative interview-based research methods by AP, the head research assistant (NC), a resident of Rukungiri District and native Runyankore/Rukiga and fluent English speaker, obtained practical training, experience and feedback while piloting the interview guide. NC conducted all interviews with AP present. AP was not involved in providing clinical care. NC, a Ugandan community member not involved in clinical care, approached patients to gauge interest in study participation, limiting pressure to participate. Interviews, including only the participants and these 2 researchers, took place in a private room outside of the ED to ensure confidentiality and lasted 15–60 minutes. Interviews were conducted in local dialects, English, or a combination, based on participant preference. Interviews were audio recorded, transcribed verbatim, and translated into English within one week of the interview by NC, then checked for accuracy by a second bilingual member of the GEC Research Team (KA). If there was a discrepancy in translation/interpretation the audio recording was reviewed by NC and KA together and discussed until consensus was reached. Due to the nature of the sample—patients utilizing the ED—typically participants were not available on site when transcription and translation were complete. The IRB limited collection of contact information to ensure privacy, therefore transcripts were not returned to interviewees for review.

Analysis

An interdisciplinary, multicultural data analysis team composed of ECPs, Ugandan members of the GEC Research Team, and AP was assembled to reduce the possible bias of a single investigator’s interpretation and to enhance the cultural depth of the data analysis [18-21]. The Framework Method, a rigorous and transparent process that produces a clear audit trail of each step of data analysis carried out by multidisciplinary research teams, was utilized [19, 21]. As little is known about delays in accessing available EC, in 2017 when the study was conducted, a thematic analysis based on descriptive phenomenology was utilized. The data analysis team met weekly, utilizing the Framework Method to inductively identify:

Concepts of delay present in multiple de-identified interviews (subthemes)

Quotes depicting the subthemes

Quotes depicting interesting outliers

After the first 25 interviews were conducted and analyzed, the data analysis team identified that no additional large-scale concepts of delay were emerging; thematic saturation was reached. The team convened to review subthemes across these first 25 interviews simultaneously and organize them, yielding the 4 themes described in Results. These themes and subthemes constituted a codebook, included as S2 File, facilitating coding of an additional 25 interviews completed to capture a broader sample of demographic groups and chief complaints across both dry and rainy seasons, given differences in EC accessibility and subsistence agriculture activities. AP and NC then coded the identified themes, subthemes, and quotes in the transcripts using QSR International’s NVivo software, version 10, which was also used to analyze themes of delay by patient and interviewee demographics. The Consolidated Criteria for Reporting Qualitative Research (COREQ) was utilized to ensure comprehensive reporting [22].

Results

Demographics

The 50 ED patients for whom care was sought (13 of whom were interviewed directly) had a mean age of 33 years. Patients had an approximately even distribution of gender, as well as occupation (none, subsistence farmers and small business owner). Demographics and CCs for the ED patients are illustrated in Table 1. A full list of CCs is included as S3 File. Fifty interviews were completed, 37 with caregivers of ED patients and 13 with the ED patients themselves. Interviewees had a mean age of 38 years, and the majority were women, worked as subsistence farmers, and had received primary education. The demographics of interview participants are presented in Table 2.

Table 1

Patients’ demographic data and CC.

Gender, no. (%)	Male: 27 (54)
	Female: 23 (46)
Age (years)	Mean: 33.2
Intervals, no. (%)	< 1: 5 (10)
	1–4: 5 (10)
	5–17: 7 (14)
	18–55: 22 (44)
	> 55: 11 (22)
Occupations (Common), no. (%)	None: 12 (30)
	Subsistence farmer: 16 (32)
	Employed or business owner: 13 (26)
Highest level of education	Median: Primary (n = 20; 40%)
	Range: None-University
CC by body system and/or mechanism of injury, no. (%) *
Medical illness: 38 (76)		Injury: 12 (24 )
Gastrointestinal/genitourinary: 12 (32)		Blunt trauma/fall: 5 (42)
Other medical complaints: 10 (26)		Burn: 3 (25)
Cardiopulmonary: 9 (24)		Road traffic accident: 3 (25)
Neurologic: 7 (18)		Bite: 1 (8)

*See S3 File for full list of CCs.

Abbreviations: CC, chief complaint.

Table 2

Interview participants’ demographic data.

Relationship to patient, no. (%)	Parent: 19 (38)
	Self:13 (26)*
	Adult Child: 9 (18)
	Other family member: 9 (18)**
Gender, no. (%)	Female: 36 (72)
	Male: 14 (28)
Age (years)	Mean: 38.1
	Range: 20–82
Occupation, no. (%)	Subsistence farmer: 33 (66)
	Employed or business owner: 14 (28)
	Student: 3 (6)
Highest level of education	Median: Primary (n = 32; 64%)
	Range: None-University

* Patient interviewed directly.

** Other family member including grandchildren, siblings, cousins, daughter or son in-law, nieces, and nephews.

Delay in ED presentation

All patients included in the study reported the acute onset of new symptoms or acute worsening of previously recognized symptoms leading to ED presentation. Delays were quantified as the duration of the patient’s CC and the time from recognition of a severe condition necessitating EC (by the patient or their caregiver) to presentation at the ED. Both are presented in Table 3. The most notable trend related to treatment delay at the ED was history of chronic disease. ED patients with acute presentations and history of underlying chronic conditions experienced substantially longer delays in reaching the ED than otherwise healthy patients. Among chronically ill patients, recognition of acute symptoms needing EC ranged from 12 hours– 14 days prior to arriving at the ED, with an average of 6 days. For all patients included in the study the mean duration from recognition of symptoms necessitating EC to presentation at the ED as 1 day. Patients with chronic disease and their caregivers reported challenges to decision-making including perceptions that:

Table 3

Patients’ delays in ED presentation.

	All patients	Medical Illness	History of Chronic Disease	Injury
Duration of CC	Median: 5.5 d	Median: 7 d	Median: 112 d	Median: 2 d
	Range: 12 h-3 y	Range: 1 d-3 y	Range: 7 d-3 y	Range: 12 h-30 d
Time from recognition of need for EC to ED presentation	Median: 1 d	Median: 1 d	Median: 6 d	Median: 1.5 d
	Range: 2 h-17 d	Range: < 1 h-17 d	Range: 12 h-14 d	Range: < 1 h-7 d

Abbreviations: CC, chief complaint; d, day; EC, emergency care; ED, Emergency Department; h, hour; y, year.

acute decompensation of chronic disease may be self-limited

the patient’s medication regimen for chronic illness would also treat acute decompensation

the incurable nature of the disease meant acute stabilization and treatment of complications was futile and therefore a waste of limited resources

No other major trends in length of delay were identified based on patient demographics, or medical illness vs injury.

Themes in delay

Delay in ED presentation provides the quantitative explanation of delays, whereas themes in delay—presented below—seek to further understand and explain these delays in ED presentation. An overview of the themes and subthemes of delay identified is presented in Table 4.

Table 4

Themes and subthemes in delay identified.

Themes	Subthemes
Cultural Factors and Limited Knowledge of Emergency Signs and Initial Actions to Take	Difficultly recognizing “danger signs”• Patient’s condition becoming severe before identifying the need for and/or seeking EC• Perception of “cultural illness”
Use of Local Facilities for Stabilization and Advice Despite Perception of Inadequate Services	• Limited first aid or home care• Use of local facilities:1. Assessable on foot or bicycle2. For advice and symptom management3. While gathering funds needed for EC4. Despite perception of poor care at high cost• Use of multiple facilities before seeking EC• Perception of delays in referral to EC
Lack of Resources to Cover the Direct, Indirect, and Opportunity Costs of EC	• Perception of EC as unaffordable• Need for funds for transport prior to seeking EC• Opportunity costs of EC
Inadequate Transportation Options, Especially at Night and in Inclement Weather	• Inadequate vehicles available for the patient’s condition• Dirt roads virtually unpassable in rain• Safety concerns when traveling at night

Cultural factors and limited knowledge of emergency signs and initial actions to take

Approximately half of study participants reported some specific symptom, such as confusion or malaise, as a “danger sign” and increasing symptom severity as a primary rationale for their decision to seek health care. Participants also reported that they did not consider utilizing EC until they feared death was imminent.

On Saturday night around 2:00 am the symptoms became very severe. I couldn’t breathe well; I didn’t sleep, and I was like “I am going to die here” … I said, “let me call a Boda [motorcycle taxi] to bring me so that I die [at] the hospital.” It is from then that I decided to come to Nyakibale [Hospital].

—37-year-old (yo) male patient

I made the decision after she had developed severe diarrhea, vomiting, and wasting, so it is (at) that point that I decided to bring her here for care, otherwise she would have died by now. Because she could not even eat or turn in bed…we thought she was going to die.

—60-yo seeking care for 4-month-old (mo) granddaughter

A substantial number of participants, approximately one-quarter, delayed going to the ED because of concern that they might have a “cultural illnesses” requiring treatment by a traditional practitioner or priest. Cultural illnesses are thought to be caused or exacerbated by factors allopathic medicine cannot treat. In these cases, the ED is utilized only when traditional treatment is thought to have been ineffective, or when traditional causes of the ailment have been sufficiently treated, enabling allopathic medicine to be effective. This approach was often advised by elders in the family or community.

There are [cultural illnesses] that attacked the patient, [which] came from my husband’s family. We first thought that she was suffering from cultural illness. Therefore, we had to first try treating with traditional herbs and then [go] to the clinic.

—20-yo seeking care for her 11-mo daughter

The majority of study participants noted that they received advice before making the decision to seek EC. Most were advised to go directly to the ED, while some were advised to treat the underlying cultural disease before seeking allopathic care. Some participants noted that they did not tell ED staff about seeking traditional treatment for fear the staff might scold them for delaying seeking EC.

Use of local facilities for stabilization and advice despite perception of inadequate services

Over three-quarters of participants reported consulting a nearby allopathic or traditional medical care practitioner before presenting at the ED; approximately half described visiting multiple health care facilities and/or practitioners before ED presentation. Participants reported that obtaining care from multiple facilities is a common practice. Most providers utilized were accessible by walking or bicycling. These nearby practitioners were often consulted to assess the seriousness of the ailment and/or as an initial step in managing symptoms while gathering funds to cover the costs of higher-level care.

In some cases, participants chose to wait until this initial care was deemed ineffective before considering other options. However, most perceived the ED to offer superior care. Private clinics were reported by participants to lack necessary medication and materials, communicate poorly about the patient’s condition and treatment, and generally provide inadequate care at high cost. Participants reported spending an average of $8 on care—a large sum in rural Uganda—before presenting at the ED.

It was lack of money, if you don’t have money then you stay home or go to private clinics because you calculate the money for transport to getting here (the ED) …. then you decide to go to nearest clinic.

—64-yo women seeking care for herself

Some participants felt that private clinics postponed referral to capture revenue, even if they were unable to provide adequate care. Despite this, once a referral was made, participants often had to search for funds or gather supplies before seeking higher-level care.

… First, I sought health care in many private clinics…[Then], I decided to take the patient to [an outside] hospital and reaching there, they referred us here… I took the patient back home [for 2 days] so that I could prepare and come here…. I wanted to prepare what we were to use here [at the hospital], like food, and other things.

—20-yo seeking care for her 11-mo daughter

Lack of resources to cover the direct, indirect, and opportunity costs of EC

While most participants understood that payment was not required before ED care, they still felt they did not have the resources needed to seek EC. The majority of participants reported that lack of finances was the main reason for delay once danger was recognized. We queried perceptions regarding how much cash on hand participants felt they needed going to the ED. Speculations ranged broadly ($10-$150) and seemed unrelated to severity of patient complaint. Whatever the dollar amount mentioned, participants felt that receiving care at Karoli Lwanga Hospital would be “somewhat” to “extremely” unaffordable. In many cases it was not the direct cost of the hospital that caused delays once a health problem was deemed severe, but indirect costs, including the expense of traveling to the hospital and paying for nonmedical supplies that would be needed by the patient and caregiver during their stay.

On average, patients traveled 12.6 km from their parish of residence to the ED (based on the distance from the center of the identified parish to the hospital determined through GIS measurement from a prior study [23]) at a mean cost of $6. Half reported that finding funds to pay for transportation was the greatest factor associated with delay. Once they obtained funding for transport, participants presented to the ED for evaluation, stabilization, and treatment and continued to search for funds to pay their hospital bill after obtaining EC.

Right now, we are still soliciting for money. We thought about her life first, before thinking about money for medical care here… We were like “we will look for money later but let us first take her to the hospital.”

—25-yo seeking care for her 61-yo mother

The opportunity costs of health care seeking on work and ability to perform essential household activities impacted the decision to seek care. This especially impacted women who had to find alternative childcare or substitute agricultural labor in their absence from home. Extended family and neighbors were often asked to assist in supervising children, tending gardens and livestock, and ensuring the security of properties.

“No one was left (at home). (The children) are at school, but on returning home they will prepare food for themselves. Their grandmother (lives) at a distance but sometimes she comes and helps…You can’t leave the home alone. Also, farming being our business and it being a season of planting is hard.”—28-yo seeking care for her 9-yo daughter

I am the one responsible for my farms. I was harvesting millet and eventually got so sick I couldn’t stand. Actually, this caused a long delay, for 4 days, because I was thinking about who will take care of my farms. I had even reached the point of refusing to be taken for medical care. I was worried about what I will feed my children after being discharged from the hospital, because I was leaving all my harvests in the farm…This disturbed [me] a lot.

—41-yo male patient

Inadequate transportation options, especially at night and in inclement weather

Transportation is a common source of delay for several reasons. Participants reported difficulty reaching a main road to access vehicles for hire. Once at the road, many reported difficulties locating a car or truck adequate for transporting patients with severe illness. Inclement weather, especially rain, further constrained participants’ transport options. During monsoon season, motorcycle taxis—the most common form of transportation—were often unusable as dirt roads turned to mud following torrential rains.

Nyakibale is very far away from us, otherwise I wouldn’t seek medical care at [local] clinics…Poor roads and inadequate transport means delay us and make patients severe.

—60-yo seeking care for her 4-mo granddaughter

Traveling at night was seen as an additional safety hazard. A third of participants reported waiting until morning to seek EC, despite a patient’s worsening condition. Most patients who arrived at the ED at night did so by car, a far more expensive and difficult to arrange option.

The first barrier was that the patient got burnt at night and I couldn’t get means of transport to bring us here quickly. Also, I was the only one at home with other young children, therefore I wouldn’t get someone to watch the children under my care. Even on the following morning it rained heavily until 2:00pm, that’s when I decided to come.

—40-yo women seeking care for her 4-yo daughter

Discussion

Extensive literature is available on factors influencing illness-specific (eg, HIV, malaria) and demographic-specific (eg, pediatrics, maternity) health care seeking in Uganda [24-31]. However, only sparse research exists on challenges facing patients experiencing acute illness or injury, and little of this pertains to when formal EC is available in a rural area.

This study presents data on the first stage of multistage formative research centered on gaining insights into factors influencing treatment delay to EC by rural community members [32, 33]. A forthcoming study will explore possible interventions to increase timely EC access. This study identified four broad areas where interventions are warranted. It was informed by the Three Delays Model initially developed for obstetric emergencies [17] and applied more broadly to EC by Calvello et al [34] and Broccoli et al [35]. In the Three Delays Model [17]:

The first delay is the decision to seek care, including factors such as:

Availability of decision makers

Cultural perceptions that predispose community members to visit a traditional healer and/or self-treatment prior to seeking allopathic care

Distance to EC

Costs of accessing care

Availability of a caregiver to accompany the patient

Perceptions of the etiology and severity of an illness or injury, and the quality of care available

The second delay includes challenges in identifying and reaching a health facility, such as:

Distance to EC

Availability and cost of transport

Availability of a caregiver to accompany the patient

The third delay relates to receiving adequate and appropriate treatment at a health facility, including factors such as:

Appropriateness and quality of care offered

Resources at hand

Competent referral, as needed

We studied the first 2 delays in depth. Although participants alluded to the third delay, we did not explicitly include this delay in the study aims or directly investigate the quality of care or timeliness of referrals offered by other facilities. The interrelated nature of delays in the Three Delays Model and our own themes of delay are illustrated in Fig 1.

Fig 1

Themes of delay of this study correlated with the three delays model.

Abbreviations: EC, emergency care.

Our results generally support the applicability of the three delays model to accessing EC at the community level. While we did not research delay in receiving adequate care at health facilities directly from interview participants’ perceptions of delay in obtaining adequate treatment and/or referral at outlying facilities it can likely be implied that this delay is equally relevant, however further research is needed.

In this study, we found that, on average, 4 days elapsed between onset of the CC and the illness/injury being recognized as needing emergency treatment by the patient or caregiver. Another day passed between recognition of the need for EC and ED presentation.

The first delay–Decision to seek care

Delays often occurred in part because of a failure to recognize danger signs soon enough. Many participants traveled to the ED only when the patient’s condition increased in severity to the point it was considered “life-threatening.” Lack of knowledge or confidence regarding the danger signs heralding medical emergencies is consistent with earlier studies from rural areas of SSA [30, 35–37], as well as the findings of a recent systematic review of barriers to seeking EC in LMICs globally [38].

Home care (dietary change, bathing, bleeding control, and use of medicines purchased at informal pharmacies) is well documented in Ugandan studies of pediatric fever [27] and obstetric emergencies [37]. In this study, we found that once severe illness was recognized, few participants continued home care except as a stopgap measure while gathering resources for consultation with an allopathic or traditional practitioner. In the case of injury, our findings are consistent with other East African studies reporting that few injury victims receive care at the scene of accidents or en route to a health facility, despite lengthy transports [39-41]. In keeping with prior studies, participants reported lack of first aid skills and supplies and expressed a desire for training [35, 36, 42, 43].

In the case of culturally salient folk illnesses, it is deemed pragmatic to seek traditional medicine before allopathic treatment. Traditional medicine is thought to improve efficacy of allopathic medicine, by removing cultural factors. Thus, not trying traditional medicine first might lead to a waste of resources, if only allopathic treatment is used and is ineffective. When the patient recovers after traditional treatment, before or in concert with allopathic treatment, dual causality of the illness is presumed [44, 45].

Based on this data we will conduct community based research which:

Investigates how best to educate community members to recognize the danger signs of acute illness/injury and administer basic first aid.

Assesses self-care practices.

Seeks to better understand existing traditional medicine and explores the possibility of establishing collaborative relationships with traditional healers as a means of promoting rapid referral for EC.

Before presenting to the ED, most study participants sought care from at least one local practitioner offering either traditional or allopathic treatment. Notably, we found that, participants did not choose to go to local allopathic clinics because they had confidence in, or good rapport with, local practitioners. Rather, they resorted to these clinics because of their accessibility and the perception that obtaining care at the ED would prove costly, even if funds were not demanded before treatment.

Similar to the findings of other studies conducted in Uganda, the general impression of participants was that government health stations and private clinics offer low-quality care and are best consulted as a stopgap measure until funds can be secured to travel to a hospital if the patient’s condition does not improve [24, 27, 28, 30]. Interim treatment measures to allow time to monitor symptoms, identify possible etiologies, and select treatment options have also been documented in other settings [46].

Our study suggests that local clinics are resorted to for practical reasons, mainly as a stopgap measure while gathering funds needed for transportation. This suggests a need for both emergency transportation and referral systems in which local practitioners—both allopathic and traditional—feel they are members of a community of health care practice [32]. As members of such a community of practice, practitioners will need to receive training in danger signs warranting immediate referral to a EC.

The second delay–Reaching care

Cost is a major factor contributing to delays in seeking health care in rural Uganda and similar settings [30, 35, 36, 41, 42, 47, 48]. Three distinct types of cost need to be recognized when considering delays in obtaining EC: direct fees for health care services, indirect costs (eg, transport, food, lodging, and medical supplies/medications required for the accompanying caregiver), and opportunity costs of seeking care. Opportunity costs include both the financial impact of health care seeking on income-generating activities and loss of time to perform essential household activities such as childcare [49]. Interviewees tended to focus on indirect and opportunity costs. The most prevalent indirect costs discussed were transport and the need for money to cover necessities like food while at the hospital. Caregivers repeatedly mentioned opportunity costs of accompanying patients to the hospital. Patient caregivers needed to find substitute labor to attend to essential household tasks in their absence. Concern about the status of one’s children and farm were reported by nearly all participants interviewed. Asking kin or neighbors to watch children or farms has been noted to strain social relations, especially if assistance is needed during peak agricultural seasons [50].

Searching for a loan with little disposable income was a common source of delay and one that has been well described by several other studies in SSA [37, 51, 52]. While community saving and lending programs do exist in the Rukungiri District, they are not typically structured to provide emergency funds. Participants spoke of the difficulties entailed in borrowing money, especially between harvest seasons, as most are subsistence farmers. This elucidates the need for future formative research on implementation of EC specific loan schemes that build upon already-thriving rotating credit and microfinance programs within the community.

Difficulties obtaining transportation to the hospital was a theme in nearly all interviews. Interviewees often did not have cash on hand for transport, which is generally paid in advance and not available on credit. Once participants were finally able to gather funds, finding transport was often a challenge. The dangers of poorly maintained roads worsened the challenge, especially at night and during inclement weather. Many participants were left with no other option but to use less-safe forms of transport like motorcycle taxis. Multiple studies on health care seeking in Uganda and other countries in SSA have also identified transportation as a key cause of delay [35, 36, 39, 42, 47, 53, 54]. An unpublished study by the GEC Research Team conducted at this study site found that the distance from the patient home to the ED significantly impacted patient outcomes. In this study, compared to patients living < 5 km from the hospital, patients living 5–10 km away had 1.7 times the rate of mortality at 3 days, and those living >10 km away, 2.2 times [23].

This study indicates the need for an emergency transport system to be set up within the hospital catchment area. Following the period of this study, Nyakibale Hospital has collaborated with Humanitarian Aid for Uganda, a German nongovernmental organization, to set up an ambulance system in the district. Currently, 2 ambulances are staffed by volunteer ECPs and midwifes who carry an emergency phone. Their phone number has been widely distributed in the community. To facilitate timely ED access, payment for ambulance service is not required at the time of transport. Given that mobile phone service is rapidly increasing in rural Uganda, future formative research will investigate whether arrangements might be made with vehicles for hire to pick up patients for a set fee, if and when an ambulance service cannot be accessed.

The third delay—Health care services

Referral delay from local healthcare facilities was beyond the scope of this study and will require a study of practitioners’ recognition of danger signs, the quality of care available at local clinics, and financial motivations for delaying a patient before referral. The collection of such data in the future is needed to inform health service interventions which provide bidirectional rewards for both local practitioners and the ED while promoting timely EC referral; in order to be successful, local practitioners will need to gain status by facilitating timely referrals.

Limitations

Four limitations of this study may be noted. First, interviews were held at the ED. While this allowed access to participants currently utilizing EC and likely reduced recall bias through timely data collection, participants may have been uncomfortable sharing negative perceptions of ED services while receiving treatment. Second, the study did not include patients who were unable to present at the ED and therefore may not have identified additional barriers to EC access and utilization by the most vulnerable people in the hospital’s catchment area. We plan to address this limitation in forthcoming research utilizing community-based focus groups. Third, we did not study delayed referral to ED from local health facilities, nor the quality of care offered at these facilities. Forth, this study was not designed to quantitatively assess household level factors potentially influencing delay in presentation to the ED, such as, age, gender, birth order, and household structure.

A final limitation of this study relates to sampling and the generalizability of the data collected. The objective of this qualitative study was to identify the range of factors that affect care seeking behavior at a relatively newly established ED. A purposeful sample was employed toward this end. As such, this data is not generalizable to other EDs although it is reasonable to hypothesize that similar factors leading to treatment delay may apply.

Conclusions

EC can substantially reduce preventable complications and mortality in LMICs when the signs of acute illness and injury are recognized early, referral networks are established, and community members view EDs as accessible and affordable. Setting up EDs and training staff in LMICs is only the first step in ensuring that availability of EC services. In Uganda, a well-functioning ED was established, and a health care team trained; however, patient presentations to the ED are often delayed, reducing its potential benefits. This qualitative research study identified major factors contributing to delays in visiting the ED. It corroborates and builds on the findings of other non-EC related health service delivery studies utilizing the Three Delays Model and recommends the need for the next phase of formative research to identify and weigh the advantages and disadvantages of various community-based EC educational, financial, transport, and referral network interventions.

Interview guide.

(PDF)

Click here for additional data file.

Code book with supplemental quotes.

(PDF)

Click here for additional data file.

ED patients’ chief complaints.

(PDF)

Click here for additional data file.

6 Aug 2021

PONE-D-21-13731

Building It so They Will Come: Interviews With Emergency Department Patients and Their Caregivers on the Challenges to Emergency Care Utilization in Rural Uganda - a Grounded Theory Approach

PLOS ONE

Dear Dr. Pickering,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Congratulations on a very significant and impactful project conducted with sound qualitative science. The utilization of frameworks and rigorous methods are excellent, the but integration of these frameworks throughout the manuscript more fundamentally and streamlining the discussion to interpretation of the results and support of next steps rather than reiteration of the results are needed. Please mention and utilize a reporting guideline in your methods. These changes while listed as major given the foundational aspect of the framework integration, can also be viewed as minor as they are less about the methods/analysis and more about the description of the results and interpretation. Please also see the comments from Reviewer #1 about data sharing.

Please submit your revised manuscript by Sep 20 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Catherine A Staton, M.D., MSc

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

"Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: N/A

Reviewer #3: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: 1.Is the manuscript technically sound, and do the data support the conclusions?

Overall a well written manuscript on an important topic. I would suggest considering adding additional quotes to support your findings.

2. Has the statistical analysis been performed appropriately and rigorously?

Methodologically sound. Mention in your methods what reporting guideline was used for this manuscript.

3. Have the authors made all data underlying the findings in their manuscript fully available?

You state that the data is available in the supporting files, however this is only the codebook and not the transcripts. If the IRB approves releasing the full transcripts, that would qualify as all data is fully available without restriction.

If not, explain the reasoning for restricted/no access, and consider adding quotes to the full codebook in the supporting information.

Overall comments:

Interesting and important study, with robust qualitative methods, and integrating ongoing local efforts to increase access to EC.

Discussion could benefit from some minor restructuring. Lots of re-iteration of the results. The three delay framework gets introduced, but throughout the discussion it is not clear how it was applied. Consider adding the headings from the identified delays you are discussing, and how those could be addressed (the 4 main points for action you mention in the Conclusion)

It is not clear to me from this paper why community based interventions should be the next focus, instead of provider centered interventions and Health systems strategies.

The Corresponding and first author is affiliated with GEC, the setting the interviews took place and patients were treated, and was present during the interviews. How was a potential power dynamic between the interviewer team and patients managed, so that participants did not feel pressured to participate?

Feedback by line:

Line 36: How was data saturation determined?

Line 40-42: Does this include patient and caregivers? The remainder of the participant characteristics is centered around the patients. It would be helpful to separate gender and occupation for patient and caregivers

Line 40: How have you identified that interventions in the community are necessary instead of possible HS based interventions/ provider interventions etc.

Line 66: remove brackets,

Line 75-83: Good summary of the need for action/ research

103-107: references available?

96-123: What information is essential for the reader to understand your research? Consider shortening this section

Line 139: Why >12 hours after presenting to the ED?

Line 141: How was determined who has significant experience with challenges in EC access and utilization?

Line 154: Was assent obtained from the patients, if the caregivers were interviewed on their behalf, since the interview questions were directly related to the patient experience?

164: add the reference to the three delays model

166/67: some additional detail on the translation process and validation. How was consensus reached?

179: Why are patients utilizing the ED at the time of interview not appropriate for member checking? Please clarify

186: growing field of research now. Consider clarifying, that in 2017, little was known about delays in accessing care

195: refer to the supporting information or where the codebook can be found

203: Are some caregivers of patients also included in the study? Highlight the most important characteristics in text instead of just referring to the table.

206: Table 1: Not immediately clear what the participant sample looks like. Consider splitting the table up in patient and caregiver characteristics

207: Table 2: What are other medical complaints? (Given that it is 26% of all medical cases.

234: remove “a”

260: is that a quote? “Failed” How were failed treatments perceived by the participants?

264: if 8$ is a large sum, provide some context, eg average net income

286: was this data provided by the participants during the interview? Consider clarifying that this is based on their recall and no additional information sources.

295: What is meant with opportunity costs?

344: Was the interview guide developed based on the three delay model or applied afterwards?

347: Figure 1: make sure to demonstrate which your own themes of delay are based on the data.

Reviewer #2: The authors present an important manuscript qualitatively evaluating reasons for delays in seeking emergency care in rural Uganda as elicited from patients and caregivers. A sound framework and underlying theoretical model were used.

- 50 surveys for a qualitative project is a large number, and data saturation was already reached at 25. Were there any additional themes identified in the last 25?

- Table 2 separates CC by medical and injury, yet in table 3 and elsewhere, CC is separated by medical illness, injury, and chronic disease. This should be consistent across tables.

- Further clarification of chief complaint is needed. The authors indicate that only acute presentation of chronic conditions were included in the sample, however, in Table 3 there are chief complaints that have a duration of 3 years noted. This does not make sense for an acute presentation of an illness or a medical illness.

- Given the distinct differences in trauma and medical emergency complaints, it would be interesting to note if there were any differences in subthemes between these 2 groups. For example, "cultural illnesses" requiring treatment by a traditional healer, seem to be more for primarily medical emergency conditions.

- Limitations section notes subjectivity of the data on the quality of local clinics. This is not a limitation and should be removed. Given that this is qualitative research, these are the noted perceptions of the interviewees and important as such in terms of understanding behaviors behind using or not using these services.

Other errors:

Line 164: need to cite The Three Delays model

Line 265: missing a period after ED

Line 351: missing the word "needing" before emergency treatment?

Reviewer #3: The goal of the study was to assess/explore the barriers for accessing and utilizing emergency care in rural Uganda. Which I think the study achieved. It was done by using semistructured interviews. 50 patients and caregivers were interviewed. Good description of method with construction of the interview questions, how the authors tried to avoid bias, ethical approvoals etc. Four major themes of delay appeared: 1) "Limited knowledge of emergency signs and of initial actions to take"; 2) "Use of local facilities for stabilization and advice despite perception of in adequate services"; 3) "Lack of resources to cover the direct, indirect and opportunity costs of EC"; 4) "In adequate transportation options, especially at night and inclement weather". In the discussion the theory of the Three Delays model (where focus was on the two first ones) was discussed and applied to the results and where a correlation was found which I find was good. Four issues (1) teaching programs, 2) EC loans, 3) emergency transport and communicating systems, 4) referral system) are pointed out in the conclusion that needs immediate attention which are all good and in line with the findings but I would have liked them to have been motivated a little bit more and maybe already pointed out in the discussion section.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

22 Sep 2021

Catherine A Staton, M.D., MSc

Academic Editor

PLOS ONE

September 20, 2021

Dear Dr Staton,

Thank you for your in-depth review of our work. We are thrilled that the reviewers find this work to be as important as we do. The insights and reviewer comments helped to clarify and strengthen the paper.

Overall, reviewer comments helped us to clarity the scope of our work both internally and for the reader. The aim of this study was increasing understanding of the challenges and delays to EC access in the community. While we are aware that there is significant work to be done to facilitate timely, quality emergency care at the health facility level this was not the focus of our aims or methods. Ultimately, this study is an initial formative work seeking to understand the emergency care seeking behavior of community members utilizing the established ED. As such the focus was on understanding challenges to emergency care access and not on specific interventions to address these challenges which would be acceptable to and effective for the community. While we make several general recommendations for areas of focus for ongoing research on interventions, this is not the focus of our paper. Rather, as this one of the first studies to describe challenges to seeking care at an established emergency department we contextualize our results within available literature on care seeking in acute illness and injury, drawn largely from the pediatrics and obstetrics literature. We believe that this increases the value of our results in adding to understanding of emergency care seeking behavior and in providing a basis for ongoing research. We have clarified our aims within the discussion section and edited our title to help the reader better understand the scope of the work. We integrated the three delays model into the discussion section and added additional discussion around the support our results lend to the model’s applicability to emergency care access.

We utilized the COREQ guideline in conducting our study and drafting our paper to ensure compressive reporting. This has been added to the methods section.

Our IRB does not allow for data sharing, so we will be unable to publish the interview transcripts. However, as suggested we have added a supplement including our code book with further representative quotes to ensure readers have as broad an access as ethically possible.

We have addressed all reviewer comments inline below.

We look forward to your review of our revisions and final decision regarding publication.

Thank you,

Ashley Pickering, MD, MPH

Corresponding Author

University of Maryland Medical Center

Department of Emergency Medicine

110 S. Paca St, Suite 600

Baltimore, MD 21201

Department Phone: 667-214-2181

Reviewer comments addressed in-line (line numbers refer to markup):

Reviewer #1: 1.Is the manuscript technically sound, and do the data support the conclusions?

Overall a well written manuscript on an important topic. I would suggest considering adding additional quotes to support your findings.

Several additional quotes were added to the results section. We also added supplemental information containing our codebook with many additional quotes by subtheme.

2. Has the statistical analysis been performed appropriately and rigorously?

Methodologically sound. Mention in your methods what reporting guideline was used for this manuscript.

Added line 286/287 - The Consolidated Criteria for Reporting Qualitative Research (COREQ) was utilized to ensure comprehensive reporting.

3. Have the authors made all data underlying the findings in their manuscript fully available?

You state that the data is available in the supporting files, however this is only the codebook and not the transcripts. If the IRB approves releasing the full transcripts, that would qualify as all data is fully available without restriction.

If not, explain the reasoning for restricted/no access, and consider adding quotes to the full codebook in the supporting information.

As above - Our IRB does not allow for data sharing, so we will be unable to publish the interview transcripts. However, as suggested we have added a supplement including our code book with further representative quotes.

Overall comments:

Interesting and important study, with robust qualitative methods, and integrating ongoing local efforts to increase access to EC.

Discussion could benefit from some minor restructuring. Lots of re-iteration of the results. The three delay framework gets introduced, but throughout the discussion it is not clear how it was applied. Consider adding the headings from the identified delays you are discussing, and how those could be addressed (the 4 main points for action you mention in the Conclusion)

These changes were made in the discussion section. The Three Delays model was applied as a framework for structuring discussion and contextualization of key results (as is common in most discussion sections) within a review of the limited available literature and ideas of how these could be addressed were moved from the conclusions to the discussion.

It is not clear to me from this paper why community based interventions should be the next focus, instead of provider centered interventions and Health systems strategies.

As above - The aim of this study was increasing understanding of the challenges and delays to EC access in the community. While we are aware that there is significant work to be done to facilitate timely, quality emergency care at the health facility level this was not the focus of our aims or methods.

The Corresponding and first author is affiliated with GEC, the setting the interviews took place and patients were treated, and was present during the interviews. How was a potential power dynamic between the interviewer team and patients managed, so that participants did not feel pressured to participate?

Added line 242-244 - AP was not involved in providing clinical care. NC approached patients to gauge interest in study participation, limiting pressure to participate.

Feedback by line:

Line 36: How was data saturation determined?

This statement as deleted as the abstract is not the appropriate word count for an in-depth discussion of data saturation.

Question is addressed 282-283 - Data saturation was determined by additional interviews failing to demonstrate additional themes of delay.

Line 40-42: Does this include patient and caregivers? The remainder of the participant characteristics is centered around the patients. It would be helpful to separate gender and occupation for patient and caregivers

Addressed 65-67: The 50 ED patients for whom care was sought (mean age 33) with approximately even distribution of gender, as well as occupation (none, subsistence farmers and business owner).

Line 40: How have you identified that interventions in the community are necessary instead of possible HS based interventions/ provider interventions etc.

Addressed line 90 – 92: Interventions are warranted to address each of the four major reasons for treatment delay. The next stage of formative research will generate intervention strategies and assess the opportunities and challenges to implementation with community and health system stakeholders

Line 66: remove brackets,

Removed

Line 75-83: Good summary of the need for action/ research

Thank you

103-107: references available?

Reference added

96-123: What information is essential for the reader to understand your research? Consider shortening this section

Removed non-essential data

Line 139: Why >12 hours after presenting to the ED?

≥12 hours was set as a benchmark based on advice from the Ugandan clinicians working in the ED

Line 141: How was determined who has significant experience with challenges in EC access and utilization?

Addressed lines 205/206 - based on conversations with ED staff

Line 154: Was assent obtained from the patients, if the caregivers were interviewed on their behalf, since the interview questions were directly related to the patient experience?

Addressed lines 220-221: In the case of caregiver interviews, assent was obtained from patients old enough, or alert enough to provide it. In most instances this applied to the young children, the elderly and those with severe illness.

164: add the reference to the three delays model

Reference added

166/67: some additional detail on the translation process and validation. How was consensus reached?

Addressed lines 250-252 - If there was a discrepancy in translation/interpretation the audio recording was reviewed by NC and KA together and discussed until consensus was reached.

179: Why are patients utilizing the ED at the time of interview not appropriate for member checking? Please clarify

Addressed lines 254-255: Due to the nature of the sample — patients utilizing the ED — typically participants were not available on site when transcription and translation were complete. The IRB limited collection of contact information to ensure privacy, therefore transcripts were not returned to interviewees for review.

186: growing field of research now. Consider clarifying, that in 2017, little was known about delays in accessing care

Added this clarification on line 261-262

195: refer to the supporting information or where the codebook can be found

Codebook with additional quotes included as Appendix S2

203: Are some caregivers of patients also included in the study? Highlight the most important characteristics in text instead of just referring to the table.

Addressed lines 290-298- The 50 ED patients for whom care was sought (13 of whom were interviewed directly) had a mean age of 33 years. Patients had an approximately even distribution of gender, as well as occupation (none, subsistence farmers and small business owner). Demographics and CCs for the ED patients are illustrated in Table 1. A full list of CCs is included as Appendix S3. Fifty interviews were completed, 37 with caregivers of ED patients and 13 with the ED patients themselves. Interviewees had a mean age of 38 years, and the majority were women, worked as subsistence farmers, and had received primary education. The demographics of interview participants are presented in Table 2.

206: Table 1: Not immediately clear what the participant sample looks like. Consider splitting the table up in patient and caregiver characteristics

Clarified relationship of interview participant to patient. Switched the ordered or Tables 1 and 2 to help clarify the demographics for all patients, prior to discussing the overlap between patients and caregivers.

207: Table 2: What are other medical complaints? (Given that it is 26% of all medical cases.

Added all list of all chief complaints in Appendix S3

234: remove “a”

Removed

260: is that a quote? “Failed” How were failed treatments perceived by the participants?

Addressed line 274 - treatment is thought to have been ineffective

264: if 8$ is a large sum, provide some context, eg average net income

As address in the introduction, subsection The Rukungiri District, 78% of the population are subsistence farmers. We have not been able to identify any source of average net income for the district, the western region, or even rural areas of Uganda. Overall average income for Ugandan is significantly higher than in the Rukungiri District due to the impact of urban areas with increased individuals with formal employment.

286: was this data provided by the participants during the interview? Consider clarifying that this is based on their recall and no additional information sources.

Clarification added lines 444-448 - based on the distance from the center of the identified parish to the hospital determined through GIS measurement for a prior study (23)

295: What is meant with opportunity costs?

Clarification added lines 457-458– The opportunity costs of health care seeking on work and ability to perform essential household activities impacted the decision to seek care

344: Was the interview guide developed based on the three delay model or applied afterwards?

Addressed line 510-511 - The interview guide and a framework for the themes we identified were informed by the Three Delays Model

347: Figure 1: make sure to demonstrate which your own themes of delay are based on the data.

Figure updated to clarify

Reviewer #2: The authors present an important manuscript qualitatively evaluating reasons for delays in seeking emergency care in rural Uganda as elicited from patients and caregivers. A sound framework and underlying theoretical model were used.

- 50 surveys for a qualitative project is a large number, and data saturation was already reached at 25. Were there any additional themes identified in the last 25?

Addressed lines 270-282: These themes and subthemes constituted a codebook, included as Appendix S2, facilitating coding of an additional 25 interviews completed to capture a broader sample of demographic groups and chief complaints across both dry and rainy seasons, given differences in EC accessibility and subsistence agriculture activities. Data saturation was determined when interviews failed to demonstrate additional themes related to treatment delay.

- Table 2 separates CC by medical and injury, yet in table 3 and elsewhere, CC is separated by medical illness, injury, and chronic disease. This should be consistent across tables.

We separated CC by medical and injury. History of chronic disease was found to significantly increase delay for acute illness or injury, regardless of if it was related to the chronic disease, so this data as pulled on in table 3 to illustrate this finding.

- Further clarification of chief complaint is needed. The authors indicate that only acute presentation of chronic conditions were included in the sample, however, in Table 3 there are chief complaints that have a duration of 3 years noted. This does not make sense for an acute presentation of an illness or a medical illness.

Addressed lines 314-320 - The most notable trend related to treatment delay at the ED was history of chronic disease. ED patients with acute presentations and history of underlying chronic conditions experienced substantially longer delays in reaching the ED than otherwise healthy patients. Among chronically ill patients, recognition of acute symptoms needing EC ranged from 12 hours – 14 days prior to arriving at the ED, with an average of 6 days. For all patients included in the study the mean duration from recognition of symptoms necessitating EC to presentation at the ED as 1 day. Patients with chronic disease and their caregivers reported challenges to decision-making including perceptions that:

- Given the distinct differences in trauma and medical emergency complaints, it would be interesting to note if there were any differences in subthemes between these 2 groups. For example, "cultural illnesses" requiring treatment by a traditional healer, seem to be more for primarily medical emergency conditions.

Any differences identified have been highlighted in the paper

- Limitations section notes subjectivity of the data on the quality of local clinics. This is not a limitation and should be removed. Given that this is qualitative research, these are the noted perceptions of the interviewees and important as such in terms of understanding behaviors behind using or not using these services.

This was removed from the limitations and is addressed in the discussion section.

Other errors:

Line 164: need to cite The Three Delays model

Citation added.

Line 265: missing a period after ED

Addressed.

Line 351: missing the word "needing" before emergency treatment?

Addressed.

Reviewer #3: The goal of the study was to assess/explore the barriers for accessing and utilizing emergency care in rural Uganda. Which I think the study achieved. It was done by using semistructured interviews. 50 patients and caregivers were interviewed. Good description of method with construction of the interview questions, how the authors tried to avoid bias, ethical approvoals etc. Four major themes of delay appeared: 1) "Limited knowledge of emergency signs and of initial actions to take"; 2) "Use of local facilities for stabilization and advice despite perception of in adequate services"; 3) "Lack of resources to cover the direct, indirect and opportunity costs of EC"; 4) "In adequate transportation options, especially at night and inclement weather". In the discussion the theory of the Three Delays model (where focus was on the two first ones) was discussed and applied to the results and where a correlation was found which I find was good. Four issues (1) teaching programs, 2) EC loans, 3) emergency transport and communicating systems, 4) referral system) are pointed out in the conclusion that needs immediate attention which are all good and in line with the findings but I would have liked them to have been motivated a little bit more and maybe already pointed out in the discussion section.

The four broad intervention strategies identified by the reviewer were moved to the discussion section. We provide limited discussion of interventions as this is largely outside the scope of this work and requires further research is addressed in the letter above.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

18 Feb 2022

3 May 2022

Reviewer Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for addressing our comments to the previous version of the manuscript. I enjoyed reading the article and think you have made the aims of this project and placing it into the larger research focus of your team much clearer. The integration of the three-delay framework and discussion especially are very well written.

I think the abstract is currently not reflecting the changes you made to the manuscript appropriately. Additionally, I suggest some restructuring and clarifications in the methods section, to further increase the rigor of the study. Please see more specific suggestions below.

Line5-7: In the response letter to the editor, you very clearly state the aims of this study. I suggest using that in the abstract. “Implementation of effective EC requires assessment of socioeconomic, cultural, and structural factors leading to treatment delay.” Might lead the reader to believe you will be focusing on the implementation of strategies, instead of first understanding the challenges.

Line 6-7 changed to: “This study seeks to understand the emergency care seeking behavior of community members utilizing the established ED.”

Line 13-14: not clear how translating and transcribing the interviews allow for a grounded theory approach and thematic analysis. I’d suggest simply stating what analysis approach you chose. If there is space, I suggest stating why you chose 50 participants as a sample size or if it was driven by saturation.

Addressed on lines 11-14: “Semistructured interviews addressing actions taken before seeking EC and delays to presentation once the need for EC was recognized were conducted until theoretical saturation and a diverse sample were obtained. An interdisciplinary and multicultural research team conducted thematic analysis based on descriptive phenomenology.”

18: I know space is very limited and you address it later in text, but some readers might wonder why there are almost three times as many caregivers interviewed as patients? Were they used as a proxy because the patients were not available? Could be as simple as stating “patients or caregivers of patients being unavailable were interviewed on their behalf.”

Addressed lines 18-19: “Interviews were conducted with 13 ED patients and 37 caregivers, on the behalf of patients when unavailable.”

16-19: you addressed the age questions in the revisions, however I still find it slightly confusing. The mean ages are not clear. All patients had an average age of 33, but all interviewees were 38 years on average? This sentence seems out of place, as you switch back to the patients’ experiences afterwards. I suggest removing that sentence altogether.

This was removed and helped to make space for other clarifications.

Line 58-60. Much clearer description of the aims, using a more qualitative language than in the abstract

Thank you.

Line 61-108: appreciate the explanations and giving context to the reader. Well written and informative.

Thank you.

Methods:

- I suggest using sub-headings for the methods section to make it more reader friendly.

These have been added.

- Provide a brief statement at the beginning of the methods that outlines the study design and analysis approach. This will be especially beneficial for readers that are not well versed in qualitative research designs.

This was added on lines 122-128: “To better understand the challenges rural Ugandans face in accessing and utilizing EC we purposefully sampled and interviewed patients presenting to the ED more than 12 hours after onset of chief complaint or their caregiver when the patient themselves was unavailable. Semistructured interviews addressed actions taken before seeking EC and delays once the need for EC was recognized. We purposefully interviewed a diverse sample of patient demographics and chief complaints. An interdisciplinary and multicultural research team conducted thematic analysis based on descriptive phenomenology, using the Framework method.”

- You state in the title that a grounded theory approach was used, however that is a very specific iterative approach, that requires more details on how the approach was followed, vs an analysis based in phenomenology for example. Make sure you truly used a complete grounded theory approach, rather than utilizing certain elements based in grounded theory. Specify how the Framework Method helps to follow a grounded theory approach.

This is a very valid insight. While a true grounded theory approach was the intention, we used more pragmatic approach which does not fully adhere to grounded theory. The mention of grounded theory has been removed throughout.

Line 111: you justify the choice for the 12 hour time mark in your response letter, I suggest adding that here. It provides additional rigor to the methods.

This was added in lines 125-127: “≥12 hours was set as a benchmark based on advice from the Ugandan clinicians working in the ED due to their observation that patients arriving outside of this timeframe are subjectively noted to have poor outcomes attributable, at least in part, to delayed presentation.”

Line 114: information rich cases. Briefly state how you determined who’d have significant experience with challenges in EC access prior to having interviewed them. What were the characteristics the ED staff suggested you to include to meet this criteria?

This was added on lines 139-140: “These conversations often focused on challenges with identifying the need for EC, obtaining EC from local facilities, finance, or transport.”

Line 129-131: I am curious if severe illness of the patient was a common reason for caregivers to refuse the interviews. I imagine this could be a very stressful and potentially challenging interview for caregivers and patients to have. Where any resources put in place, in case they needed support?

For exactly these reasons we did not approach caregivers when patients were severely ill. A description was added on lines 145-146: “Recruitment and interviews did not interfere with patients’ clinical care and were not conducted when the patient was severely ill but rather once they were stabilized, if at all.”

Line 132-145: good description of the interview guide development. Was piloting of the questions performed? If not, what was the rationale?

Added on lines 173 -175: “The guide was initially piloted with community members not seeking EC followed by approximately ten ED patients or their caregivers.”

Line 146: briefly describe how the training in qualitative interview-based methods looked like. Didactics only, practice sessions, etc.

Added on lines 177-179: “After didactic training in qualitative interview-based research methods by AP, the head research assistant (NC), a resident of Rukungiri District and native Runyankore/Rukiga and fluent English speaker, obtained practical training, experience and feedback while piloting the interview guide.”

Lines 172- 183: the data saturation statement is still not clear. I understand the rationale about adding more interviews to have a broader demographic profile, but it still sounds like saturation was reached after 25 interviews and not after 50. How exactly was saturation determined? Provide more details on the type of saturation (thematic vs data saturation)

Revised lines 204 -211 for clarity: After the first 25 interviews were conducted and analyzed, the data analysis team identified that no additional large-scale concepts of delay were emerging; thematic saturation was reached. The team convened to review subthemes across these first 25 interviews simultaneously and organize them, yielding the 4 themes described in Results. These themes and subthemes constituted a codebook, included as Appendix S2, facilitating coding of an additional 25 interviews completed to capture a broader sample of demographic groups and chief complaints across both dry and rainy seasons, given differences in EC accessibility and subsistence agriculture activities.

Line 194: I appreciate switching the Tables 1 and 2. Much clearer

Thank you for the useful suggestion.

198: the distinction of “delay in ED presentation” and “Themes in Delay” is not very clear. Delay in ED presentation provides the quantitative explanation of delays, whereas themes in delay further explains and seeks to understands the delays.

Although you understood and stated the difference perfectly this is a very helpful clarification to make. This was added on lines 237-240: “Delay in ED presentation provides the quantitative explanation of delays, whereas themes in delay — presented below — seek to further understand and explain these delays in ED presentation.

220: A brief overview of themes and subthemes in table form or some visual could be helpful

A very helpful idea. Table 4. Themes and subthemes in delay identified was added on lines 244 – 245:

Table 4. Themes and subthemes in delay identified

Theme Subthemes

Cultural Factors and Limited Knowledge of Emergency Signs and Initial Actions to Take • Difficultly recognizing “danger signs”

• Patient’s condition becoming severe before identifying the need for and/or seeking EC

• Perception of “cultural illness”

Use of Local Facilities for Stabilization and Advice Despite Perception of Inadequate Services • Limited first aid or home care

• Use of local facilities:

1. Assessable on foot or bicycle

2. For advice and symptom management

3. While gathering funds needed for EC

4. Despite perception of poor care at high cost

• Use of multiple facilities before seeking EC

• Perception of delays in referral to EC

Lack of Resources to Cover the Direct, Indirect, and Opportunity Costs of EC • Perception of EC as unaffordable

• Need for funds for transport prior to seeking EC

• Opportunity costs of EC

Inadequate Transportation Options, Especially at Night and in Inclement Weather • Inadequate vehicles available for the patient’s condition

• Dirt roads virtually unusable in heavy rain

• Safety concerns when traveling at night

239: Could you include the specific word used for “cultural illness” in the local dialects?

I discussed this with Charles Ndyamwijuka and the Ugandan members of our research team all of whom are native speakers. There is no one specific word for “cultural illness” these were identified through patient or caregivers’ descriptions of the presumed etiologies and treatment strategies.

259: Do you have information on what the maximum number of facilities visited before the ED presentation was?

Unfortunately, I do not. We queried types of facility and estimated costs but did not ask for a specific number of facilities. The use of multiple facilities was a somewhat unexpected result and was identified as multiple interviews were compared.

288: I understand you mentioned in your response letter that it is difficult to provide context on how much a dollar sum is in the Ugandan context, specifically in this region, where a national net income would likely be an overestimate. But I think it is important to emphasize somehow, that 150$ is a significant sum and potentially catastrophic healthcare expenditure. Maybe you could add some small section in the description of the setting (lines 68-69)

This is included on lines 80-81: “The vast majority of inhabitants are subsistence farmers (78%), and most lack regular access to money, especially between harvest seasons [6].”

324: change to past tense, check for consistency throughout

This was changed and the rest of the manuscript reviewed.

343: Discussion makes the objectives and planned next steps much clearer, positioning this paper in the ongoing research efforts. Good integration of the available literature.

Thanks!

Reviewer #2: (No Response)

Reviewer #3: I think you have addressed and covered the reasons for delay in careseeking of this area in Uganda well and got results that are in line with the Three Delays Model. These results are discussed thoroughly in the discussion as well as suggestions and thoughts on how to move forward are pointed out. With all studies there are limitations and I think you are aware of them and address them wisely.

Submitted filename: Response to reveiwers 5_22.docx

Click here for additional data file.

19 Jul 2022

Getting to the Emergency Department in Time: Interviews With Patients and Their Caregivers on the Challenges to Emergency Care Utilization in Rural Uganda

PONE-D-21-13731R2

Dear Dr. Pickering,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

George Vousden

Staff Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The manuscript has been appropriately modified. authors have adequately addressed the suggestions from the reviewers.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

**********

27 Jul 2022

PONE-D-21-13731R2

Getting to the Emergency Department in Time: Interviews With Patients and Their Caregivers on the Challenges to Emergency Care Utilization in Rural Uganda

Dear Dr. Pickering:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. George Vousden

Staff Editor

PLOS ONE