Data collection and study instruments.

Data were collected at initial hospitalization as a part of a prospective trauma registry (S1 Data) and again at a 2-week follow-up via administration of a standardized survey tool that was developed as a part of a larger study on patient care transitions into the community after injury (S2 Data). Data were collected between August 2018 and February 2020. Participants were screened for participation at the time of entry into the trauma registry.

The 2-week follow-up survey was developed with input from two physicians from the United States, a physician from Tanzania, a physical therapist and the local community advisory board (CAB). The tool was then reviewed for content validity, format, and cultural appropriateness by the local Tanzanian emergency medicine research team of trained research nurses and research assistants. The survey was piloted by the local research team prior to implementation in order to ensure correct language translation, cultural appropriateness, and to be fully inclusive of all local means of transport.

The survey tool was initially developed in English and then translated into Swahili by bilingual research team members. The tool was administered in Swahili by trained research assistants at the patient’s home, hospital clinics, or via phone interview.

Study participants.

Patients enrolled in the prospective trauma registry were screened for a two-week post-discharge follow-up survey. Inclusion criteria for the initial trauma registry were: 1) age ≥ 18 years, 2) had an injury requiring hospitalization within 24 hours of presentation, 3) initial visit for the injury. Patients enrolled in the trauma registry were then further screened for participation in the two-week follow-up study if they resided in Moshi Urban, Moshi Rural, or nearby (defined as an estimated cost of 5,000 Tanzanian shillings for transportation from their home to KCMC). Patients who were discharged to a location other than their usual home environment, such as prison or a substance abuse rehabilitation center, were excluded.

Variables.

Demographic information and variables specific to the injury and prehospital course were collected from the prospective trauma registry and included the following: age, sex, district of residence, insurance, employment status, mechanism of injury, mechanism of arrival, initial health centre to which the patient presented and injury intent. Data on hospital course included: ICU stay, length of stay in the ICU and hospital, number of surgical interventions, Glasgow Coma Score (GOS), and a modified Functional Independence Measure (FIM) (S1 Data), designed to measure disability. The modified FIM was completed by either the patient or a family member with results categorized as “dependent” (score in the lowest quartile, <51) or “independent” (any score greater than 51).

The two-week follow-up survey collected data from both open and closed-ended questions on the patients’ ability to travel alone, means of transport to appointments, the need for an escort to travel, social support and functional independence. The primary outcome, difficulty in traveling since the injury, was defined as barriers the patients encountered in traveling after hospitalization and was measured through the patient’s self-reported answer (Yes or No) to the following question: “Is travel more difficult since your injury?”. Additional questions were asked to further specify if the challenges were due to financial, logistical, or other reasons. Social support was measured through the modified Medical Outcomes Study Social Support survey (mMOS-SS) (S2 Data), an eight-question survey designed to determine emotional and tangible social support in patients with chronic illnesses. Results from the mMOS-SS were dichotomized, with scores in the lower quartile (less than 21) considered “poor support” and any scores above this categorized as “good support”.

Data analysis.

Statistical analysis was performed using R statistical software, Version 4.0.2. Categorical variables were expressed as frequencies and percentages, continuous variables were expressed as mean and standard deviation (SD). The associations between the outcome of interest, difficulty in traveling after injury, and the predictors described earlier were assessed using logistic regression. Univariable and multivariable models were developed in an exploratory analysis approach. The association between the selected predictors and outcomes are reported in odds ratios and 95% confidence intervals. No elimination methods were applied and models were built by building predictors that were deemed to theoretically be associated with difficulty in traveling since the injury. Missing data were analyzed and deemed to be missing at random, thus a complete case analysis was performed in the multivariable models.

Study design and theoretical framework.

We used a phenomenological approach to focus on the individual experiences of the patients, caregivers, and healthcare personnel. The study adopted a longitudinal qualitative research using a multi-perspective approach in order to acquire dynamic experience on barriers for transportation for injury patients post hospitalization.

Data collection.

Qualitative data via focus groups and in-depth interviews were collected between January 20, 2019 and December 12, 2019. Five Tanzanian research assistants, three females and two males with either a bachelor’s or master’s degree and a registered nurse all with training in qualitative data collection, collected the data through in-depth interviews and focus group discussions. A sample size of 45 patients and caregiver interviews was estimated. If thematic saturation of qualitative data was reached prior, data collection was terminated.

Interviews lasted approximately 30 minutes on average, but ranged between 10 and 70 minutes. Interviews were conducted in Swahili in a private room at KCMC or at the patient’s home; only the participant and interviewer was present when the interview was taking place. The researchers informed the participants on the study process and goals at the beginning of data collection. The researchers also emphasized that they were not involved in the clinical care of the patient, nor would their participation in the study affect the patient’s treatment or hospital care. Interviews were audio-recorded on encrypted devices and uploaded onto a secure database behind a firewall on a university database. Audio recordings were deleted from the devices after being uploaded.

Study instruments.

Semi-structured interview guides for both the in-depth interviews and focus groups (S1 and S2 Text) were developed iteratively with input from mixed methods researchers from both institutions. The interview guides were initially created in English and then translated and back-translated by bilingual Tanzanian research assistants. The local Tanzanian research team evaluated the interview guides for cultural appropriateness and piloted the instruments prior to implementation. The interview guides were reviewed by the research team mid-way through data collection and revised based on preliminary results.

Study participants.

Convenience sampling was used for in-depth interviews with patients as they were identified at the time of entry into the trauma registry. Additional data were collected on the barriers to transportation through interviewing the patients’ family members and two focus group discussions with health care providers, and Community Advisory board (CAB) members. Inclusion criteria for patients included the following: 1) age ≥ 18 years, 2) medically stable, 3) living in or near Moshi Urban or Moshi Rural and 4) injuries requiring hospitalization. Patients who were not discharged to their usual home environment, such as patients who were incarcerated or admitted to a substance use rehabilitation center, were excluded. Patient caregiver inclusion criteria were, 1) age ≥ 18 years, 2) residing close to the patient after discharge. Any CAB member 18 years or older was eligible for inclusion.

Data analysis.

Interviews were transcribed ad verbatim in Swahili to ensure consistency and then translated to English by bilingual Tanzanian research assistants. The initial codebook was developed by a qualitative researcher following a mix of deductive and inductive analysis approaches using the first round of interviews. The coding was done by three data analysts, and the emergent codebook was developed and updated throughout the data collection process through discussions between data analysts and the interviewers, to validate the interpretation of our findings and to validate the developed themes.

The de-identified transcripts were coded and analyzed using google spreadsheets, which allowed for close collaboration and communication between the research team in different locations by providing comments and posing questions throughout the coding process.

Financial challenges and transportation costs.

Challenges in obtaining transportation due to financial constraints were reported by participants in both the interviews and focus groups as barriers to transportation in recently injured patients. Patients, family members and health care providers mentioned financial challenges and transportation costs as limiting factors in continuing treatment even with the knowledge that it was necessary.

Participants in both interviews and surveys reported that most injury patients had to find extra money for hiring means of transportation resulting in increased financial burden to the patients. They indicated that an inability to use public transportation forced the patient to hire bajajis and taxis for transportation to follow up clinics. Also, the fact that the patient was not able to work, resulting in a loss of income made it even harder to afford the cost of transportation. These challenges are reflected in the following quote:

"[…] not much because the way we bring him here we must use a means, and that transportation is taxi, and that is the only thing that can carry him otherwise nothing carry him and the taxi has to be paid for, it is not free and that also is a challenge" (Family IDI 111)

Physical abilities.

Participants in interviews and surveys identified physical limitations as a barrier to transportation. Examples included patients not being able to bend their leg after injury or leg swelling when travelling on long routes. In one interview, a patient stated that he had to send someone to bring a bajaji to him as he was physically unable to hail one. The following quotes demonstrate examples of these types of barriers:

“I can’t go very far because if I walk for long my arm aches" (Patient IDI 567)

“Yeah, because I have neighbors, the day that my wife hasn’t gone out, I just send someone, the bajaji pass here often, the person goes to stand there and wave and come with it, even our neighbors here have three bajaji” (Patient IDI 593)

Ability to use regular mode of transportation.

Most of the participants in interviews, surveys and focus groups identified an inability to use their regular mode of transportation as a barrier to transportation for injury patients. A patient in an interview said he used to rely on his motorcycle for transportation and now he has to hire a bajaji. Several patients had to change from their usual means of transportation to more comfortable means. The following quote represents this challenge.

“Transportation is the problem because I was relying on my motorcycle and that’s not working anymore, here I just hire a Bajaji when I am going for wound dressing at Msaranga Dispensary.” (Patient IDI 423)

“Transportation, we had to be using the public transport, it also affects income because you find sometimes he used to go alone and come back but now you have to pay someone to go with him or sending someone to look for needs in the market.” (Family 240)

Access to health care.

Health care providers and CAB members mentioned that access to health care was affected by transportation challenges. One CAB member believed that transportation to receive medical care was the first challenge in accessing healthcare, since medical services are not close enough to be accessed without modes of transportation. Additionally, one patient mentioned that transportation was required in order to access specialized care. This was reiterated by a healthcare provider focus group participant who stated that a lack of healthcare specialists, particularly for those with chronic or long-term rehabilitation or medical needs, the cost of public transportation and the time commitment required in order to travel long distances to access specialized care were all barriers to access. Another healthcare provider mentioned the need to educate patients on nearby health facilities to limit the need for transportation and thereby improve access to care and follow up appointments. Overall, healthcare providers identified a lack of proximity to healthcare and the need for specialized services as barriers that ultimately limit access to care.

“…the relevant institutions should ensure that we have the enough staff to serve those patients who come to us instead of getting access to other low level centres, for example the expansion and understanding of the service available to the patient who has suffered injuries” (healthcare provider)

Compliance to doctors’ orders.

Healthcare providers stated that when patients live too far to properly follow up with them, there is an increased risk of patients not adhering to medication schedules, resulting in complications. Lack of transportation was also mentioned as a limiting factor to accessing medications by both patients and caregivers.

Patients in surveys also mentioned that they were instructed not to move or travel and had to comply with their doctor’s order for full time bed rest, thus limiting their ability to travel.

“Doctor restrictions (full time bed rest)” (patient survey)

“I was told not to move this leg so much” (patient survey)

“[…] a doctor must require you when you sit you have to put your legs straight, even when I will be in the car legs must be kept straight, I use the elongated back seat as precaution […] I can’t sit like other normal passengers.” (P577)

Assistive devices.

Patients in interviews described the need to use assistive devices for movement, which in some cases served as barriers in using certain modes of transportation. Patients in surveys stated that they needed to use assistive devices like crutches or wheelchairs for moving around and sometimes had difficulties using them.

“I have to sit on a wheelchair and get into the car with a lot of pillows” (Patient survey)

“Using the crutches was hard at first, getting tired and falling” (Patient survey)

Social support.

Participants in most of the interviews and surveys mentioned social support as an important factor in transportation for injury patients. Dependence on family support and community as a whole was brought up by both patients and relatives in interviews and surveys when transportation was being discussed.

Participants in interviews and surveys mentioned that transportation required support from both the community and family. Some patients indicated that they were unable to travel to the clinic without the help of family members or friends. These family members or friends would assist the patients in ambulating by holding their hands when they experienced pain or dizziness. Some required assistance in holding their legs during transportation to avoid excessive movement on rough roads. The following quotes highlight this need:

“Asking for transportation help from friends” (Patient survey)

“…it is the bajaji, it comes to pick me up and my wife takes me there direct, we get to the gate and she takes my hand and we go to the eye ward” (Patient IDI 593)

Personal factors.

Several patients in surveys mentioned that they had difficulties travelling due to personal concerns. Two patients identified a fear of reinjuring themselves as a barrier to transportation.

“I have to get private transportation because the public transports are usually so congested and so it is easier to re-injure my leg.” (Patient Survey)

“I don’t travel because I have to be careful not to reinjure myself” (Patient survey)