Setting

Interviews were conducted at three institutions in two metropolitan areas. One site was a large university hospital in Bogotá. Another was a large public teaching hospital in Cali with an emergency department, operating theaters, and dedicated neurotrauma intensive care unit (ICU). The third was a private clinic for the care of patients of road traffic accidents in Cali. Interviews were conducted with researchers and trainees at Fundacion Meditech, an educational and research foundation based in Neiva and Cali that serves as the coordinating center for the registry.

Key Informant Interviews

The interview questions were designed using the Theoretical Domains Framework (TDF) which is a set of 12 domains used to guide the implementation of health care initiatives.13 14 15 16

Our research group previously used this framework to investigate the transition from paper to electronic surgical records in a low- and middle-income country (LMIC).15 We adapted the focus group questions used in this effort to fit the context of a multisite neurotrauma registry. Two researchers with experience in trauma systems research in Latin America reviewed this revised question set. The questions were translated. The translation was again reviewed for accuracy and relevance by the research team.

Interviews took place over a 6-week period in 2018. Participant names and other identifying information were not recorded. Interviews were conducted in either English or Spanish. A translator fluent in both languages was available to assist during interviews. Interviews were audio recorded, then transcribed for coding. Spanish interviews were translated into English and were compared with the Spanish transcripts and reviewed for accuracy by a member of the team.

Participant Observation

The ethnographic portion of this study aimed to understand the current state of implementation of the neurotrauma registry at each study site. Participant observation is a form of ethnographic research where a researcher observes and records data while engaging directly with the sample population.17 18 This method was selected to encourage communication between the researcher and those working locally on the registry. Observations were conducted in research and administrative offices in participating institutions, as well as in emergency departments, outpatient clinics, ICUs, and operating rooms.

Qualitative Coding

Our analysis was focused on identifying barriers and facilitators to the implementation and use of the neurotrauma registry. Qualitative findings from interviews and observation logs were organized into themes. The data were coded independently by two of the research team and were grouped into relevant themes. Any disagreements were resolved by consensus.

Summary of Implementation Efforts and Identification of Stakeholders

At the time of the study, pilot data collection had been active at three study sites for approximately 6 months. Researchers had developed a targeted set of registry data domains and data elements through a process of iterative review by regional content experts, such as physicians, health care providers and researchers working in neurosurgery, trauma and emergency services in several Latin American countries, and developed to fit regional needs and differences in health care delivery and management. The study team had selected and tested an online data collection tool. Two trainees developed the online platform with the support of their university. These two trainees were also primarily responsible for the support and maintenance of the registry platform. The registry itself was maintained on servers at a large, private university in Bogotá, Colombia, and administered through a secure online data collection system. Pilot data collection had commenced at three clinics associated with a private hospital for the treatment of road traffic injuries in Cali, Colombia. Physicians working in the emergency department, outpatient clinics, surgical services, wards, and ICU of the pilot sites identified eligible patients and entered patient data into the online data collection tool. Clinicians and researchers at the two other participating health care institutions had identified personnel to perform patient identification and data entry, received training in the use of the registry tool, and worked with their local IRBs to obtain ethical approval for the project.

We identified clinicians specializing in emergency medicine, critical care, general internal medicine, and neurosurgery as stakeholders in the registry. Nonclinicians, such as researchers, hospital administrators, policymakers, patients, members of the public, and trainees, including medical students, interns, residents, fellows, and graduate students, were also identified as stakeholders.

Identification of Themes

Throughout the study, we conducted 20 interviews with key stakeholders in the neurotrauma registry. Participants identified time constraints, funding and resource limitations, information technology (IT) and software issues, challenges in obtaining ethical approvals, deficiencies in clinical data collection and lack of long-term follow-up as barriers to implementation efforts. Factors that helped with efforts to create the neurotrauma registry included the ease of use of the registry tool, the ability to integrate registry data collection into the clinical workflow, the presence of local champions, and individual engagement with the project. Support from the administration was cited as both a barrier and a facilitator, depending on the perceived level of support for the project from institutional administration.

Stakeholders frequently discussed their goals for registry data. Many participants stated that using registry data to improve the quality of patient care at their institution was a central goal of the project.

Barriers to Implementation: Incomplete Clinical Data

Most stakeholders described limited or incomplete clinical data collection as a significant barrier to the success of the project. One respondent said, “we know the importance of cranioencephalic trauma, but we do not record it in the clinic. We know it, and we see it, but we do not write it down. For example, a patient has amnesia, but they didn’t record for how long, what is the time course of the amnesia.” Another said, “what makes it difficult is that the data as such is not complete. What would help us to create the registry? First, to have complete data to be able to do it. To know statistically how many patients are admitted each month, what was their evolution over time, what was their treatment; I think that would be important.” Occasionally, data were missing due to issues with equipment and other resources. One stakeholder mentioned, “for example, they wrote, ‘pupils equal and reactive’ in the electronic record, but they don’t calculate pupil size, because we don’t have an established method to measure the pupils.” Many stakeholders were concerned with the lack of long-term follow-up in their patient population. One clinician pointed out, “the study’s follow-up period is 24 months, and the patients here are vulnerable patients, and in some cases, outpatient follow-up is difficult.” Another provider described loss to follow-up because of insurance issues. “When the patients come here, they come here with insurance support, but this insurance has a limit, more or less $5,000. When the patient’s cost of care surpasses this amount, the patient should go on to receive medical assistance with his insurance company. And then we lose the patient.”

Limited Resources

Limited resources were another frequently described barrier. One stakeholder said, “well, the Wi-Fi here does not serve us much (laughs) because it fails. Then there is the issue of the computers: there are times that using the clinic computers is problematic. Here there are only two computers, so that part is difficult.” Most stakeholders described computer and IT issues. For example, one stakeholder mentioned, “The only problem is that sometimes the platform falls off. Sometimes we load some data, we record the data, and suddenly when we next open the platform, the data were lost. Suddenly, all of the data are lost.” Other issues with the platform include errors in the database. One stakeholder said, “I notice a lot of errors with the branch logic. I don’t know if the other person (IT support) has much time to correct those errors.” One individual remarked, “another issue is the lack of dedicated IT support staff at the university; most of these guys handle a number of different projects, not just the registry. When there are problems with the registry, we often have to wait until they have time to address them.”

Time Constraints

Many stakeholders mentioned that time constraints posed a significant challenge. One said, “within the time part, it is quite difficult. We are two doctors here in the hospital part, and we have many patients. Here at (our clinic) is where there is the highest volume of hospitalized patients. Then there are 40, 50 to 60 patients, in a 2-hour shift, it is necessary to type everything, absolutely everything, then the time is quite limited.” Another talked about the time to enter patient data into the registry tool: “the most difficult thing could be the number of items, but it is something that has to be evaluated. If more information is required, obviously much more time is required to fill it; that comes into play.”

Finally, ethical approval at the many clinical sites was described as a significant barrier. “It is a matter of ethics, and above all, of going through all the committees of each institution in order to have the approval ….” Another researcher said, “then, there is also the issue of how to obtain IRB approval at all local sites–it can be difficult and time-consuming. Currently, we are focusing our attention on sites where approval is already in place. For other sites in general, the process is to give them the IRB applications submitted already and to act as resource if there are any questions.”

Positive Factors: Ease of Use and Integration of Data Collection into Clinical Workflow

A commonly cited strength of the neurotrauma registry was the simple, easy to navigate online data collection system. One stakeholder said, “the form is something that is quite (user-) friendly. It can be filled out very easily.” Others felt that the flexibility of the platform facilitated data entry. “You can enter (data) anywhere. You don’t have to only use the computer. You can do it on your laptop or in your office, anywhere.” According to many stakeholders, the perceived strength of the project was that the data collected for the registry was often collected during routine care, and they could easily integrate it into their workflow. One individual stated, “The other thing is that the information we collect is basic, it is data that are collected in the evaluation and tests that are done with the patient.” Another said, “here in this institution, for the registry, we can work with the medical personnel, the technical personnel in nursing, who are the ones helping us with the collection of information. It is simple to integrate the information and the items that you require.” He continued, “I do believe there is a way to focus all the health personnel to keep a record. At this time, we do have more administrative and billing (data collection) processes than clinical processes. You would have to start working with the clinical record …. So, they can include data collection for the registry in what they are doing every day.”

Local Support and Motivation

Several stakeholders highlighted that local support for the registry was a key facilitator of their efforts. One said, “we have a lot of support for the program, which is demonstrated by funds we have been provided from hospital leadership to support travel to conferences and meetings.

Then, we have some passionate physicians and residents who are dedicated to getting the registry up and running.”

Still, others stated that motivation and engagement were critical to the success of the registry.

One participant pointed out, “you need to let people know why it is important, and then they will be motivated to do it.” Another concluded, “I think that the first step is for people to be in love with idea, you know, and after that, the people will make the compromise to make it happen.” Several respondents discussed the importance of engaging support by highlighting the impact of the project. One participant stated, “So, if I could understand why registries are so important, then I would like to be involved. Rather than it just being a part of my work, I would be involved.” Another said, “I think that if you explain … the magnitude of the project, that there is no data (on neurotrauma) in Colombia, in Latin America, then they will be like, ‘Ok, let’s do it.’ That is important, to explain the project. For example, in the ICU, when I presented the project to them, all the, ‘Oh, I want to help … how can I get involved?’”

Future Use

The majority of stakeholders interviewed emphasized that their primary aims in working with the project were to improve the quality of care offered to their patients and improve patient outcomes. One said, “I believe that the registry may help, may serve to better treat the patient with neurotrauma. You can look for ways to generate better management than what was used initially to treat those patients.” Another stated, “the information will ultimately allow institutions to evaluate themselves and to draw conclusions about how we can optimize care. So, I think the registry will do just that.” One clinician said, “by knowing the information, we can start taking action, more precise measures can be taken based on the real needs we have; then we would have a record that will allow us to make a detailed analysis about how to improve our performance and what to focus on promptly.”

Researchers and clinicians at participating institutions saw several other potential uses for registry data. One researcher said, “first, I would like to do an epidemiological description of neurotrauma impact and burden. Second, we could do it in our country, and then in Latin America, our whole region. That doesn’t exist yet. We don’t know the real burden of neurotrauma in our country.” Two respondents emphasized the need to build “a culture of research” in Latin America. Three others hoped to use registry data to drive health policy. One said, “first we have to know the real burden of epidemiological data, and then work with that on public policies.”