Continuing professional development challenges in a rural setting: A mixed-methods study

This study was conducted at “Angel Albino Corzo” (HAAC) community hospital in Chiapas, one of Mexico’s most economically disadvantaged regions, where inadequate health infrastructure has translated into lower health standards [10] and socioeconomic inequalities are replicated in health professions education [11]. This facility provides care for ~ 28,883 inhabitants (in collaboration with an adjacent birth center), with 72 employees. Nursing students from a local school complete their clinical learning requirements in this hospital. However, their curriculum and classes are supervised by their university and not open to the rest of the hospital staff. At the time of this study, there was no formal CPD curriculum in the hospital, with irregular educational sessions being conducted.

This exploratory sequential mixed-methods study [12] was approved by the Harvard Medical School (IRB19-1138) and Chiapas state (5003/8784) institutional review boards and conducted from August 2019 to January 2020. Taking advantage of the pragmatism research paradigm that a mixed-methods methodology provides, we acknowledge health professionals may perceive CPD differently. Therefore, we gathered information from multiple qualitative and quantitative sources to optimize our understanding of their experience [12].

Qualitative data was collected via semi-structured interviews, focus group discussions (FGDs), and observation. Participants were recruited using purposeful sampling [13] of all hospital employees, including clinical staff, non-clinical staff, and hospital leadership (ensuring the inclusion of multiple perspectives). Recruitment and data collection were done by MC‑Z, a bilingual healthcare professional and researcher unknown to research participants, who did not undertake any clinical activities locally. Participants were approached, flyers were placed in multiple hospital areas, and the hospital director sent an invitation via email and text message to the entire hospital staff. Verbal consent was obtained for all study participants. An interview guide developed by four researchers (MC‑Z, HG, VD, MIL-M) was piloted with two independent native Spanish-speaking healthcare providers for clarity and coherence (see Electronic Supplementary Material [ESM], part A, “Qualitative instrument”). Semi-structured interviews and focus group discussions were conducted during the same data collection period to ensure research captured both individual and collective information and to accommodate participants’ preferences as they selected which modality they participated in. Interviews and focus groups took place in a private area (classroom) in Spanish by MC‑Z, who audio-recorded and later transcribed and translated them into English in a single-step process.

The primary author (MC-Z) also conducted 40 h of field observations of formal and informal educational activities to understand the context and nuances of information provided by participants. Observation time was distributed across different weekdays and times to capture variation among shifts and days, and field notes with descriptions and reflections were recorded in a journal. These observations were merged with the qualitative and quantitative dataset during the analysis when field notes were iteratively reviewed.

De-identified qualitative data were analyzed using conventional content analysis methods [14]. The dataset was read meticulously. Two authors (MC‑Z, RE-P) open-coded a subset of transcripts to identify key concepts that were developed into a draft codebook that was then revised. The resultant final codebook was subsequently used to code the complete dataset using Dedoose qualitative data management software (SocioCultural Research Consultants, CA, USA). MC‑Z, RE‑P, HG, and VD analyzed the data using an inductive, iterative process in which transcripts and codes were continuously reviewed to develop a set of draft themes; this process continued until saturation was achieved (no new themes were identified). All differences were resolved by the group (MC‑Z, RE‑P, HG and VD) through discussion until consensus was reached. This process yielded seventeen themes that were further refined into a final set of four main descriptive categories.

Throughout the qualitative study phase, reflexivity was maintained by the lead author (MC-Z) through an approach intentionally designed to challenge her own assumptions. While conducting interviews, discussion groups, and translating the data, conscious efforts were made to distinguish her thoughts and hypotheses from the data collected by reviewing field notes and through multiple conversations systematically conducted throughout the data collection stage with VD and MIL‑M, in order to identify and acknowledge assumptions and positionality. The remainder of the research team (a medical anthropologist, health professionals, educators, and researchers) had no relationship with participants or study setting and intentionally left assumptions out while analyzing the dataset. Discussions included acknowledgment and allowance of views, values, and beliefs that may have influenced the analysis.

The quantitative portion of the study was designed to allow participants to share their experiences anonymously and to complement qualitative findings. The literature review did not identify any suitable preexisting quantitative instruments for our aims and context. Therefore, a questionnaire was developed following a structured approach described by Artino et al. [15]. Three researchers (MC‑Z, VD, RE-P) with context knowledge and familiarity with the study aims reviewed the qualitative dataset and discussed the most common interview themes that emerged. They subsequently designed potential questionnaire items and multiple-choice questions specific to each one of the constructs to explore. The options for each question were developed based on information shared in the interviews and focus groups (e.g., participants mentioned that the classroom was inadequate for learning, hence that was included as an option when asking about factors limiting participation in CPD activities).

The questionnaire was piloted with four Spanish-speaking health professionals using cognitive interviews to assess the clarity of questions and answers. During the cognitive interviews, we examined thought processes of responses by asking participants to rephrase and explain interpretation of questions. The final version of the questionnaire (see ESM, part B, “Quantitative instrument”) explored four essential constructs within six main questions (48 individual items). The constructs were: (1) perceived importance, (2) motivation (both rated on a 5-point Likert scale), (3) relevant barriers (nine options identified from interviews), and (4) preferred educational strategies (five educational strategies rated on a 1–10 scale). There were 32 educational topics that participants identified as important for future CPD activities included and rated using a 5-point Likert scale (these topics were identified from the qualitative dataset as themes participants mentioned they needed to learn). Content validity was established through an iterative development process plus cognitive interviews. Distribution was done via an anonymous link distributed by mobile message from the hospital director. Demographic data are presented as mean (standard deviation), median (interquartile range), or number (%), and were analyzed using Stata statistical software (Stata Corp, TX, USA).

The quantitative and qualitative datasets were merged to inform the analyses to better understand participants’ experience with CPD activities. Triangulation was conducted using qualitative (transcripts and field notes) and quantitative data to consider different aspects of reality and examine points of convergence or divergence between datasets [13]. Qualitative and quantitative results are reported merged later in this article.

Twenty-two semi-structured interviews, three FGDs, and 40 observational hours were conducted. Participant demographic characteristics are described in Tab. 1. From these qualitative data sources, we identified 35 initial codes, and 17 themes that were refined into four main categories (Fig. 1): 1) recognized need for and interest in learning, 2) discrepancy between existing didactic activities and desire for practice, 3) institutional barriers, and 4) adverse interprofessional environment.

Out of 72 eligible hospital employees, 20 completed the questionnaire with only thirteen respondents answering all questions (18% response rate). All respondents were individuals whose primary workplace was the hospital (no responses from the birth center). All respondents considered CPD as important (100%, n = 20); 65% (n = 13) described it as very important and 35% (n = 7) as extremely important. However, only 20% (n = 4) felt very frequently motivated to participate in existing CPD activities, 35% (n = 7) frequently motivated, 35% (n = 7) occasionally motivated, and 10% (n = 2) rarely motivated. When questioned about the most important factors that limit participation in CPD, they identified lack of motivation (65%, n = 13), inconvenient scheduling of sessions (75%, n = 15), an inadequate classroom (65%, n = 13), high workload (60%, n = 12), unappealing speakers (60%, n = 12), and boring sessions (55%, n = 11).

Participants rated their preferred learning strategies for future CPD activities on a scale of 1–10. Practical workshops (mean 9.06 ± 1.39), simulations (mean 8.73 ± 1.48), and case analysis (mean 8.44 ± 1.85) received the highest ratings. The questionnaire also explored the topics that health professionals in the HAAC felt were needed in their interprofessional CPD curriculum (results in ESM part C and ESM Fig. S1).

Overall, the four main qualitative categories were substantiated by participants’ questionnaire responses, as demonstrated in Tab. 2, where we display mixed-methods integration of the results.

This study was conducted at a single healthcare facility, and our results may not be generalizable to other facilities or regions. However, the major findings of this study may be relevant for similar settings and could be used as an initial framework for future research.

The qualitative data collection was conducted during a pre-defined 5‑week period entirely by one author, potentially introducing bias. This is an intrinsic situation in qualitative research and intentional efforts were carried out to maintain reflexivity. The quantitative portion of our study faced several limitations, including small sample size and a response rate lower than expected. Likewise, evaluation of the psychometric properties of this instrument may be required for future use. Birth center staff were not included in the questionnaire due to internal communication issues (illustrating some of the administrative and communication challenges that CPD faces in this setting). We do not have demographic information for non-responders, and it was unclear why seven individuals left questions unanswered. Therefore, our quantitative results might not accurately represent the study population perceptions. However, responders were proportionally representative of the demographic distribution of the hospital staff, and we found convergence between the identified themes and the questionnaire data, suggesting that survey respondents had similar opinions to those collected in the qualitative data.