Study design

This is an explanatory sequential mixed-methods study protocol to be conducted in Thailand. The purpose of an explanatory sequential design is to expand and elaborate on the study findings from the quantitative strand through qualitative methods in the second strand to better understand the social determinants of health and postnatal wellbeing among women with type 2 diabetes (Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Explanatory sequential mixed methods flow chart diagram.

https://doi.org/10.1371/journal.pone.0354735.t001

Research setting

This study will be conducted across two public hospitals in Thailand—one urban and one suburban. The urban site is a tertiary care center in Bangkok, serving a high-density population with diverse socioeconomic backgrounds, ranging from low to high income. In contrast, the suburban hospital in Samut Prakan serves a more stable, residential, and primarily middle-income community. This dual-site approach utilizes maximum variation sampling to capture a diverse range of experiences, drawing on participants from both urban and rural settings to reflect their distinct contexts. By selecting sites with contrasting geographical and socioeconomic characteristics, the study aims to account for the varying SDoH that influence maternal well-being during the postpartum transition.

This study is divided into two phases: 1) translation of the instrument and 2) data collection for the mixed-method study.

Eligibility criteria

Quantitative phase

As this study is descriptive rather than comparative in nature, statistical significance does not drive the determination of target sample size. This study will recruit 50 participants (Maghalian et al., 2024) through both online methods and face-to-face outreach at postnatal units in two public hospitals.

Qualitative phase

A total of 12 women will be contacted for qualitative interviews: six participants (n=6) with the lowest mean total scores (indicating lower well-being) and six participants (n=6) with the highest mean total scores (indicating higher well-being). Semi-structured phone interviews will be conducted to gain a deeper understanding of postnatal well-being during the transition to motherhood. The sample size of qualitative arm was guided by the previous study among Thai pregnant women as data saturated at 12 participants [8]. If data saturation is not reached by the 12th interview, we will continue interviewing participants from both the lower and higher well-being groups accordingly until no new themes emerge.

The qualitative component utilized three types of purposeful sampling: nested. Nested sampling, a feature of explanatory sequential mixed methods designs, involves re-sampling the same individuals who participated in the quantitative phase for the qualitative phase. As a result, the same inclusion and exclusion criteria from the quantitative phase were applied to the qualitative phase [16]. The use of a nested approach is intentional, as this inter-relationship is a recognized strength of the explanatory sequential design; it allows for a deeper understanding of the extreme cases identified in the first phase.

This study’s qualitative research used purposeful sampling, supported by nested sampling, to select information-rich cases. Purposeful sampling involves intentionally choosing specific units or cases based on a clear objective, rather than random selection, and it can be used to achieve representativeness or comparability [16]. This method allowed for a broader range of cases to be selected and facilitated comparisons across different types. Results from the quantitative phase were examined to inform the qualitative sampling strategy, with the potential to apply stratified and theoretical sampling techniques. Stratified sampling helped identify subgroups and enabled comparisons between them, which aligned with the qualitative research question. It was particularly valuable as it captured variations in the phenomenon of interest. To apply this, participants from the quantitative phase were stratified based on postnatal well-being scores during their transition. This approach aims to explore the challenges these women experience that contribute to their reduced well-being during the postnatal period as they adjust to motherhood.

Quantitative data collection

Three questionnaires are as follows:

First, we will collect demographic data using a questionnaire, which will include the mother’s age, marital status, duration of diabetes, parity, date of delivery, gestational age at delivery, and current medications.

Second, we will assess the social determinants of health across five domains [3]. The questionnaire for this measure is based on previous research and evidence [6]. This 20-item tool consists of two open-ended questions and 18 frequency-based items designed to quantify women’s experiences with social determinants of health (SDoH). There is no formal scoring system for this tool; the responses will be rated as frequency of each item. Validation has been completed; the I-CVI and S-CVI were both 1.0, as determined by a panel of three experts.

1. Domain 1: Economic Stability (4 items)– Family income and employment status.
2. Domain 2: Education Access and Quality (1 item)– Educational attainment.
3. Domain 3: Health and Healthcare (4 items)– Health insurance, tobacco use, alcohol use, and comorbidities.
4. Domain 4: Neighborhood and Built Environment (7 items)– Transportation and access to healthy foods
5. Domain 5: Social and Community Context (4 items)– Unsafe neighborhoods and community interaction

Third, we will measure postnatal well-being using the Postnatal Well-being in Transition questionnaire [18]. This 30-item questionnaire, which uses for each item a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree).

Data analysis

For quantitative data, information from questionnaires will be entered and analyzed using IBM SPSS version 25.0. The analyses include calculating means, standard deviations, and ranges for continuous variables and frequencies and percentages for categorical variables. Missing values will be replaced using item mean imputation; any missing value for a participant will be substituted with the mean value calculated from all available responses for that variable [19].

Qualitative data collection

We have developed the interview guide which aims to deepen the understanding of women’s experiences during their transition to motherhood. The interviewer is a female nursing faculty member with 12 years of clinical nursing practice and expertise in qualitative data collection. At the start of each interview, she explicitly states that she is acting in a research capacity, rather than a clinical one, to encourage honest disclosure. Interviews are conducted over the phone at times most convenient for the participants. A semi-structured interview guide is used to allow participants to lead the narrative, ensuring their lived experiences take precedence over the researcher’s clinical assumptions. The interviewer records field notes regarding the conversational tone, date, and time of each interview.

Data analysis

Since the sampling is sequential, using quantitative scores to identify participants, the analysis of the qualitative interviews will be conducted independently. Coders will be blinded to the specific individual quantitative scores of the participants during the coding phase.

Interview transcripts will be analyzed using Atlas.ti version 9 for qualitative data analysis, which will facilitate data organization and joint analysis of the text. We will identify recurring patterns in participants’ responses based on their relevance to the predetermined domains of inquiry. Descriptive coding will be employed to assign a single code to each segment of data (2–3 lines) that conveys a complete, standalone meaning. The analysis will involve a continuous process of moving back and forth between the entire dataset and individual transcript coding. The coders will work systematically through the dataset, giving equal attention to each transcript and identifying interesting aspects that could form the basis of recurring patterns across the dataset. They will then describe a clear definition and name for each theme by identifying its essence and determining the aspects that capture information. A table will be developed to present the themes, subthemes, and quotations.

Data management

Quantitative data will be stored in the REDCap database. Participants will be tracked using ID numbers. The file linking participant IDs to their contact information will be stored in a separate secure location. A research assistant will transcribe the audiotape verbatim in the Thai language. All information, such as pauses and laughter, will be noted in each transcribed interview. All identifiable information, such as participants’ names and hospital settings, will be removed. Each interview file and questionnaire will be labeled with a three-digit number in order (e.g., ID001). Participants will be tracked using case identification numbers. Interview data (audio files) and transcribed interviews will be saved on a secure server. Assurances of protection of individual participant privacy and confidentiality will be provided. Data results will be reported only in aggregate form.

New data will be obtained through surveys and interviews conducted with Thai postnatal women with T2DM. This data will include demographic details, survey responses, interview transcripts, and any other relevant information related to the study’s objectives. To ensure quality control and reproducibility, the data will be systematically organized, meticulously documented, and clearly described. Standardized formats and protocols will be followed throughout. Data will be securely stored in password-protected folders on a Mac computer, with regular backups conducted to reduce the risk of data loss. The principal investigator will be responsible for managing the data throughout the research process. All data will be retained for 3 years, after which it will be deleted from password-protected folders on a Mac computer and its backup files.

The findings of this study will be published in an international peer-reviewed journal. To protect participant confidentiality and privacy, the de-identified datasets generated during the study are not publicly archived but are available from the corresponding author upon reasonable request.

Mixed-methods approach: Triangulation of qualitative and quantitative data

Triangulation is desirable in mixed-methods research because it serves to validate and confirm the phenomena being studied [20]. To conduct triangulation, qualitative and quantitative analyses will be conducted separately. The results will then be compared and contrasted to identify data convergence and divergence. To ensure unbiased triangulation, a side-by-side comparison will be conducted during the final integration phase. We will identify exemplary quotations where qualitative themes either converge with or contradict quantitative trends.

Statistics-by-themes joint displays organize quantitative results (e.g., high and low well-being) across qualitative themes in a matrix format, resembling a crosstabulation. These displays feature qualitative themes in the first column, paired with corresponding statistical data (means and standard deviations). This layout enables the research team to triangulate data, identifying areas of convergence or divergence between the two sources.

To transition from a descriptive joint display to an analytical integration, the research team will follow the three-step interpretive framework [21] to derive integrated meta-inferences.

1. Step 1: Systematic Assessment of Fit. Once quantitative metrics (means, percentages) and qualitative data (themes, quotes) are populated side-by-side in the joint display, the team will evaluate their relationship using three analytical criteria: Confirmation (Convergence): Qualitative findings directly validate or mirror statistical trends. Expansion: Qualitative data explains or nuances a statistical trend (e.g., high quantitative well-being scores contrasted by qualitative insights showing mothers suppress pain to prioritize childcare). Discordance (Divergence): Quantitative metrics and qualitative findings present conflicting realities.
2. Step 2: Procedural Resolution of Discordance. In instances of discordance, the team will initiate an analytical reconciliation process (Fetters et al., 2013) by: (1) checking raw quantitative data for outliers or measurement biases; (2) re-examining the qualitative matrix for sub-demographic patterns; and (3) conducting peer debriefings to explore whether the instruments measured different dimensions of the same phenomenon.
3. Step 3: Synthesis into Meta-Inferences. Finally, the team will synthesize the reconciled findings into overarching meta-inferences. Documented in the final column of the joint display, these meta-inferences ensure the discussion yields a fully integrated, multi-dimensional understanding of postnatal well-being rather than parallel independent findings.

Ethics and dissemination

This protocol has been reviewed and approved by the Faculty of Medicine Ramathibodi Hospital, Mahidol University, Thailand (MURA.2025/648).

Quantitative phase

Qualitative phase

Limitation

A key limitation of this study is the lack of formal cross-cultural psychometric validation of the adapted Social Determinants of Health (SDoH) instrument within the Thai population. While the tool demonstrated strong content relevance via a Content Validity Index (CVI) from an expert review panel, content validity alone does not guarantee cultural equivalence. Because this instrument was adapted from Western framework domains, it may not fully capture distinct Thai sociocultural nuances. Future research should prioritize the formal translation, cultural adaptation, and psychometric validation of standardized SDoH metrics tailored specifically to Southeast Asian populations.

Research significance, impact, and outcomes

The protocol outlines a study designed to improve our understanding of the social determinants of health and well-being for Thai women in the postnatal period. A mixed-methods approach is crucial for this, as it allows us to fully capture the complexity of women’s experiences by combining qualitative and quantitative results.

The findings have significant implications for the future of nursing research. They highlight the need for future studies on pregnant women with T2DM to use mixed methods designs, which can illuminate the evolution of the postnatal journey. By gaining a better understanding of a woman’s unique healthcare needs, we can create better tools for assessing women’s postnatal health based on their specific social determinants. This knowledge of sociocultural factors will eventually help us design patient-centered, culturally tailored care to improve women’s health outcomes. Furthermore, our findings could be used as a model for developing new healthcare guidelines in Asian regions and will eventually inform policy decisions, leading to a more integrative care model that benefits both mothers and infants in Thailand.