This descriptive qualitative study design was employed. The study adhered to the Standards for Reporting Qualitative Research (SRQR) guidelines (50). Data was collected from four primary health centres (PHCs) in Kerala, namely Mynagapally, Punnapra, Kalloorkad, and Engandiyur. In the Indian health system, PHCs serve as the first point of contact for most patients, particularly in rural and semi-urban areas. They provide a comprehensive range of essential health services, including maternal and child health, immunizations, management of communicable and non-communicable diseases, health education, and referral services. In Kerala, PHCs play a pivotal role in chronic disease management through dedicated non-communicable disease (NCD) clinics, operated under the state’s “Amrutham Arogyam” program. This initiative focuses on early detection, screening, treatment, and the promotion of healthy lifestyles to address the rising burden of conditions such as diabetes and hypertension.

A purposive sampling strategy was employed to achieve maximum variation, ensuring representation across age, gender, socioeconomic background, and type of multimorbidity. This approach facilitated the inclusion of diverse perspectives and experiences related to shared decision-making in primary care.

Inclusion criteria were: (1) adults aged 60 years and above; (2) presence of multimorbidity, defined as the coexistence of two or more chronic non-communicable conditions (e.g., diabetes, cardiovascular disease, cancer, stroke, chronic obstructive pulmonary disease, or chronic kidney disease); and (3) ability to converse in either Malayalam or English. Exclusion criteria included: (1) diagnosis of dementia; and (2) communication impairments hindering meaningful participation.

Potential participants were identified by healthcare providers at the PHCs and approached by the research team after initial eligibility screening. Four interviews were conducted from each district, which sums up to a total of 16 interviews. Data collection proceeded iteratively until the lead researcher, in collaboration with academic supervisors, determined that data saturation was achieved (51). This decision was based on a systematic analysis of the interview data, regularly examining emerging themes, codes, and patterns. Saturation was reached when no new information or insights emerged, indicating further interviews would yield redundant data. The evaluation included monitoring recurring themes, verifying data redundancy, and confirming rich, comprehensive narratives from diverse participant perspectives.

Individual in-depth semi-structured interviews were conducted with participants between June and December 2024. The interview guide, developed with reference to the SEM, included questions designed to elicit participants’ experiences and perspectives on SDM at the individual (e.g., health beliefs, literacy), interpersonal (e.g., family involvement), organizational (e.g., provider practices, continuity), and sociocultural/policy (e.g., cultural norms, health system factors) levels. The guide was pilot tested with five interviews and refined accordingly, the interviews were designed to last between 30 to 60 min, depending on the flow of conversation and the level of detail provided by the participants.

The interview questions followed a hierarchical structure, beginning with open-ended questions to allow participants to describe their general experiences. For example, an initial question asked: “What are the difficulties caused by multimorbidity in your life?” Subsequent follow-up questions were tailored to explore specific issues raised by participants, such as, “What obstacles do you face in managing the problems caused by your health conditions?” or “What kind of support would be most helpful from your family, community health centers, or other organizations in primary care settings?” To ensure thorough understanding and contextual validation, prompts were used when participants provided insufficiently detailed responses, such as, “How would the support you mentioned help you specifically?” This approach helped to validate the participants’ priorities and needs in their own words.

The interviews were held in person at the primary health centers (PHCs) or at the home of the participant, depending on the participant’s preference and convenience. Before each interview, the researcher thoroughly explained the purpose of the study and the interview process to the participants, ensuring they understood the content and objectives. The duration of each interview ranged from 45 to 60 min. During the interviews, the researcher remained respectful, objective, and flexible, adjusting the order of questions as needed and asking follow-up questions to probe deeper into specific responses. This approach ensured a more natural and fluid conversation, allowing participants to express their thoughts and experiences fully. The interviewer ensured privacy and that no other persons were present during interviews. Field notes were taken during and immediately after each interview to capture non-verbal cues and contextual information. The interviews were audio-recorded and transcribed verbatim. The Malayalam transcription then translated to English. Transcripts were not returned to participants, however, the researcher documented key points during the interviews and discussed these summaries with the participants to verify that they accurately reflected their experiences and perspectives. Participants were given the opportunity to provide feedback, clarify any misunderstandings, and offer additional insights. This process helped to enhance the credibility and authenticity of the study’s findings.

The research team recognized the importance of reflexivity throughout the study. Data collection was conducted by the lead researcher, a male PhD scholar with a background in social work in public health and expertise in older adult care and management of non-communicable diseases. He has prior experience as a research associate on a qualitative study of health system preparedness for road traffic injuries. He has completed a four-credit course on social science research and qualitative methods. As a native of Kerala, the researcher’s familiarity with the local language, culture, and healthcare context facilitated rapport-building and contextual understanding during interviews with older adults. There was no prior relationship between the interviewer and participants.

To enhance the credibility and trustworthiness of the findings, the lead researcher maintained a reflexive journal to document personal assumptions, preconceptions, and observations throughout the research process. Regular supervisory meetings were held with three academic supervisors, who provided critical feedback and guidance on data collection, analysis, and interpretation. These discussions encouraged the researcher to critically examine how their background and positionality might influence interactions with participants and the interpretation of findings. This reflexive approach aimed to ensure that the analysis authentically represented participants’ perspectives, rather than being shaped by researcher expectations or prior experiences.

Thematic analysis was employed to analyze the interview data, following the six-step framework outlined by Braun and Clarke (52). These steps included: (1) familiarization with the data, (2) generating initial codes, (3) searching for themes, (4) reviewing themes, (5) defining and naming themes, and (6) producing the final report. Within 24 h of each interview, audio recordings were transcribed verbatim by the researcher. The interviews conducted in Malayalam were initially transcribed verbatim in the original language and subsequently translated into English by the lead researcher. To ensure the accuracy and quality of the translation, a second independent bilingual reviewer cross-checked the English transcripts against the original Malayalam audio recordings, resolving any discrepancies through consensus discussion. This process ensured that the nuances and meanings of participants’ expressions were faithfully preserved in the English translations.

The transcripts were then imported into NVivo 12 (53) software to facilitate systematic data management, coding, and retrieval. SEM was used as a sensitizing framework to organize and interpret themes across individual, interpersonal, organizational, and sociocultural/policy levels.

To ensure analytic rigor and consistency, the lead researcher conducted the initial coding of all transcripts. A subset of transcripts was independently reviewed by the three academic supervisors, who provided critical feedback and suggestions. Discrepancies or differences in coding were addressed through a process of contextual verification by revisiting the original transcripts, followed by collaborative discussions between the lead researcher and supervisors. Where necessary, consensus was reached through further deliberation. The codes were then refined and systematically organized into broader themes and sub-themes. The research team-comprising the lead researcher and supervisors jointly reviewed the relevance and coherence of the themes in relation to the study objectives and finalized their naming. This collaborative and reflexive process ensured a transparent and robust interpretation of the data, allowing for nuanced insights into the shared decision-making experiences of older adults with multimorbidity in primary care settings.

This study was approved by the Institutional Ethics Committee (IEC) of Kasturba Medical College (KMC) and Kasturba Hospital (KH), Karnataka, India (Ethical Review No. IEC1: 05/2023, dated 27 July 2023). Administrative permission to conduct the study and collect data from primary health centers was obtained from the Directorate of Health Services, Kerala (Order No. MC4-48885/2022/DHS dated 09/2022). Written and verbal informed consent was obtained from all participants, ensuring that they understood their rights and voluntarily agreed to participate. All data collected was kept confidential and anonymous, and no identifying information was published in this paper.

Several participants described feeling uncertain or reluctant to disclose to their allopathic doctor that they were also receiving Ayurvedic or other alternative treatments. This hesitancy often stemmed from concerns about being judged, misunderstood, or admonished by the allopathic provider.

In contrast, participants felt more comfortable telling their Ayurveda practitioner that they were also seeing an allopathic doctor, reflecting a perception that Ayurveda is more accepting of integrative approaches.

This selective disclosure highlights a subtle but important aspect of patient-provider relationships in Kerala: while patients value the expertise of allopathic doctors, they may withhold information about alternative treatments to avoid conflict or disapproval. Such practices can have implications for safety, drug interactions, and the effectiveness of care. At the same time, the coexistence of multiple medical systems in Kerala fosters a unique environment where patients navigate between traditions, sometimes blending approaches to suit their needs.

Participants who felt excluded from decision-making reported a range of negative outcomes, including dissatisfaction with care, confusion about their treatment plans, and poor adherence to prescribed regimens. When patients did not understand the rationale for their treatments or felt their concerns were not addressed, they were less likely to follow through with recommendations.

Others described feeling anxious, isolated, or disempowered, which affected their willingness to seek care or follow through with recommendations. The lack of explanation or involvement in decisions often led to uncertainty and disengagement, potentially undermining the effectiveness of care.

The experiences of older adults with multimorbidity in Kerala’s primary care settings reveal a complex interplay of individual knowledge, family involvement, provider practices, and cultural norms shaping shared decision-making. While some participants described positive, inclusive encounters, many faced significant barriers at multiple SEM levels. These findings underscore the need for multi-level interventions to promote person-centered care and empower older adults to participate actively in decisions about their health.

At the individual level, limited health literacy and low self-efficacy emerged as critical barriers to active engagement in SDM. Many participants described difficulties in understanding medical terminology, treatment options, and the implications of various choices. This uncertainty often resulted in a passive approach, with patients deferring decisions to healthcare providers rather than actively participating in their own care. These findings align with previous research in LMICs, which has consistently highlighted low health literacy as a key obstacle to SDM (28, 54–56). Additionally, some participants expressed apprehension about asking questions, fearing they might be seen as challenging the provider’s authority or wasting their time. This lack of confidence and perceived power imbalance further limited their willingness to engage in meaningful dialogue. To address these barriers, targeted health education interventions and the use of patient-friendly decision aids are essential. Providing information in clear, accessible language and encouraging patients to voice their concerns can empower older adults to take a more active role in their healthcare decisions (57). Decision aids and pictorial tools have been shown to effectively reduce decisional conflict and support shared decision-making in LMICs, with evidence from Malaysia highlighting the importance of addressing cultural paternalism and role boundaries among healthcare providers for successful implementation. In India, a self-administered, adaptive decision aid significantly lowered decisional conflict among early breast cancer patients, demonstrating feasibility and cultural adaptability in improving patient engagement and preference-concordant surgical decisions (58, 59).

Family involvement was a defining feature of SDM in Kerala. Family members, especially adult children, frequently accompanied older adults to appointments, mediated interactions with providers, and sometimes made decisions on the patient’s behalf. While this support was often valued, particularly by those with physical or cognitive limitations, it could also overshadow the patient’s own preferences. Notably, in contrast to findings from other Indian states and some LMICs, this dynamic in Kerala was not strongly gendered, reflecting the state’s relatively high levels of gender equity and autonomy among older women.

Our findings that family members can both enable, and hinder decision-making are consistent with broader research on family-centered care (60, 61). On one hand, family members can help patients articulate concerns, interpret complex information, and provide emotional support. On the other, their involvement may inadvertently suppress the patient’s voice, especially if family members dominate discussions or make decisions without fully consulting the older adult. Effective SDM in such contexts requires a nuanced approach that respects patient autonomy while constructively integrating family input. Training providers to facilitate inclusive conversations and explicitly inviting the patient’s perspective can help balance these influences.

At the organizational level, provider communication and continuity of care emerged as critical enablers of SDM. Participants who consistently saw the same provider and experienced empathetic, clear communication described feeling more included in decisions about their care. These positive encounters fostered trust, comfort, and a sense of partnership, making it easier for patients to express their concerns and preferences. However, high patient loads, limited consultation times, and system pressures often resulted in rushed, transactional encounters that limited opportunities for meaningful engagement. Many participants described feeling like “just another number” in a busy health center, with little time or space to discuss their preferences or ask questions. These challenges are widely reported in LMIC settings, where resource constraints and workforce shortages are common (62, 63). Such organizational realities reinforce passive patient roles and undermine the principles of SDM. Interventions at this level should focus on providing training in SDM techniques, restructuring clinic workflows to allow more time for patient engagement, and promoting continuity of care. Studies have shown that even brief SDM interventions, when consistently applied, can improve patient satisfaction and health outcomes (64, 65).

Kerala’s community norms strongly emphasize respect for medical authority, which can discourage patients from questioning providers or expressing preferences. This cultural expectation of difference to expertise was evident in many participants’ accounts, regardless of gender or educational background. While such respect can foster trust, it may also limit open dialogue and reduce opportunities for SDM (66, 67).

A unique finding in this context was the prevalence of pluralistic health-seeking, with many patients consulting both allopathic and Ayurvedic practitioners. Several participants hesitated to disclose their use of ayurveda to allopathic doctors, fearing disapproval or negative reactions. This selective disclosure may have implications for treatment safety, drug interactions, and the effectiveness of SDM, highlighting the need for open, nonjudgmental communication and provider awareness of pluralistic practices (68, 69). At the policy level, Kerala’s health reforms have strengthened primary care and promoted patient-centered approaches, yet the real-world implementation of SDM remains inconsistent. Ongoing system-level support and investment are needed to institutionalize SDM practices and create an environment conducive to participatory care (70).

Exclusion from SDM was associated with dissatisfaction, confusion about treatment plans, and poor adherence to prescribed regimens. When patients did not understand the rationale for their treatments or felt their concerns were not addressed, they were less likely to follow through with recommendations. This finding is supported by a substantial body of evidence that SDM improves adherence, self-management, and health outcomes in chronic disease management (28). For older adults with multimorbidity, SDM is not only an ethical imperative but also a practical strategy to improve outcomes, as patients’ willingness to adhere to treatment is closely linked to their understanding and involvement in care decisions.

Individual level and Interpersonal level: Implement educational programs and initiatives to promote and encourage the adoption of shared decision-making among both patients and primary caregivers.

Organizational level: Training initiatives for primary healthcare providers should emphasize communication and skills to encourage SDM tailored to the specific context of Kerala’s healthcare system. Additionally, clinic workflows should be redesigned to extend consultation durations, helping to alleviate the effects of high patient volumes and brief appointments. Enhancing continuity of care can also build patient trust and improve adherence to treatment plans.

Sociocultural and policy levels: State health policies should formally integrate SDM practices into primary care within the framework of Kerala’s ongoing healthcare reforms. Encouraging open communication about diverse health-seeking behaviors, including both allopathic and traditional medicine, can help address patient concerns and enhance transparency in treatment. Additionally, policies should ensure creation of a space for patients to openly share their use of alternative medicine.