Determinants and Experiences of Care‐Seeking for Childhood Pneumonia in a Rural Indian Setting: A Mixed‐Methods Study

Pneumonia is the leading infectious cause of death among children under five worldwide, claiming over 800,000 lives or approximately two lives every minute [1, 2, 3, 4, 5], and most of these deaths occur in low‐ and lower‐middle‐income countries [2, 6]. In India, pneumonia causes 30 million annual episodes among young children, accounting for millions of hospital admissions and 15% of under‐five mortality [3].

Reducing mortality from childhood pneumonia hinges on streamlining the care pathways [7, 8]. Prompt recognition of danger signs, appropriate management and effective antibiotic treatment are crucial for saving lives [9]. However, in resource‐limited settings, accessing these vital interventions can be a challenge for caregivers. Financial constraints, limited knowledge about pneumonia and long distances to healthcare facilities often deter families from seeking timely medical care [10, 11, 12, 13, 14, 15]. A 2015 report by the Lancet Global Health revealed that only 54% of children exhibiting symptoms of pneumonia in low‐ and middle‐income countries (LMICs) are taken outside the home for care [16]. Current coverage of treatment with effective antibiotics for children under five with pneumonia is 31% globally and 25% in India. Delayed medical attention from an appropriate provider is associated with substantial morbidity and mortality among childhood pneumonia cases. Unfortunately, studies conducted in LMICs have found that healthcare workers lack expertise in initiating appropriate treatment due to insufficient training, lack motivation due to delayed wages and face a shortage of medical supplies. This also results in improper management [17, 18, 19], affects the caregivers' and community members' trust in health services and can affect future care‐seeking behaviours [20].

India has developed the Childhood Pneumonia Management Guidelines (CPMG) to streamline the treatment of pneumonia in children [4]. The primary aim of these guidelines is to facilitate the effective management of pneumonia symptoms such as cough, rapid breathing and/or breathing difficulties in children under 5 years through community‐ and facility‐based approaches [4]. We have previously explored barriers and facilitators to implementing the new guidelines across various tiers of the health system. Current data on barriers, behaviours and perceptions that influence care‐seeking for childhood pneumonia at the community level is currently scarce [15, 21, 22]. Such information is essential to plan effective strategies to promote healthcare‐seeking for childhood pneumonia. This study, therefore, aims to explore experiences with care‐seeking for under‐five children with suspected pneumonia and identify socio‐demographic and economic characteristics associated with seeking care from appropriate sources.

As part of a broader implementation research (IR) project, details can be found elsewhere [23]. To enhance the coverage of effective pneumonia management, this study employs a mixed‐methods design, integrating quantitative insights on care‐seeking patterns with qualitative perspectives on underlying socio‐cultural and systemic barriers, as well as experiences. The study was conducted 1 May to 30 August 2023, in Palwal district, Haryana, covering a population of 107,440 across 42 villages, which included 10 HWCs, 5 primary health centres, 4 community health centres and 1 district hospital. Other healthcare providers included private care providers, community‐level health workers, chemists or pharmacies, and informal providers [7]. This study employed an Explanatory Sequential Mixed‐Methods Design as part of the formative phase of a broader IR study. The study was conducted in two phases: the quantitative phase included a cross‐sectional survey of households to identify potential under‐five pneumonia cases and assess care‐seeking behaviours. Following this, the qualitative phase was conducted, where in‐depth interviews were carried out with purposively selected caregivers of under‐five pneumonia cases. The qualitative component of this study integrated Qualitative Description (QD) and Participatory Action Research (PAR) approaches [24, 25]. The QD approach provided a clear, factual representation of caregivers' experiences with care‐seeking, capturing their perceptions and challenges without extensive theoretical interpretation [24]. Additionally, elements of PAR were embedded through active community engagement in shaping research directions, refining study tools and informing the subsequent phases of implementation [25]. The iterative involvement of caregivers, healthcare providers and community stakeholders, as a part of the primary IR, in providing feedback on pneumonia management strategies ensured that the research findings were not only descriptive but also action‐oriented, contributing to policy and practice improvements in pneumonia care‐seeking [26].

For quantitative data, we obtained a list of under‐five children from community health workers and conducted a cross‐sectional survey using a complete enumeration technique. Relevant data were collected through house‐to‐house visits, covering 9593 households with under‐five children. At the outset, the trained research assistants (who were university graduates by qualification) used an under‐five pneumonia identification checklist (Supporting Information: Annexure). Children who were reported by their caregivers to have had any signs or symptoms listed in the checklist within the past 28 days were identified as potential pneumonia cases, and this served as the sample for the quantitative study. For these potential under‐five pneumonia cases, data were collected on household socio‐demographic and economic factors along with data on possession of a ‘Below poverty’ card, health scheme and insurance (any government health scheme or insurance, e.g., Ayushman Bharat, Employees' State Insurance scheme, etc. or any private company insurance), details of pneumonia signs and symptoms, time from symptom onset to seeking care, sources of care, (Supporting Information: Annexure) and caregivers' previous knowledge about pneumonia (Supporting Information: Annexure). The Below poverty card provides several benefits to households living below the poverty line. The scheme ensures that such families have free access to essential commodities, healthcare, education, housing and employment opportunities. The information was collected using a structured questionnaire with a few open‐ended questions about treatments or remedies used at home and the symptoms caregivers believed to be indicative of pneumonia and so forth (Supporting Information: Annexure). This quantitative data was collected once per caregiver of under‐five children, with each interview taking approximately 40–45 min on average. All questionnaires were translated into the local language (Hindi), piloted among caregivers from 2 April to 16 April 2023 and refined based on their feedback (Supporting Information: Annexure). The quantitative data collection team was fluent in both Hindi and the local Haryanvi dialect and clarified any terms in Haryanvi during the survey, as needed, to ensure participants fully understood the questions. I II III II IV

The qualitative component of this study was conducted following the Standards for Reporting Qualitative Research (SRQR) guidelines. The study coordinators and investigators, who were medical doctors with specialisations in public health/PhD and had extensive training and experience in qualitative research methods, carried out the qualitative component using in‐depth interview guides (Supporting Information: Annexure). All guides were translated into Hindi language (Supporting Information: Annexure), and these investigators were proficient in the Hindi language and well‐versed in the Haryanvi dialect. These guides focused on perceptions of pneumonia, barriers causing delays in care‐seeking and reasons for choosing specific sources of care. Purposive sampling was employed to select participants who could provide rich insights into care‐seeking behaviours. Caregivers were eligible if they were fluent in the local language, Hindi or Haryanvi dialect, to ensure effective communication. Recognising the influence of family dynamics in decision‐making, particularly in joint or three‐generation households, interviews were conducted with 12 mothers, 5 fathers and 5 grandparents. This approach aimed to capture the diverse perspectives of key family members involved in healthcare decisions, offering a more comprehensive understanding beyond the primary caregiver's (mother's) viewpoint. Before participation, all selected individuals were informed about the qualitative component of the study, its purpose and their voluntary involvement. Informed consent was obtained from all participants before data collection. Data saturation was achieved when no new themes emerged from additional interviews, ensuring a comprehensive exploration of the research questions. V VI

The trustworthiness of the qualitative data was ensured through multiple strategies to enhance the credibility and reliability of the findings. Investigator triangulation was employed, meaning that multiple researchers independently reviewed and analysed the data to reduce individual biases and ensure that interpretations were consistent and well‐founded. This approach helped cross‐verify findings and minimise any single researcher's subjectivity influencing the conclusions. Additionally, member checking was conducted with a subset of participants, where key findings were shared with them to confirm whether the interpretations accurately represented their experiences and perspectives. This process helped ensure that the data remained true to the participants' actual responses rather than being influenced by researchers' assumptions. Reflexivity was also maintained throughout the research process. This involved researchers critically reflecting on their own perspectives and potential influence on data collection and analysis. By being mindful of their preconceptions, the researchers aimed to ensure that findings were derived from the participants' narratives rather than their own expectations. These combined measures strengthened the credibility and authenticity of the study findings.

After surveying 9593 children under five, we identified 231 cases of suspected pneumonia occurring within the 28 days preceding the household visit. Out of these 231 cases, 83% (193 cases) suffered a cough, 53% (123 cases) had difficulty breathing and 54% (125 cases) had fast breathing. Additionally, 69% (159 cases) had a temperature above 37.5°C, 25% (58 cases) suffered from vomiting and 9% (20 cases) were unable to breastfeed. Mothers were the primary informants in 79% of cases, while both parents contributed information in 16% of the households. Two‐thirds of the households possessed Below‐poverty‐level cards, and 35% had some form of health insurance. Around two‐thirds of the pneumonia cases were male children, with the average child age being 22.7 months. Educational levels varied, with 28% of mothers and 11% of fathers having less than primary education. Most mothers (97%) were housewives, while fathers primarily held private jobs (38%) or worked as daily wagers (30%) (Table 1).

Among the 231 under‐five pneumonia cases, 97% of the caregivers sought medical care from government or private healthcare facilities or non‐RMPs. Only 3% (7 cases) did not seek any medical care. The distribution of health insurance and healthcare‐seeking as per the availability of insurance in different socio‐economic tertiles is detailed in Supporting Information: Tablesand. S1 S2

Most of the caregivers sought care from non‐registered practitioners, and one‐fourth utilised private facilities, which included private hospitals/nursing homes. Government healthcare options were less frequently used, and 4 out of 224 caregivers, that is, around 2%, directly sought assistance from chemists to obtain medicines (Table 2).

The duration within which caregivers sought care for under‐five children with suspected pneumonia is detailed in Table 3. Out of the children with suspected pneumonia, children in the poorest families were most likely to receive no treatment (Supporting Information: Graph 4). Across all economic groups, families most often turned to non‐RMPs for pneumonia treatment. This tendency was particularly pronounced among the poorest families. The poor families were more likely to use private healthcare than the poorest and very poor families, and the government facilities were the least preferred facilities for care in all the wealth categories.

In the multivariable model, higher maternal education (diploma level or above) was associated with an increased likelihood of seeking appropriate care for under‐five pneumonia (Table 4). The least poor group were also more likely to seek care from appropriate sources compared to the poorest group. Maternal age was inversely associated with appropriate care‐seeking, but the wide confidence intervals suggest that these results were uncertain (Table 4). An overview of the qualitative findings is provided in Table 5. Caregivers often misinterpreted symptoms such as difficulty in breathing or chest in‐drawings as signs of common cold or seasonal ailments, resulting in delays in seeking timely medical care. Some families, particularly mothers, perceived these symptoms as indicative of a disease but believed they were unlikely to lead to serious illness, which contributed to delays in care‐seeking. More concerningly, certain families, especially those with grandparents as household heads, viewed these signs and symptoms as natural adaptations of the child's body to seasonal changes, such as the transition from autumn to winter, and did not recognise them as indicators of illness. Grandmothers often held strong beliefs that such symptoms, like those experienced by their own children in the past, could be adequately managed with home remedies, negating the need for professional medical attention. Furthermore, findings from a subset of households indicated that poor household environmental conditions, such as exposure to indoor smoke, exacerbated respiratory symptoms, and the team observed that there was limited awareness of strategies to mitigate these risks.

Caregivers demonstrated a strong reliance on home remedies and traditional healers, often delaying formal healthcare interventions until traditional methods had failed. Among many households, initial treatment involved rituals and superstitions, stemming from a belief that illnesses were caused by the ‘evil eye’ or negative influences from others in the community. This practice was widespread. Additionally, significant cynicism towards modern medicine was observed in some families. This scepticism stemmed from a belief that allopathic treatments were inherently associated with side effects, leading to hesitation in fully trusting modern medical care. Government healthcare facilities were often viewed as inefficient due to long waiting times and poor service quality. The perceived unavailability of medicines further deterred families, who believed essential medicines for children would also be lacking. While ASHA workers disseminated information about childhood illnesses and appropriate care options, mistrust in their advice undermined community confidence in public health services. Additionally, sporadic reports of societal‐class‐based discrimination discouraged families from seeking care at government facilities. These factors collectively led families to favour private providers, despite the higher financial burden and availability of government healthcare services. Seeking care from private practitioners for childhood illnesses could often cost a family approximately 6–7 dollars per visit per child. The perceived efficiency, responsiveness and individualised attention offered by private practitioners were strong factors driving their preference.

Logistical barriers, such as limited transportation options and financial constraints, were significant factors delaying timely care. Many caregivers reported being unaware of government health centres providing paediatric care, assuming these facilities catered only to antenatal and postnatal services.

Family dynamics significantly influenced care‐seeking decisions, with grandparents often advocating for traditional remedies, contributing to delays in seeking appropriate medical care. In joint family structures, the involvement of multiple members in decision‐making further slowed the process of determining when and where to seek care. Social networks, including neighbours and community members, also shaped healthcare choices within households. Traditional gender roles played a pivotal role, as women frequently deferred healthcare decisions to male family members. Additionally, caregivers with higher educational attainment exhibited better health literacy, enabling them to recognise symptoms promptly and seek timely medical intervention. Gender biases also played a role in shaping care‐seeking behaviour, with boys often receiving quicker medical attention compared to girls. Families perceived boys as being more vulnerable to illness and prioritised their health needs over those of girls. In contrast, girls were often considered to be inherently stronger and more resilient, leading to delayed care‐seeking for female children.

Of the 231 suspected cases of under‐five pneumonia, 97% of caregivers sought medical care, but the majority relied on non‐RMPs, with only a minority accessing appropriate sources. Socio‐economic factors, such as higher maternal education and being in the least poor wealth group, were associated with a greater likelihood of seeking care from appropriate sources. The qualitative findings revealed that delays in treatment were often due to a perception that pneumonia is not a life‐threatening disease, reliance on home remedies, consultations with traditional healers and financial constraints. Additional barriers included mistrust in government healthcare services regarding the quality of services provided, social dynamics complicating decision‐making, and, in some cases, gender biases that delayed care‐seeking.

Logistical barriers and dependence on male family members for transportation, significantly delayed care, as highlighted in previous studies [32]. In this community, women making healthcare decisions independently or taking their children to healthcare practitioners without male or any other family‐member accompaniment are often perceived as culturally inappropriate behaviour. To avoid social repercussions and maintain their own reputation, women typically conform to established socio‐cultural norms, which can sometimes hinder timely care‐seeking. Additionally, most women in this community were found to be less educated and unemployed, which may have contributed to their lack of confidence and hesitation in independently approaching the tertiary level or distant healthcare centres [33]. The Ayushman Bharat programme in India aims to address such challenges by establishing Health and Wellness Centres (HWCs) for decentralised healthcare [34]. However, many caregivers reported that these centres were predominantly utilised for antenatal care services. Caregivers expressed willingness to use HWCs for under‐five care if such services were made available. Raising awareness about the services offered at HWCs and ensuring their functionality remain critical for enhancing access, especially in rural areas where transportation challenges persist. In addition, empowering women to make independent healthcare decisions could help.

Financial constraints emerged as a significant barrier, with many families delaying treatment until they could borrow money. This burden was particularly acute for those relying on private practitioners and hospitals due to perceived gaps in government hospitals, such as the frequent unavailability of essential medicines and the lack of diagnostic facilities. Although healthcare services and drugs for children are intended to be free in government facilities, these shortcomings force families to incur out‐of‐pocket expenses. Similar patterns have been reported in other LMIC settings where financial limitations significantly delayed care‐seeking for childhood illnesses [7, 8, 10, 35]. Financial restrictions may also lead to prioritising treatment for male children, as observed in this study. While boys were often perceived as more vulnerable than girls, this perception may reflect deeper socio‐cultural norms. In many Indian settings, boys are traditionally regarded as more valuable to the family due to their perceived role in ensuring future welfare, performing death rites and upholding family prestige. Studies from India have documented that families often prioritise healthcare and financial expenditures for male children over females, leading to delayed or inadequate care for girls [36, 37]. Such findings have also been reported in studies conducted in other LMICs [32, 38]. The preference for home remedies as the initial choice of care and reliance on spiritual healers probably reflected deeply rooted cultural norms and spiritual explanations for illness but may also be partially attributed to financial constraints. With an average monthly household income of around 158 dollars and limited awareness of the availability of free medicines and consultations at nearby government facilities, families are naturally inclined to choose cost‐effective options. Consequently, financial limitations indirectly shape and restrict their primary healthcare choices. Targeted interventions to address financial barriers by expanding the reach and awareness of government healthcare programmes, such as Ayushman Bharat, to ensure that essential medicines and diagnostic facilities are consistently available in government facilities is critical [34]. Also, increasing awareness of free transportation facilities to hospitals for sick under‐five children available under the national ambulance policy could alleviate the economic burden on families [39]. Strengthening financial literacy and providing clear information about available health insurance options can further mitigate financial constraints. Additionally, introducing community‐based health financing schemes or micro‐loans to support families during medical emergencies could reduce delays in seeking care. These measures could not only improve equitable access to care for both boys and girls but also reduce the delays in seeking timely healthcare from appropriate healthcare facilities across socio‐economic groups.

There was a clear socio‐economic gradient in health insurance coverage, but our findings also indicated that health insurance coverage did not eliminate all financial barriers. Several factors may contribute to more than half of the insured caregivers opting for care from non‐RMPs, like the perceived complexity of insurance schemes, indirect costs like transportation and lost wages, and mistrust in government facilities, discouraging families from using insured services [40, 41].

Another critical issue identified was the nature of counselling provided at government healthcare facilities. Counselling sessions were often one‐sided and lacked meaningful interaction. Healthcare providers of government centres frequently fail to dedicate sufficient time to understand the perspectives and concerns of beneficiaries, resulting in suboptimal quality of care. Lack of engagement and personalised attention further dissuades beneficiaries from utilising public health services and drives families toward private providers despite higher costs, as also seen in studies from Peru [35]. The preference for non‐RMPs may stem from their accessibility, convenience and established community relationships. Enhancing the quality and reliability of government services is crucial to rebuilding trust and encouraging utilisation. However, the qualitative findings revealed that mistrust in government services also often arose from community attitudes and rumours rather than direct personal experiences. The erosion of trust in ASHAs is partly driven by the perception that they prioritise national health programmes that offer financial incentives over those that do not. Since many ASHAs focus more on activities linked to incentives, community members may feel that their engagement is transactional rather than driven by a commitment to overall health promotion. This perception can undermine the credibility of ASHAs and affect participation in non‐incentivised health initiatives. Additionally, rumours casting doubt on their commitment to community welfare could have long‐term detrimental effects on the healthcare system. Efforts to improve public healthcare delivery and build community trust, for example, through clearly demonstrating efforts to be responsive and accountable and ensuring the consistent availability of essential medications, can make formal healthcare more accessible and appealing to the most vulnerable households. This may be combined with addressing gender biases through community awareness programmes, such as engaging with frontline health workers like ASHAs in counselling about girl children's health, to achieve more equitable health outcomes for all children.

The perception that pneumonia is not life‐threatening, as observed in this study, significantly contributes to delays in seeking formal care. Similar findings from other resource‐limited settings, where misbeliefs about illness severity led to delayed healthcare‐seeking, have been found [7, 8, 42]. The interplay of cultural beliefs, mistrust of modern medicine and perceptions of healthcare quality highlights the complex factors shaping care‐seeking behaviour in these communities. Addressing these challenges requires educational campaigns emphasising the severity of pneumonia and the risks of delayed care. Such interventions should target entire households, recognising the influential role of family members, particularly grandparents, in healthcare decision‐making.

This study's mixed‐methods design is a major strength, providing an understanding of many aspects of care‐seeking behaviour. The inclusion of multiple family members, including mothers, fathers and grandparents, offered diverse perspectives that enhanced our understanding of healthcare decision‐making. Additionally, this study, as part of the formative research within a broader IR, integrates structured community engagement in shaping implementation strategies. The findings from this phase will inform the development of the initial implementation model, while the broader IR will facilitate the ongoing involvement of caregivers, family members, community representatives and healthcare providers in refining and adapting strategies using structured feedback mechanisms (Supporting Information: Annexures III and VII). Additionally, the findings will support the preparation of a logic model [43] developed using the IR principles and tools [44, 45]. This strong engagement of beneficiaries, alongside active participation from community representatives and healthcare providers, enhances the study's relevance, sustainability and future impact by ensuring community‐driven, contextually appropriate strategies for improving pneumonia care‐seeking behaviours and effective management in rural India.

The reliance on self‐reported data introduced the possibility of recall bias, especially concerning the timing and source of care‐seeking. Additionally, symptoms may have been over‐reported or inaccurately recalled, inflating the number of suspected pneumonia cases. This issue is compounded by the low specificity of the definition of pneumonia used, which includes a broad range of symptoms without using investigations like chest X‐rays or laboratory tests. Consequently, it might capture many cases of other respiratory illnesses, leading to an overestimation of pneumonia cases. Furthermore, the qualitative component relied on purposive sampling, which enabled an in‐depth exploration of diverse perspectives. However, this approach may not have fully captured the experiences of less vocal caregivers or those facing local language barriers, who might encounter different challenges in care‐seeking. Social desirability bias may have also influenced responses, as some caregivers could have provided answers they perceived as favourable rather than their actual experiences. Additionally, while this study primarily focused on community‐level care‐seeking experiences, incorporating data from healthcare providers and health system stakeholders could have further enhanced validity and facilitated the triangulation of findings. However, this was beyond the scope of this manuscript. To address this gap, a separate manuscript exploring health system challenges and healthcare provider perspectives on pneumonia management has been published [26]. The findings of this study are also specific to the Palwal district, limiting their generalisability to regions with different healthcare infrastructures or cultural contexts.

This study highlights the complex socio‐economic, cultural and systemic factors that shape care‐seeking behaviour for childhood pneumonia in rural North India. Delays in seeking appropriate care are driven by poverty, reliance on non‐RMPs, social dynamics, inadequate awareness and knowledge of pneumonia, gender biases and mistrust in public healthcare systems. These findings contribute to the growing body of evidence on care‐seeking behaviour in LMICs and underscore the need for multifaceted approaches to improving healthcare access.

Barsha Gadapani Pathak: conceptualisation, methodology, software, data curation, investigation, validation, formal analysis, funding acquisition, visualisation, project administration, writing – original draft, writing – review and editing. Sarmila Mazumder: conceptualisation, investigation, funding acquisition, project administration, writing – review and editing, methodology, validation, visualisation, supervision. Yasir Bin Nisar: resources, writing – review and editing. Aditya Bhatt: data curation, project administration, software. Mandeep Singh: project administration, data curation. Tarun Madhur: project administration. Ingvild Fossgard Sandøy: methodology, supervision, writing – review and editing, validation.

The authors alone are responsible for the views expressed in this article, and they do not necessarily represent the views, decisions or policies of the institutions with which they are affiliated.

Ethical approval was granted by the ethical committees of the Society for Applied Studies (SAS/ERC/IR Pneumonia/2021), the Regional Ethics Committee of Western Norway (REK/2022/531608) and the World Health Organisation (WHO/ERC.0003652). Additionally, this study has obtained approvals from the Government of Haryana state (Memo no. HSHRC/2022/505) and the health ministry steering committee (approval date: 19 December 2022, proposal id 2022‐17596).

Consent was obtained from primary caregivers and their families after reading out the information sheet to them. The participants could withdraw or request to stop recording the interviews at any time.

The authors declare no conflicts of interest.

The authors would like to express sincere gratitude to all the under‐five caregivers who participated in this study and shared their experiences. They would also like to thank Mr Hashmi and Dr Manish Kumar for coordinating research activities and Ms Arti, Ms Kavita and Ms Dipti for assisting in the data generation process. The study was funded by the Bill & Melinda Gates Foundation (#INV008068) through a grant to the World Health Organization. The funders had no role in the study design or in the collection, analysis, or interpretation of the data. The funders did not write the report and had no role in the decision to submit the paper for publication.

Barsha Gadapani Pathak, Email: bapat2595@uib.no, Email: barsha.pathak@sas.org.in.

Sarmila Mazumder, Email: sarmila.mazumder@sas.org.in.

The data used and/or analysed during the current study are available from the corresponding author/s Barsha Gadapani Pathak () and/or () and Sarmila Mazumder () upon reasonable request. bapat2595@uib.no barsha.pathak@sas.org.in sarmila.mazumder@sas.org.in