Spiritual Care in Serious Illness: A Narrative Review of Low- and Middle-Income Country Evidence

Introduction

Spiritual care (SC) is increasingly recognized as essential to holistic healthcare and involves “care that is respectful of and responsive to individual patient preferences, needs, and values” [1–3]. While religion and spirituality have intertwined with health and healing until recent history, and robust evidence links SC to improved outcomes in serious illness, provision remains inconsistent [4–6]. Concerned about this gap, the World Health Organization has aimed to integrate spirituality and faith‑based organizations into health [7, 8]. However, SC provision goals remain unmet. Balboni et al. [1] synthesized 371 serious‑illness studies and concluded that 71% to 99% of patients view spirituality as necessary, spiritual needs are common (23–98%), and SC provision is associated with better quality of life and end‑of‑life outcomes [1]. Yet clinicians addressed spirituality in only 9% to 51% of cases. The panel recommended implementing routine spiritual assessments, clinician training, and engaging SC specialists.

The evidence base is skewed toward North America and high‑income countries (HICs), yet Asia‑Pacific, African, and Latin American populations frequently report that religion plays a central role in daily life [9]. Prior syntheses draw mostly on Christian‑majority or secular contexts, leaving Muslim‑, Hindu‑, and Buddhist‑majority regions and indigenous spiritual systems understudied. Between 2020 and 2022, the share of new studies from non‑American populations rose to 63%, up from 33% between 2000 and 2020 [1]. As research expands globally, it is essential to examine how existing methods are applied across diverse cultural and religious settings.

This narrative synthesis of 46 low‑ and middle‑income countries (LMICs) studies, drawn from the Balboni et al. [1] parent review, makes three contributions. First, it presents the first systematic synthesis of spirituality and serious illness restricted to LMIC settings, spanning South and Southeast Asia, the Middle East and North Africa, sub‑Saharan Africa, and Latin America. Second, it documents an inverse association between national Human Development Index (HDI) and clinician‑provided SC, with implications for how SC delivery may shift as health systems formalize and traditional family and faith networks recede. Third, it identifies the dominance of HIC‑developed measurement instruments in the LMIC evidence base (used in 38 of 46 studies) and proposes a both/and framework that retains validated global tools for cross‑cultural comparison while incorporating locally developed instruments to capture context‑specific spiritual phenomena.

Methods

Results

Of the 442 serious‑illness studies identified in the parent review (373 identified through April 2020 and 69 added in the 2020 to 2022 update), 61 were conducted in LMICs, with 46 studies rated as having low or moderate risk of bias (Figure 1) and 15 studies excluded for serious or critical risk of bias (Supplementary S1). 43 included studies (Supplementary S2) were patient‑related, while three were caregiver‑related (Supplementary S3). Therefore, we focused on patient‑related insights, unless otherwise stated. The studies were partitioned into two groups: (1) Those examining the role of or need for SC, including SC in decision‑making, and (2) Studies evaluating SC interventions. We also examined insights parsed by religious affiliation and the role of spirituality in medical decision‑making.

Figure 1

SC in Serious Illness

Spiritual Care Interventions in Serious Illness

Religious Affiliation and Spirituality in LMICs

In Muslim‑majority settings, prayer and a relationship with God were the most common sources of hope and coping. Greater spiritual well‑being was associated with better mood and greater hope [11, 31, 35]. An Iranian ICU study found an association between higher spirituality and lower depression, but greater post‑traumatic stress, emphasizing the psychological complexity of faith under serious illness [32]. Although SC was normatively supported, clinician‑led assessments in these settings were rare, largely due to limited training [19, 20, 48].

In Christian‑majority cohorts, including Latin America and South Africa, SC was associated with less pain, more home deaths (vs. hospitalized), and better psychosocial outcomes [26]. Across Brazilian and multinational samples, positive religious coping correlated with higher QoL and lower distress, whereas religious struggle predicted anxiety and impaired well‑being [34].

In Buddhist‑majority Thailand, spirituality centered on existential preparation for death and the cultivation of meaning. Higher spiritual well‑being was associated with greater perceived effectiveness of palliative care [17], and patients most valued “preparing for death” and “finding life purpose” [29]. Finally, in non‑religious or mixed settings (e.g., China), spiritual well‑being was typically lower. Nevertheless, individuals reported benefits from interventions that fostered peace, connection, and meaning rather than explicit religiosity [18, 36, 54]. Overall, spirituality and religion were salient, measurable, and generally associated with improved mental, social, and existential outcomes across traditions, with the strongest associations observed for meaning‑related and affective dimensions of well‑being.

SC Assessments Rely on Methods Developed in HIC Settings

Of the 46 studies, 38 (83%) used assessments that were originally developed in HIC settings, most of which were translated, validated, or culturally adapted for local use. Across these, forward–backward translation, expert review, pilot testing, and reliability checks were standard. Several studies reported cognitive interviewing (e.g., Satija et al. [46] for the Hindi FACIT‑Sp). A smaller subset employed locally developed measures (e.g., the Iranian Palliative‑Care Needs Scale, the Thai Spiritual Needs Scale for terminal cancer, and the Nigerian Spiritual‑Issues Checklist) [29, 38, 45]. Among the HIC tools, 13 studies explicitly reported additional modifications, including tradition‑specific terms or restructuring items to improve conceptual fit (Figure 2). For example, in Ghana, the SNAP was administered in Twi with contextualized wording (e.g., “visits from clergy/imam of your own faith community”) [12]. SNAP comprises 23 items grouped into psychosocial, spiritual, and religious domains and was designed to be multi‑faith (e.g., item examples reference the Torah, Qur’an, Bible, Analects, and Tibetan Book of the Dead). Other examples of adaptation include the SWBS, edited for the Thai Buddhist context [17]; the Arabic Beliefs and Values Scale, revised with faith‑specific wording and item deletion to suit Jordanian culture [21]; and the Brazilian Brief RCOPE, in which the “demonic reappraisal” item did not load cleanly and was handled separately due to cultural semantics [34]. Several teams employed measures developed to be religion‑agnostic, such as the EORTC QLQ‑SWB32 [18], or followed standardized translation protocols across countries [24].

Figure 2

In contrast, few studies explicitly incorporated context‑specific practices within their research processes (Figure 2). Although excluded from our primary analysis due to quality concerns, Sankhe et al. [55] drew on Hindu practices (e.g., austerities/mercy toward animals), and Han et al. [56] applied combined Naikan and Morita therapies in China (Naikan derives from Jodo Shin Buddhist practice but was secularized by its founder for general use) [55, 56].

Finally, a South African study combined Brief RCOPE items with African Palliative Care Association Palliative Outcome Scale (APCA POS) existential items (“felt at peace,” “life worthwhile”) and directly asked about traditional healers/medicine, but also noted gaps in assessing ancestral beliefs, suggesting these require bespoke tools [26]. Similarly, cross‑regional work faced adaptation constraints (e.g., resource‑limited translations), highlighting that even well‑validated instruments may miss salient, culture‑bound constructs [47].

Discussion

This narrative synthesis of 46 empirical studies with low or moderate risk of bias constitutes, to our knowledge, the most rigorous and comprehensive summary to date of SC in serious illness across LMICs. Across diverse regions, illnesses, and religious traditions, three findings stand out. First, spiritual concerns are pervasive and clinically salient, with spiritual well‑being and positive religious coping consistently associated with better QoL, psychological adjustment, and family outcomes. Second, spiritual needs remain substantially unmet despite strong endorsement of SC by both patients and clinicians. Third, SC provision in LMICs is shaped by distinctive patterns in delivery systems, religious landscapes, and measurement practices. Since most of the evidence is cross‑sectional and several regions are represented by only a few studies, the associations summarized here describe consistent patterns rather than established causal effects.

In many LMICs, family, religious leaders, and community networks, rather than clinicians, provide SC, reflecting cultural strengths but potentially disadvantaging patients lacking such networks [12, 16, 19]. This pattern invites reflection on how SC is conceptualized. In much of the HIC empirical literature, SC is framed as a professional service located within the health system, delivered by trained clinicians and chaplains. In many LMIC settings, SC instead functions as an existing social practice embedded in family life and faith communities, one that precedes and often operates independently of formal services. An implication for health systems is not merely to introduce SC as part of care delivery where it is absent, but to simultaneously recognize, support, and coordinate with care already occurring (e.g., through community structures), while ensuring that patients without strong family or faith networks are not overlooked. We suggest that neither model is inherently superior. Each carries distinct risks. For example, professionalized SC is often overlooked by physicians, undervalued by hospital systems, and may feel institutional or inaccessible to patients, all of which make trusted community structures more critical. Community‑dependent SC, on the other hand, may leave isolated patients unserved and its quality unexamined. Recognizing SC provision as varying along this professional–communal continuum, shaped by health system development and the local religious community, may be more useful than treating HIC delivery models as the default endpoint toward which LMIC systems should converge. Additionally, clinicians’ beliefs and training influenced SC provision, with those in lower‑HDI settings offering SC more often, although this finding comes from a study in a single region [13, 14]. Notably, as services shift toward allopathic, medicalized structures, communal structures that provide SC may weaken or be overlooked [13, 14]. This pattern follows HICs, which have moved away from historical affiliations with religious bodies and are now seeking to reintegrate SC [57]. Rather than supplanting traditional SC sources, structured systems could preserve their relational ethos while ensuring equitable reach [58]. Embedding some training or exposure to SC in medical and nursing education would also enable clinicians to provide SC in routine practice. Beyond clinician training, culturally responsive models of care could formally incorporate the actors who already deliver most SC in these settings. Family‑integrated intervention formats, which achieved the most sustained benefits in our synthesis [52, 54], offer one potential avenue. Others may involve structured partnerships in which faith leaders receive basic serious‑illness communication training, and lay or community health workers conduct spiritual screening with referral pathways to clinicians or chaplains where available.

Limitations

This review has several limitations. As a secondary analysis of a parent systematic review, it inherits previously identified limitations, including reliance on the original study selection, data extraction, and risk‑of‑bias judgments. The LMIC evidence base remains thin and uneven, with many regions and religious traditions represented by single, relatively small, hospital‑based samples. Most studies were cross‑sectional surveys conducted in specialist or tertiary‑care settings, using self‑report instruments. This limits causal inference, constrains generalizability, and under‑represents patients managed in primary care, community, or home‑based services, as well as family caregivers and people with non‑cancer serious illness. Sex‑disaggregated analyses were not conducted in this review because the parent systematic review did not extract sex‑stratified data in a form that would permit systematic synthesis across the included studies. Heterogeneity in populations, measures, and outcomes precluded meta‑analysis, so our narrative synthesis is vulnerable to selective reporting and publication bias.

Conclusions

Despite these limitations, coherence with Balboni et al. [1] strengthens the conclusion that SC is a fundamental dimension of serious‑illness care. The LMIC literature reviewed here, though still developing and unevenly distributed across regions, demonstrates methodological rigor and practical feasibility. Together with the HIC literature, our findings suggest that the association among spiritual well‑being, religious coping, QoL, and psychological distress is consistent across settings, although largely based on cross‑sectional evidence. However, patterns such as the dominance of family and faith networks in providing SC, the inverse gradient in HDI‑SC provision, and the link between spiritual concerns and structural disadvantage may characterize LMIC contexts. We recommend building on this foundation by developing culturally grounded, scalable SC models co‑designed with clinicians, patients, families, and faith/community partners. Combining globally validated instruments with locally relevant measures, and extending research beyond cross‑sectional designs to include longitudinal and interventional studies, will strengthen the evidence base. In particular, multi‑country consortia that intentionally include sites from multiple LMIC regions and HIC comparators may distinguish context‑specific phenomena from globally shared SC features. Given the small number of studies per region in this review, our synthesis provides a map of promising directions rather than definitive statements about cultural uniqueness. Embedding indicators of spiritual well‑being within serious‑illness and palliative‑care quality frameworks and incorporating SC competencies into medical and nursing education may help align health systems with what patients and caregivers consistently value—meaning, connection, and purpose. As the evidence base advances, consolidating it through culturally responsive implementation and sustained evaluation will help make spiritual care a core element of person‑centered healthcare in LMICs.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Competing Interests

The authors have no competing interests to declare.

Author Access to the Data

All authors had access to the data and contributed to the writing of the manuscript (see below).

Author Contributions

Ryan Meachen: Investigation, Writing—original draft, Data Analysis, Project administration

Junita Henry: Investigation, Writing—original draft, Data Analysis, Project administration

Sabbi Lall: Validation, Writing—original draft, Data Analysis, Writing – review & editing

Tracy A. Balboni: Supervision, Writing—review & editing, Data curation

Constantine S. Psimopoulos: Supervision, Writing—review & editing

Tyler J. VanderWeele: Conceptualization, Supervision, Writing—review & editing

Katelyn N.G. Long: Conceptualization, Supervision, Writing—review & editing

Data Availability Statement

All authors had access to the data and contributed to the writing of the manuscript. No new data were generated for this study. The narrative synthesis draws on previously published literature. The full list of included studies, extraction tables, risk‑of‑bias assessments, and supporting data for the analyses is provided in Supplementary Material S1–S4.

Additional File

The additional file for this article can be found as follow: