Study objectives

The primary objective of this investigation was to evaluate the diagnostic accuracy of the Distress Thermometer in detecting clinically significant depressive symptoms among Nigerian women with metastatic breast cancer, using the Beck Depression Inventory-II (BDI-II) as the reference standard. Additionally, we sought to determine the optimal DT cutoff score for this population using Youden’s Index and to examine the concordance between DT-defined distress and BDI-II-defined depression. We also aimed to identify clinical and demographic factors associated with discordant screening results between the two instruments, and ultimately to provide evidence-based recommendations for psychosocial screening approaches in LMIC oncology settings. These objectives were designed to address critical gaps in the psycho-oncology literature while informing clinical practice and health policy in resource-limited environments.

Study design and setting

This cross-sectional diagnostic validation study was conducted using de-identified data from the NSIA-LUTH Cancer Centre, a premier tertiary oncology facility located within Lagos University Teaching Hospital (LUTH), Nigeria. The center serves as a regional referral hub for cancer care in West Africa and offers comprehensive multidisciplinary services including medical oncology, radiation therapy, surgical oncology, palliative care, and limited psychosocial support services. The patient population predominantly comprises individuals from diverse socioeconomic backgrounds across Nigeria and neighboring West African countries. Data collection occurred over a 17-month period, from September 2020 to February 2022, encompassing both the initial COVID-19 pandemic period and subsequent recovery phases.

Participants

Eligible participants comprised adult patients aged 18 years or older with histologically confirmed metastatic breast cancer (Stage IV disease according to AJCC 8th edition) and Eastern Cooperative Oncology Group (ECOG) performance status ranging from 0 to 3. Additional inclusion criteria required the ability to read and comprehend English or access to reliable translation services, intact cognitive functioning as assessed by clinical evaluation, and willingness to provide written informed consent. Patients were excluded from participation if they had active psychotic disorders or severe cognitive impairment that would preclude reliable self-report, were concurrently participating in psychological intervention studies, had missing data on primary outcome measures (DT or BDI-II scores), or were unable to complete questionnaires due to severe physical symptoms or medical instability. These criteria were designed to ensure a representative sample of the target population while maintaining data quality and participant safety.

Sampling methodology

Consecutive sampling was employed to minimize selection bias and ensure representativeness of the target population. All eligible patients presenting for routine oncology care during the study period were approached for participation. Sample size was calculated based on an expected AUC of 0.75, with 80% power and α = 0.05, requiring a minimum of 280 participants. To account for potential missing data and loss to follow-up, we aimed to recruit 320 participants.

Measures

1. 1. Distress thermometer (DT)

The Distress Thermometer represents a single-item, 11-point Likert-type visual analog scale ranging from 0 (“no distress”) to 10 (“extreme distress”), designed to assess overall psychological burden experienced during the preceding week [8]. The DT was administered with the complete 34-item problem checklist organized into five domains: practical problems (transportation, insurance, work/school), family problems (dealing with children, dealing with partner), emotional problems (worry, fears, sadness, depression), spiritual/religious concerns, and physical problems (various symptoms). Participants were asked to indicate which problems they had experienced in the past week by checking applicable items across the five domains.

A threshold score of ≥4 has been widely endorsed by international guidelines, including those of the NCCN and the European Society for Medical Oncology (ESMO), to indicate clinically significant distress warranting further evaluation [9]. However, given documented variations in optimal cutoffs across populations and settings, we evaluated the full range of possible thresholds rather than assuming the appropriateness of this established cutoff.

The DT was selected for this investigation due to its brevity, ease of administration, widespread international adoption, and particular suitability for resource-limited environments where time constraints and limited mental health expertise may preclude the use of more complex screening instruments [20].

1. 2. Beck depression inventory-II (BDI-II)

The Beck Depression Inventory-II represents a well-validated, 21-item self-report instrument designed to assess the severity of depressive symptoms over the preceding two-week period [21]. Individual items are scored on a 4-point scale (0–3), yielding total scores ranging from 0 to 63. The instrument demonstrates excellent psychometric properties, with reported Cronbach’s alpha coefficients typically exceeding 0.90 in both clinical and non-clinical populations.

For this study, we employed a cutoff score of ≥20 to define clinically significant depressive symptoms, corresponding to moderate-to-severe depression severity. This threshold aligns with established guidelines for clinical practice and research applications, represents the level of symptom severity most likely to benefit from professional intervention, and has been validated in previous oncology studies [22,23].

We acknowledge that some validation studies in advanced cancer populations have identified lower optimal BDI-II cutoffs (e.g., ≥ 16) for screening purposes, which may increase sensitivity at the cost of specificity. However, we selected the ≥ 20 threshold to focus on moderate-to-severe depression most clearly warranting clinical intervention in resource-limited settings. Supplementary analyses using alternative BDI-II cutoffs (≥14, ≥ 19) are presented in Results to assess robustness of findings across different depression severity definitions.

We selected the BDI-II as our reference standard with full acknowledgment of the inherent construct mismatch in using a depression-specific instrument to validate a broad distress screening tool. This represents a significant methodological limitation, as discussed extensively in our limitations section. However, we pursued this design for specific pragmatic and clinical reasons, recognizing that our findings speak to the DT’s performance for one particular clinical application (identifying patients with depression) rather than validating its full intended scope of measuring general distress. We made this methodological choice for several reasons. First, depression represents the most prevalent and clinically significant form of psychological morbidity in advanced cancer populations, and is the primary target for psychological intervention in most LMIC oncology settings where mental health resources are scarce. Second, while other DT validation studies have used broader measures such as the BSI-18, HADS total score, or DSM-IV interviews capturing multiple forms of distress, the BDI-II offers superior feasibility for implementation in resource-limited environments, requiring no specialized training for administration. Third, despite the DT’s intended broad scope, clinical guidelines frequently recommend it for identifying patients with depression requiring referral for mental health services, making depression-specific validation clinically relevant. We explicitly acknowledge that our focus on depression as the outcome represents a narrower evaluation than the DT’s full intended purpose, and that poor performance for depression detection does not necessarily indicate failure to capture other forms of distress. However, if the DT cannot adequately identify depression, the most common and treatable form of psychological morbidity in cancer, its clinical utility remains limited regardless of performance for other distress dimensions.

Nonetheless, we acknowledge that the ideal reference standard for depression validation in medically ill populations is a structured clinical interview (e.g., SCID, MINI) conducted by trained mental health professionals, which can distinguish depressive symptoms from somatic symptoms of cancer. The use of a self-report measure (BDI-II) as our reference standard, particularly one containing substantial somatic content, represents a significant limitation that may have contributed to the observed poor concordance. Our findings should be interpreted as reflecting the challenges of depression screening using brief self-report tools in advanced cancer populations, rather than definitively establishing the DT’s cross-cultural invalidity.

1. 3. EORTC QLQ-C30 and BR23 modules

To comprehensively characterize participants’ health-related quality of life and contextualize psychological findings, all participants completed the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) and its breast cancer-specific module (BR23) [24]. These extensively validated instruments evaluate multiple domains including physical functioning, role functioning, emotional well-being, cognitive functioning, social functioning, global health status, and disease-specific symptoms.

Scoring followed the official EORTC guidelines, employing linear transformation to generate standardized scores on a 0–100 scale. Higher scores on functional scales represent better functioning, while higher scores on symptom scales indicate greater symptom burden.

Statistical analysis

Statistical software and significance

All analyses were conducted using R statistical software version 4.4.1 (R Core Team (2024). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria). Specific packages included pROC for ROC analysis, psych for reliability analysis, and corrplot for correlation matrices. Statistical significance was set at α = 0.05 for all tests. Given the exploratory nature of subgroup analyses, we applied Bonferroni correction where appropriate to control for multiple testing.

Missing data management

Missing data were minimal (< 5% for any variable) and appeared to occur at random based on Little’s MCAR test. Complete case analysis was performed for the primary diagnostic accuracy analysis to preserve the integrity of sensitivity and specificity estimates. Sensitivity analyses using multiple imputation were conducted to assess the robustness of findings.

Ethical considerations

The study protocol received approval from the Lagos University Teaching Hospital Health Research Ethics Committee (LUTH-HREC, IRB ADM/DCST/HREC/APP/7058). All participants provided written informed consent after receiving comprehensive information about study procedures, potential risks and benefits, and their right to withdraw at any time without affecting their medical care. Special considerations were made for participants with limited literacy, including the availability of witnessed verbal consent procedures when appropriate.

This study represents a secondary analysis of de-identified patient data originally collected for a broader investigation of quality of life and psychosocial outcomes in Nigerian women with metastatic breast cancer. All identifying information was removed prior to data sharing, and researchers had access only to anonymized data. Data were first accessed for this secondary analysis on 27/10/2023.

Data confidentiality and security were maintained in accordance with international guidelines, including the Declaration of Helsinki and Good Clinical Practice standards. All data were de-identified and stored on password-protected servers with restricted access limited to authorized study personnel.

Participant characteristics and study flow

A total of 313 patients were initially screened for study participation, of whom 309 (98.7%) met all eligibility criteria and completed the required assessments (Fig 1). Four participants (1.3%) were excluded due to incomplete data on primary outcome measures. The high retention rate reflects the robustness of our recruitment strategy and the acceptability of study procedures to participants.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. STROBE-compliant participant flow diagram.

Study participant recruitment and enrollment flow. A total of 313 patients with metastatic breast cancer were screened for eligibility at NSIA-LUTH Cancer Centre, Lagos, Nigeria (September 2020–February 2022). Four patients were excluded due to incomplete primary outcome data (DT or BDI-II scores), resulting in 309 participants included in the final analysis. The high retention rate (98.7%) demonstrates the feasibility and acceptability of study procedures in this population.

https://doi.org/10.1371/journal.pone.0341221.g001

The study population comprised predominantly female participants (98.7%, n = 305), with a median age of 53.0 years (interquartile range: 44.0–63.0 years). The age distribution was slightly right-skewed, reflecting the inclusion of some younger patients with aggressive disease. Educational attainment varied considerably, with 34.6% having completed secondary education and 22.3% having obtained post-secondary qualifications. Approximately 45.0% of participants reported being unemployed or unable to work due to their illness, highlighting the substantial socioeconomic impact of metastatic disease.

Primary outcome measures

Beck depression inventory-II findings.

BDI-II scores averaged 12.3 (SD ± 7.2, range: 0–39), with the distribution demonstrating positive skewness consistent with a predominantly non-depressed sample. Applying our predetermined cutoff of BDI-II ≥ 20 for clinically significant depressive symptoms, only 49 participants (15.9%) met this criterion (Table 1). An additional 95 participants (30.7%) scored in the mild depression range (BDI-II 14–19), while 130 participants (42.1%) demonstrated minimal or no depressive symptoms (BDI-II 0–13).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Baseline demographic, clinical, psychosocial, and quality of life characteristics of participants (N = 309).

https://doi.org/10.1371/journal.pone.0341221.t001

Clinical and disease characteristics

Regarding metastatic disease patterns, 213 participants (69.0%) presented with visceral metastases, 151 (48.9%) had spinal involvement, and 40 (12.9%) had brain metastases. Multiple metastatic sites were common, with 156 participants (50.5%) having ≥3 distinct sites of disease. Brain metastasis was present in 40 (12.9%), pleural metastasis in 57 (18.4%), while liver metastases occurred in 42 (13.6%) participants.

Treatment history revealed that 176 patients (57.0%) had received chemotherapy, 149 patients (48.2%) had undergone surgery, and 62 patients (20.1%) had received neoadjuvant therapy. Hormonal therapy utilization was surprisingly low at 4.2% (n = 13), likely reflecting the predominance of triple-negative breast cancer in this population, consistent with epidemiological patterns observed in sub-Saharan Africa. No participants had received immunotherapy, reflecting financial barriers and limited access to novel therapeutic agents in this LMIC setting.

EORTC QLQ-C30 assessments revealed substantial symptom burden, with mean Physical Symptom Scores of 31.3 (SD ± 17.4) and Emotional Symptom Scores of 19.0 (SD ± 13.3). The Global Quality of Life Score averaged 62.7 (SD ± 11.1), indicating moderate impairment relative to population norms. Problem checklist analysis revealed that participants reported an average of 5.0 physical problems (SD ± 3.6), 2.7 practical problems (SD ± 1.6), and 2.3 emotional problems (SD ± 1.9), with physical symptoms representing the predominant source of distress (Table 1).

Diagnostic performance analysis

ROC curve analysis.

The ROC curve analysis revealed extremely poor diagnostic performance of the DT in identifying participants with clinically significant depressive symptoms (Fig 2). The AUC was 0.414 (95% CI: 0.326–0.503), significantly below the threshold for acceptable diagnostic accuracy (p < 0.001 for comparison with AUC = 0.5). This poor concordance between the DT and BDI-II in identifying participants with clinically significant depressive symptoms likely reflects that these instruments measure different psychological constructs in this population. As discussed in our limitations section, this pattern is consistent with the DT capturing primarily physical and practical distress while the BDI-II, containing substantial somatic content, is confounded by cancer-related symptoms in this advanced-stage population.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Receiver operating characteristic (ROC) curve showing the performance of the Distress Thermometer (DT) in predicting clinically significant depressive symptoms (BDI ≥ 20).

ROC curve analysis demonstrating the diagnostic performance of the Distress Thermometer for detecting clinically significant depressive symptoms (BDI-II ≥ 20) in Nigerian women with metastatic breast cancer (n = 309). The AUC was 0.414 (95% CI: 0.326–0.503), indicating performance significantly worse than chance (diagonal reference line, AUC = 0.5). The curve’s position below the reference line confirms the DT’s inadequate discriminatory ability for depression screening in this population.

https://doi.org/10.1371/journal.pone.0341221.g002

Optimal cutoff analysis.

Using Youden’s Index to optimize the balance between sensitivity and specificity, the optimal DT cutoff was identified as ≥7.5. However, this threshold yielded extremely poor sensitivity (2.0%) while achieving high specificity (98.5%). The positive predictive value was 20.0%, and the negative predictive value was 84.2%. The positive likelihood ratio was 1.33 and the negative likelihood ratio was 1.00, both indicating minimal diagnostic utility (Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Sensitivity and specificity by DT cutoff scores.

Sensitivity and specificity curves for the Distress Thermometer across all possible cutoff values (0.5–9.5) for detecting clinically significant depressive symptoms (BDI-II ≥ 20). The red line represents sensitivity and the blue line represents specificity. The vertical dashed line indicates the Youden-optimal cutoff (DT = 7.5), which achieved 98.5% specificity but only 2.0% sensitivity. The curves demonstrate the classic inverse relationship between sensitivity and specificity, with no threshold providing clinically acceptable performance (≥80% sensitivity and ≥70% specificity simultaneously).

https://doi.org/10.1371/journal.pone.0341221.g003

Performance across multiple thresholds.

Examination of diagnostic performance across all possible DT cutoffs revealed no clinically acceptable threshold (Table 2). The conventional DT ≥ 4 cutoff demonstrated sensitivity of 34.7% and specificity of 50.8%, representing suboptimal performance by any clinical standard. Even lowering the threshold to DT ≥ 1.5 achieved sensitivity of only 83.7% with specificity of 10.0%, resulting in an unacceptably high false-positive rate.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Diagnostic performance of distress thermometer (DT) cutoffs for predicting clinically significant depression (BDI ≥ 20).

https://doi.org/10.1371/journal.pone.0341221.t002

The diagnostic odds ratio (DOR) for the optimal cutoff (DT ≥ 7.5) was 1.34 (95% CI: 0.18–10.2), indicating minimal discriminatory ability. Across all evaluated thresholds, no combination achieved both clinically acceptable sensitivity (≥80%) and specificity (≥70%) simultaneously.

Subgroup and supplementary analyses

Subgroup analyses by patient characteristics.

Comprehensive subgroup analyses revealed remarkably consistent poor diagnostic performance across all major patient demographic and clinical characteristics, suggesting that the DT’s limitations in this population are not confined to specific patient subgroups (Fig 4). Among younger patients (≤50 years), the AUC was 0.38 compared to 0.44 in older patients (>50 years), with no statistically significant difference between groups (p = 0.45). This finding indicates that age-related factors such as coping mechanisms, disease perception, or communication styles do not substantially influence the DT’s diagnostic utility. Similarly, performance status did not meaningfully affect diagnostic accuracy, with patients having better functional status (ECOG 0–1) demonstrating an AUC of 0.41 compared to 0.40 in those with poorer performance status (ECOG 2–3, p = 0.89). This suggests that physical functioning and symptom burden, while potentially influencing psychological well-being, do not alter the fundamental measurement properties of the DT in this population.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Forest plot showing the area under the receiver operating characteristic curve (AUC) for the Distress Thermometer’s ability to detect clinically significant depressive symptoms (BDI-II ≥ 20) across different patient subgroups.

Error bars represent 95% confidence intervals. The horizontal dashed line at AUC = 0.5 indicates performance no better than chance, while the dotted line at AUC = 0.7 represents the threshold for acceptable diagnostic accuracy. All subgroups demonstrated poor diagnostic performance with AUC values well below the acceptable threshold, indicating consistent failure of the DT across patient characteristics. P-values compare AUC between subgroups within each category using DeLong’s test for correlated ROC curves. Reference lines: Dashed line (AUC = 0.5) = chance performance; Dotted line (AUC = 0.7) = acceptable diagnostic accuracy threshold. Statistical note: No subgroup achieved clinically acceptable diagnostic performance (AUC ≥ 0.7). The consistently poor performance across all demographic and clinical characteristics suggests fundamental limitations of the DT in this population rather than subgroup-specific issues.

https://doi.org/10.1371/journal.pone.0341221.g004

Analyses stratified by metastatic disease patterns revealed that patients with visceral metastases had an AUC of 0.39 compared to 0.46 in those with non-visceral disease (p = 0.34), again demonstrating no clinically meaningful differences. The consistently poor performance across disease severity indicators suggests that the DT’s limitations are not attributable to specific disease characteristics but rather reflect more fundamental issues with the instrument’s validity in this cultural and clinical context.

Relationship between distress, depression, and quality of life

To better understand the discordance between DT and BDI-II measurements, we examined their relationships with quality of life domains assessed by the EORTC QLQ-C30. The DT demonstrated moderate positive correlation with physical symptom burden (r = 0.34, p < 0.001) and weak correlation with emotional functioning (r = 0.19, p = 0.001). In contrast, the BDI-II showed stronger correlation with emotional functioning (r = 0.42, p < 0.001) and weaker correlation with physical symptoms (r = 0.28, p < 0.001). Both instruments showed similar modest correlations with global quality of life (DT: r = −0.26; BDI-II: r = −0.31; both p < 0.001). Analysis of DT problem checklist domains revealed that participants with elevated BDI-II scores (≥20) endorsed significantly more emotional problems (mean 3.8 vs 2.0, p < 0.001) but not significantly more physical problems (mean 5.4 vs 4.9, p = 0.34) compared to those with BDI-II < 20. These patterns suggest that the DT and BDI-II capture distinct aspects of patient experience, with the DT more heavily influenced by physical symptomatology while the BDI-II more specifically reflects emotional distress.

Principal findings and clinical implications

This study represents the first comprehensive validation of the Distress Thermometer in sub-Saharan African patients with metastatic breast cancer and reveals concerning findings regarding its utility for depression-specific screening. Our results demonstrate poor concordance between the DT and BDI-II for identifying depressive symptoms, likely reflecting methodological limitations rather than systematic failure of the DT as a screening instrument. With an AUC of 0.414, the instrument performed significantly worse than random classification, challenging fundamental assumptions about its universal applicability in cancer care.

The magnitude of diagnostic failure observed in our study is particularly striking when contextualized against international validation studies. While previous investigations in high-income countries typically report AUC values ranging from 0.70 to 0.85 for the DT [27,28], our findings suggest a complete breakdown of the instrument’s psychometric properties in this LMIC setting. The optimal cutoff of DT ≥ 7.5, despite achieving high specificity, demonstrated very poor sensitivity (2.0%), indicating inadequate performance for depression-specific screening in this methodological context.

Perhaps most concerning is the substantial discordance between DT-defined distress (47.0% of participants) and BDI-II-defined depression (15.9% of participants). This finding suggests that nearly half of patients reporting significant distress on the DT do not meet criteria for clinically significant depression, while conversely, many patients with depression are not identified as distressed by the DT.

Somatic symptom overlap and reference standard limitations

Beyond the construct mismatch between broad distress and depression-specific measurement discussed above, a critical methodological factor that may substantially explain our findings is the confounding effect of somatic symptoms in advanced cancer populations. The most parsimonious explanation for our findings, including the extraordinary AUC below 0.5, is the fundamental methodological artifact created by using a somatically-confounded, depression-specific self-report instrument to validate a broad distress screening tool in an advanced cancer population. This methodological issue likely contributes more substantially to the observed poor concordance than cultural factors or cross-cultural invalidity of the DT itself. The BDI-II, while an excellent depression screening instrument in general populations, contains seven items (33% of total) that assess somatic/neurovegetative symptoms: fatigue/loss of energy (Item 15), changes in sleeping pattern (Item 16), irritability (Item 17), changes in appetite (Item 18), concentration difficulty (Item 19), tiredness or fatigue (Item 20), and loss of interest in sex (Item 21). These symptoms overlap extensively with the direct physiological effects of advanced cancer and its treatment.

In our MBC population experiencing uniformly advanced disease with high physical symptom burden (mean EORTC Physical Symptom Score 31.3 ± 17.4), participants likely endorsed many BDI-II somatic items due to cancer-related symptoms rather than depression per se. This phenomenon, namely, somatic symptom inflation of depression scores in medically ill populations, is well-documented in the psycho-oncology literature but particularly pronounced in advanced cancer. A patient with severe cancer-related fatigue, chemotherapy-induced sleep disruption, pain-related concentration difficulties, and disease-related appetite loss may score in the moderate depression range on the BDI-II (total score 15–20 or higher) without experiencing the core mood and cognitive symptoms of clinical depression.

Critically, this somatic confounding may have created a systematic bias in our reference standard that explains the inverse relationship we observed (AUC < 0.5). If many participants had elevated BDI-II scores driven primarily by somatic symptoms rather than true depression, while the DT (being a single-item global rating) was more accurately capturing psychological distress specifically, this would produce the pattern we observed, DT and BDI-II measuring different phenomena and showing poor concordance or even inverse relationships at certain cutoffs.

This interpretation is supported by our finding that the DT correlated more strongly with physical symptom burden than with emotional functioning, suggesting participants were primarily rating physical/somatic distress when completing the thermometer. Meanwhile, the BDI-II, intended to measure depression, was paradoxically also capturing substantial somatic symptomatology. The result is two instruments that, in this advanced cancer population with overwhelming physical symptoms, were both contaminated by somatic burden but in different ways, producing poor concordance that reflects methodological artifact rather than true diagnostic failure of either instrument.

This sobering realization highlights a fundamental limitation of our study design: we compared one screening tool (DT) against another screening tool (BDI-II), rather than against a gold-standard diagnostic interview conducted by a mental health professional who could determine whether somatic symptoms represented depression, cancer effects, or both. A diagnostic interview allows clinical judgment to weigh the relative contribution of mood versus disease burden to reported symptoms, a nuance impossible with self-report questionnaires alone. Our AUC of 0.414 therefore reflects, at least in part, the discordance between two imperfect instruments both affected by somatic confounding, rather than necessarily indicating poor validity of the DT per se.

This methodological explanation does not entirely exculpate the DT, which still failed to identify depression adequately in our sample, but it substantially reframes the interpretation. Rather than representing a catastrophic failure of cross-cultural validity, our findings may primarily reflect the methodological challenges of depression assessment in advanced cancer using self-report screening tools, a challenge that would likely affect any brief instrument in this population. Future validation studies in advanced cancer populations must employ diagnostic clinical interviews as gold standards to disentangle these confounded relationships.

Construct mismatch and measurement specificity

A critical consideration in interpreting our findings concerns the fundamental difference between what the DT and BDI-II measure. The Distress Thermometer was designed as a broad screening tool to capture multiple dimensions of distress including anxiety, depression, existential concerns, practical problems, spiritual distress, and physical symptom burden. In contrast, the BDI-II specifically measures depressive symptoms. This construct mismatch raises the question of whether our findings reflect cultural invalidity of the DT or simply the expected consequence of using a depression-specific reference standard to validate a broad distress measure.

We acknowledge this methodological tension explicitly. The poor AUC we observed (0.414) may partly reflect the DT’s capture of non-depressive forms of distress that are not measured by the BDI-II. Our supplementary analysis of EORTC QLQ-C30 relationships provides evidence for this interpretation: the DT correlated more strongly with physical symptom burden (r = 0.34) than emotional functioning (r = 0.19), while the BDI-II showed the opposite pattern (r = 0.42 for emotional functioning vs r = 0.28 for physical symptoms). This suggests the DT may be more responsive to somatic distress while the BDI-II captures psychological distress more specifically.

However, we contend that this construct mismatch, rather than invalidating our conclusions, actually underscores a critical clinical limitation of the DT in LMIC oncology settings. Current clinical guidelines, including those from NCCN and ESMO, recommend the DT not merely as a general distress screening tool but specifically for identifying patients with anxiety and depression requiring mental health referral. In resource-limited settings where comprehensive psychological assessment is often unavailable, the DT is frequently used as the primary or sole mechanism for identifying patients needing psychological intervention, and depression represents the most prevalent and treatable form of psychological morbidity in these contexts. If the DT cannot adequately discriminate patients with clinically significant depression from those without, its utility as a practical clinical tool is fundamentally compromised regardless of its ability to capture “distress” in some broader, less clinically actionable sense.

Moreover, the pattern of discordance we observed suggests the DT may be identifying distress driven primarily by physical symptom burden and practical problems rather than psychological morbidity per se. This is evidenced by the high proportion of participants with elevated DT scores but low BDI-II scores, and by the stronger correlation between DT and physical symptoms compared to emotional functioning. While physical and practical concerns certainly merit clinical attention, they require different interventions than depression (symptom management, social services) and should not be conflated with psychiatric morbidity. A screening tool that cannot distinguish between somatic distress and depression has limited utility for guiding referral decisions in settings where mental health resources must be carefully allocated.

Therefore, while we acknowledge our study evaluated the DT against a depression-specific rather than broad distress reference standard, we maintain that our findings reveal clinically significant inadequacies in the instrument’s performance for its primary intended purpose in cancer care settings.

Construct non-equivalence and cross-cultural measurement challenges

Beyond the immediate diagnostic accuracy findings, our results raise fundamental questions about construct validity and measurement equivalence in cross-cultural psycho-oncology research. The substantial discordance observed between DT-defined “distress” and BDI-II-defined “depression” in our Nigerian cohort (47.0% vs. 15.9%) may reflect not merely measurement error or tool inadequacy, but rather fundamental differences in how these psychological phenomena are conceptualized, experienced, and expressed across cultural contexts.

In Western psychological frameworks, “distress” and “depression” represent related but distinguishable constructs, both falling within an individualistic model of emotional suffering as internal psychological states amenable to numeric quantification. However, in many African cultural contexts, emotional suffering may be understood through fundamentally different frameworks, as spiritual experiences, social disruptions, or somatic manifestations, that do not map neatly onto Western psychological categories. What Western instruments label as “distress” or “depression” may be experienced and articulated by Nigerian patients as ancestral displeasure, family disharmony, or bodily imbalance.

This construct non-equivalence has profound implications for the validity of cross-cultural research in psycho-oncology. Our findings suggest that simply translating and administering Western-developed instruments, even those with strong psychometric properties in their original contexts, may yield measurements of uncertain meaning when the underlying psychological constructs themselves may not exist or may be structured differently across cultures. The poor correlation between DT and BDI-II scores in our sample (r = 0.23) supports this interpretation, indicating that these instruments may be tapping into qualitatively different dimensions of experience in this population.

Cultural and contextual factors influencing DT performance

Several interconnected factors likely contribute to the DT’s poor performance in this Nigerian cohort. While the cultural interpretations we offer below are plausible and consistent with existing literature on mental health in African contexts, we acknowledge that without formal cognitive debriefing or qualitative assessment of how participants understood and responded to the DT, these explanations remain speculative. Direct evidence from patients about their subjective experience of completing the instrument would be necessary to definitively determine the mechanisms underlying the DT’s poor performance. First, cultural conceptualizations of psychological distress may differ fundamentally from Western frameworks underlying the instrument’s development. In many African societies, emotional suffering is often understood through spiritual, communal, or somatic lenses rather than individualistic psychological models [29,30]. The concept of rating one’s internal emotional state on a numeric scale may be culturally unfamiliar or inappropriate, leading to inconsistent or invalid responses.

Second, pervasive mental health stigma in Nigerian society may lead to systematic underreporting of psychological symptoms [31]. Participants may fear discrimination, family rejection, or perceived weakness if they acknowledge significant emotional distress, particularly in a medical setting where physical symptoms are prioritized. This phenomenon could explain why many participants with clinically significant depression (as measured by the more detailed BDI-II) report low distress scores on the single-item DT.

Third, linguistic and semantic challenges may compromise the validity of distress measurement. While our study was conducted in English, many participants likely think and feel in their native languages (Yoruba, Igbo, Hausa, among others), where concepts of “distress” may not have direct equivalents or may carry different connotations. The translation of internal emotional experiences into English-language numeric ratings may introduce systematic measurement error.

Fourth, the collectivist cultural orientation prevalent in Nigerian society may influence how individuals conceptualize and report personal distress. In collectivist cultures, emotional experiences are often understood in relational and community contexts rather than as individual psychological states, potentially rendering single-item self-report measures less valid.

Comparison with international literature

Our findings align with a growing body of evidence questioning the universal applicability of Western-developed psychosocial screening instruments. A systematic review of DT validation studies across diverse populations revealed substantial heterogeneity in optimal cutoff scores and diagnostic performance metrics [32]. Studies conducted in non-Western populations consistently reported lower sensitivity and altered optimal thresholds compared to investigations in high-income Western countries.

Specifically, while DT validation studies using broad distress measures (BSI-18, HADS total score) in high-income countries have reported AUC values of 0.70–0.85 [33], depression-specific validations show more variable performance with AUC values frequently in the 0.60–0.70 range. our AUC of 0.414 for depression detection remains substantially lower than even these more modest benchmarks, suggesting fundamental challenges beyond the typical variance observed across validation studies. Studies in other LMIC settings have shown intermediate performance with AUC values of 0.60–0.70 for various psychological outcomes [8,34]. This pattern suggests a gradient of diagnostic utility that correlates with cultural and socioeconomic distance from the instrument’s development context.

The discordance between our findings and those from high-income countries cannot be attributed solely to differences in disease stage or treatment setting. Previous studies of advanced cancer patients in Western settings, while showing somewhat reduced DT performance compared to early-stage populations, have not demonstrated the complete diagnostic failure observed in our cohort. This suggests that cultural and contextual factors, rather than disease-related variables, represent the primary drivers of poor performance.

Comparison with prior Nigerian validation studies

Our findings present a striking contrast to prior Nigerian DT validations and require careful interpretation. Lasebikan et al. [13] and Obiajulu et al. [14] both reported good-to-excellent diagnostic performance (AUCs of 0.87 and 0.82 respectively) in a mixed-stage Nigerian cancer population and a non-cancer cohort, respectively, while our study found diagnostic performance significantly worse than chance (AUC 0.414). This dramatic discordance demands systematic exploration of potential explanatory factors.

Several factors may explain the discordance between these prior studies and our findings. Regarding implementation methodology, both Lasebikan et al. [13] and Obiajulu et al. [14] administered instruments in controlled research settings with standardized protocols, though neither study provided detailed descriptions of specific administration procedures, interviewer training protocols, or quality control measures beyond their use of HADS as reference standard. Our study similarly employed standardized administration protocols with trained research personnel, suggesting that differences in basic implementation methodology are unlikely to fully explain the divergent findings. First, both previous studies used HADS, which measures combined anxiety and depression, whereas we used the depression-specific BDI-II. The DT may perform better at detecting general psychological distress (anxiety plus depression) than depression specifically, consistent with its design as a broad screening tool. Second, our population consisted exclusively of patients with metastatic breast cancer, a uniformly advanced-stage cohort facing terminal prognosis, whereas Lasebikan et al. [13] included mixed cancer stages and Obiajulu et al. [14] studied non-cancer patients. Disease severity and proximity to death may influence how distress is experienced and reported. Third, our sample was 98.7% female, while the prior studies included more balanced gender distributions; cultural factors around emotional expression may differ between men and women in Nigerian society. Fourth, the different reference standards may be measuring genuinely different psychological constructs, our focus on moderate-to-severe depression (BDI-II ≥ 20) represents a more specific and severe outcome than general psychological distress measured by HADS.

Rather than contradicting these prior Nigerian studies, our findings complement them by providing the first depression-specific validation in an advanced cancer population. Taken together, the three Nigerian studies suggest that while the DT may have some utility for detecting general psychological distress in mixed Nigerian populations, its performance deteriorates substantially when (a) the outcome of interest is depression specifically rather than combined anxiety/depression, and (b) the population consists of patients with advanced, terminal illness. These nuanced interpretation has important clinical implications. In Nigerian oncology settings, the DT may retain utility for initial broad screening in mixed cancer populations, identifying patients with any form of significant distress who warrant further evaluation. However, it cannot be relied upon as a depression-specific screening tool, particularly in advanced cancer populations where somatic symptoms confound assessment. If depression identification is the clinical goal, as it often is in resource-limited settings where treatment is targeted at the most prevalent and treatable condition, then depression-specific instruments or clinical interviews are necessary following positive DT screens.

Methodological strengths and study rigor

This investigation possesses several methodological strengths that enhance confidence in our findings. First, to the best of our knowledge, our study represents the largest validation of the DT in a sub-Saharan African cancer population and the only investigation specifically focused on metastatic breast cancer in this region. The homogeneous disease population eliminates confounding from cancer stage or treatment heterogeneity that has complicated previous multi-cancer validation studies.

Second, we employed rigorous diagnostic accuracy methodology, including comprehensive ROC curve analysis, multiple cutoff evaluation, and optimal threshold determination using established statistical methods. Our use of Youden’s Index, while conventional, was supplemented by examination of clinical utility indices and likelihood ratios to provide a comprehensive assessment of diagnostic performance.

Third, the BDI-II represents an appropriate and well-validated reference standard for depression screening in cancer populations. Unlike some previous DT validation studies that used clinical interviews or other distress measures as reference standards, our choice of a depression-specific instrument allows for clearer interpretation of discordant results.

Fourth, our comprehensive assessment of potential confounding variables, including detailed clinical and sociodemographic data, allows for robust interpretation of findings and identification of factors that might influence screening performance.

Study limitations and interpretive considerations

Several limitations warrant acknowledgment in interpreting our findings. First, and most critically, we did not conduct formal cognitive debriefing or qualitative assessment of how participants interpreted and responded to the DT scale. This represents a critical gap in our investigation, as cognitive interviews could have revealed whether patients struggled with the numeric rating format, misunderstood the concept of “distress,” interpreted the scale anchors differently than intended, or experienced cultural barriers to reporting psychological symptoms. Without this qualitative component, we cannot definitively determine why the DT failed in this population, whether due to semantic/linguistic issues, cultural inappropriateness of the rating format, stigma-related response bias, or fundamental construct non-equivalence. While our cultural interpretations are plausible and consistent with existing literature on mental health in African contexts, there is need for direct evidence from participants about their subjective experience of completing the instrument. Future validation studies in LMIC settings should employ mixed-methods designs that combine quantitative diagnostic accuracy assessment with qualitative cognitive interviewing to understand the mechanisms underlying tool performance or failure. Such approaches would not only identify whether an instrument works but also why it succeeds or fails, providing crucial insights for cultural adaptation efforts and tool development.

Second, an important limitation concerns our English language inclusion criterion. While we allowed access to translation services, requiring English comprehension may have systematically excluded patients with lower educational attainment or from rural areas who might express and report psychological distress differently. This language requirement may have resulted in a sample with greater familiarity with Western conceptual frameworks and numeric rating scales, potentially underestimating the cultural barriers to valid DT administration in the broader Nigerian population. Our findings may therefore represent a “best-case scenario” for DT performance, with even poorer diagnostic accuracy likely in populations with more limited English proficiency or formal education. Future studies should include participants assessed in their native languages to provide a more comprehensive picture of the DT’s cross-cultural validity.

Third is the limitation of using one screening instrument (BDI-II) as the reference standard for validating another screening instrument (DT), rather than employing a diagnostic gold-standard clinical interview conducted by a mental health professional. While the BDI-II represents a well-validated depression screening instrument, it remains a self-report measure rather than a structured clinical interview. However, the BDI-II’s strong psychometric properties and widespread validation in cancer populations support its appropriateness as a reference standard for this investigation. Additionally, we acknowledge that our use of a depression-specific reference standard to validate a broad distress measure represents a methodological limitation, as discussed extensively in our construct mismatch section. The poor performance we observed may partly reflect this measurement incongruence rather than solely cultural invalidity.

Fourth, the high potential for somatic symptom overlap in the BDI-II to have confounded our results represents a critical interpretive limitation. Seven of the 21 BDI-II items (33%) assess somatic/neurovegetative symptoms (fatigue, sleep disturbance, appetite changes, concentration difficulty, loss of interest in sex, irritability, tiredness) that overlap extensively with the direct physiological effects of advanced cancer and its treatment. In our metastatic breast cancer population with uniformly high physical symptom burden (mean EORTC Physical Symptom Score 31.3 ± 17.4), BDI-II scores were likely inflated by cancer-related symptoms independent of mood disturbance. This somatic confounding may have created a reference standard that systematically misclassifies patients, some with high BDI-II scores driven by somatic symptoms without true depression, others with low BDI-II scores despite depression because psychological symptoms are overshadowed by overwhelming physical symptoms. If the DT was more accurately capturing psychological distress while the BDI-II was confounded by somatic symptoms, this would produce the inverse relationship we observed (AUC < 0.5) through methodological artifact rather than true DT failure. The absence of a diagnostic interview prevents us from determining the true depression prevalence in our sample and from validating whether the BDI-II ≥ 20 threshold identified genuine depression or somatic symptom burden. This fundamentally limits our ability to conclude that the DT “failed”, as it may have been responding to a different (and potentially more valid) construct than our confounded reference standard. Future studies must employ clinical interviews as gold standards in advanced cancer populations to disentangle these relationships.

Fifth, our cross-sectional design precludes assessment of how screening performance might vary across different points in the cancer trajectory or in response to treatment changes. Longitudinal investigations would provide valuable insights into the temporal stability of DT performance and its relationship to disease and treatment dynamics. Moreso, while our sample was drawn from Nigeria’s largest cancer center and likely represents the broader population of patients with MBC in the region, generalizability to other African countries or LMIC settings cannot be assumed. Cultural, linguistic, and healthcare system differences across LMICs may influence screening instrument performance in unpredictable ways. Moreover, our findings are specific to MBC and may not be generalized to earlier-stage disease or other cancer types without validation in those populations.

Sixth, our focus on depression as the primary outcome may not capture the full spectrum of psychological morbidity relevant to cancer care. The DT was designed as a broad screening tool for multiple forms of distress, and its performance in identifying anxiety, adjustment disorders, or existential distress was not evaluated in this study.

Clinical practice implications and actionable recommendations for LMIC oncology settings

The clinical implications of our findings are profound and immediate. Oncology providers in LMIC settings should exercise extreme caution when using the DT as a stand-alone screening tool for psychological morbidity, particularly depression. Our results suggest that reliance on the DT with standard cutoffs will result in missing the majority of patients with clinically significant depressive symptoms while potentially over-identifying patients without such symptoms.

Based on our findings, we propose a structured, multi-stage screening approach for psychological morbidity in LMIC cancer care settings. The first stage involves implementing universal brief screening using locally validated instruments administered to all patients at key clinical timepoints, including diagnosis, treatment initiation, treatment completion, and disease progression. Until culturally appropriate alternatives are developed, clinicians may use the DT cautiously as a conversation starter rather than a diagnostic tool, recognizing its profound limitations for identifying depression. The DT may retain utility for initiating discussions about sources of distress, particularly when used alongside the problem checklist to identify practical, physical, and social concerns that require attention regardless of depression status.

The second stage requires that all patients screening positive on brief tools, as well as those identified through clinical observation as potentially distressed, undergo semi-structured clinical interviews conducted by trained oncology staff. These interviews should focus on culturally relevant expressions of distress rather than relying exclusively on Western psychological terminology such as “depression” or “anxiety.” Specifically, interviewers should explore somatic complaints and physical symptom preoccupation, changes in social and family functioning, spiritual concerns and religious coping mechanisms, practical problems including financial hardship and transportation barriers, and disturbances in sleep and appetite patterns. This culturally grounded approach acknowledges that emotional suffering may manifest through physical, social, and spiritual channels in African contexts rather than solely through psychological symptomatology.

In the third stage, patients identified through clinical interview as requiring intervention should receive comprehensive psychological assessment using culturally adapted, multi-dimensional instruments. When feasible and culturally appropriate, collateral information from family members can provide valuable contextual understanding, given the collectivist orientation prevalent in many African societies where individual distress is understood within relational and communal frameworks. Referral to mental health specialists should occur when such resources are available, though task-shifting models utilizing trained nurses or social workers may be necessary in resource-limited settings where psychiatric expertise remains scarce.

Successful implementation of such multi-stage screening requires substantial investment in training and capacity building. Healthcare systems must train oncology nurses, social workers, and physicians to recognize culturally specific expressions of psychological distress, understand how somatization, spiritual attributions, and collectivist coping may manifest in African cancer patients, and conduct brief clinical interviews that explore distress through culturally relevant frameworks. Training should emphasize cultural humility and the importance of avoiding the imposition of Western psychological categories onto patient experiences. Staff must also develop competence in identifying patients requiring referral for specialized mental health care while recognizing the limitations of available resources.

Screening programs must be designed for feasibility within existing cancer care workflows to ensure sustainable implementation. This necessitates conducting brief screening at routine clinic visits rather than requiring separate appointments that burden already overwhelmed patients and healthcare systems. Screening documentation must be seamlessly integrated into existing medical records, with clear protocols established for positive screen follow-up and referral. Regular team meetings should be scheduled to review screening results, discuss complex cases, and ensure continuity of care across the oncology team.

Critical to the success of any screening program is the consideration of resource allocation and ethical implementation. Given the limited mental health infrastructure in many LMIC settings, screening programs must be coupled with realistic referral pathways and adequate intervention capacity. Implementing screening without available treatment resources raises profound ethical concerns and may paradoxically increase patient distress by identifying problems without offering solutions. Healthcare systems should therefore conduct comprehensive resource mapping to identify available mental health services, develop task-shifting models to expand intervention capacity beyond specialized psychiatry, train oncology staff in basic psychological support techniques, establish formal partnerships with psychiatric services and community mental health resources, and consider group interventions and peer support programs as cost-effective approaches to expanding treatment access. Only through such comprehensive planning can screening programs fulfill their promise of improving patient outcomes rather than merely documenting unmet needs.

Healthcare systems should also invest in training oncology staff to recognize cultural expressions of psychological distress that may not be captured by Western-developed screening tools. This includes understanding how patients from different cultural backgrounds may express emotional suffering through somatic complaints, spiritual concerns, or social disruption rather than direct psychological terminology.

The substantial discordance between DT and BDI-II results in our study highlights the non-interchangeable nature of different psychological constructs. Clinicians should recognize that “distress” and “depression” represent distinct albeit related phenomena, and that screening approaches should be tailored to the specific psychological outcomes of greatest clinical concern.

Health policy and global guidelines implications

Our findings have significant implications for international cancer control guidelines and health policy in LMICs. Currently, major organizations including the NCCN, ESMO, and WHO recommend routine distress screening using the DT with relatively uniform cutoff recommendations [35,36]. Our results challenge the evidence base underlying these recommendations and suggest that blanket adoption of Western-developed tools without local validation may be ineffective or potentially harmful.

We recommend that international cancer organizations fundamentally revise their psychosocial screening guidelines to explicitly acknowledge that instruments developed and validated in high-income Western populations may not perform adequately in LMIC or non-Western settings, thereby removing prescriptive recommendations for uniform tool adoption across diverse cultural contexts.

Guidelines should mandate local validation studies before implementing any screening program, providing clear frameworks that specify minimum methodological standards, recommended sample sizes, appropriate reference standards for different cultural contexts, and statistical approaches for determining optimal cutoffs. Rather than promoting reliance on single screening instruments, guidelines should emphasize multi-stage, multi-method approaches that combine brief screening tools with clinical interviews and culturally relevant assessment methods, particularly in LMIC settings where Western psychological frameworks may not apply. International organizations should actively support the development of regional and national psychosocial screening guidelines that reflect local cultural contexts, healthcare system capacities, and patient populations, while providing technical assistance and funding for guideline development processes in LMICs. Training recommendations must include specific guidance on cultural competence and recognition of culturally specific expressions of psychological distress.

Corresponding shifts in funding priorities are essential to support these guideline revisions. Research funding should prioritize indigenous tool development and validation studies in LMICs, community-based participatory research that engages local stakeholders in screening tool design, and implementation science examining the integration of culturally appropriate screening into existing cancer care workflows. International agencies should establish collaborative networks for psycho-oncology research across diverse cultural settings and develop standardized protocols for cross-cultural validation studies that can be adapted to local contexts while maintaining methodological rigor.

LMICs should consider developing regional or national guidelines for psychosocial screening that reflect local cultural contexts, healthcare system capacities, and patient populations. Such guidelines should emphasize flexible, multi-modal approaches to psychological assessment rather than reliance on any single screening instrument.

Research implications and future directions

This investigation opens several critical avenues for future research in psycho-oncology and global health. First, urgent efforts are needed to develop and validate culturally appropriate psychological screening tools for African cancer populations using community-based participatory research approaches. Such development should engage patients, families, survivors, and community members throughout the process, while including traditional healers and religious leaders who often serve as frontline mental health resources in these contexts. Oncology healthcare providers familiar with local contexts must be involved in identifying culturally relevant indicators of psychological distress through qualitative methods including focus groups and in-depth interviews. Development processes should explore preferred assessment formats ranging from numeric scales to pictorial representations and narrative approaches, testing comprehension and acceptability across diverse educational and linguistic groups to ensure broad applicability.

Second, future validation research must employ integrated mixed-methods designs that address the critical limitation identified in our study regarding the absence of cognitive debriefing. Such studies should incorporate cognitive interviewing to understand how patients interpret screening questions and whether their interpretation matches intended meaning, focus groups exploring cultural conceptualizations of psychological distress and barriers to reporting symptoms, ethnographic observation of patient-provider communication to identify conversational patterns and cultural norms, think-aloud protocols where patients verbalize their thought processes while completing instruments, and member checking to validate researcher interpretations with community stakeholders. These qualitative components are essential for understanding not merely whether screening tools work, but why they succeed or fail in specific cultural contexts.

Third, comparative effectiveness research should systematically evaluate different screening strategies, including single-instrument approaches, multi-stage screening models, and clinician-based assessment without formal instruments, and hybrid models combining cultural and Western screening methods. Outcomes should encompass detection rates for various forms of psychological morbidity, false-positive and false-negative rates, referral patterns and treatment initiation, cost-effectiveness and resource requirements, patient satisfaction and acceptability, and ultimate impact on quality of life and clinical outcomes. Longitudinal studies investigating the temporal stability of screening instrument performance across the cancer trajectory would provide crucial insights into optimal timing and frequency of psychological assessment, how screening performance varies with disease progression or treatment changes, and the natural history of psychological distress in LMIC cancer populations.

Fourth, the ultimate value of improved screening depends on the availability of effective interventions. Research must test whether enhanced psychological screening translates into improved clinical outcomes, quality of life, and patient satisfaction, while evaluating culturally adapted psychological interventions for African cancer populations. Studies should examine task-shifting models where non-specialists deliver mental health support and assess the feasibility and effectiveness of group-based and peer support interventions as scalable approaches.

Finally, implementation science research should examine broader healthcare system requirements including training needs and curricula for oncology staff, referral pathway development and optimization, integration strategies with existing cancer care workflows, barriers and facilitators to screening program implementation, and considerations for sustainability and scalability in resource-constrained settings.

Theoretical implications for psycho-oncology

Our findings contribute to broader theoretical discussions about the universality of psychological constructs and measurement instruments in psycho-oncology. The failure of a widely validated Western instrument in our African population raises fundamental questions about whether psychological distress represents a universal human experience that can be measured consistently across cultures, or whether distress is culturally constructed in ways that render cross-cultural measurement problematic.

The substantial discordance between DT and BDI-II results in our study, despite both instruments ostensibly measuring psychological morbidity, highlights the multidimensional and culturally contingent nature of emotional suffering. Rather than representing a unitary construct, “distress” appears to encompass multiple distinct phenomena including depression, anxiety, existential concerns, practical problems, social disruption, and spiritual crisis that may require different assessment approaches and may be differentially salient across cultural contexts.

Our findings align with anthropological critiques of psychiatric universalism that question whether diagnostic categories developed in Western contexts represent genuine natural kinds versus culture-bound constructs that emerge from specific historical, philosophical, and social frameworks. The poor correlation between DT and BDI-II in our Nigerian sample (r = 0.23) raises the possibility that these instruments, rather than measuring the same underlying construct with different degrees of precision, may actually be tapping into qualitatively different dimensions of experience that are structured differently across cultural contexts.

This has profound implications for the validity of cross-cultural research in psycho-oncology. Simply translating and administering instruments cross-culturally, even when accompanied by careful linguistic validation, may yield measurements of uncertain meaning if the underlying psychological constructs themselves are not equivalent across cultures. The concept of “measurement invariance”, the assumption that an instrument measures the same construct in the same way across groups, may not hold when comparing Western and non-Western populations, necessitating fundamental rethinking of how we assess psychological phenomena across cultures.

These findings support calls for more nuanced, culturally grounded approaches to psycho-oncology research and practice that recognize the diversity of human responses to cancer across different cultural contexts [37,38]. Rather than seeking to validate Western instruments across cultures with the goal of achieving measurement equivalence, psycho-oncology may need to embrace a more pluralistic approach that develops culture-specific frameworks and assessment methods while maintaining rigorous scientific standards.