The Hidden Toll of Occupational Health Hazards: A Qualitative Study of Workers from Small-Scale Illegal Miners in Ghana

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript presents findings from a qualitative study done among 16 miners (men and women) working in "illegal" small-scale gold mines in Ghana. The results illustrate and confirm the main adverse health consequences of these jobs - and their "fatalistic" perception by workers.  

The study appears to have been done correctly, the presentation and interpretation of the findings is fine, and the manuscript is well written.

The following comments are mainly intended to improve the paper.

Title:

• “invisible toll”? One could argue that the toll is in fact very visible or audible (as shown by the paper!), but it remains “hidden” to outsiders and the world.
• “small-scale illegal miners”: I suggest changing this to “workers from illegal small-scale gold mines” (this avoids stigmatizing the workers as being illegal!)

Abstract: specify “gold” mining, specify numbers of men and women, the range of ages and work experience, and the types of jobs.

Background

• Line 56 (and elsewhere): the term “alluvial” is surprising in view of the descriptions made of the work processes (especially excavation in underground pits), since alluvial gold is found mainly in riverbeds; this must be checked
• Line 59: it does not occur often that both mercury and cyanide are used on the same sites; did all 16 participants workers come from the same mine?
• If possible, representative photographs of the work processes in the studied mines would be useful and desirable (to increase the credibility of the statements).

Methods

• 4. Interviews: it is important to indicate how medical symptoms were elicited: were participants asked to describe their symptoms freely, were they asked to indicate whether they ever had had symptoms among a list of symptoms, or did they give answers to specific questions ? (e.g. “do you sometimes bring up phlegm from your chest?”, “have you ever had blood in your phlegm?”, “do your hands sometimes shake?”, “do you have problems remembering things?"); some of the quotes are remarkably consistent with “textbook” descriptions of respiratory or neurological conditions!
• 5. Exclusion criteria: it is unfortunate (and, hence, a limitation of the study) that no individuals from the legal mining sector were included because this would have allowed comparisons to be made between the illegal and licenced sectors, thereby providing specificity (or not) to the findings. [the last sentence (line111) is unnecessary]
• Line 163:what is “Member checking”?

Results

• • Line 187: “most were men” yes, but specify 10 men and 6 women (Table 1)
  • Table 1: unless this is critical (it is not discussed in the paper), the geographic origin of the participants doesn’t need to be specified (this will simplify the table)
  • Lines 262 and below: ore crushing done by ball mill crushers?
  • Line 292: washing/panning: panning does not lead to much dust exposure !

Discussion

• • The discussion is generally adequate
  • Some potential limitations indicated above (e.g. no “control” miners) should be acknowledged.

References

• • Look appropriate, except #41/42!

Author Response

Comment: "One could argue that the toll is in fact very visible or audible, but it remains “hidden” to outsiders."

Response: Agreed. The title has been revised to: "The Hidden Toll of Occupational Health Hazards: A Qualitative Study of Workers from Illegal Small-Scale Gold Mines in Ghana." "Hidden" more accurately captures the paper’s central argument that these health consequences are concealed from formal health systems and policy attention.

 

Comment R2-T2 (Title — Stigmatising Language)

Comment: "I suggest changing “small-scale illegal miners” to “workers from illegal small-scale gold mines”."

Response: Revised throughout the manuscript. Consistent language now positions workers as individuals employed in illegal operations, not as illegal persons themselves.

 

Comment R2-A1 (Abstract — Specify Gold, Demographics, Job Types)

Comment: "Specify “gold” mining; specify numbers of men and women, age range, work experience, and job types."

Response: All additions have been made. The abstract now reads: “A total of 16 purposively selected workers,  ten men and six women, aged 22 to 58 years, with experience ranging from 8 months to 23 years, working as excavators, washers/panners, chemical processors, ore grinders, and support workers, were interviewed using semistructured in-depth interviews.”

 

Comment R2-B1 (Background — “Alluvial” Terminology)

Comment: "The term “alluvial” is surprising in view of underground pit excavation; alluvial gold is found mainly in riverbeds."

Response: Corrected. The revised text now reads “both alluvial and hard rock (eluvial/residual) gold deposits,” with a clarifying note that surface alluvial deposits exist alongside hard rock primary deposits that require underground excavation.

 

Comment R2-B2 (Background — Mercury and Cyanide at Same Sites)

Comment: "It does not occur often that both mercury and cyanide are used on the same sites; did all 16 participants come from the same mine?"

Response: Clarified. Participants were recruited from three distinct communities (Pakyi No. 2, Toase, and Keniago). Mercury is used for amalgamation at some sites; cyanide is used for leaching at others. A clarifying sentence has been added explaining that both substances are practiced within the district across different sites and processing stages, not co-administered at a single workstation.

 

Comment R2-B3 (Background — Photographs)

Comment: "Representative photographs of the work processes would be useful."

Response: Acknowledged as a limitation. The illegal status of operations meant workers and gatekeepers were understandably unwilling to permit photography. A statement to this effect has been added to the Limitations section.

 

Comment R2-M1 (Methods — How Symptoms Were Elicited)

Comment: "It is important to indicate how medical symptoms were elicited."

Response: A new paragraph in Section 4 (Data Collection) explains that symptoms were elicited via open-ended narrative prompts, not a symptom checklist or leading questions. Research assistants were trained to probe only symptoms the participant had already spontaneously raised.

 

Comment R2-M2 (Methods — No Legal Mining Comparison)

Comment: "No individuals from the legal mining sector were included; this is a limitation."

Response: Acknowledged. A sentence has been added to the Exclusion Criteria acknowledging this limitation, and the Limitations section now explicitly discusses the absence of a comparison group and its implications.

 

Comment R2-M3 (Methods — Define Member Checking)

Comment: "What is “member checking”?"

Response: A definition has been added: “Member checking — a trustworthiness strategy in qualitative research whereby preliminary findings are returned to participants for verification and correction.”

 

Comment R2-R1 (Results — Men and Women Numbers)

Comment: "Specify 10 men and 6 women."

Response: Corrected. The line now reads “Ten participants were men and six were women.”

 

Comment R2-R2 (Table 1 — Geographic Origin)

Comment: "Geographic origin of participants doesn’t need to be specified."

Response: The column has been simplified to Yes/No for migrant status, with specific regions removed.

 

Comment R2-R3 (Results — Ball Mill Crushers)

Comment: "Ore crushing done by ball mill crushers?"

Response: Confirmed and specified. The revised text now reads: “Ore crushing was performed using ball mills (stamp mills), which generated particularly intense exposures to fine, respirable silica-containing dust.”

 

Comment R2-R4 (Results — Panning and Dust Exposure)

Comment: "Panning does not lead to much dust exposure."

Response: Acknowledged and clarified. The revised text now notes that respiratory symptoms reported by washer/panners were attributable to proximity to crushing and excavation operations, not to the panning activity itself, which is wet and generates minimal airborne dust.

 

Comment R2-D1 (Discussion — Acknowledge Limitations)

Comment: "Some potential limitations (e.g., no “control” miners) should be acknowledged."

Response: The Strengths and Limitations section has been substantially expanded. Five explicit limitations are now discussed: absence of a licensed mining comparison group; self-reported health conditions without clinical confirmation; inability to obtain photographs; possible under-disclosure due to the illegal context; and limited geographic scope.

 

Comment R2-Ref1 (References — References 41/42 Merged)

Comment: "References look appropriate, except #41/42."

Response: The merged reference has been corrected and separated into two properly formatted entries: Reference 41 (Stewart et al., 2020, Review of African Political Economy) and Reference 42 (Zafar et al., 2024, Mercury Toxicity Mitigation). All subsequent reference numbers have been renumbered accordingly.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The article is quite interesting, covering a topical issue of great importance to the mining sector.
There are, however, a few areas that could be improved:

 the sample size: the study presents 16 interviewees, which is insufficient to provide robust data for a critical analysis of the results;
no technical or scientific information is provided regarding the interview guide, nor is it stated whether it has been validated or adapted from validated references;
the methodology is not described in sufficient detail to understand the results and carry out a critical analysis of them in line with current best practice

Comments for author File: Comments.pdf

Author Response

Comment: "What kind of interview? Is it a scientifically validated tool? Should this information be provided?"

Response: We appreciate this observation. We have revised the abstract and methods section to explicitly state that semistructured, in-depth individual interviews were used as the primary data collection instrument. We have also clarified in the Methods (Section 4) that the interview guide was developed following an established framework for qualitative health research [Green & Thorogood, 2018; Smith, Flowers & Larkin, 2022], incorporating domains derived from a systematic review of occupational health literature on artisanal mining. Pilot testing with three eligible participants ensured face validity and cultural appropriateness prior to use. These details are now stated both in the abstract and in the data collection section.

 

Comment R1-2 (Abstract — Insufficient Detail in Each Section)

Comment: "I would try to provide more information on each of the points covered in your summary. It is on the basis of this that another researcher will decide whether or not to include it in their research."

Response: We have substantially revised the Abstract to provide richer detail in each section. Background now specifies the Ghanaian context and health burden. Methods now states the sample composition (10 men, 6 women), age range (22 to 58 years), experience range (8 months to 23 years), job types, and the nature of the interview instrument. Findings now describes each of the four themes with specific health conditions identified. Conclusions now include actionable policy recommendations. These additions give prospective readers sufficient information to assess the study’s relevance.

 

Comment R1-3 (Abstract — Inconsistent Capitalisation of Themes)

Comment: "Standardise the formatting. Some points begin with a capital letter and others with a lower-case letter."

Response: We have standardised the formatting of all four theme names in the abstract and throughout the manuscript. All themes now begin with a capital letter and follow a consistent syntactic structure. The revised list reads: (1) Normalisation of Risk and Fatalistic Beliefs; (2) Respiratory Manifestations and Dust-Related Ailments; (3) Chemical Exposures and Dermatological Conditions; (4) Traumatic Injuries and Musculoskeletal Disorders.

 

Comment R1-4 (Abstract and Methods — Sample Size of 16)

Comment: "In such a vast field, why does the sample consist of only 16 interviews? Statistically, this will not provide robust data for a critical analysis of the subject."

Response: We acknowledge this concern and recognise that it stems from an expectation more typical of quantitative research. This study is explicitly qualitative and phenomenological in design. In interpretative phenomenological analysis (IPA), which underpins this work, the aim is not statistical representation but depth of understanding of lived experience [Smith, Flowers & Larkin, 2022]. Established IPA guidelines recommend samples of between six and sixteen participants per study to allow sufficient analytical attention to each case [Smith & Osborn, 2015]. A sample of 16 workers is thus methodologically appropriate and consistent with published IPA studies in occupational health. Furthermore, data saturation was reached, with no new themes emerging after the 13th interview. We have added an explicit justification of the sample size in the Methods section and have also acknowledged, in the Limitations section, that the findings reflect depth rather than statistical generalisability.

 

Comment R1-5 (Background — Existing Literature, Statistics, and Contribution)

Comment: "Workplace accident rates in the extractive industry have been extensively studied worldwide. It would be helpful if you could include further information to justify the relevance of your article. Given the existing state of the art, what contribution does your study make, and why? What is the percentage of workplace accidents and occupational illnesses? What are the main existing reports on this subject?"

Response: We have significantly expanded the Background section to situate the study within the existing global and regional literature. We have added: (1) global statistics from the International Labour Organization (ILO) estimating that the mining sector accounts for approximately 8% of global fatal occupational accidents despite employing less than 1% of the global workforce; (2) regional data indicating that ASM in sub-Saharan Africa is associated with injury rates five to ten times higher than formal mining operations; (3) references to key existing reports including WHO and ILO documents on ASM health hazards; and (4) an explicit statement of the study’s contribution, namely, that while quantitative prevalence data on health conditions exist, the lived experiences, subjective perceptions, and structural barriers to care among workers in illegal small-scale gold mining in Ghana remain poorly documented, and that this study addresses that gap directly.

 

Comment R1-6 (Background/Methods — Mine Employment Numbers and Legalisation)

Comment: "How many workers are employed at these illegal mines? Why aren’t mines legalised?"

Response: We have added a paragraph to the Background section addressing both questions. The revised text notes that, while Ghana’s Small-Scale Gold Mining Law (PNDC Law 218) provides a legal framework for licensed artisanal mining, the licensing process is complex, expensive, and requires land ownership documentation that most informal workers cannot access. Political, economic, and bureaucratic barriers have historically impeded formalisation, as documented by Hilson and colleagues [2013]. Estimates from the Ghana Minerals Commission suggest that illegal galamsey operations directly engage between 500,000 and 1 million workers nationally, with tens of thousands operating in the Ashanti Region alone, though precise district-level figures are unavailable due to the clandestine nature of operations.

 

Comment R1-7 and R1-8 (Methods — Sample Size and Scientific Rigour)

Comment: "What is the sample size? What criteria were used to select the sample?" and "The sample size is too small to allow for a statistically robust critical analysis of the results."

Response: The sample size (n = 16) and selection criteria are now stated clearly in the Methods section. As noted in response to Comment R1-4, qualitative phenomenological research does not seek statistical robustness but rather theoretical saturation and depth of understanding. IPA studies of comparable scope in occupational health have used similar or smaller samples [Smith & Osborn, 2015; Larkin et al., 2019]. The sample was selected purposively to maximise variation across mining roles, experience levels, gender, and age, thereby strengthening the analytical range of the findings.

 

Comment R1-9 (Methods — Interview Guide Validation and Scientific Basis)

Comment: "Has the script used been validated? What is the reference for this script? Or is it your own adaptation of various interview scripts? If so, you must indicate the scientific basis from which the questions were taken."

Response: The interview guide was not a pre-existing validated instrument but was specifically developed for this study, as is standard practice in phenomenological qualitative research [Creswell & Poth, 2018]. It was constructed following a systematic review of the occupational health literature on artisanal and small-scale mining, in consultation with an occupational health specialist, and guided by IPA methodology [Smith, Flowers & Larkin, 2022]. The domains covered (daily work activities, perceived health risks, experienced symptoms, and health-seeking behaviours) are consistent with established frameworks for occupational health research. Pilot testing with three eligible workers confirmed clarity, cultural sensitivity, and relevance. We have elaborated this in the revised Data Collection section and added a new subsection titled “Interview Guide: Scientific Basis and Development.”

 

Comment R1-10 (Methods — Interview Questions Not Provided)

Comment: "The list of questions should be made available so that the content of the questions asked and their critical analysis can be understood. Your methodology should be described in such a way that it can be replicated by other researchers in similar contexts."

Response: We have added Appendix A to the revised manuscript, which presents the full semistructured interview guide, including opening questions, domain-specific probes, and follow-up prompts. This enables full methodological transparency and replicability. The domains and representative questions are also summarised in the revised Data Collection section.

 

Comment R1-11 (Results — Methodology and Presentation Order)

Comment: "In order to analyse the results, there must be a methodology that explains how they were obtained. This must be described in detail in the same order in which the results are presented."

Response: We have revised the Data Analysis section to explicitly map the thematic analysis process onto the results structure. A transitional paragraph at the start of the Results section now explains that the four themes emerged from the analysis and are presented in the order of their prominence and theoretical significance. The relationship between the analytical process (Braun and Clarke’s six-phase model) and the structure of the results is now made explicit.

Comment R1-12 (Results — Insufficient Participant Details)

Comment: "I cannot understand what you are trying to say here. No specific details are provided."

Response: We have substantially expanded the Participant Characteristics subsection to provide specific sociodemographic details: the precise numbers of men and women, the full age range, the range of mining experience, the range of educational attainment, migration status, and the full range of occupational roles represented. The revised section now gives a clear profile of the sample and contextualises the diversity of experiences reported in the thematic findings.

 

Comment R1-13 (Results — Statistical Data vs. Qualitative Excerpts)

Comment: "It would be more interesting if the data were analysed by theme, as is currently the case, but instead of presenting excerpts from the interviews, the statistical results from all the questionnaires should be included. I would also suggest increasing the number of interviews so that the analysis has some scientific rigour."

Response: We respectfully clarify that this study is phenomenological and qualitative in design; it did not use questionnaires or collect statistical data. In IPA research, verbatim interview excerpts are the primary evidence presented because they allow the reader to directly evaluate the interpretive claims made by the authors,  this is a methodological requirement, not a limitation [Smith, Flowers & Larkin, 2022; Braun & Clarke, 2006]. The excerpts are selected not for representativeness but for their exemplary quality and richness in illustrating the themes. We have added a methodological clarification at the start of the Results section that explains this approach, so that readers with a quantitative background can situate the evidence appropriately. We acknowledge, in the Limitations, that a mixed-methods study incorporating both qualitative and quantitative components would constitute a valuable future extension of this work.

 

Comment R1-14 (Conclusions — Underdeveloped Contributions and Literature Link)

Comment: "The conclusions could be developed further to highlight not only the main contributions of your article in light of the study you conducted, but also how they relate to the current state of the art."

Response: We have substantially expanded the Conclusions section. The revised text now: (1) explicitly states the four main findings and what each contributes to the occupational health literature; (2) situates these findings in relation to existing ASM health literature, noting specifically what this study adds (the subjective, lived dimension that prevalence studies cannot capture); (3) highlights the policy implications for Ghana’s illegal mining sector; and (4) identifies specific directions for future research, including longitudinal mixed-methods studies and biomonitoring programmes.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have addressed my comments satisfactorily. I commend the clarity of their responses and the way these were incorporated in the final manuscript.

Note: line 223 Results should be 3. Results and subsequent sections should be corrected accordingly (4. Discussion etc).

Author Response

 

Comment R2-1 (Section Numbering — Results and Subsequent Sections)

Comment: "The authors have addressed my comments satisfactorily. I commend the clarity of their responses and the way these were incorporated in the final manuscript. Note: line 223 Results should be 3. Results and subsequent sections should be corrected accordingly (4. Discussion etc)."

 

Response: We sincerely thank the Reviewer for the kind commendation and for the careful attention to the manuscript’s structure. We are pleased to confirm that this has been corrected. The section numbering has been revised throughout the manuscript as follows:

• Section 1: Background
• Section 2: Methods
• Section 3: Results (previously unnumbered/incorrectly numbered at line 223)
• Section 4: Discussion
• Section 5: Conclusions

All in-text cross-references to section numbers have also been updated accordingly to ensure internal consistency throughout the manuscript. We thank the Reviewer for identifying this oversight.

We trust that these revisions fully satisfy the concerns of both Reviewers. We believe the manuscript is now ready for publication and respectfully request favourable consideration. All changes are bolded in the revised manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

A significant effort has been made to improve the article.
However, I am concerned about the sample size and believe that, statistically speaking, we do not have sufficient information to conclude the study. I would suggest that you try to increase the number of interviews conducted

Author Response

Comment: "A significant effort has been made to improve the article. However, I am concerned about the sample size and believe that, statistically speaking, we do not have sufficient information to conclude the study. I would suggest that you try to increase the number of interviews conducted."

Response: We respectfully but firmly address this concern, as it reflects a misapplication of quantitative statistical standards to a study that is explicitly qualitative in design and purpose. This study employs Interpretative Phenomenological Analysis (IPA), a well-established qualitative methodology, which operates on fundamentally different epistemological and methodological principles from quantitative research.

Statistical concepts such as sample size adequacy, statistical robustness, and generalizability do not apply to qualitative phenomenological inquiry. In qualitative research, the guiding criterion for sample size is not statistical power or representativeness, but rather theoretical saturation, the point at which no new themes, categories, or meaningful insights emerge from additional data collection [Guest, Bunce & Johnson, 2006; Fusch & Ness, 2015].

As reported in our manuscript, data saturation was reached after the 13th interview, with the remaining three interviews conducted to confirm saturation, consistent with best practice in qualitative inquiry. Adding further interviews beyond the point of saturation would not strengthen analytical conclusions; it would, in fact, violate the very logic of purposive sampling and inductive analysis that underpins qualitative methodology (Braun & Clarke, 2006; Creswell & Poth, 2018).

In IPA specifically, published methodological guidance recommends samples of between six and sixteen participants per study to allow sufficient depth of analytical attention to each case (Smith, Flowers & Larkin, 2022; Smith & Osborn, 2015). Our sample of 16 workers is therefore not only methodologically appropriate, it is at the upper end of what IPA standards recommend. Numerous high impact qualitative studies in occupational health, published in peer reviewed journals including this one, have used comparable or smaller samples to reach robust and credible findings.

We also note that the Reviewer's earlier concerns regarding the sample size (Comments R1-4, R1-7, and R1-8 from the previous round of review) were thoroughly addressed in our prior response, with explicit references to the methodological literature, and those responses and revisions were incorporated in the manuscript. We respectfully submit that the sample size is both theoretically justified and methodologically sound, and that no additional interviews are warranted.

To further support the transparency of our approach, we have ensured that: (1) the rationale for the sample size and the use of saturation as the stopping criterion are clearly stated in the Methods section; (2) the Limitations section explicitly acknowledges that qualitative findings reflect depth of understanding rather than statistical generalizability; and (3) a recommendation for future mixed-methods or larger-scale quantitative studies has been included in the Conclusions as a direction for future research.

We trust this clarification satisfactorily addresses the Reviewer’s concern and respectfully ask that the manuscript be evaluated within the appropriate qualitative research paradigm.

 

Author Response File: Author Response.pdf