Applicability of infrared thermography in occupational health for multidisciplinary clinical decision-making: a systematic review with narrative synthesis

Abstract

Objectives

To identify and synthesize evidence on the applications of infrared thermography in the assessment of work-related conditions and to examine its contribution as a noninvasive, adjunctive, and complementary tool to support multidisciplinary clinical assessment and decision-making in occupational health.

Methods

We conducted a systematic review with narrative synthesis following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 statement; it was registered in the International Prospective Register of Systematic Reviews. Nine bibliographic databases were searched without date or language restrictions. Critical appraisal used Joanna Briggs Institute tools selected according to study design.

Results

Database searches identified 317 records, and 7 additional reports were identified through other methods. Forty-nine studies published from 1975 to 2025 were included. Infrared thermography characterized thermal patterns, monitored occupational exposures and physiological responses, documented changes following interventions or across assessed conditions, and contributed to exploratory screening or predictive models. Applications addressed peripheral vascular responses to cold and vibration, musculoskeletal loading, heat stress, fatigue and drowsiness, respiratory protection, and emerging computational technologies. Observational designs predominated, with heterogeneous protocols and several small or convenience samples. No included study was classified as a diagnostic accuracy study.

Conclusions

Overall, infrared thermography was used primarily as a noninvasive, adjunctive, and complementary imaging method to document skin-surface temperature distribution and thermal patterns in occupational contexts. When interpreted alongside occupational history and clinical, physiological, ergonomic, and environmental information, thermographic findings may contribute to multidisciplinary clinical assessment and decision-making. Methodological limitations and heterogeneity warrant caution when transferring findings across work-related conditions and occupational settings.

Key points  

• What is already known on this topic: Infrared thermography has been investigated in occupational health as a noninvasive, adjunctive, and complementary imaging method for documenting skin-temperature patterns and supporting the assessment of work-related conditions when integrated with clinical, physiological, ergonomic, and environmental information.

• What this study adds: This systematic review synthesizes 49 studies published from 1975 to 2025 and clarifies the main applications of infrared thermography, the scope of the findings, and the influence of study design, sample size, methodological critical appraisal, and protocol heterogeneity on their interpretation.

• How this study might affect research, practice, or policy: The review may inform the cautious integration of thermographic findings into multidisciplinary clinical assessment and decision-making in occupational health, while emphasizing the need for protocols and study designs suited to each intended application.

1. Introduction

Work-related conditions represent a major challenge for occupational health and require approaches capable of supporting their identification, prevention, and follow-up. These conditions include musculoskeletal disorders, injuries associated with repetitive exertion, extreme environmental exposures, and thermal strain, with consequences for workers’ health and for production and health systems.¹ In this context, noninvasive technologies have been investigated as additional resources for occupational assessment.^2,3 Infrared thermography (IRT) makes it possible to record and represent the distribution of body-surface temperature in real time,^4,5 providing thermal information that can be interpreted alongside clinical, physiological, ergonomic, and environmental data.

Skin temperature and thermographic patterns may vary according to physiological, environmental, and technical factors, including peripheral circulation, muscle activity, environmental conditions, participant preparation, and image-acquisition and analysis procedures.^3–7 For this reason, IRT has been investigated in different areas of health and human performance,^8–10 predominantly as an adjunctive and complementary tool. Its potential contribution lies in documenting thermal asymmetries and variations, monitoring responses to exposures or interventions, and supporting the interpretation of other data, without replacing clinical examination, established diagnostic methods, or relevant environmental and ergonomic assessments.

In occupational health, the available studies encompass different populations, exposures, conditions, purposes of use, and thermographic protocols. This diversity makes it difficult to obtain an integrated understanding of the applications investigated, the scope of the findings, and the methodological limitations of the evidence produced. Accordingly, this systematic review aimed to identify and synthesize the available scientific evidence on the applications of IRT in the assessment of work-related conditions and to examine its usefulness as an adjunctive and complementary tool supporting multidisciplinary clinical assessment and decision-making in occupational health.

2. Methods

2.1. Study design

This was a systematic review of the literature with narrative synthesis and without meta-analysis. This design was adopted to identify, critically appraise, and synthesize evidence on the applicability of IRT in the assessment of work-related conditions and in supporting multidisciplinary clinical assessment and decision-making in occupational health. Meta-analysis was not performed due to clinical, methodological, and statistical heterogeneity among the included studies, which encompassed different occupational populations, work settings, conditions investigated, thermographic acquisition protocols, anatomical regions, study designs, and outcomes.

The review followed the methodological guidance established by the Joanna Briggs Institute (JBI),¹¹ which informed all stages of the process: formulation of the research question, definition of eligibility criteria, searching, data extraction, critical appraisal, and synthesis of findings. The review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines,¹² using the corresponding checklist and flow diagram to ensure transparency, rigor, and reproducibility. The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) in 2024 under registration number CRD42024563096.

2.2. Study location and period

The database searches were conducted between September 2024 and December 2025 by researchers based in Rio Grande and Pelotas, Rio Grande do Sul, Brazil, and Montevideo, Uruguay.

2.3. Eligibility criteria

The PICO framework was used to formulate the research question for this systematic review, as follows: (P) Patient/Population/Problem of interest (workers); (I) Intervention/Area of interest (infrared thermography); (C) Comparison (not applicable because of the nature of the investigation); and (O) Outcome (support for multidisciplinary clinical assessment and decision-making in occupational health). Based on these elements, the following research question was formulated: What are the applications of IRT in the assessment of work-related conditions, and what evidence supports its use as an adjunctive and complementary tool in multidisciplinary clinical assessment and decision-making in occupational health?

The inclusion criteria encompassed studies that used IRT in an occupational health context, involving different professional groups and addressing the assessment of the risk of work-related conditions, including diagnostic-accuracy studies, randomized clinical trials, case–control studies, and observational studies.

No restrictions were imposed on publication period or language to maximize search sensitivity. Theoretical studies, publications outside the thematic scope, and studies that did not clearly describe the use of thermography in relation to work-related conditions, exposures, or occupationally associated circumstances were excluded. Reports whose full text could not be obtained after retrieval attempts were recorded separately as not retrieved in the PRISMA flow diagram.

2.4. Information sources and search strategies

The 9 bibliographic databases were PubMed, CINAHL, Embase, the Cochrane Library, PEDro, PsycINFO, Scopus, Web of Science, and the Virtual Health Library. They were selected to broaden coverage across medicine, nursing, physiotherapy, occupational therapy, psychology, safety engineering, and public health. The Cochrane Library, PEDro, and PsycINFO yielded no eligible studies but are reported for transparency. Additional reports identified through other methods were assessed using the same eligibility criteria. A librarian experienced in systematic reviews contributed to strategy development, and targeted tests were conducted to assess search sensitivity.

The search strategy combined controlled vocabulary and synonyms using the Boolean operator “OR” within concepts and “AND” between the concepts of thermography, occupational groups, and work-related conditions or exposures. It was developed in PubMed using Medical Subject Headings and Emtree terms, and adapted to each database’s indexing system and syntax. Complete strategies for the 9 bibliographic databases are provided in Table S1.

2.5. Study selection

The bibliographic search results were exported to Rayyan® (Qatar Computing Research Institute, online version), which was used to identify and remove duplicate references. The studies then underwent initial screening by 2 independent reviewers working in a blinded manner, based on titles and abstracts and the predefined eligibility criteria. Studies that met the criteria at this stage were subsequently assessed through full-text review. When the reviewers disagreed on the inclusion or exclusion of a study, a third reviewer was consulted to reach a consensus decision.

The electronic database search was conducted on September 26, 2024, with 2 subsequent updates on March 11, 2025, and December 4, 2025, to include more recent publications and ensure the breadth and currency of the findings.

2.6. Data extraction and analysis

Data from the selected studies were extracted using a previously structured form designed to ensure standardization and consistency of the information collected. The form included: author, year of publication, and country; authors’ professional field; objective; methodological design; workers and context; number of participants; condition, exposure, or circumstance assessed; application of thermography; reported contribution to occupational or multidisciplinary assessment; thermographic technique used; and main findings. Based on this information, the studies underwent narrative synthesis to identify patterns of application, occupational contexts, conditions assessed, possibilities for complementary use, gaps, and relevant contributions.

In response to the heterogeneity observed across thermographic applications, an additional descriptive characterization of the available information on image-acquisition and analysis protocols was conducted. Four general domains were examined in full texts and, when indicated by the authors, in explicitly referenced protocols, standards, or methodological publications: environmental conditions and participant preparation; equipment and acquisition parameters; anatomical region and timing of assessments; and image processing and analysis. This characterization was used to describe variability and transparency in technical reporting, without constituting a new risk-of-bias assessment or assuming that procedures not described had not been performed.

In the narrative synthesis, the contribution of each study was interpreted by jointly considering study design, sample size, presence of comparison groups, the purpose assigned to thermography, and identified methodological limitations. Studies with very small samples, including case reports, case series, and exploratory or quasi-experimental investigations, were retained because they documented relevant clinical, physiological, ergonomic, or technological applications. However, their findings were given predominantly descriptive, exploratory, or feasibility-related weight and were not used independently to support population-level generalizations, claims regarding diagnostic accuracy, or conclusions about intervention effectiveness. Repeated images, anatomical regions, or measurement occasions were not treated as independent participants.

The methodological quality of the included primary studies was assessed using the JBI critical appraisal tools corresponding to each study design. Item responses were recorded as yes, no, unclear, or not applicable, and disagreements between the 2 reviewers were resolved by consensus or, when necessary, by a third researcher. Because the JBI tools do not establish universal cutoffs for classifying methodological quality, interpretation was not based exclusively on summed scores. The domains relevant to each design were considered, including participant selection, measurement of exposure and outcomes, identification and control of confounding, presence of comparison groups, and appropriateness of the analysis. Studies with substantial limitations or insufficient reporting were retained when they contributed historical, clinical, or contextual information, but were given less interpretive weight.

2.7. Ethical considerations

This systematic review used only secondary data from previously published and publicly available studies. Therefore, approval by a Research Ethics Committee was not required. The search, selection, extraction, critical appraisal, and synthesis stages were conducted rigorously and with faithful reporting of the findings, in accordance with good practices in research integrity. Generative artificial intelligence tools (ChatGPT, based on GPT-5.5 Thinking by OpenAI) were used to support improvements in textual clarity, paragraph organization, and grammatical revision in Portuguese and English, as well as the schematic organization of Figure 2. Grammarly Pro, version 1.2.277.1921, was also used for grammar checking. These tools were not used to formulate the review question, conduct the search, select or extract studies, perform the methodological appraisal, analyze the data, or interpret the scientific findings. All suggestions were reviewed and approved by the authors, who assume full responsibility for the final content of the manuscript.

Figure 2.

Summary of the applications of infrared thermography in occupational health. Infrared thermography was used predominantly as a noninvasive, adjunctive, and complementary imaging method integrated with other information relevant to occupational assessment. The blocks represent application purposes rather than hierarchical levels of evidence.

3. Results

3.1. General characteristics of the study set

The database searches identified 317 records, distributed as follows: PubMed (n = 42), CINAHL (n = 8), Scopus (n = 73), Web of Science (n = 99), Embase (n = 41), and the Virtual Health Library (n = 54). After removal of 159 duplicates, 158 records underwent title and abstract screening, of which 41 were excluded. A total of 117 reports were sought for retrieval; despite the attempts made, 24 full texts could not be obtained. Of the 93 reports assessed for eligibility from the database pathway, 51 were excluded because of an ineligible population (n = 16), inappropriate study design (n = 16), publication type (n = 13), or absence of a description of IRT (n = 6), resulting in 42 included studies. Seven additional reports identified through other methods were retrieved, assessed, and included, bringing the total to 49 studies in the review. Figure 1 presents the selection process according to the PRISMA 2020 model.¹²

Figure 1.

PRISMA 2020 flow diagram of study identification, screening, eligibility assessment, and inclusion.

The 49 studies showed broad geographical distribution. Brazil accounted for 15 publications (31%), followed by the United States (n = 5; 10%), Italy (n = 4; 8%), Japan (n = 3; 6%), and Norway (n = 3; 6%). The United Kingdom, Sweden, Portugal, India, Mexico, and Canada contributed 2 studies each (4% each). Australia, Belgium, Finland, Romania, Spain, and Malaysia, as well as 1 multinational study involving Iran, Finland, and Japan, contributed 1 study each (2% each). The studies included in the systematic review were published between 1975 and 2025, with a greater concentration from 2016 onward. Five studies were published in 2020,^13–17 followed by other concentrations in 2024^18–20 and in 2025.^21–26 Earlier clusters occurred in 2017,^27–31 2018,^32–35 2019,^36–38 and 2016^39–41 (Table 1).

Table 1.

Characteristics of the 49 studies included in the systematic review.

Study, year, and country	Study design	Workers/occupational context, and sample	Condition, exposure, or outcome assessed	IRT application and main findings	JBI appraisal a
Sakaguchi S et al, 2008, 42  Japan	Cross-sectional	Forestry workers; n = 80 (38 exposed workers and 42 controls; 159 hands analyzed)	Hand-arm vibration syndrome	Post-warming dorsal-hand thermography generated a quantitative thermal-distribution index that discriminated exposed workers from controls in this sample; the authors proposed it as an objective adjunct for assessing vibration-related peripheral circulatory impairment	6/8
Soranso DR et al, 2022, 43  Brazil	Cross-sectional	Furniture industry workers; n = 9	Musculoskeletal overload of the lumbar spine and shoulder girdle	Lumbar and scapular thermography before and after shifts following different recovery intervals showed no significant between-condition differences. The authors considered IRT potentially useful only alongside other ergonomic methods	6/8
Yano and Kinjo, 2020, 13  Japan	Case report	Construction worker; n = 1	Hand-arm vibration syndrome with Raynaud phenomenon	Lower right fingertip temperatures were interpreted with clinical examination and angiography in a case of hand-arm vibration syndrome. IRT added noninvasive thermal information but did not determine diagnosis or management	7/8
Amaro AM et al, 2019, 36  Portugal	Quasi-experimental	Recreational tennis players; n = 6	Upper-limb musculoskeletal overload associated with racket vibration	Upper-limb thermography before and after standardized tennis serves with different grips and devices showed contralateral thermal differences in 6 participants; the findings were exploratory and were interpreted as possible signs of injury risk	8/9
Cabrera IN et al, 2001, 44  United States	Case report	Professional clarinetist with performance-related pain; n = 1	Right upper-limb pain related to musical performance	Posterior upper-limb images obtained before, after pain-limited clarinet playing, and after recovery documented asymmetries corresponding to painful regions; IRT was presented as an objective adjunct, not a first-line diagnostic procedure	5/8
Ramos L et al, 2020, 14  Brazil	Cross-sectional	Office workers; n = 24	Injury or symptoms in the wrist and finger-extensor region among workers performing typing tasks	Bilateral wrist and finger-extensor thermography before and during a 10-minute typing task showed between-group and side differences, but no significant temperature change over time; IRT was proposed as a complementary assessment method	6/8
Bourlai T et al, 2012, 45  United States	Cross-sectional	Wildland firefighters; n = 6	Heat stress assessed through thermal facial biometrics	High-resolution facial thermal imaging during precooling, treadmill exertion, and recovery was compared with ingestible-capsule core temperature; estimates were most accurate during exertion in the heated environment	6/8
Nag PK et al, 2013, 46  India	Cross-sectional	Foundry, pottery, and quarry workers; n = 1947	Occupational heat stress	Regional thermography estimated mean skin temperature alongside core temperature and environmental measures across heat-exposed industries, characterizing thermal responses and reduced heat-dissipation capacity at more intense exposure	6/8
Fioravanti A et al, 1995, 47  Italy	Cross-sectional	Textile-industry workers; 104 workers examined; 28 with hand and wrist complaints underwent ultrasonography and thermography	Degenerative osteoarticular lesions and carpal tunnel syndrome	Among 28 symptomatic workers, hand thermography and ultrasonography complemented clinical assessment of tendon and joint abnormalities and carpal tunnel syndrome; 19 received a clinical diagnosis of carpal tunnel syndrome	4/8
Williams DMJ et al, 1977, 48  United Kingdom	Cross-sectional	PVC industry workers exposed to vinyl chloride; n = 199	Acro-osteolysis and Raynaud phenomenon related to PVC exposure	Pre- and post-cold-provocation hand thermography showed little group differentiation in vinyl chloride–exposed workers and controls; the authors considered only a possible complementary role	5/8
Castro P et al, 2020, 15  Brazil	Cross-sectional	Household waste collectors; n = 9	Musculoskeletal symptoms affecting the limbs and spine	Serial thermography during waste collection showed end-of-shift changes corresponding to reported symptoms, especially in the lumbar and thoracic regions. Organizational recommendations were based on the combined thermographic and questionnaire assessment	5/8
Guertler C et al, 2016, 39  Brazil	Cross-sectional	Oyster-farm workers; n = 10	Multiple occupational risks in oyster production, including musculoskeletal overload, solar radiation, and electrical hazards	Pre- and post-task thermography recorded shoulder and back temperature changes and complemented observation, interviews, postural assessment, and environmental measurements; prevention recommendations were based on the multimethod assessment	5/8
Hirata H, 1984, 49  Japan	Cross-sectional	Railway-system workers; n = 2680	Hand-arm vibration syndrome	Pre- and post-cold-immersion hand thermography was associated with season and clinically confirmed vibration disease, performing better in colder periods; its usefulness depended on testing conditions	8/8
Holm CA et al, 2016, 40  United States	Cross-sectional	Male participants performing hot work in thermal protective clothing; n = 10	Physiological heat strain during recovery from hot work	Facial and skin thermal measurements, together with heart rate, were evaluated against gastrointestinal temperature for calculation of the Physiological Strain Index. Agreement was insufficient during recovery from mild-to-moderate heat strain, although it improved at higher core temperatures	8/8
Amaravathi T et al, 2016, 41  India	Cross-sectional	Female seafood-processing workers; n = 140	Occupational cold exposure, respiratory and dermatological symptoms, and reduced skin temperature of the hands and feet	Finger and toe thermography before and during seafood-processing work documented significant skin-temperature reductions; together with other assessments, it characterized cold-related occupational health problems	6/8
Gold JE et al, 2009, 50  United States	Cross-sectional	Office workers with and without upper-extremity musculoskeletal disorders; n = 45	Upper-extremity musculoskeletal disorders	Pre- and post-typing dorsal-hand thermography at 3 ambient temperatures showed that lower pre-typing temperature at 18°C-22°C was associated with greater disorder severity; post-typing temperature did not differentiate severity	8/8
Stankovic SJ et al, 2011, 51  Belgium	Cross-sectional	Vibration-exposed workers and an unexposed comparison group; n = 45	Vibration-induced white finger	Dynamic hand thermography after cold provocation differentiated symptomatic exposed workers, asymptomatic exposed workers, and controls through rewarming parameters, providing complementary vascular information	8/8
Norheim AJ et al, 2018, 32  Norway	Cross-sectional	Healthy young Norwegian Army recruits; n = 255	Variability in peripheral rewarming response after cold provocation	Dynamic hand thermography after standardized cold provocation showed wide variation in digital rewarming among healthy recruits and identified slow, intermediate, and rapid patterns	8/8
Tirloni AS et al, 2018, 33  Brazil	Cross-sectional	Poultry-slaughterhouse workers; n = 143	Occupational cold exposure of the hands and wrists	Finger thermography and thermal-sensation interviews in 143 poultry-slaughterhouse workers showed that 66.4% had at least one finger at ≤15 °C and 99.3% at ≤24 °C; multiple overlapping gloves were insufficient to maintain finger temperatures.	8/8
Estevam FNL et al, 2019, 38  Brazil	Cross-sectional	Agricultural walking-tractor operator; n = 1	Heat stress	Serial infrared surface-temperature measurements were analyzed with environmental variables and a discomfort index. Heat-stress conditions were identified, but surface temperature did not correlate with discomfort	4/8
Tirloni AS et al, 2017, 27  Brazil	Cross-sectional	Swine-slaughterhouse workers; n = 106	Cold exposure of the hands	Hand thermography and thermal-sensation assessment in 106 swine-slaughterhouse workers showed lower temperatures in the product-handling hand and greater cold sensation among knife users; the findings supported the need for adequate glove insulation	8/8
Brändström H et al, 2008, 52  Sweden	Longitudinal study with pre- and post-training assessments	Men participating in prolonged military training in a cold climate; n = 77 at baseline; 45 reassessed after 15 months	Hand rewarming response and occurrence of local cold injury during military training	Dominant-hand thermography monitored 30-minute rewarming after cold-water immersion before and after 15 months of cold-climate training. Cold injury was more frequent among initial slow rewarmers; selected thermal parameters improved after training	10/11
Takeda F et al, 2018, 34  Brazil	Cross-sectional	Refrigerated-slaughterhouse workers; n = 20	Occupational cold exposure of the hands and wrists	Hand thermography was used to relate skin temperatures to environmental conditions and tasks in refrigerated sectors, visualizing extremity cooling and supporting discussion of thermal protection and workplace conditions	6/8
Tirloni AS et al, 2024, 18  Brazil	Cross-sectional	Poultry-slaughterhouse workers; n = 571	Combined cold exposure and musculoskeletal overload	Thermography of the hands, wrists, and forearms, combined with symptom and cold-perception questionnaires, identified sectors and activities associated with greater thermal burden and musculoskeletal complaints	6/8
Takeda F et al, 2017, 28  Brazil	Cross-sectional	Refrigerated-slaughterhouse workers; n = 19	Occupational cold exposure of the hands	Controlled hand thermography documented low digital and wrist temperatures, particularly in product-handling fingers, contributing to characterization of discomfort and pain risk in refrigerated work	6/8
Baldini A et al, 2015, 53  Italy	Quasi-experimental	Airforce pilots with bruxism and temporomandibular disorders; n = 11	Temporomandibular disorders related to occupational stress	Masticatory-muscle thermography with and without temporary occlusal-splint use showed small temperature increases interpreted as compatible with relaxation; thermal symmetry changed little	8/9
Chapman D et al, 2023, 54  Australia	Quasi-experimental/test method	Healthcare professionals attending a respirator fit-testing clinic; n = 48	Inadequate respirator fit and leakage risk	Infrared images of the thermal gradient across flat-fold P2 respirators during inhalation were analyzed using machine-learning models to predict quantitative fit-test pass/fail outcomes. The models showed high predictive accuracy in this sample, although further validation across other respirator styles was recommended	8/9
Cosoroaba MR et al, 2019, 37  Romania	Quasi-experimental	Dentists; n = 4	Musculoskeletal overload of the cervical and lumbar spine and upper limbs	Thermography of cervical, triceps, and lumbar regions in 4 dentists showed lower surface temperatures after ergonomically guided posture; larger samples are needed	8/9
Enander A et al, 1980, 55  Sweden	Quasi-experimental	Cold-exposed workers and office workers; n = 20	Thermal response to cold and risk of cold injury	Palmar thermography before and after cold-water immersion showed no significant differences in cooling or rewarming between cold-exposed and office workers; differences were mainly subjective	8/9
Lasanen R et al, 2018, 35  Finland	Quasi-experimental crossover	Office workers; n = 14	Neck and low-back discomfort under different workstation configurations	Cervical and upper-back thermography, surface electromyography, and disability ratings compared traditional and upright sitting postures. Upright posture reduced upper-back temperature variation, muscle activity, and neck-disability scores	8/9
Soares AL et al, 2020, 16  Brazil	Quasi-experimental	Members of an academic community performing simulated repetitive tasks; n = 32	Risk of upper-limb musculoskeletal disorders	Forearm thermography during simulated repetitive tasks across 5 OCRA risk levels supported development of a predictive equation; the small sample, simulated setting, and model limitations require validation	8/9
Vitale E et al, 2024, 19  Italy	Quasi-experimental	Agricultural tractor operators using an assisted-driving system; n = 10	Musculoskeletal overload of the neck, shoulders, and lumbar spine	Pre- and post-operation thermography, electromyography, and discomfort ratings were used to compare display positions during tractor operation. Integrated findings supported ergonomic recommendations; IRT was not used alone	8/9
Brioschi ML et al, 2012, 56  Brazil	Quasi-experimental/exploratory	Two workers in an indoor industrial environment; n = 2	Changes in skin temperature during work and after hand exposure to cold water	Continuous infrared imaging during work, compared with thermocouples and a cold-water hand-immersion experiment, demonstrated noncontact skin-temperature monitoring in 2 workers; findings were exploratory	8/9
Cázares-Manríquez MA et al, 2021, 57  Mexico	Analytical cross-sectional/predictive modeling	Maquiladora workers performing manual tasks; n = 33	Upper-limb musculoskeletal disorders	Upper-limb thermography, the Strain Index, and worker variables were modeled to predict cumulative trauma disorders across 2 company areas. High in-sample fit requires validation in larger, external settings	8/8
Clemente MP et al, 2020, 17  Portugal	Cross-sectional	Wind- and string-instrument musicians; n = 77	Musculoskeletal overload of the cranio-cervico-mandibular complex	Cranio-cervico-mandibular thermography identified group differences and asymmetries in wind- and string-instrument musicians; findings support adjunctive assessment but not stand-alone diagnosis or causal inference	6/8
House R et al, 2017, 29  Canada	Case–control	Workers with the vascular component of hand-arm vibration syndrome and controls; n = 85	Hand-arm vibration syndrome, vascular component	Post-cold-provocation hand and foot thermography showed some case–control differences, but sensitivity and specificity were insufficient for use in isolation	9/10
Youakim S, 2010, 58  Canada	Case–control	Workers with hand-arm vibration syndrome and healthy controls; n = 30	Hand-arm vibration syndrome with Raynaud phenomenon	Serial infrared finger thermometry after cold-water immersion differentiated confirmed vibration-related Raynaud phenomenon from controls in this sample; use was proposed as complementary and dependent on clinical context	8/10
Cherniack M et al, 2008, 59  United States	Observational study involving 2 occupational cohorts with repeated assessments	Shipyard workers and dental hygienists; baseline: 308 (214 and 94); follow-up: 201 (135 and 66)	Carpal tunnel syndrome case definitions and neurosensory measures in vibration-exposed occupational cohorts	Skin temperature was assessed by point thermistor and regional thermal imaging under controlled conditions during sensory nerve-conduction and vibrotactile testing. IRT was not evaluated as a diagnostic test; the principal findings concerned the limited agreement and temporal stability of the CTS case definitions	5/11
Falappa P et al, 1982, 60  Italy	Case series	Vinyl chloride–exposed workers with vascular disease; n = 5	Vinyl chloride disease, including acro-osteolysis and Raynaud phenomenon	Thermography and other noninvasive vascular methods in 5 vinyl chloride–exposed workers showed substantial agreement with angiography; IRT contributed to complementary vascular assessment	7/10
Norheim AJ et al, 2017, 30  Norway	Case report	Worker with chronic sequelae of hand frostbite; n = 1	Frostbite sequelae, pain, and vasospasm	Dynamic thermography and quantitative sensory testing before and after botulinum-toxin treatment in 1 frostbite case indicated improved perfusion, cold hypersensitivity, and pain; further long-term study was recommended	8/8
Stewart JD et al, 1975, 61  United Kingdom	Case report	PVC-industry worker with acro-osteolysis; n = 1	Acro-osteolysis and peripheral vasculopathy associated with vinyl chloride	Thermography complemented clinical, radiographic, and angiographic investigation of vinyl chloride–associated acro-osteolysis in 1 worker, contributing descriptive evidence on vascular and bone abnormalities	7/8
Martins AB et al, 2025, 21  Brazil	Cross-sectional	Forest-harvester operators engaged in mechanized timber harvesting; n = 7	Musculoskeletal overload of the hands and wrists due to repetitive movements and continuous use of control levers	Serial palmar and carpal thermography across 6 shifts in 7 forest-harvester operators showed thermal changes, especially in the palms and wrists; larger longitudinal studies and ergonomic-intervention evaluation were recommended	5/8
Cázares-Manríquez MA et al, 2025, 22  Mexico	Cross-sectional	CNC machine operators in the metalworking industry; n = 36	Work-related upper-limb musculoskeletal disorders, including tendinopathies, suspected carpal tunnel syndrome, and other disorders	Dorsal and palmar hand thermography at rest and during recovery was integrated with clinical, physiological, and occupational data in regression models. Models identified suspicious thermal changes, but small sample size and variable fit limit generalizability	7/8
Chau PKT et al, 2025, 23  Norway	Cross-sectional	Seafood handlers working in a cold environment and administrative workers from the same plant; n = 44	Occupational cold exposure of the hands, including white fingers, pain and/or numbness, skin changes, and inflammatory markers	Repeated hand thermography before and after breaks, alongside environmental, symptom, and biomarker data, showed task-, shift-, and break-related differences in cold exposure and thermal recovery	7/8
Moreno FC et al, 2025, 24  Spain	Cross-sectional	Vessel traffic service operators in maritime control centers; n = 23	Fatigue and mental workload associated with shift work in a high-responsibility environment requiring continuous vigilance	Facial thermography, subjective fatigue and workload scales, and cognitive tests during actual vessel-traffic shifts demonstrated feasibility for integrated fatigue assessment; replication in larger, multicenter samples is needed	6/8
Aghamalizadeh A et al, 2024, 20  Iran/Finland/Japan	Quasi-experimental/experimental driving-simulator study	Adult drivers assessed in a driving simulator under different levels of drowsiness; n = 20	Driver drowsiness and fatigue risk in driving activities, with implications for occupational accidents and road safety	Facial thermography during simulated driving supported machine-learning classification of alert versus drowsy states, with poorer multilevel performance; real-world occupational implementation was not assessed	8/9
Dias NF et al, 2025, 26  Brazil	Cross-sectional	Swine-slaughterhouse workers; n = 40	Occupational cold exposure of the extremities (hands) and thermal discomfort	Pre- and post-break hand thermography showed finger thermal recovery after breaks of at least 10 minutes, with effects influenced by outdoor temperature; findings inform the assessment of break duration and thermal conditions	5/8
Nomura EY et al, 2017, 31  Brazil	Cross-sectional	Asymptomatic workers; n = 306	Parameters of facial thermal symmetry in carotid-branch territories	Facial thermography in asymptomatic workers provided descriptive temperature and symmetry values across carotid territories; screening, diagnostic, and referral performance were not evaluated	5/8
Osman B et al, 2025, 25  Malaysia	Cross-sectional	National-institute personnel, including trainers, administrative staff, and cleaners with different activity levels; n = 216	Thermal discomfort and changes in the facial microclimate associated with respirator use, with possible implications for adherence to personal protective equipment	Serial facial thermography after 4 and 8 hours of KF94, surgical, and cloth mask use showed increased facial skin temperature after 8 hours for all 3 types. Changes in penetration and breathing resistance were most evident for surgical masks worn by cleaners	6/8

Abbreviations: CNC, computer numerical control; CTS, carpal tunnel syndrome; IRT, infrared thermography; OCRA, Occupational Repetitive Actions; PVC, polyvinyl chloride.

^aThe Joanna Briggs Institute (JBI) critical appraisal tools were selected according to the methodological classification adopted in this review and the design characteristics confirmed through full-text assessment. Table 1 provides a summary of the critical appraisal results, whereas item-level judgments are presented in Tables S2-S7, organized by study design and corresponding JBI checklist. The 2 longitudinal cohort studies were appraised using the JBI checklist for cohort studies and are presented in Table S6.

Table 1 presents, for each included study, the study design, occupational context and sample, condition or exposure assessed, main thermographic application and findings, and a summary of the critical appraisal conducted using the JBI checklist appropriate to the study design. Item-level judgments (Yes, No, Unclear, or Not applicable) are presented in Tables S2–S7.

3.2. Study designs and occupational contexts

Cross-sectional studies predominated in the sample, totaling 30 publications (61.2%) across industry, refrigerated workplaces, services, the armed forces, office environments, and other occupational contexts.^{14,15,17,18,21–28,31–34,38–43,45–51,57} Ten studies (20.4%) were classified as quasi-experimental or related crossover, exploratory, modeling, simulator-based, or method-evaluation designs and assessed tasks, interventions, posture changes, equipment, work organization, or experimental conditions without randomized allocation.^{16,19,20,35–37,53–56} Four case reports (8.2%) contributed exploratory and descriptive accounts of individual work-related clinical situations,^13,30,44,61 and 1 case series (2.0%) included 5 workers with vinyl chloride disease and severe Raynaud phenomenon.⁶⁰ Two case–control studies (4.1%) compared groups with and without vascular manifestations related to vibration.^29,58 Two longitudinal cohort studies (4.1%) comprised a study with assessments before and after military training in a cold climate,⁵² and a study involving 2 vibration-exposed occupational cohorts assessed at 2 time points, in which thermal imaging was integrated into regional skin-temperature measurement during neurosensory testing.⁵⁹

The occupational contexts investigated included forestry, railway, and construction workers exposed to vibration^{13,29,42,49,51,58,59}; workers in refrigerated plants and slaughterhouses exposed to intense cold^{18,23,26–28,33,34,55}; waste collectors and mariculture workers^15,39; agricultural tractor and microtractor operators^19,38; women working in seafood processing⁴¹; industrial operators at risk of repetitive strain injuries^22,57; office workers and typists^14,35,50,56; healthcare professionals undergoing respirator fit and seal testing⁵⁴; musicians and pilots^17,44,53; dentists³⁷; and military personnel exposed to cold climates.^30,32,52 The review also included forest harvester operators,²¹ vessel traffic service operators,²⁴ drivers in a driving simulator,²⁰ and workers who wore masks for prolonged shifts, with assessment of the facial microclimate.²⁵

Some included studies used very small sample sizes, particularly case reports, case series, and exploratory or quasi-experimental studies. Their findings contributed primarily to documenting the technical feasibility of IRT, characterizing thermal patterns, or describing responses under specific conditions. Accordingly, these studies were interpreted separately from larger observational and comparative investigations and were not considered sources of population-level estimates or conclusive evidence of diagnostic performance or intervention effectiveness.

3.3. Applications of IRT in the assessment of work-related health risks

IRT was used in different occupational contexts to characterize and identify thermal alterations, including asymmetries and variations associated with neuropathies, vasculopathies, and musculoskeletal disorders.^{13,14,16,17,22,29,42,47,50,51,56,57,59} Another recurrent use of IRT involved monitoring thermal exposure and workload. In studies of cold and heat, IRT was applied to record segmental temperatures across the workday, under different environmental conditions and, in some cases, in comparison with reference methods for core temperature.^{32,38,40,45,46,52,55} These studies documented, for example, very low finger temperatures among workers in refrigerated plants, frequently associated with discomfort and pain complaints,^18,27,28,34 and were complemented by recent investigations in cold environments and work routines involving breaks or shift work,^23,26 as well as operational tasks with repeated measurements during work.^14,21,24,35

In studies focused on musculoskeletal conditions, IRT was used to record thermal differences across body regions, tasks, postures, and assessment time points, and to monitor changes after interventions such as stretching, postural changes, use of an occlusal splint, or ergonomic adjustments.^{15,17,19,35–37,39,43,44,53,56} The direction and magnitude of thermal variations differed across studies and were interpreted alongside symptoms, muscle activity, posture, and other assessments reported in those studies.

Some studies incorporated thermographic measurements into regression models or computational algorithms for risk classification, prediction of discomfort or musculoskeletal injury, and identification of failures in protective barriers.^{16,20,22,54,57} These applications generally remained exploratory or were linked to specific development and validation contexts. A synthesis of the identified purposes, their contributions, and their respective interpretive limits is presented in Table 2.

Table 2.

Synthesis of the purposes of infrared thermography use and the interpretive scope of the included evidence.

Purpose of IRT	Applications/contexts represented	Predominant nature of evidence	Contribution to occupational health	Limits of interpretation
Peripheral vascular characterization and monitoring of cold or vibration exposure	Thermal distribution of the hands, asymmetries, responses to cold tests, and rewarming curves among workers exposed to cold, vibration, or vinyl chloride13,23,26–30,32–34,42,48–49,51–52,55,58,60–61	Observational and comparative studies, including a longitudinal cohort study, supplemented by case reports and a case series	Characterizes peripheral thermal responses and may complement clinical, vascular, functional, and exposure assessment	Heterogeneity in cold tests, rewarming times, and thermal criteria limits direct comparisons and restricts diagnostic inferences based on thermography alone
Identification of thermal patterns associated with musculoskeletal workload and ergonomics	Localized variations, asymmetries, and thermal changes in the upper limbs, spine, shoulder girdle, and cranio-cervico-mandibular complex during tasks, postures, or work shifts14–17,19,21,22,35–40,43,44,50,53,56,57	Cross-sectional and quasi-experimental studies, frequently with repeated measures and small samples	Documents thermal patterns associated with the conditions assessed and may complement ergonomic, functional, and symptom-based analysis	The findings represent associations or changes under the conditions studied; alone, they do not demonstrate causality, clinical effectiveness, or diagnostic validity
Monitoring of thermal stress, physiological load, fatigue, and drowsiness	Body or facial skin temperature associated with heat exposure, exercise, recovery, fatigue, mental workload, or drowsiness20,24,28,41,45,46	Observational and exploratory evidence, including real and simulated environments	Indicates potential for monitoring physiological responses and supporting identification of situations requiring additional assessment	Surface temperature is sensitive to the environment, recovery, fluid intake, and other factors and should not automatically be interpreted as a measure of core temperature or work incapacity
Monitoring responses to clinical, functional, or ergonomic interventions	Assessments before and after botulinum toxin, occlusal splints, postural changes, devices, equipment configurations, or other strategies19,30,35–40,53,55,56	Case reports and quasi-experimental studies, generally without randomization and with few participants	Enables documentation of within-participant thermal changes over the course of an intervention or across assessed conditions	Thermal change alone does not establish therapeutic or ergonomic effectiveness; small samples and absence of parallel groups limit generalizability
Screening, modeling, and emerging technological applications	Regression models, machine-learning algorithms, respirator-seal assessment, and classification or prioritization systems16,20,22,54,57	Quasi-experimental method evaluation, cross-sectional predictive modeling, initial validation, and exploratory applications	Explores the integration of thermographic parameters into screening, classification, and decision-support tools	External validation, calibration, reproducibility, and performance in independent populations remain limited
Physiological and methodological support for interpretation of other tests	Regional assessment of skin temperature as part of a neurosensory protocol in vibration-exposed occupational cohorts59	Observational study involving 2 occupational cohorts with repeated assessments	Contributes to documenting and controlling a physiological condition relevant to interpretation of nerve-conduction and vibrotactile-perception tests	Thermal imaging was part of the assessment protocol; the study did not aim to estimate the technique’s specific diagnostic performance

Abbreviation: IRT, infrared thermography.

The table organizes the purposes identified in the included studies and the interpretive scope adopted in the narrative synthesis. The categories do not constitute a new eligibility classification and do not imply equivalence among studies with different designs and sample sizes.

The descriptive characterization of protocols showed variation in both acquisition conditions and the extent of technical reporting. Studies of cold exposure and vascular alterations frequently used serial assessments, thermal provocation tests, and analysis of cooling or rewarming curves. Musculoskeletal and ergonomic studies predominantly compared body sides, times during the workday, tasks, postures, or intervention conditions. Environmental conditions and participant preparation were described directly in some articles and, in others, by reference to previously published protocols, standards, or methods. The studies also varied in the body regions examined and the metrics used, including absolute temperatures, asymmetries, temporal variations, recovery curves, and predictive models. This variability contributed to methodological heterogeneity and limited direct comparison of thermal values, without interpreting the lack of detail as evidence that a given procedure had not been adopted (Table 3).

Table 3.

Aggregate characterization of reporting on thermographic protocols in the included studies.

Domain examined	Information considered	Synthesis observed across studies	Implication for the synthesis
Environmental conditions and participant preparation	Ambient temperature and humidity, acclimatization/thermal equilibration, and prior controls, when reported	Directly described in some articles; in others, referenced through earlier protocols, standards, or methods; extent of reporting varied	Limits direct comparison across acquisition conditions, without allowing an inference that unreported controls were not adopted
Equipment and acquisition parameters	Camera or system used, settings, emissivity, distance, and positioning, when reported	Level of detail varied according to period, purpose, and reporting practices	Makes it difficult to establish technical equivalence across studies and compare absolute temperature values
Body region and timing	Regions of interest, static or dynamic measurements, and timing in relation to exposure, task, or intervention	Generally defined according to the condition, exposure, or response investigated	Explains part of the contextual and methodological heterogeneity and guides contextual interpretation of findings
Image processing and analysis	Absolute temperatures, asymmetries, temporal changes, recovery curves, and analytical models	Wide diversity of metrics, thresholds, and analytical procedures	Precludes direct pooling of quantitative results and supports the appropriateness of narrative synthesis
Methodological reference	Standard, protocol, or methodological publication cited to supplement the procedural description	Identified in some studies and considered in the characterization	Avoids interpreting omission of repeated details in the article as absence of procedural standardization

Descriptive synthesis based on the full texts and, when explicitly indicated by the authors, on cited protocols, standards, or methods. Failure to locate an element was not interpreted as evidence that the procedure had not been performed, and the table does not constitute an additional risk-of-bias assessment tool.

3.4. Contribution of IRT to occupational and multidisciplinary assessment

Across the included studies, IRT was used in different ways as a complementary resource alongside other occupational health assessment methods. In vascular disorders and neuropathies associated with occupational exposure to vibration and vinyl chloride, it was used alongside clinical, radiological, vascular, or neurophysiological examinations to characterize hand temperature distribution, asymmetries, and rewarming patterns after cold tests.^{13,29,42,47–49,51,58–61} These findings added thermal information to the assessment of workers with hand-arm vibration syndrome, Raynaud phenomenon, or acro-osteolysis, without replacing the other methods used in the studies.

IRT was used in settings involving therapeutic or rehabilitation interventions to document changes in thermal parameters before and after treatment. Similarly, in vasospastic disorders resulting from occupational cold injuries, IRT enabled monitoring of changes in digital perfusion after botulinum toxin treatment.³⁰ In studies of musculoskeletal disorders, thermal imaging was used to compare temperature patterns before and after stretching, posture changes, use of an occlusal splint, or other management strategies.^{17,19,36,37,44,53,56} In these contexts, the authors used IRT to illustrate reductions or redistributions of temperature in overloaded regions and interpreted these findings as compatible with possible beneficial effects of the interventions evaluated.

Thermographic findings also informed ergonomic and organizational recommendations aimed at preventing work-related health problems. Findings of very low finger temperatures and discomfort among workers in refrigerated plants and cold rooms were used by the authors to support the need to improve gloves and environmental conditions and to organize breaks and cold-protection strategies.^{18,27,28,33,34,38} Similarly, studies involving waste collectors, mariculture workers, agricultural workers, and other sectors used thermal information to support suggestions for equipment adjustments, task redistribution, or additional protective measures.^{15,16,19,35,39,41,43} In these cases, from the authors’ perspective, IRT contributed to proposals for complementary interventions aimed at reducing exposure to thermal and biomechanical risks. Other studies in healthcare settings explored IRT as a tool to support training activities and surveillance of protective practices.⁵⁴ IRT was also used to assess the seal of P2 respirators; when coupled with machine-learning algorithms, it identified leaks around the equipment and classified fits as adequate or inadequate.⁵⁴ In these investigations, the authors indicated the potential of IRT to support training, feedback, and audit processes related to respiratory-protection practices in workplaces.

Taken together, this evidence indicates that IRT was used predominantly as a noninvasive, adjunctive, and complementary tool. Its contribution was to add thermal information to the clinical, physiological, ergonomic, environmental, or organizational assessments conducted in the studies; findings should therefore be interpreted in conjunction with the other methods used and the occupational context investigated.

3.5. Methodological quality of the studies

The critical appraisal showed methodological heterogeneity both across designs and within each study group. In cross-sectional studies, the most frequent limitations involved identification and control of confounders, participant selection, and, in some cases, insufficient description of measurement procedures. Case reports and the case series contributed specific clinical and occupational descriptions, but their scope remained predominantly descriptive because of the small number of participants and absence of comparison groups. The case–control studies^29,58 provided relevant comparisons but also presented limitations related to control selection and adjustment for confounders. The 2 longitudinal cohort studies^52,59 enabled assessment of changes over time but presented limitations related to participant follow-up and control of confounding. In Cherniack et al,⁵⁹ thermal imaging formed part of the neurosensory assessment protocol through regional measurement of skin temperature, without formal evaluation of the diagnostic performance of IRT. Across all designs, sample size was considered separately from completeness of the JBI items because favorable checklist responses do not remove limitations in precision and external validity arising from small samples. The synthesis by design, including limitations that guided the interpretive role of the studies, is presented in Table 4.

Table 4.

Synthesis of study designs, main methodological limitations, and the interpretive role of the included studies.

Study design/description	n	Contribution to the review	Limitations guiding interpretation	Role in the narrative synthesis
Analytical cross-sectional studies	30	Characterization of thermal patterns, exposures, and associations in different occupational contexts	Convenience sampling and small samples in some studies; identification and control of confounders frequently incomplete; diversity of thermographic protocols	Predominantly associative and descriptive evidence; useful for characterization and comparison, without supporting causality or isolated diagnostic performance
Quasi-experimental and related crossover, exploratory, modeling, simulator-based, or method-evaluation studies	10	Assessments before and after tasks, interventions, postural adjustments, equipment, or simulated conditions, and evaluation of infrared imaging during respirator fit testing	Absence of randomization and, in several studies, of a parallel comparison group; small samples, specific settings or equipment, and limited external validity	Evidence on thermal changes under the assessed conditions and feasibility of monitoring or method application; cautious interpretation regarding effectiveness and generalizability
Case–control studies	2	Comparison of thermal responses between workers with and without vascular manifestations related to vibration	Potential limitations in control selection and control of confounding factors	Comparative evidence for complementary use of IRT, with limitations for diagnostic inferences based exclusively on IRT
Case reports	4	Detailed description of specific clinical and occupational applications and individual follow-up	Single participant, absence of comparison, and variable information across reports	Descriptive evidence; illustrates possible applications without permitting generalization
Case series	1	Description of vascular alterations in a small group of exposed workers	Small number of participants, absence of a comparison group, and limitations inherent to the design	Descriptive and contextual evidence
Longitudinal cohort studies	2	Assessment of changes over time in hand cold-recovery responses among military personnel undergoing training in a cold climate and in clinical and neurosensory outcomes among vibration-exposed occupational cohorts, with regional skin-temperature measurement by thermal imaging	Attrition or losses to follow-up between assessments; limitations in the control of confounding; absence of an untrained concurrent comparison group in one study; and use of thermal imaging as a component of the neurosensory assessment protocol in the other	Longitudinal evidence on changes over time and on the complementary use of thermal measurement in physiological and neurosensory assessment, without supporting conclusions regarding the isolated diagnostic performance of IRT

Abbreviation: IRT, infrared thermography; JBI, Joanna Briggs Institute.

This synthesis complements, but does not replace, the item-by-item appraisal presented in Tables S2-S7. The 2 longitudinal cohort studies are presented in Table S6 and were appraised using the JBI checklist for cohort studies. Sample size, study design, and study-specific limitations were considered jointly when interpreting the evidence.

Figure 2 provides an integrated synthesis of the main purposes for which IRT was applied across the included studies.

4. Discussion

The 49 included studies, published between 1975 and 2025, show a broadening over time in the occupational health applications investigated using IRT. Earlier studies focused mainly on peripheral vascular conditions related to vibration, cold, and vinyl chloride exposure, particularly Raynaud phenomenon, hand-arm vibration syndrome, and perfusion changes in the extremities.^{29,42,47–49,51,52,58,60,61} From 2016 onward, and more markedly between 2020 and 2025, the range of investigated applications expanded to include musculoskeletal loading, heat stress, occupational fatigue, monitoring of ergonomic interventions, and approaches involving modeling and algorithms.^{14–26,35–40,43,53–57} Within the body of evidence reviewed, this shift should not be interpreted as uniform consolidation of the technique across all settings; rather, it reflects technological and conceptual expansion of its use in diverse occupational contexts.

Consistent with the narrative synthesis, the discussion is organized into four thematic domains reflecting the main clusters of work-related conditions and IRT applications identified in the included studies. Within each domain, the findings were interpreted in relation to the purpose assigned to IRT, the occupational context, and the methodological characteristics of the contributing studies. The domain-specific discussion was followed by an integrated synthesis of the design-specific JBI appraisal and its implications for interpretation, and by consideration of the implications for occupational health and the limitations of the review.

4.1. Peripheral vascular conditions, occupational vibration, and cold exposure

Peripheral vascular conditions and cold exposure accounted for one of the largest groups of studies and reported findings physiologically related to thermal responses in the extremities. This domain included cross-sectional studies with varying methodological completeness, case–control studies that enabled group comparisons, 1 longitudinal cohort study with repeated assessments, and case reports or a case series with predominantly descriptive contributions. The cross-sectional studies included Sakaguchi et al,⁴² Hirata,⁴⁹ Stankovic et al,⁵¹ Norheim et al,³² Tirloni et al,^27,33 and Chau et al.²³ The longitudinal cohort study by Brändström et al⁵² assessed hand cold-recovery responses before and after military training in a cold climate. The case–control studies by House et al²⁹ and Youakim⁵⁸ added comparisons between cases and controls. Case reports and case series described specific clinical situations but offered limited generalizability.^13,30,60,61

In studies with comparative designs, larger samples, or more clearly described protocols, IRT was used mainly to assess hand cooling and rewarming patterns after cold provocation, thermal asymmetry, and thermal findings interpreted in relation to digital perfusion in workers exposed to vibration, occupational cold, or vinyl chloride. Hirata,⁴⁹ Stankovic et al,⁵¹ and House et al,²⁹ for example, presented comparative data on thermal patterns associated with hand-arm vibration syndrome and Raynaud phenomenon, contributing to understanding of the complementary use of IRT in these settings. Similarly, studies conducted in refrigerated workplaces and cold environments, including Tirloni et al,^27,33 Norheim et al,³² and Chau et al,²³ used IRT to monitor extremity temperature and relate the findings to exposure conditions, rest breaks, and protective equipment. Even within this domain, however, interpretation must account for differences in sampling, control of confounding, and thermographic protocols.

Other studies broadened the description of these applications but had limitations related to participant selection, control of confounding, sample size, or completeness of reporting, including Takeda et al,^28,34 Dias et al,²⁶ and Williams et al.⁴⁸ The longitudinal cohort study by Brändström et al⁵² contributed evidence on hand cold-recovery responses over time but was limited by attrition between assessments and the absence of an untrained concurrent comparison group.

Peripheral vascular conditions and cold exposure were therefore recurring IRT applications in the reviewed evidence, with predominantly complementary use in occupational assessment. The contribution in these settings should be understood in conjunction with clinical examination, occupational history, functional testing, and assessment of working conditions.

4.2. Musculoskeletal loading and work-related disorders

The musculoskeletal domain was both the largest and the most heterogeneous. It included cross-sectional studies with differing methodological limitations, such as Ramos et al,¹⁴ Castro et al,¹⁵ Martins et al,²¹ Clemente et al,¹⁷ Guertler et al,³⁹ Soranso et al,⁴³ Fioravanti et al,⁴⁷ Gold et al,⁵⁰ and Cázares-Manríquez et al,⁵⁷ the latter incorporating predictive modeling based on thermographic and worker-related variables. Quasi-experimental studies included Soares et al,¹⁶ Lasanen et al,³⁵ Amaro et al,³⁶ Cosoroaba et al,³⁷ Vitale et al,¹⁹ and Baldini et al.⁵³ Amaro et al³⁶ examined a sports and recreational context, and the transferability of its findings to occupational settings should therefore be interpreted cautiously. As a case report, Cabrera et al⁴⁴ made a descriptive contribution to the discussion of localized musculoskeletal pain related to professional activity.

In cross-sectional studies, IRT was used to characterize localized increases in skin temperature or thermal asymmetry in body regions subjected to functional demand, including the upper limbs, cervical spine, shoulder girdle, and lumbar region. These findings indicate the potential of IRT to complement the assessment of body segments subjected to greater functional demands; however, interpretation was often limited by the absence of multivariable analyses, small sample sizes, and the cross-sectional nature of the observations. This was observed, for example, in Ramos et al,¹⁴ Castro et al,¹⁵ Soranso et al,⁴³ and Martins et al.²¹ Gold et al,⁵⁰ despite greater methodological completeness, also reported modest findings regarding the contribution of IRT as a complementary tool in the assessment of upper-extremity musculoskeletal disorders.

The quasi-experimental studies provided evidence on thermal changes observed before and after interventions or across compared conditions. Soares et al¹⁶ related thermal patterns to Occupational Repetitive Actions (OCRA) -classified risk levels; Lasanen et al³⁵ examined the effects of working posture on cervical and scapular temperature; Cosoroaba et al³⁷ compared thermal patterns across dental working postures; and Vitale et al¹⁹ assessed agricultural cab configurations. These studies indicate the potential complementary contribution of IRT to ergonomic assessment and biomechanical monitoring but do not, on their own, establish causality or the effectiveness of the assessed interventions or conditions, particularly given the small samples and the absence of randomization or parallel control groups in some investigations.

The most cautious interpretation for this domain is therefore that IRT may be useful for documenting thermal patterns associated with musculoskeletal loading and thermal changes following interventions or across assessed conditions, particularly when combined with electromyography, self-report measures, ergonomic assessment, and clinical examination. Methodological heterogeneity and protocol variability nevertheless limit direct comparisons and the formulation of generalizable recommendations across occupational populations and settings.

4.3. Heat stress, occupational heat exposure, fatigue, and physiological monitoring

Studies addressing occupational heat exposure, heat stress, drowsiness, physiological strain, and fatigue were fewer in number and investigated distinct objectives using different methods. Cross-sectional investigations included Bourlai et al,⁴⁵ Nag et al,⁴⁶ Holm et al,⁴⁰ Estevam et al,³⁸ Moreno et al,²⁴ and Osman et al.²⁵ Among the quasi-experimental studies, Aghamalizadeh et al²⁰ explored facial thermal patterns associated with drowsiness in a driving simulator.

Within this domain, Holm et al⁴⁰ indicated that IRT may contribute to physiological monitoring while also demonstrating limitations in measurement precision related to environmental influences and the post-exertion recovery period. Nag et al⁴⁶ and Bourlai et al⁴⁵ indicated the potential complementary use of IRT under conditions of intense heat exposure. Moreno et al²⁴ and Aghamalizadeh et al²⁰ extended this field by relating facial thermal patterns to fatigue and drowsiness. However, the specific settings, differences among protocols, and, in some studies, small samples limit comparability and the direct transferability of the findings to broader occupational programs.

The reviewed evidence therefore suggests a potential role for IRT in monitoring heat exposure, fatigue, and physiological responses to work. Although several studies were conducted in occupational settings, this remains an exploratory field and requires greater protocol standardization, further evaluation across diverse workplaces and routine working conditions, and integration with established physiological, environmental, and occupational measures.

4.4. Personal protective equipment, respiratory protection, and emerging modeling and algorithm-based applications

A fourth domain comprised emerging IRT applications related to respiratory protection, facial personal protective equipment, predictive modeling, and screening-support systems. Rather than representing an entirely separate set of occupational conditions, these studies extended some of the applications discussed above through more recent technological and analytical approaches, including thermal pattern recognition, classification algorithms, respirator-seal assessment, and facial microclimate monitoring. This group included Chapman et al,⁵⁴ Osman et al,²⁵ Aghamalizadeh et al,²⁰ and the studies by Cázares-Manríquez et al.^22,57

In respiratory protection, Chapman et al⁵⁴ developed and evaluated the use of IRT as a complementary visual resource for identifying air leakage during respirator fit testing, without proposing that it replace standardized fit-testing methods. In a cross-sectional study, Osman et al²⁵ investigated facial microclimate and thermal discomfort associated with prolonged mask use. These findings may be relevant to the assessment of mask tolerability and working conditions during prolonged use, but should be interpreted as contextual and exploratory.

In modeling and algorithm-based applications, Aghamalizadeh et al²⁰ combined facial IRT with machine-learning methods to classify drowsiness states in a driving simulator. The simulated setting limits the direct transferability of these findings to routine occupational settings. The studies by Cázares-Manríquez et al^22,57 incorporated thermographic measurements into regression models related to the screening or prediction of musculoskeletal conditions among industrial workers. Their findings indicate the potential contribution of IRT to screening-support approaches when combined with ergonomic, individual, and functional variables. However, these applications remain exploratory and require external validation, assessment of reproducibility and model performance, and evaluation across different occupational populations and working conditions.

These studies reflect the technological expansion of IRT in occupational health, particularly through the integration of thermal imaging with algorithms, predictive modeling, and screening- or monitoring-support tools. Technological innovation, however, should not be interpreted as established evidence of clinical or occupational effectiveness. Most of these applications remain exploratory and are based on specific samples, controlled or simulated settings, or limited validation. In this domain, IRT was investigated as a complementary technology for expanding the assessment of functional and thermal responses and contributing to screening, training, or monitoring processes alongside relevant clinical, ergonomic, environmental, or standardized methods.

Although thermography has also been used in contexts adjacent to occupational health for the initial identification of individuals with elevated body temperature in hospitals, airports, and other public or institutional settings, none of the included studies examined fever detection or infection surveillance as a primary occupational health outcome. Literature outside the included evidence indicates that system performance varies according to the device, facial region assessed, environmental conditions, procedural standardization, and integration with broader screening and risk-management measures.^62–71 This application was therefore acknowledged only as an adjacent context and was not incorporated into the thematic synthesis of the findings from the included studies.

4.5. Critical synthesis of the methodological appraisal and implications for interpretation

The integrated appraisal using the JBI tools showed a heterogeneous body of evidence in terms of study designs, samples, protocols, and methodological completeness. Interpretation was based on design-specific methodological domains rather than on thresholds derived solely from summed item scores. Cross-sectional studies often adequately described exposures and outcomes but remained subject to limitations involving participant selection, identification and control of confounding, and, in many cases, small samples. Quasi-experimental and related studies provided evidence on thermal changes across tasks, interventions, or assessed conditions, as well as on the technical application of IRT during respirator fit testing, but were limited by the absence of randomization or parallel comparison groups and by restricted external validity. The 2 longitudinal cohort studies contributed evidence on changes over time, although participant attrition and control of confounding limited interpretation. The case–control studies added comparative evidence on vibration-related vascular manifestations, whereas case reports and the case series broadened the description of specific clinical and occupational applications without supporting broad inference. Across designs, thermography was used predominantly as an adjunctive resource integrated with clinical, physiological, ergonomic, environmental, or standardized assessment methods.

Studies with larger samples or comparative and observational designs focused mainly on cold exposure, peripheral vascular responses, and musculoskeletal loading, although differences in populations, thermographic protocols, and analytical approaches limited direct comparison of their findings. Studies with small sample sizes or substantial methodological limitations contributed to the documentation of specific applications, thermal patterns, and feasibility under particular occupational or clinical conditions. Their findings were therefore interpreted primarily as descriptive or exploratory evidence and did not support generalized conclusions regarding effectiveness or the routine incorporation of IRT into occupational health protocols.

Comparability was also influenced by the diversity of thermographic protocols and variability in the extent of technical reporting. To examine this source of heterogeneity, the descriptive synthesis considered environmental conditions; acclimatization and participant preparation; camera settings, emissivity, acquisition distance, body positioning, and operator-related procedures, when reported; the body regions examined; the timing of assessments in relation to recent physical activity, occupational exposure, tasks, interventions, or circadian period; and image-processing and analytical procedures. Some studies reported these elements directly, whereas others referred to standards or previously published protocols or methods. When a procedural element was not described in the primary report or in a methodological source explicitly cited by the authors, this was treated as incomplete reporting that limited direct comparison and independent replication, rather than as evidence that the procedure had not been performed. This structured characterization allowed protocol heterogeneity to be discussed systematically without converting reporting completeness into an additional risk-of-bias assessment.

4.6. Implications for occupational health and review limitations

Taken together, the included evidence supports the use of IRT as a complementary source of information on skin-temperature patterns in occupational assessment. Depending on the study purpose, thermographic findings were integrated with clinical examination, functional or neurofunctional testing, ergonomic analysis, environmental characterization, and information provided by workers. This complementary role was more consistently represented in studies of occupational cold, vibration exposure, peripheral vascular responses, and musculoskeletal loading than in emerging applications involving fatigue, drowsiness, predictive modeling, or computational classification. The findings therefore do not support a broad claim of overall effectiveness or the standalone use of IRT for diagnosis, surveillance, or occupational decision-making.

Systematic and scoping reviews covering clinical, sports, and rehabilitation applications have similarly described the potential of IRT to monitor thermal responses associated with physical load and, in some contexts, mental workload during exertion and recovery.^3,8,62,63 Methodological reviews of human thermography, however, emphasize that recorded skin temperature is influenced by multiple individual, environmental, and technical factors and that reproducible interpretation requires clearly described and standardized image-acquisition and analysis procedures.^3,64 Technical documents in medical thermology and thermography may guide specific aspects of acquisition and interpretation. In contrast, thermal-environment ergonomics standards such as ISO 7243 address heat-stress assessment through the wet bulb globe temperature (WBGT) index and should be understood as complementary references for characterizing environmental exposure rather than as standards for thermographic image acquisition.^65–68 International guidance on occupational heat exposure also emphasizes monitoring environmental conditions and protecting workers through measures involving rest breaks, clothing, hydration, and protective equipment.¹

The contribution of thermography should therefore remain limited to its specific role within the assessment processes examined. Occupational health encompasses working conditions and work organization, occupational exposures, health outcomes, and interdisciplinary actions related to prevention, surveillance, and care. Within this broader field, IRT may add information on skin-temperature distribution and variation when used contextually and interpreted together with clinical, physiological, neurofunctional, ergonomic, environmental, and worker-reported information. Thermographic changes should not be interpreted in isolation as direct evidence of altered core temperature or perfusion, diagnosis, causal relationships, or intervention effectiveness unless these properties have been specifically evaluated and supported by the study design.

Limitations in the primary studies require caution in interpreting the review findings. Diversity in thermographic equipment, acquisition protocols, body regions, timing of measurements, analytical procedures, and technical reporting limited direct comparison across studies. The predominance of observational designs, variable identification and control of confounding, and the absence of randomization or concurrent comparison groups in several investigations also restricted causal and effectiveness inferences. In addition, several studies used small or convenience samples, reducing statistical precision, external validity, and the transferability of findings across occupational populations and settings. These limitations do not negate the descriptive, comparative, and exploratory contributions of the studies, but they restrict the strength and scope of the conclusions that can be drawn from them.

At the review level, the substantial heterogeneity of study designs, occupational contexts, thermographic procedures, outcomes, and analytical approaches precluded meaningful quantitative pooling and required a narrative synthesis. Despite retrieval attempts, 24 reports selected for full-text assessment could not be obtained, and their eligibility could therefore not be confirmed. The nonretrieval of these reports may have reduced the breadth of the evidence base, although the direction and magnitude of any resulting effect cannot be determined. The review consequently provides a structured characterization of the available evidence rather than pooled estimates of diagnostic performance, intervention effectiveness, or exposure-related effects.

Future research may benefit from thermographic protocols that clearly report environmental conditions, participant preparation and acclimatization, equipment and camera settings, emissivity, acquisition distance and positioning, anatomical regions of interest, timing in relation to exposure or task, and image-processing and analytical procedures. Samples, study designs, comparison groups, and analytical strategies should be aligned with the specific research objective. For established applications, studies may examine repeatability, reproducibility, validity, and the additional contribution of IRT beyond clinical, physiological, ergonomic, and environmental assessments. Emerging applications involving predictive models or computational algorithms require calibration, external validation, and assessment in independent occupational populations and routine working conditions. Research should also examine how thermographic information can be incorporated into occupational-health assessment and surveillance processes without replacing established methods or extending interpretation beyond the evidence supported by each application.

5. Conclusions

This systematic review identified applications of IRT across occupational health contexts, including cold and vibration exposure, peripheral vascular changes, musculoskeletal loading, heat stress, fatigue, respiratory protection, and computer-modeling applications. The included studies used IRT primarily to document and analyze skin-surface temperature distribution and thermal patterns, characterize responses under specific occupational conditions or exposures, and monitor changes over time or following interventions.

Taken together, the findings support the interpretation of IRT as a noninvasive, adjunctive, and complementary imaging method within multidisciplinary clinical assessment in occupational health. Thermographic findings may contribute to related decision-making when integrated with occupational history, clinical examination, and relevant physiological, ergonomic, environmental, or neurofunctional assessments. However, the available evidence does not permit generalized conclusions regarding diagnostic performance, causal relationships, intervention effectiveness, or applicability across occupational populations and settings. The predominance of observational designs, small or convenience samples, variable control of confounding, and protocol heterogeneity limits the precision, comparability, and transferability of the findings. Conclusions should therefore remain confined to the purposes, populations, conditions, and settings examined in the included studies.

Further research may strengthen this evidence base through clearly described protocols and study designs suited to the intended thermographic application. Depending on the purpose investigated, studies may examine reproducibility, validity, and the contribution of IRT when integrated with established clinical, physiological, neurofunctional, ergonomic, and environmental methods used in occupational health.

Author contributions

Marta Regina Cezar-Vaz (Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Resources, Writing—original draft, Writing—review & editing, Funding acquisition, Project administration), Aline da Rosa Pizarro (Methodology, Writing—original draft, Writing—review & editing, Supervision), Joice Simionato Vettorello (Formal analysis, Data curation, Writing—review & editing), Gabriela Laudares Albuquerque de Oliveira (Formal analysis, Writing—review & editing), Sérgio Martin Gonzalez Montaño (Formal analysis, Writing—review & editing), Clarice Alves Bonow (Conceptualization, Methodology, Writing—original draft, Writing—review & editing, Supervision), All authors read and approved the final version of the manuscript, including the revisions made during the peer review process, and attest that they meet the current International Committee of Medical Journal Editors (ICMJE) authorship criteria.

Funding

This study was supported by the National Council for Scientific and Technological Development (CNPq), Call No. 09/2023—Research Productivity Fellowships (PQ-1A), process No. 316678/2023-6.

Conflicts of interest

The authors declare no conflicts of interest.

Data availability

The data supporting the findings of this review are presented in the article and its supplementary material.