オフィス環境と人間工学的リスク：非製造業企業に適用される事項

📌 Key takeaways

• The "office-only workplace" exemption in Korea is narrower than its name suggests. It removes the management-structure chapter, the safety and health training chapter, and part of the contractor-duty chapter. It does not remove Article 39 health measures which is where ergonomic risk lives.
• Work environment measurement is decided by exposure, not by headcount or industry. There is no size threshold and no industry list. If none of the 192 statutory agents is present, your office is out of scope but the way you demonstrate that is a preliminary survey, not an assertion.
• Where contractor personnel work at your premises, Article 125(2) puts the measurement duty on you, the principal, not on the cleaning or facilities vendor. For dispatched workers, the user company is the employer.
• The musculoskeletal category that catches offices four hours a day of concentrated keyboard or mouse work turns on three practical tests, not on job title: whether input volume is pre-set at an excessive level, whether the worker can adjust their own work and rest time, and whether actual accumulated hours reach four.
• Ergonomic standards for screen work sit in three tiers with different legal force: an enforceable Rule that contains no numbers, a Public Notice that contains numbers but no penalty, and a KOSHA technical regulation whose numbers differ from the Public Notice. A global ergonomics policy has to be mapped against the right tier.

---

🔎 The exemption global HR teams over-read

Almost every foreign-invested entity we speak to has, at some point, been told that its Korean office is an "office-only workplace" and therefore outside the Occupational Safety and Health Act. That sentence is half right, and the half that is wrong is the expensive half.

Under Article 3 of the Act and Schedule 1 of the Enforcement Decree, a workplace using only office workers is excluded from a listed set of provisions. On the reading confirmed across four independent Korean sources, that set covers Chapter 2 Sections 1 and 2 (the safety and health management structure the responsible person, safety managers, health managers, the committee), Chapter 3 in full (safety and health education, including special training), and Chapter 5 Section 2 (the principal's safety and health measures, save for the welfare-facility limb).

Chapter 4 is not on that list. Chapter 4 contains Article 38 (safety measures) and Article 39 (health measures), and Article 39(1)5 is the provision that carries the musculoskeletal duty. Risk assessment is likewise unaffected. General health examinations are unaffected they are simply on a different cycle for office workers.

The practical posture, then, is not "we are exempt." It is: assume Article 39 applies, and treat the exemption as something to be established with your local counsel if you intend to rely on it for the training and management-structure duties.

---

🧪 Work environment measurement: probably out of scope, and how to show it

This is the duty most global EHS teams expect to be the heavy one for a Korean entity, and for an office it is usually the light one.

The test under Article 125 and Enforcement Rule Article 186(1) is a single question: does a worker in this workplace or work process get exposed to one of the hazardous agents listed in Schedule 21? There is no headcount threshold and no industry list. A workplace with one exposed worker is in scope; a workplace with four hundred unexposed workers is not.

Schedule 21 runs to 192 agents: 114 organic compounds, 24 metals, 17 acids and alkalis, 15 gaseous substances, 12 permit-required substances, and one metalworking fluid, plus two physical agents noise at an eight-hour time-weighted average of 80 dB or above, and high heat and seven dusts. Older circulating material states 190; that figure predates the current schedule and should not be quoted.

An ordinary office generates none of these. The conclusion is usually straightforward. What is not straightforward is the evidence for it.

In Korean practice, the determination is made through a **preliminary survey **normally the first stage a licensed measurement institution performs, which identifies the processes and agents present and establishes whether the exclusion grounds in the proviso to Article 186(1) apply. Those grounds include consumption of controlled hazardous substances below the permitted quantity, temporary work (under 24 hours a month), short-duration work (under one hour a day), and dust work falling within the Schedule 16 exclusions. Note the anti-avoidance limbs: temporary work repeating monthly at 10 hours or more, or short-duration work repeating daily, does not qualify, and neither does anything involving carcinogens on the Minister's list.

For an office-only entity the useful output of that exercise is a dated document, produced by a competent third party, recording that no Schedule 21 agent is present. That is a far better answer to an inspector than a manager's recollection.

Where an office does come into scope, it is normally through something attached to the premises rather than through the office floor. Laboratory or R&D space is the clearest case a manual for research-facility measurement exists precisely because those spaces are in scope. Equipment producing noise at or above the 80 dB eight-hour average is another, on the face of the schedule. 

🧭 Ergonomic scope: the determination is the deliverable

If the work environment measurement duty usually falls away for an office, the musculoskeletal duty usually does not or rather, it does not fall away by itself. Someone has to determine that it falls away, and that determination is the document an inspector will ask for.

The governing category is the first of the eleven in MOEL Public Notice No. 2020-12: four or more hours a day of concentrated keyboard or mouse operation for data entry and similar work. Three provisions of that notice do most of the work in an office context.

Hours are actual, accumulated hours. Article 2 measures the four-hour threshold by the time the worker actually spends performing the burdening work, aggregated across the day. Not scheduled hours, not hours logged into a system, not the length of the working day.

Short-term and intermittent work is excluded, with definitions. Short-term work means one-off work completed within two months. Intermittent work means work whose total working days in a year do not exceed 60. A quarterly stock-take or an annual audit cycle sits outside the category on this basis.

"Concentrated" is a substantive test. Administrative interpretation reads the word as work where keyboard or mouse operation is continuous and where either the input target is pre-set at an excessive level or the worker cannot adjust their own working and rest time beyond a certain point. Intermittent input, search and retrieval work, two-way communication and information-gathering, and work where the employee controls their own input timing, fall outside it. 

Put together, these give a three-part test that a global HR team can actually apply.

Most white-collar roles fail at least one test and fall outside the category. That is a real conclusion, not a convenient one. But it is a conclusion about individuals doing particular work, and it splits populations that an org chart treats as identical. Two people reporting to the same manager with the same title can land on opposite sides of it if one works to daily throughput targets and the other does not.

The functions that most often reach category 1 in a non-manufacturing entity are the ones where volume is managed: transaction processing in a shared service centre, high-volume data capture, and contact-centre work where handle time is a measured target.

Documenting a negative

Here is the point that Korean-language content on this board does not cover, and that matters most to a headquarters audience.

There is no statutory requirement to document a conclusion that no burdening work exists. Practitioner sources are explicit that the law does not mandate it. But there is also no other way to prove you performed the assessment. If the investigation duty is disputed at an inspection, or after an occupational disease claim the employer's only defence is a record showing which tasks were examined, on what basis, and when.

Three features of the investigation regime make that record more useful than it first appears.

The investigation itself must cover three subject areas, and commentary from a former MOEL district office head is direct about the consequence: an investigation missing any one of the three cannot be recognised as a valid hazard investigation. Those areas are workplace conditions (equipment, process, workload, work pace); work conditions (hours, posture, method); and the presence of work-related musculoskeletal signs and symptoms. A scope determination built on the same three headings maps onto the investigation you will run if the conclusion changes.

---

💻 Screen work: three tiers, three different legal characters

This is where a global ergonomics standard most often collides with Korean law, and the collision is usually a misreading of which document binds.

Tier one — the Rule.

 Article 667 of the Rules on Occupational Safety and Health Standards ("Measures for computer terminal operation work") requires the employer, where workers operate computer terminals, to avoid harsh light contrast and direct sunlight, use low-luminance lighting and non-reflective wall and window materials, provide desks and chairs individually height-adjustable, and give workers engaged in continuous terminal work appropriate rest during working hours.

Article 667 contains no numbers at all. What it does have is enforcement: it implements Article 39 health measures, and a violation is punished under Article 168 imprisonment of up to five years or a fine of up to KRW 50 million. The undefined term "appropriate rest" carries criminal exposure.

Tier two — the Public Notice.

 The VDT Worker Work Management Guidelines (MOEL Public Notice No. 2020-17, issued 6 January 2020, effective 16 January 2020) supply the numbers. But the Notice is made under Article 13 of the Act "standards concerning technology or the working environment" which authorises the Minister to guide and recommend. There is no penalty attached to the Notice itself.

Tier three — KOSHA technical regulation.

 E-G-3-2025, on managing office environments using visual display terminals (revised 3 February 2025, published 26 March 2025), describes itself as a technical recommendation. Its numbers differ from the Public Notice's in several places.

One warning about a figure still circulating widely. The "10 to 15 minutes of rest every hour" rule comes from the 1997 version of the guidelines. The current Public Notice contains no numeric rest interval. Writing "Korean law requires a 10-minute break every hour" into a global policy states something that is not in the current text.

The workable structure for a headquarters framework is straightforward once the tiers are separated. Adopt the Public Notice figures as the baseline your Korean entity is measured against, since they are what an inspector and a court will treat as the yardstick for whether Article 667 was satisfied. Where your global standard is more protective, keep it. Where KOSHA's figures are more protective the seat-height range, the hourly rest recommendation treat them as the better practice they are, and document the choice. What you should not do is import a European or US display-screen-equipment standard, assume equivalence, and stop.

---

💬 How EAP supports employees and HR

The office-based entity has a particular blind spot, and it is not a shortage of ergonomic equipment.

An entity with no factory tends to have no EHS function either. Ergonomic risk is handled by whoever also handles the lease and the laptop refresh, and it surfaces as a facilities request a chair, a monitor arm, a sit-stand desk. Those requests get fulfilled, sometimes generously. What is missing is anyone whose job is to notice that three people on the same team have asked for the same equipment in the same quarter, or that a workload change six months ago is the common factor.

That gap matters more in a non-manufacturing setting than the equipment budget suggests, because the risk factors that drive musculoskeletal symptoms in office work are largely organisational. WHO guidance on mental health at work identifies workload, work pace, and control over how work is done as primary determinants of worker health outcomes, and recommends manager training as a complement to worker-directed support. ISO 45003 places psychosocial risk inside the occupational health and safety management system rather than beside it the same system that holds physical hazards. The two are not parallel programmes; high demand and low control produce musculoskeletal symptoms as well as psychological ones, which is why the scope tests for category 1 are themselves written around pacing and autonomy rather than around posture.

Read that against the three-part test above and the alignment is unusually direct. Pre-set input targets and an inability to adjust one's own rest are, at the same time, the statutory markers of burdening work and the textbook organisational risk factors. An entity that has genuinely eliminated both has narrowed its ergonomic exposure and improved its psychosocial conditions with one decision.

Under the EAPA Core Technology, consultation with managers and organisations on employee and organisational issues, and confidential problem identification and assessment for employees whose personal concerns affect work performance, are core EAP functions. In a Korean entity with no in-house EHS capability, that translates into three specific contributions: a confidential channel in Korean and English through which discomfort can be raised without going through the line manager who sets the workload; support for managers receiving those reports, who in an office setting have usually never handled one; and aggregated, de-identified pattern data that gives a scope determination something to work from beyond a headcount and a floor plan.

The determination remains the employer's. What changes is whether it is made with information.

---

🔗 Related guides

---

❓ FAQ

Q2. We have never done work environment measurement. Should we be worried?

Probably not, and you can find out cheaply. Measurement is required only where a worker is exposed to one of the 192 agents in Schedule 21 there is no size or industry threshold. An ordinary office generates none of them. The right step is a preliminary survey by a licensed measurement institution, which produces a dated record that no listed agent is present. Two things to check while you are at it: whether any attached space (laboratory, R&D area, or equipment producing noise at 80 dB or above as an eight-hour average) changes the answer, and whether contractor personnel work at your premises — because under Article 125(2) their measurement is your duty, not the vendor's.

Q3. How do we decide whether our shared service centre staff perform "burdening work"?

Apply three tests to the actual work of actual individuals. Is the input volume pre-set at a level that is excessive? Can the worker adjust their own working and rest time? Do actual accumulated keyboard and mouse hours reach four a day, counting only time genuinely spent on that work? Where throughput targets are set and monitored, the first two tests point toward the category, and a shared service centre is a population to examine rather than one to assume is outside. Record which tasks you examined and what you observed. Nothing requires that record, and nothing else will serve if the determination is ever questioned.

Q4. Our global ergonomics standard specifies desk and chair dimensions and a rest schedule. Can we run it in Korea unchanged?

Only after mapping it. Korean screen-work requirements come in three tiers. Article 667 of the Rules is enforceable a violation is an Article 39 health measure breach, punishable under Article 168 by up to five years' imprisonment or a fine of up to KRW 50 million but it contains no numbers, only "appropriate rest" and adjustable furniture. Public Notice 2020-17 supplies figures (illuminance 300–500 or 500–700 lux by screen background, viewing distance 40 cm or more, seat height 35–45 cm) but is guidance with no penalty. KOSHA's E-G-3-2025 recommends different figures again, including a seat range to 52 cm and 5–10 minutes of rest each hour. Keep your global standard where it is more protective, adopt the Public Notice figures as the baseline you are measured against, and document where you chose KOSHA's more protective numbers. Do not carry across a "10 to 15 minutes per hour" rest rule as Korean law — that is from the 1997 guidelines and is not in the current text.

---

👉 Next step

For a non-manufacturing Korean entity, the whole exercise reduces to four documents. None of them is expensive; all four are hard to produce retrospectively.

A preliminary survey record

 from a licensed measurement institution, establishing whether any Schedule 21 agent is present and covering attached spaces and contractor activity at your premises, not only the office floor.

A scope determination for burdening work, built on the same three subject areas the statutory investigation requires: workplace conditions, work conditions, and the presence of musculoskeletal signs and symptoms. Applied to individuals, with the pacing and autonomy tests answered explicitly, and dated.

A mapping of your global ergonomics standard

 against the three Korean tiers, showing which figure governs where and why any divergence was chosen.

A coverage list

 naming dispatched workers and contractor personnel working at your premises, with the duty holder identified for each item because the allocation is not intuitive and it is not the same across duties.

---

NOTE: This article is intended for general informational purposes only. Specific legal, medical, clinical, or employment-related matters may require review by an appropriately qualified professional. Reliance on the office-only workplace classification, and any conclusion that a work population falls outside the burdening-work categories, should be confirmed with qualified Korean counsel — a violation of Article 39 health measures carries criminal rather than administrative liability.

👉 Go to Nudge EAP Implementation Consultation →

---

📚 Sources

1. 산업안전보건법 제125조 (작업환경측정) — 제1항 측정 대상, 제2항 도급인 사업장에서 관계수급인 근로자 작업 시 도급인이 측정, kopshe — https://kopshe.co.kr/business/sub14
2. 작업환경측정 대상 유해인자 (시행규칙 [별표 21]) — 화학적 인자 183종(유기화합물 114·금속류 24·산알칼리 17·가스상 15·허가대상 12·금속가공유 1), 물리적 인자 2종(소음 8시간 TWA 80dB 이상, 고열), 분진 7종, 총 192종, hseworld — https://hseworld.co.kr/126
3. 작업환경측정 대상 유해인자, 산업안전보건법 시행규칙 별표 21 — ㈜우리안전지도원 — https://www.woorisafety.co.kr/bbs/board.php?bo_table=table43&wr_id=284
4. 작업환경측정 및 정도관리 등에 관한 고시 — 측정 주기별 최소 간격(3개월·45일·6개월), 제4조의2 주유소 특례, U-LEX 법률우주 — https://www.ulex.co.kr/법률/2100000186111-22235-작업환경측
5. 작업환경측정 모든 것 — 시행규칙 제186조제1항 단서 측정 제외 사유(허용소비량 미만, 임시작업 월 24시간 미만, 단시간작업 1일 1시간 미만, 분진작업 적용제외, 주유소), 제188조 최초 30일 이내·정기 6개월, 산업보건이야기 — https://blog.occupationalhealthblog.net/작업환경측정-모든-것/
6. 작업환경측정 대상 주기 유해인자 및 과태료 기준 — 과태료 세부 항목, soonyguide — https://soonyguide.com/80/
7. 작업환경측정 사업장 및 대상 유해인자가 궁금합니다 — 규모 기준 없음, 유해인자 발생 여부로 판단, 대한산업보건협회 월간 산업보건 2023.7 — https://kiha21.or.kr/monthly/2023/07/2023_07_20_s423.pdf
8. 안전·보건 관계서류 목록 (2024) — 작업환경측정 서류 보존기간(결과 기록 5년 / 발암성 물질 등 30년), 과태료 부과 기준, 근거 조문(법 제164조·제175조), 시흥상공회의소 — https://shiheungcci.korcham.net/file/dext5uploaddata/2024/안전·보건 관계서류 목록_.pdf
9. 작업환경측정 결과 보고 — 측정 완료일부터 30일 이내 지방고용노동관서장 제출, 예비조사 단계에서 제186조제1항 단서 해당 여부 판단, 무사퇴근연구소 — https://musa-lab.com/kor/board/issue?viewMode=view&idx=329
10. 파견근로자가 있는 경우 산업안전보건법 준수 주체는? — 파견법 제35조, 사용사업주를 산안법상 사업주로 봄, 무사퇴근연구소 — https://musa-lab.com/kor/board/issue?viewMode=view&idx=121
11. 사무직만을 사용하는 사업장에 대한 판단과 해석 (정재호 특별기고) — 시행령 별표1 제5호다목 적용제외 범위(제2장 제1절·제2절, 제3장, 제5장 제2절), 사무직 정의(시행규칙 제197조제1항), 행정해석 산재예방정책과-5529·산재예방정책과-1026, 안전신문 — https://www.safetynews.co.kr/news/articleView.html?idxno=235487
12. 사무직 사업장에 대한 판단기준과 안전 관리 방안에 대하여 — 사무직 판단 2요소(한국표준직업분류 / 직접적 산업활동 아닌 경영지원), 사무직 사업장도 위험성평가 실시 의무, 법무법인(유한) 율촌 — https://www.yulchon.com/ko/resources/publications/legal-update-view/36101/page.do
13. 행정해석 산업안전보건정책과-1439 (2021. 9. 17.) — 게임 개발사 직원 775명 중 737명이 기획·프로그래머·그래픽디자이너인 경우 주된 산업활동 직접 수행으로 보아 사무직 전용 사업장 불인정, 무사퇴근연구소 — https://musa-lab.com/_subpage/kor/board/administration.php?viewMode=view&idx=166
14. 사무직으로만 구성된 본사에서는 안전보건교육이 배제될까? — 사무직 전용 사업장 요건 4가지, 비사무직 혼재 시 전체 탈락, 무사퇴근연구소 — https://musa-lab.com/_subpage/kor/board/issue.php?viewMode=view&idx=94
15. 산업안전보건법과 중대재해처벌법 준수 감사 — 사무직만 고용한 사업장도 제39조 보건조치 의무 적용, 건강진단 2년 1회, k-labor — https://k-labor.co.kr/main/klabor_04_view.html?pgubun=55&lang=ko&chapter_idx=2051&items_idx=8132
16. 근골격계부담작업의 범위 및 유해요인조사 방법에 관한 고시 (고용노동부고시 제2020-12호) — 제2조 4시간 이상 산정(실제 수행시간 합산)·단기간작업(2개월 이내 1회성)·간헐적 작업(연간 총 작업일수 60일 이하), 제3조 제1호 집중적 자료입력 작업, 제4조 별지 서식, U-LEX 법률우주 — https://www.ulex.co.kr/법률/2100000185793-34925-근골격계부
17. 사무직도 근골격계 부담작업에 해당될 수 있을까? — 행정해석 산재예방정책과-2416(2019. 5. 21.) "집중적" 해석(입력 목표량 과도 사전 설정 또는 휴식·작업시간 임의 조절 불가), 검색·해독·쌍방향 통신 작업 제외, 직종이 아닌 실제 작업상황으로 판단, 무사퇴근연구소 — https://musa-lab.com/kor/board/issue?viewMode=view&idx=372
18. 근골격계 유해요인조사, 누가 할 수 있고 언제까지 보관해야 하나 — 조사자 자격 제한 없음(행정해석 근로자건강보호과-4778, 2009. 12. 22.), 조사 관련 문서 5년 보존, 시설·설비 개선계획은 설비 존치 기간, 무사퇴근연구소 — https://musa-lab.com/kor/board/issue?viewMode=view&idx=366
19. 근골격계부담작업 유해요인조사 모든 것 — 규칙 제657조 조사 3요소, 미해당 판단 근거의 법정 문서화 의무 없음(실무상 보존 권고), 산업보건이야기 — https://blog.occupationalhealthblog.net/근골격계부담작업-유해요인조사-모든-것/
20. 정진우, 「작업관련 근골격계질환 예방조치 의무」 — 조사 3요소 중 하나라도 빠지면 유해요인조사로 인정받을 수 없음, 제39조 위반의 형사처벌, 한국산업보건학회 월간 산업보건 2014.11 — https://kiha21.or.kr/monthly/2014/11/SOBGBO_2014_s319_37.pdf
21. 산업안전보건법 제168조 (벌칙) — 제39조제1항 위반 시 5년 이하 징역 또는 5천만원 이하 벌금, goodlabor — https://www.goodlabor.com/산업안전보건법
22. 영상표시단말기(VDT) 취급근로자 작업관리지침 (고용노동부고시 제2020-17호, 발령 2020. 1. 6. / 시행 2020. 1. 16.) — 제4조 작업시간·휴식시간(수치 없음), 제5조 작업대·의자 치수, 제6조 시거리 40cm 이상·시야각·팔꿈치 내각, 제7조 조도 기준, U-LEX 법률우주 — https://www.ulex.co.kr/법률/2100000185801-22229-영상표시단
23. 산업안전보건법 제13조 (기술 또는 작업환경에 관한 표준) — 고용노동부장관의 지도·권고 근거 조항, CaseNote — https://casenote.kr/법령/산업안전보건법/제13조
24. 기술상의 지침의 '비정상의 정상화' (정진우 칼럼) — 제13조에 근거한 고시는 지도·권고 성격이며 위반 시 벌칙 없음, 안전저널 — https://www.anjunj.com/news/articleView.html?idxno=24359
25. KOSHA GUIDE E-G-3-2025 영상표시단말기를 사용하는 사무환경 관리에 관한 기술지원규정 (개정 2025. 2. 3. / 공표 2025. 3. 26.) — 의자 좌면 높이 35~52cm, 작업대 68cm 전후, 연속작업 1시간 초과 금지, 매 1시간 5~10분 휴식 권장, 안전보건공단(APOSHO 미러) — https://www.aposho.org/kosha/info/koshaGuideData.do?mode=download&articleNo=453895&attachNo=261182
26. KOSHA GUIDE H-9-2022 근골격계부담작업 유해요인조사 지침 — 법 제39조 이행 지침임을 명시, 인스도잇 미러 — https://insdoit.co.kr/board/board.php?bo_table=reference&wr_id=897
27. 2025년도 근골격계부담작업 유해요인조사 실시 안내 — 행정직원 대상 실시, 근거로 법 제39조·규칙 제12장 제2절·고시 제2020-12호 명시, 실시 절차(사전교육 → 체크리스트·설문 → 분석 → 현장조사·작업자세 분석 → 개선계획), 가톨릭대학교 산학협력단 — https://iacf.catholic.ac.kr/board/BoardDetail?_menuCd=001100003000&part=REDO&noticeYn=N&idx=12
28. 사무실 공기관리 지침 (고용노동부고시 제2020-45호, 발령 2020. 1. 15. / 시행 2020. 1. 16.) — 제1조 지도·권고 목적, 제3조 환기기준(1인당 분당 0.57㎥ 이상, 시간당 4회 이상), 라돈은 지상 1층 포함 지하 사무실에만 적용, 제9조 신축·개수 시 건축자재 방출기준, U-LEX 법률우주 — https://www.ulex.co.kr/법률/2100000186112-22233-사무실공기
29. 사무실에서의 실내 공기질 관리 직업건강가이드 — 환기기준 0.57㎥/min·시간당 4회 교차 확인, 안전보건공단 산업안전보건연구원 — https://oshri.kosha.or.kr/kosha/data/healthPromotionData.do?mode=download&articleNo=241364&attachNo=118025
30. 산업안전보건기준에 관한 규칙 제3편 제11장 사무실에서의 건강장해 예방 — 제646조~제654조 구성, 제646조 사무실오염물질 정의, 제649조 사무실공기 평가의 조건부 성격, 세종경제뉴스 칼럼 — https://www.seenews365.com/news/articleView.html?idxno=38098
31. 실내공기질 관리법 시행령 제2조 — 다중이용시설 적용대상, 업무시설 연면적 3,000㎡ 이상, U-LEX 법률우주 — https://www.ulex.co.kr/법률/261009-005137-실내공기질관리법시행령
32. 건강검진 미실시 시 과태료 — 일반건강진단 사무직 2년 1회·그 밖의 근로자 1년 1회(시행규칙 제197조제1항), 사무직 범위, 과태료 1인당 10/20/30만원, 노동법률사무소 희수 — https://hrbyhs.com/연말-전-반드시-점검-건강검진-미실시-시-과태료-주의/
33. World Health Organization, Guidelines on Mental Health at Work (2022) — https://www.who.int/publications/i/item/9789240053052
34. World Health Organization, Mental health at work fact sheet — https://www.who.int/news-room/fact-sheets/detail/mental-health-at-work
35. ISO 45003:2021, Occupational health and safety management — Psychological health and safety at work — https://www.iso.org/standard/64283.html
36. Employee Assistance Professionals Association, Definition and Core Technology of Employee Assistance Programs — https://eapassn.org/page/definitionandcoretechnology