Table 1 Characteristics of included studies

Cohort Studies

 Agaliotis et al., 2013, Australia [24]

To determine: (1) reduced work productivity (absenteeism/ presenteeism) and (2) associated risk factors, in symptomatic knee OA

1 year

Q + D

Older workers (aged ≥ 45 years); completing 2-year “Long-term Evaluation of Glucosamine Sulfate” study

Diagnosis based on medial tibiofemoral joint space narrowing (≥ 2 mm joint space) on x-ray

Knee

n = 360

Mean age (SD) 57.5 (7.2) y

54.0% F

Work productivity:

1) Absenteeism (past 2 months): How many days off due to knee problems?

2) Presenteeism (daily for 1 week): self-reported ‘estimated capacity from 0% (unable to do usual work/ activities) to 100% (fully functioning in usual role).’ Questions derived from WPAI:OA

 Hubertsson et al., 2017, Sweden [30]

To determine association between occupation and risk for absenteeism and disability pension due to knee or hip OA

5-years

R

Residents Skåne region, Sweden; Skåne Health Care Register data linked with SSIA data

ICD-10 code for knee OA (M17); hip OA (M16)

Knee and hip

n = 165,179

Age: 40-70y.; (average age not reported)

58.3% F

Absenteeism and disability pension data from SSIA register

 Kontio et al., 2018, Finland [31]

To examine association of education and physical workload factors with occupational differences in disability retirement due to knee OA

8-years

R

Population-based; register data from 70% random sample of Finnish population; aged 18–70 years living in Finland on 31/12/2004 (~ 2.5 million)

Finnish version ICD-10 (1996) code for knee OA (M17)

Knee

n = 1,135,654

Mean age (SD): F = 45.3 (8.4) y; M = 44.6 (8.3) y

49.4% F

Full-time disability retirement (either temporary or permanent) due to knee OA from 01/01/2005 to 31/12/2013

 Kontio et al., 2020, Finland [32]

To examine to what extent disabling OA, led to prolonged absenteeism, interferes with work participation and shortens working life-years

8-years

R

Population-based study; register data from 70% random sample of Finnish population; aged 18–70 years living in Finland on 31/12/2004

(~ 2.5 million)

Finnish version ICD-10 (1996) codes: M15 (polyarthritis), M16 (hip OA), M17 (knee OA), M18 [carpometacarpal (CMC) joint OA, and M19 (other OA)

For the analysis, CMC joint OA and polyarthritis were combined

Hip, knee, hand and other

n = 4,704

Mean age (SD): Knee OA = 51.1 (6.2); Hip OA = 52.0 (6.0)

Polyarthritis or CMC joint OA = 53.5 (4.6); Other OA = 51.1 (6.7)

56.2% F

Sustained return to work (i.e., returning to regular duties for ≥ 28 consecutive days immediately following sickness absence). Time spent in different work participation statuses

Potential working life–years lost (i.e., actual retirement age (whatever cause) versus working life expectancy forecast for Finland). Early exit from paid employment (i.e., permanent disability retirement or old age retirement prior to 63 years old)

 Summanen et al., 2021, Finland [33]

To determine the burden of hip and knee OA in Finnish occupational healthcare

8-years

R

Electronic medical records (EMRs) of Terveystalo (largest private and occupational healthcare provider in Finland)

Diagnosis of: hip or knee OA (ICD-10 codes M16* or M17*, respectively)

Hip and knee

N = all OA cohort: 51,068; OCH subcohort from all OA cohort = 35,109; controls = 35,101

Mean (SD) age at index: ALL OA = 56.6 (10.1); OCH subcohort = 53.3 (7.7); Controls = 52.4 (7.6)

F: all OA cohort = 56.3%; OCH subcohort hip and knee = 54.2%; Controls = 54.2%

Absenteeism days

 Wilkie et al., 2014, United Kingdom [34]

To estimate proportion of working age adults with OA predicting work limitations prior to future pension age of 69

6-years

Q

North Staffordshire OA project (NorStOP): population-based prospective cohort study

At least one consultation during study period for OA based on NHS Read codes (N05 category) for primary care consultations

Any joint(s) – not specified

n = 297. consulters for OA

Mean age (SD) 54.0 (2.34) y

54.9% F

Expected work limitations: at 6-y. follow-up, “Do you think joint pain will limit your ability to work before you reach 69 years old” (will limit or stop me/ don’t know/ won’t limit)

 Wilkie et al., 2014, United Kingdom [17]

To describe the extent of premature work loss and associated factors in OA consulters

6-years

Q

NorStOP – a population-based prospective cohort study

At least one consultation during the study period primarily for OA based on NHS Read codes (N05 category) for primary care consultations

Any joint(s) – not specified

n = 612. consulters for OA

Mean age (SD) 54.6 (2.8) y

48.2% F

Premature work loss (i.e., moving from employment to retirement prior to state retirement age or transition from employment to being off work due to ill-health or unemployment)

 Wilkie et al., 2015, United Kingdom [35]

To examine how pain leads to onset of work productivity loss in OA; and identify new intervention opportunities

3-years

Q

NorStOP—population-based prospective cohort study

At least one consultation during study period for OA based on NHS Read Codes (N05 category) for primary care consultations

Any joint(s) – not specified

n = 318

Mean age (SD): 56.2 (2.2) y

52.8% F

Work productivity loss (SF-36 item: “During the past 4 weeks, have you accomplished less than you would like in your work or other regular daily activities as a result of your physical health? yes/no)

Pain intensity (SF-36 Bodily Pain in last 4 weeks: high (i.e., moderate, severe, very severe) or low (i.e., none, very mild, mild)

Cross-Sectional Studies

 Agaliotis et al., 2017, Australia [36]

To explore personal and workplace environmental factors as predictors of reduced worker productivity among older workers with chronic knee pain

Q + D

Older workers (aged ≥ 45 years) completing 2-year “Long-term Evaluation of Glucosamine Sulfate” study

Knee pain, or taking NSAIDs/ analgesia for knee pain on most days in the past month; Knee pain of 4–10 on 10 cm Visual Analogue Scale; Medial tibio-femoral compartment joint space narrowing in symptomatic knee

Knee

n = 129

Mean age (SD) 60.0 (6.8) y

52.0% F

Single-item questions from WPAI:OA for absenteeism (past 2 months) and presenteeism (7-day diary); Multi-item at-work limitations/productivity = Work Transition scale (past 6 months); Workplace Activity Limitations Scale

 Bieleman et al., 2010, The Netherlands [13]

To examine work participation of people with early knee or hip OA and compare to data from American Osteoarthritis Initiative (OAI) cohort. The influence of health status and personal factors on work participation analysed

Q

The Cohort Hip and Cohort Knee (CHECK) study; data from the OAI

CHECK: Self-reported pain and/ or stiffness hip and/or knee; first GP consult for these symptoms ≤ 6 m; OA explained their symptoms. OAI: Baseline data participants’ clinical and joint status; risk factors for progression and development of knee OA (questionnaires and examination)

Knee and hip

Participants in paid work:

CHECK: n = 970

Mean age (SD) 56.0 (6.0) y

79% F

OAI: n = 1,578. Mean age (SD) 56.0 (6.0) y

64% F

Work participation (including: present or last job, work hours, working history, present working status, absenteeism) investigated using “Economic Aspects in Rheumatoid Arthritis” questionnaire. Labour force participation (i.e., paid job ≥ 8 h/ week)

For participants in paid employment: present health condition; whether adapted, or like to adapt, their work (tasks/hours/ workplace). Participants not in paid work: why not employed?

 Conaghan et al., 2021, Europe [37]

To assess whether impact of OA pain on health-related quality of life, work, and healthcare resource use differed by pain severity and prescription medication status

NS

National Health and Wellness Survey (NHWS) 2016–2017 (pooled self-reported data) from the 5 European countries

Self-reported physician-diagnosed OA; experienced pain in past 12 months (worst pain with or without prescription medication use [none/mild/moderate/severe])

Any joint(s) – top 3 affected joints: knees, fingers, hips

n = 2,417

Age groups (years), n (%): 18-39y = 94 (3.9%); 40-49y = 236 (9.8%); 50-59y = 550 (22.8%); 60-69y = 940 (38.9%); 70 + y = 597 (24.7%)

64.5% F

WPAI: GH (work productivity loss; non-work activity impairment). HRQoL and health status: SF-12v2; EQ-5D; SF-6D; EQ-VAS. All-cause healthcare resources utilization past 6 months: self-reported no. healthcare provider visits (primary care, emergency/ urgent care, and hospitalisations)

 daCosta DiBonaventura et al., 2011, USA [38]

To evaluate impact of OA pain on: direct medical costs; indirect costs associated with lost productivity for

any reason; healthcare provider visits; and general health status in employed population

NS

NHWS 2009

Self-reported physician-diagnosed OA

Ankles, elbows, feet, fingers, hands, hips, knees, neck, shoulders, spine, wrists, other

n = 2,173. diagnosed with OA

Age range: 20–39 y.: n = 267; 40–64 years n = 1,453; ≥ 65 y. n = 453

56.1% F

Work productivity (WPAI. Health utility scores (using SF-6D). For each respondent, percent overall work impairment (WPAI) multiplied by annual income. Direct and indirect costs summed to estimate total costs

 daCosta DiBonaventura et al., 2012, USA [39]

To evaluate impact of patient-rated OA severity on: productivity; health-related quality of life (HRQoL); healthcare resource use and costs in employed individuals compared to employed individuals without OA

NS

NHWS 2009

Self-reported physician-diagnosed OA;

Ankles, elbows, feet, fingers, hands, hips, knees, neck, shoulders, spine, wrists, other

Total n = 4,876 Mild OA n = 2,192,

Age range: 20-39y. n = 338; 40–64 y. n = 1,339; ≥ 65 y. n = 515

48.9% F; Moderate OA n = 2,240

Age range: 20–39 y. n = 298; 40-64y. n = 1,447; ≥ 65 y. n = 495

55.0% F;

Severe OA n = 444;

Age range: 20–39 y. n = 49; 40–64 y. n = 312; ≥ 65 y. n = 83. 57.2% F

Work productivity (WPAI). HRQoL (physical and mental component scores SF-12v2). Health utility scores (SF-6D)

 Gignac et al., 2018, Canada [25]

To compare health and work contexts of M:F workers ≥ 50 years to understand similarities/ differences in workplace accommodation availability, need, use, and unmet needs

Q

Born 1946 to 1964; employed ≥ 15 h/week. Data from larger project on health, accommodations, and retirement expectations in arthritis, diabetes, no chronic disabling health conditions

Self-reported physician diagnosis of arthritis; duration ≥ 1 year (to ensure has worked with arthritis). OA only data used in this systematic review

Any joint(s) – not specified

n = 273

Mean age:

F = 59.5 y

M = 59.1 y

64.8% F

Fourteen workplace accommodations: availability (Y/N), need for (Y/N), and use of (Y/N) in previous 12 m.; accommodation needs reported as (i) congruence between accommodations needed and used (i.e., accommodation needs met), (ii) needing more accommodations than used (i.e., accommodation needs unmet), or (iii) not needing some accommodations, but using anyway (i.e., accommodation needs exceeded)

 Hermans et al., 2012, The Netherlands [40]

1. To identify and quantify knee-related productivity and medical costs in conservatively treated knee OA patients in paid employment

2. To evaluate associations between knee-related productivity loss and individual, disease and work characteristics

Q

Participants in a randomised controlled trial investigating cost-effectiveness of intra-articular hyaluronic acid plus usual care

Only data acquired from baseline measurement (pre-intervention) used

Consecutive patients with knee OA at 2 hospitals out-patient Orthopaedic clinics; Kellgren/Lawrence grade 1–3 and pain score ≥ 2 (0–10 numerical rating scale); age 18–65 y.; treated conservatively for knee OA for ≥ 6 m. prior to inclusion

Knee

n = 117

Mean age (SD) 53.2 (7.4) y

43.0% F

Productivity and medical costs: Productivity and Disease Questionnaire measuring: productivity costs due to knee symptoms; knee-related absence from work past 3 m.; lost productivity due to knee symptoms while at work (using Quality and Quantity method)

 Hubertsson et al., 2013, Sweden [14]

To investigate extent of absenteeism and disability pension in knee OA; and compare to general population

Retrospective Secondary data analysis; R

Skåne Health Care Register data linked with SSIA data

Skåne Health Care Register data: diagnosis of knee OA (ICD-10 code M17)

Knee

n = 15,345

Mean age (SD): F = 55.0 (8.2) y.; M = 53.0 (9.2) y

49.6% F

Absenteeism and disability pension data retrieved from SSIA register

 Jackson et al., 2020, USA and Europe [41]

To evaluate burden of pain associated with knee and/or hip OA

Patient records; Q

Large, multinational, observational study of clinical practice (Adelphi Disease

Specific Programme (DSP)™

Patients diagnosed with knee and/or hip OA by consulting physicians

Knee and/or hip

N = 2170

Mean age (SD): 66.4 (11.2) y

57.9% F

Work productivity and daily activity impairment (WPAI:OA); HRQoL

 Laires et al., 2018, Portugal [42]

To describe impact of OA (pain and physical disability) on early exit from work

Q

National, cross-sectional, population-based study of rheumatic diseases in Portugal – the EpiReumaPt study

Clinical confirmation OA by a rheumatologist; validated by three experienced rheumatologists using American College of Rheumatology criteria

Knee, hip, hand

n = 382

Mean age 57.5y

72.3% F

Self-reported employment status: employed (i.e., part/ full-time); early exit from paid employment (including: students; homemakers; no regular salary; early retired; or with disability pension)

Quality of life (SF-36v2 total and 8 sub-scale scores); EQ-5D-3L; longstanding musculoskeletal pain (≥ 3 m.); pain interference with function (i.e., pain affecting work and domestic activities from item in SF-36v2; functional capacity (Health Assessment Questionnaire)

 Nakata et al., 2018, Japan [43]

To examine work impairment and OA; HRQoL and health status

NS

2014 Japan National Health and Wellness Survey (NHWS, Kantar Health, New York, NY, USA),

Self-reported received OA diagnosis from a healthcare provider

Ankles, elbows, feet, fingers, hands, hips, knees, neck, shoulders, spine, wrists, other

n = 233

Mean age (SD): 54.2 (12.2) y

43.8% F

Work productivity (WPAI). Presenteeism and absenteeism categorised as any vs. none. HRQoL (SF-36v2; SF-6D). Depression (Patient Health Questionnaire-9)