BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from
BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.
When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.
The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.
BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay- per-view fees (http://bmjopen.bmj.com).
If you have any questions on BMJ Open’s open peer review process please email [email protected]
http://bmjopen.bmj.com/ on September 29, 2021 by guest. Protected copyright. BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open
Predicting End of Work Disability After Cancer. A 7 years Follow-up Population-based Cohort Study Using Competing Risks Analysis
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2016-014094
Article Type: Research
Date Submitted by the Author: 30-Aug-2016
Complete List of Authors: Kiasuwa Mbengi, Régine; Scientific Insitute of Public Health, Belgian Cancer Centre ; Universite Libre de Bruxelles, Centre de Recherche en Santé Environnement et Santé au travail Nicolaie, Alina; Universiteit Gent, Statistical Department Goetghebeur, Els; Universiteit Gent, Applied mathematics, computer science and statistics Otter, Renee; Scientific Institute of Public Health, Belgian Cancer Centre Mortelmans, Katrien; KaMoCo Missinnne, Sarah; Scientific Insitute of Public Health, Belgian Cancer Centre Arbyn, Marc; Scientific Insitute of Public Health, Unit Cancer Epidemiology http://bmjopen.bmj.com/ Bouland, Catherine; Universite Libre de Bruxelles, Centre de Recherche en Santé Environnement et Santé au travail de Brouwer, Christophe; Universite Libre de Bruxelles, Centre de Recherche en Santé Environnement et Santé au travail
Primary Subject Epidemiology Heading:
Secondary Subject Heading: Public health on September 29, 2021 by guest. Protected copyright. Epidemiology < ONCOLOGY, ONCOLOGY, Organisation of health services < Keywords: HEALTH SERVICES ADMINISTRATION & MANAGEMENT
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 1 of 36 BMJ Open
1 2 3 4 1 Predicting End of Work Disability After Cancer. A 7-year Follow-up 5 2 Population-based Cohort Study Using Competing Risks Analysis. 6 7 8 3 R. Kiasuwa Mbengia,c, M.A Nicolae b, Els Goetghebeurb, R. Ottera, K. Mortlemansd, S. Missinne a, 9 4 M. Arbyne, C. Boulandc and C. de Brouwerc 10 11 5 a Belgian Cancer Centre, Scientific Institute of Public Health, Brussels (WIV-ISP), Belgium 12 6 b Stat-Gent CRESCENDO, University of Ghent 13 7 c Research Centre for Environmental and Occupational Health, Brussels School of Public Health, Université Libre de Bruxelles (ESP-ULB) 8 d Dr. Katrien Mortelmans, PhD Consulting (KaMoCo) 14 9 e Unit Cancer Epidemiology, Scientific Institute of Public Health, Brussels (WIV-ISP), Belgium 15 10 For peer review only 16 11 Corresponding author: Régine Kiasuwa Mbengi 17 12 [email protected] 18 13 14 rue Juliette Wytsman, 1050 Bruxelles - Belgium 19 14 Office: +32 2 642 57 65 20 15 Mobile: +32 479 3926 58 21 22 23 16 ABSTRACT 24 25 17 Objectives 26 18 The number of workers who are disabled due to cancer is increasing dramatically worldwide. One of the 27 19 main priorities is to preserve their quality of life (QoL) and the sustainability of financing systems for 28 20 sickness absence benefits. We have carried out this study in order to assess factors associated with the 29 21 return to ability to work after cancer-related work disability. This should help with planning of 30 22 rehabilitation needs and tailored programmes. 31 23 Participants 32 24 We conducted this register-based cohort study using individual data from the Belgian Disability http://bmjopen.bmj.com/ 33 25 Insurance. Data on 15, 543 socially insured Belgian people who entered into work disability due to cancer 34 26 in 2007-2011 were used. 35 36 27 Primary and secondary outcome measures 37 28 We estimated the time spent in disability and the cause-specific cumulative incidences of return to ability 38 29 to work stratified by age, gender, occupational class, year of entrance into work disability for 11 cancer 39 30 groups using Kaplan Meier analyses and Cox regression allowing for competing risks.
40 31 Results on September 29, 2021 by guest. Protected copyright. 41 32 The overall median time spent into work disability is 1.59 years (95%CI [1.52-1.66]), ranging from 0.75 42 33 to 4.98 across the 11 cancer groups. By the end of follow-up, more than one-third of work disabled cancer 43 34 survivors were able to RTW (35%). While larger proportions of women were able to RTW at the end of 44 35 follow-up, men who do RTW are able to do so sooner. Women, white-collar, younger and having 45 36 haematological, male genital or breast cancers were the most likely to be able to RTW. 46 37 Conclusion 47 38 Good prognostic factors for the ability to RTW are young age, being female, being a white-collar worker, 48 49 39 and having breast or haematological cancers. 50 40 Looking at our results together with the cancer incidence predictions up to 2025 has a high value for the 51 41 purposes of social security and rehabilitation planning and ascertaining patients’ perspectives. 52 42 53 43 Key words: sickness absence, cancer survivors, competing risks, predictive model, social inequalities 54 44 55 45 56 46 57 47 58 59 Version 30 août 2016 60 1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 2 of 36
1 2 3 48 Strengths and limitations of this study: 4 49 • (Good) Generalizability: we used a population-based dataset without loss of follow-up; the 5 6 50 external validity is therefore barley not limited; 7 51 • Methods: we add competing risks to the traditional survival analysis to respect the complexity of 8 52 the outcomes which is still rarely performed in disability studies; 9 53 • High value of our results (if linked with cancer incidence predictions) for the planning of 10 54 rehabilitation needs of cancer patients, up to 2025; 11 55 • Identification of cancer survivors being at risk to be socio-economically excluded; 12 56 • We miss information on treatments and job demands: these could have helped to (1) precise the 13 57 risk profile and (2) tailor RTW interventions 14 58 15 For peer review only 16 59 17 60 BACKGROUND 18 19 61 The direct and indirect effects of work disability represent an important burden for people who are absent 20 21 62 due to sickness, their families and employers [1]. Long-term work disability may lead to social exclusion, 22 23 63 deprivation or economic insecurity [2], as well as poor health [3]. The negative impact of work disability 24 25 64 on both social and health status is of high importance for public health [4] but studies identifying those 26 27 65 cancer survivors who are at risk of experiencing long-term work disability and identifying the avoidable 28 29 30 66 proportion of work disability are lagging behind. 31 32 67 http://bmjopen.bmj.com/ 33 68 Work disability imposes significant costs on society [5,6] with up to 5% of GDP in Organisation for 34 35 36 69 Economic Co-operation and Development (OECD) countries being spent on disability benefits [5]. In 37 38 70 2010, the OECD published a report describing the barriers to (re)integration in the labour market for 39
40 71 people in work disability (i.e. greater competition, heavier workload and work pressure) [5]. The report on September 29, 2021 by guest. Protected copyright. 41 42 72 also describes the underlying social and economic tragedies. Since the results for Belgium were poor, 43 44 73 with a decrease in the number of people with disabilities employed over the past decade, authorities and 45 46 74 social security administrators have been looking for measures or interventions to reverse the trend. A 47 48 49 75 number of studies have been performed to support the authorities, but these are mainly qualitative and are 50 51 76 based on a small sample of cancer survivors [7-13]. 52 53 77 54 55 56 57 58 59 Version 30 août 2016 60 2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 3 of 36 BMJ Open
1 2 3 78 Insurance medicine researchers and epidemiologists acknowledge differences between diagnoses in terms 4 5 79 6 of the duration of work disability [14,15]. Overall, the leading causes of work disability are 7 8 80 musculoskeletal disorders and mental health problems, which have been widely studied [16]. 9 10 81 In Belgium, cancer is the fifth largest cause of work disability, with 18,462 people in work disability due 11 12 82 to cancer in 2013 (6.2% of all disabled workers in Belgium) [17] (Table 1). Each year, more than 25, 000 13 14 83 Belgian inhabitants of working age (20-64 years) are diagnosed with cancer1. 15 For peer review only 16 84 During the last decade, cancer treatments in middle and high income countries have seen major 17 18 85 19 improvements, leading to increased rates of cancer survival [18,19]. Despite improved survival, the 20 21 86 disclosure of cancer still causes important distress among the people affected and their relatives [20], and 22 23 87 it is associated with work disability or death by their colleagues and supervisors [7,21-24]. 24 25 88 This automatic association of cancer with death is becoming obsolete, however, as was notably shown 26 27 89 through the study by Dal Maso et al. [25] which showed that a quarter of Italian cancer survivors have 28 29 90 reached a death rate similar to that of the general population. 30 31
91 http://bmjopen.bmj.com/ 32 Cancer survivors can experience physiological and/or psychosocial symptoms, due to side or long-term 33 34 92 effects of treatment [26] and are more likely to report fair or poor health overall in all age groups [27]. 35 36 93 For these survivors, work can represent a return to health or “normality” or a need to safeguard their 37 38 94 financial security, self-esteem and social contacts [28-33]. 39
40 95 on September 29, 2021 by guest. Protected copyright. 41 42 96 Little is known about social inequalities in relation to return to work (RTW) among cancer survivors [34]. 43 44 97 45 However, the well-established relationship between socio-economic position (SEP) and long-term 46 47 98 sickness absence predicts that returning to work will be more difficult for cancer survivors working in 48 49 99 manual occupations [35,36]. Previous research has shown that working conditions and psychosocial 50 51 100 conditions in manual occupations act as additional barriers [35,37-39]. Alongside the impact of working 52 53 54 55 1 Numbers are computed online. Official website of the Belgian Cancer Registry: 56 http://www.kankerregister.org/default.aspx?lang=EN 57 58 59 Version 30 août 2016 60 3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 4 of 36
1 2 3 101 conditions, the unequal use of cancer rehabilitation services [40] may also lead to social inequalities in 4 5 102 6 terms of returning to work. It has also been shown that cancer survivors with a low SEP more commonly 7 8 103 become unemployed [41] or take early retirement, which can act as a substitute for sickness absence 9 10 104 benefits or unemployment [41-43]. 11 12 105 13 14 106 This article describes and discusses the results of a population-based cohort study of people with long- 15 For peer review only 16 107 term cancer-related work disability, i.e. receiving sickness absence benefits for more than one year. We 17 18 108 19 will refer to this population below as ‘disabled workers’. They are followed for 3 to 7 years in order to 20 21 109 measure the outflow from work disability to either retirement, ability to RTW or death. 22 23 24 110 The Belgian context 25 26 111 In Belgium, cessation of work due to sickness has to be reported to the employer immediately. The 27 28 112 employer pays the “guaranteed salary” for 14 working days for blue-collar workers (manual workers) 29
30 2 31 113 and 28 for white-collar workers (intellectual workers) . For self-employed workers or unemployed 32 http://bmjopen.bmj.com/ 33 114 people, the SSS covers salary replacements starting after 28 working days. The absence due to sickness 34 35 115 has to be confirmed by a general practitioner or a specialist doctor. 36 37 116 After the period of guaranteed income, the social security system (SSS) takes over the provision of a 38 39 117 replacement income. These benefits due to sickness-related absences vary between 40% and 65% of the
40 on September 29, 2021 by guest. Protected copyright. 41 118 reference salary, depending on the family situation. 42 43 44 119 The SSS distinguishes between short-term and long-term work disability. Short-term work disability has 45 46 120 duration of 1 year, while long-term work disability is for periods exceeding 1 year. The division reflects a 47 48 121 different evaluation method to assess the worker’s eligibility for sickness absence benefits and also the 49 50 122 calculation of the level of sickness absence benefits. 51 52 53 54 2 The legal distinction between both OCs is based on the “nature” of the work performed: mainly manual jobs for blue-collar 55 workers and intellectual jobs for white-collar workers. Official website of the national authority of employment: 56 http://www.belgium.be/fr/emploi/contrats_de_travail/types_de_contrats/nature_du_travail. 57 58 59 Version 30 août 2016 60 4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 5 of 36 BMJ Open
1 2 3 123 Entitlement to long-term sickness absence benefits starts from the second year after stopping work (13th 4 5 124 6 month) until the age of retirement, with no limit to its duration. This applies to employees, self-employed 7 8 125 and unemployed socially insured Belgian citizens, while civil servants (almost 20% of the Belgian 9 10 126 workforce) benefit from a specific social security scheme. In Belgium, more than 90% of citizens are 11 12 127 socially insured and covered by compulsory health insurance [44]. 13 14 128 Our work reflects research on work disability due to cancer. Work disability is defined or measured as a 15 For peer review only 16 129 legal status based on administrative definitions, i.e. eligibility for benefit. In 2011, a new measure was 17 18 130 19 implemented, allowing disabled workers to resume work without prior agreement of the health insurer's 20 21 131 medical advisor. 22 23 132 24 25 133 There is an important knowledge gap in Belgium regarding quantitative assessment of the impact of 26 27 134 cancer on work disability. The following aspects need to be better understood: how long it lasts, how it 28 29 135 ends, which workers are more at risk, etc. Our research helps to fill this gap. It is based on a recent model, 30 31
136 http://bmjopen.bmj.com/ 32 developed in 2011 to study RTW after cancer [22] which proposes a comprehensive list of influencing 33 34 137 factors. Among these, we have been able to collect and analyze the following: age, gender, occupational 35 36 138 class, cancer site and work-related outcomes (ability to work, (early) retirement, death and disability). 37 38 139 39
40 140 This study is part of the scientific approach initiated in 2012 at the request of the Federal Ministry of on September 29, 2021 by guest. Protected copyright. 41 42 141 Public Health and Social Security [45,46], to provide evidence and support to the decision-making 43 44 142 45 process to improve and facilitate the professional reintegration of cancer survivors. 46 47 143 48 49 144 50 51 145 52 53 146 54 55 147 56 57 58 148 59 Version 30 août 2016 60 5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 6 of 36
1 2 3 4 149 Methods 5 6 7 150 Design & Statistical analysis 8 9 10 151 We conducted a register-based cohort study, using data from the disability register of the National 11 12 152 Institute for Health and Disability Insurance (NIHDI). Our research has three goals. Our first goal was to 13 14 153 measure the between time spent in disability and age, gender, year of entrance, occupational class and 11 15 For peer review only 16 154 cancer groups covering all cancer sites observed. To achieve this first goal, we calculated the Kaplan- 17 18 155 Meier estimate of time to event and non-parametric and model-based cause-specific cumulative 19 20 156 incidences. 21 22 23 157 Secondly, following the taxonomy set out in theories of work disability [48], our study aimed to build a 24 25 158 prognostic model to estimate the distribution of time spent in disability in the presence of three competing 26 27 159 events: death, (early) retirement, and ability to RTW. For this second objective, we fitted a Cox 28 29 160 proportional hazards model to the cause-specific hazards using a stratified (Model 1, Table 4) and 30 31 161 separated approach (Model 2, Table 5), for each age group and gender, accounting for occupational class, 32 http://bmjopen.bmj.com/ 33 162 cancer groups and year of entrance. 34 35 36 163 A third objective was to investigate social inequalities in work disability among cancer survivors, paying 37 38 164 attention to differences in age, gender and occupational class. For this, we fitted a Cox proportional 39
40 165 hazards model to the cause-specific hazards using a stratified and separated approach for each age group on September 29, 2021 by guest. Protected copyright. 41 42 166 and gender, accounting for occupational class and four categories of cancer: those with low3, medium4 43 44 167 and high5 survival rates, according to the age-standardized 5-year relative survival (ASRS), calculated by 45 46 168 the Belgian Cancer Registry [49]. The –fourth- missing category includes those cancer sites for which the 47 48 49 50 51 52 3 "Esophagus", "Stomach", "Colon & rectum", "Pancreas", "Oth. Mal. Neop. of digestive organs and peritoneum", "Mesothelioma", "Trachea and lung", "Myeloid and others", "CNS", "Oth. malignancies and undefined sites, invasive" 53 4 54 "Lip and oral cavity, nasal cavities, middle ear and accessory sinuses, pharynx, larynx","Non Hodgkin's Disease", "Uterus", "Cervix","Ovary","Oth. mal. neop. of female genitals", "Cervix", "Ovary", "Oth. mal. neop. of female genitals", "Bladder", "Urinary system other 55 than bladder", "Bone and connective tissue" 56 5 "Hodgkin Disease","Acute lymphoid leukemia and lymphoid leukemia, other", "Breast female","Uterus","Kidney","Melanoma of the skin", 57 "Thyroid and other endocrine glands" 58 59 Version 30 août 2016 60 6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 7 of 36 BMJ Open
1 2 3 169 information on the ASRS was not available6. The two rationales behind this approach were as follows: 4 5 170 6 first, it generates a parsimonious model (it avoids the lack of convergence due to the large size of the data 7 8 171 set). Second, this approach makes it possible to account for the “severity” of the disease. 9 10 172 11 12 173 Study population 13 14 174 15 We included all sociallyFor insured peer Belgian people review who were recognized only as work disabled due to cancer on 16 st st 17 175 1 January 2007 and up to 31 December 2011. It must be remembered that civil servants are not included 18 19 176 in the NIHDI database. This cohort comprises 21,701 individuals of whom 6,098 were excluded due to 20 21 177 work disability having started before 1st January 2007 (and non-equivalent follow-up time) or due to 22 23 178 inconsistent records (see Figure 1). The last update of the data was on 31st December 2013, leading to a 24 25 179 maximum of 7 years follow-up. 26 27 28 29 180 Independent prognostic variables 30 31 181 Socio-demographic characteristics included in our study are age at entrance into work disability, 32 http://bmjopen.bmj.com/ 33 182 gender and occupational class (OC). The age variable was based on the date of birth and was further 34 35 183 categorized into four groups: 17-39; 40-49; 50-59 and 60+ years. Occupational classes were based on 36 37 38 184 four categories: blue-collar workers, white-collar workers, self-employed people and assisting spouses 39
40 185 and were recoded into a 3-level variable: blue-collar workers, white-collar workers and self-employed on September 29, 2021 by guest. Protected copyright. 41 42 186 people. 43 44 187 In total, 39 cancer sites have been identified using the ‘pathology codes’ transmitted by the NIHDI and 45 46 188 registered by their ICD-9 codes (Table 2). For the sake of comparability, we translated these into ICD-10 47 48 189 codes7 and gathered them into 11 groups (Table 2). 49 50 51 190 The year of entrance in disability was a continuous variable ranging from 2007 to 2011. We decided to 52 53 191 recode the year of entrance into a two-level variable: 2007-2010 and 2011. This decision is based on an 54 55 6 "Benign tumour", "Mal. neop. of skin other than melanoma", "Oth. malignancies and undefined sites, CIS", "Tumours of uncertain and 56 unspecified behavior" 57 7 Online validated tool to translate ICD-9 codes into ICD-10 codes: http://www.icd10data.com/Convert . 58 59 Version 30 août 2016 60 7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 8 of 36
1 2 3 192 exploratory analysis that showed significant difference in survival pattern between disability acquired 4 5 193 6 before or after 2011. We recoded the year of entrance into a two-level variable: 2007-2010 and 2011 (log- 7 8 194 rank test=502, df=1, p-value < 0.001) (Figure 2). 9 10 11 195 Outcome variables: the three competing events 12 13 196 The outcome variable is the event that causes the end of work disability. We defined three mutually 14 15 197 exclusive events: Fordeath, retirement peer and ability review to return-to-work (RTW).only 16 17 18 198 The status “able to RTW” means that the cancer-disabled worker has been recognized by a health 19 20 199 insurer's occupational doctor as “able to work”. In practice, this might lead to a return to work, 21 22 200 unemployment or a decision to be a housewife/husband. “Retirement” means that the worker is definitely 23 24 201 out of the labour market due to his age and will receive social benefits until death. 25 26 202 Those long-term disabled workers, who have not experienced any event by the end of follow-up, i.e. on 27 28 203 31st December 2013, were administratively censored (31%, Table 3). 29 30 31 32 http://bmjopen.bmj.com/ 33 204 Results 34 35 36 37 205 Description of the study population 38 39 206 No observed workers were lost to follow-up. Table 3 describes the main characteristics of work disabled
40 on September 29, 2021 by guest. Protected copyright. 41 207 cancer survivors included in the study. 42 43 208 The overall mean age at entry is 48.5 years (SD=8.7 years), with a mean age among men of 50.4 years 44 45 209 46 (SD=9 years) and 47.3 years among women (S.D=8.33 years). The majority (77%) of the cancer-disabled 47 48 210 workers are aged 40–59 years. 49 50 211 Women are over-represented (62%), younger at entry (median age is 48 vs. 53 years in men) and mostly 51 52 212 white-collar workers (46% vs. 21% among men) or blue-collar workers (43% vs. 60%). The outcome for 53 54 213 the majority of cancer disabled women is ability to RTW (44.10%), while for most men the outcome is 55 56 214 death (50.86%). 57 58 59 Version 30 août 2016 60 8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 9 of 36 BMJ Open
1 2 3 215 The most frequent cancer site is breast with 35% (n= 5,949) of disabled workers, followed by 15% 4 5 216 6 (n=2,400) of digestive tract and 9% (n=1,417) of respiratory tract cancers. 7 8 217 Regarding OC, half of the disabled workers are blue-collar workers, the majority of whom (38.2% of the 9 10 218 total) die by the end of follow-up. White-collar workers (37%) have the shortest median time spent in 11 12 219 work disability (1.30 years vs. 1.79 for the others) and the majority of them (47.7%) are able to RTW by 13 14 220 the end of follow-up. Self-employed disabled workers represent 13% of the cohort and the majority of 15 For peer review only 16 221 them (38.2%) die by the end of follow-up. 17 18 222 19 By the end of follow-up, 69% of the cohort have experienced one of the three competing events (32.2% 20 21 223 died, 2.2% retired and 34.6% have been able to RTW). The other 31% remained disabled, distributed as 22 23 224 follows: 35% of those who entered in 2009 and 2010, 31% of those aged 40-49 years, 27% of women, 24 25 225 28% of blue-collar workers and 44% of those with benign or in situ tumours. 26 27 226 Figures 3a-e show the non-parametric cause-specific cumulative incidences of time to ability to RTW in 28 29 227 the presence of competing risks. Figures 4a-d show the box plots of time to any event stratified by each 30 31
228 http://bmjopen.bmj.com/ 32 prognostic variable, respectively. 33 34 229 For all prognostic variables, the curves show a steep increase in ability to RTW within the first two years; 35 36 230 later on, the curves virtually level off. 37 38 231 Young workers (17-39 years) have the highest rates of ability to RTW at the end of follow-up (Figure 3a) 39
40 232 and the shortest periods spent in work disability (Figure 4b). Older workers present the shortest time spent on September 29, 2021 by guest. Protected copyright. 41 42 233 in work disability (Figure 4b), mainly due to death and retirement (77%, Table 3). 43 44 234 45 Women have higher rates of ability to RTW compared to men (Figure 3b) but spend longer periods in 46 47 235 work disability (Figure 4a). White-collar workers have higher rates of work disability and spend shorter 48 49 236 periods in it (Figure 4c). Regarding the cancer groups, workers with breast or haematological cancer have 50 51 237 the highest rates of ability to RTW by the end of follow-up (Figure 3c), but the longest periods spent in 52 53 238 work disability (Figure 4e). Those with respiratory tract, head and neck, digestive or CNS cancers have 54 55 239 the lowest rates of ability to RTW (Figure 3c) and a short time spent in work disability (Figure 4e), 56 57 58 240 mainly due to death. 59 Version 30 août 2016 60 9
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 10 of 36
1 2 3 241 Those cancer survivors who entered into work disability in 2011 had higher rates of ability to RTW by the 4 5 242 6 end of follow-up (Figure 3e) and shorter periods of time spent in work disability (Figure 4c). 7 8 9 243 The prediction of ending work disability patterns (model 1) 10 11 244 In Table 4 we report the results of a Cox proportional hazards model on the cause-specific hazards of each 12 13 245 competing event, with stratification by age and gender and allowing interaction between gender and the 14 15 246 prognostic variables.For Good prognostic peer factors reviewfor ability to RTW for onlyboth men and women are disability 16 17 18 247 experienced before 2011 and being a white-collar worker. Regarding the 11 cancer groups, men with 19 20 248 haematological or genital organ cancers are the most likely to be able to RTW. Among women, the 21 22 249 cancer groups with the best chance for ability to RTW are haematological and breast. 23 24 250 Concerning deaths among men, disabled workers with respiratory tract, CNS, bone & connective tissue 25 26 251 cancers are most at risk. Among women, those with respiratory tract, female genital organs, digestive 27 28 252 tract and head and neck cancers are the more at risk. 29 30 31 32 253 Social inequalities in work disability of cancer survivors (model 2) http://bmjopen.bmj.com/ 33 34 254 In the second model we stratify by age and gender and allow interactions between both these variables 35 36 255 and occupational class and survival categories (Table 5). The absence of individuals in certain age 37 38 256 categories experiencing retirement (17-49 years, Table 2) leads to a convergence issue when modelling 39
40 on September 29, 2021 by guest. Protected copyright. 41 257 the cause-specific hazard for this type of event and this is therefore not reported. 42 43 258 Table 5 shows that among men, workers aged 40-49 and 50-59 years and blue-collar workers are less 44 45 259 likely (just over half as likely) to be able to RTW compared to white-collar workers. These results 46 47 260 translate larger social inequalities in the 40-49 and 50-59 age groups compared to the youngest cancer 48 49 261 survivors (blue-collar workers are 22% less likely to be able to RTW in age category 17-39 compared to 50 51 262 55% less likely in age category 40-49). There are no differences between blue-collar and white-collar 52 53 54 263 workers for cancer survivors older than 60 years. Similar results are found for females. 55 56 57 58 59 Version 30 août 2016 60 10
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 11 of 36 BMJ Open
1 2 3 264 For both genders, self-employed people are less likely to be able to RTW work than white-collar workers 4 5 265 6 from the age of 40 onwards. 7 8 266 9 10 267 DISCUSSION 11 12 13 268 In this study, we aimed to identify the factors that influence the reason for exiting from work disability 14 15 269 and the length of workFor disability peer among cancer review survivors. only 16 17 270 To achieve this, we first measured the association between the time spent in work disability and age, 18 19 271 gender, occupational class, the year of entrance and 11 cancer groups. Secondly, we estimated the 20 21 272 distribution of reasons for ending work disability in the presence of three competing events and thirdly we 22 23 273 investigated social inequalities in work disability among cancer survivors. 24 25 26 274 27 28 275 Since not many of the studies in this field are population-based, making comparisons is not easy. 29 30 276 However, the impact of these – risk - factors on labour market participation has been tested in previous 31 32 277 studies [47]. Authors report that overall, older age at entry and being male are both factors that decrease http://bmjopen.bmj.com/ 33 34 278 the chance of being economically active. Our results show that being older (>60 years) increase the risk 35 36 279 of dying or retiring and that workers aged 40-49 are the least likely to be able to RTW. Being male does 37 38 39 280 indeed reduce the likelihood of being able to RTW but women spend more time in work disability overall.
40 on September 29, 2021 by guest. Protected copyright. 41 281 42 43 282 Regarding the cancer groups, we found a strong association between respiratory tract, head and neck and 44 45 283 digestive tract cancers and death. The first two include smoking-related cancer sites [50], which represent 46 47 284 major sources of work disability and death in the working age population. 48 49 285 Haematological cancers in both genders, breast cancer among women and male genital cancers among 50 51 52 286 men, were found to be positively associated with the ability to RTW. 53 54 287 Other studies have compared different cancer sites to assess their association with employment status 55 56 288 after cancer diagnosis. In line with our results, workers with respiratory and female genital cancers 57 58 59 Version 30 août 2016 60 11
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 12 of 36
1 2 3 289 present smaller proportions of employment than workers with breast or haematological cancers, mainly 4 5 290 6 due to poor self-reported health status [26,27, 51]. 7 8 291 9 10 292 Compared to white-collar workers, blue-collar workers and self-employed disabled workers are less likely 11 12 293 to be able to RTW, especially men aged 40-59 years. This is in line with previous research showing that, 13 14 294 overall, blue-collar workers and self-employed people are less likely to be able to work compared to 15 For peer review only 16 295 white-collar workers [26]. In addition to working conditions [52], later stage at diagnosis, differences in 17 18 296 19 treatment [53] and lower participation in rehabilitation services [54] can contribute to these social 20 21 297 inequalities. However, in our competing risk analysis, such inequalities were not found among men, while 22 23 298 the opposite was found among women. OC is also strongly associated with the level of income which 24 25 299 may represent an incentive to RTW when these are significantly different (higher) than sickness absence 26 27 300 benefits [36]. 28 29 301 A different impact of the OC on the risk of work disability according to age and gender has been shown to 30 31
302 http://bmjopen.bmj.com/ 32 exist in a population in a Norwegian county, where young workers with blue-collar jobs are more at risk 33 34 303 than older men [55]. The association between of age and RTW has been reported with contradictory 35 36 304 results in the literature, but the majority found higher age to be associated with later RTW or reduced 37 38 305 chance of employment [34]. Our results show that for Belgian cancer survivors, the opposite is found with 39
40 306 a larger impact of occupational class from the age of 40 years onwards, compared to their younger on September 29, 2021 by guest. Protected copyright. 41 42 307 counterparts. 43 44 308 45 46 47 309 Demographic change and the rising retirement age will increase the number of disabled workers and the 48 49 310 length of work disability, which combined with the effects of the economic crisis (i.e. greater competition 50 51 311 and emphasis on maximum performance), will worsen the situation if we do not implement measures, 52 53 312 interventions and rehabilitation programmes to (re)integrate disabled workers better in the labour market 54 55 313 [5]. 56 57 58 59 Version 30 août 2016 60 12
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 13 of 36 BMJ Open
1 2 3 314 The measures introduced by the Belgian government by the end of 20108 seem to already have had an 4 5 315 6 impact, as those disabled workers who entered in 2011 show better outcomes than the others. Further 7 8 316 studies need to be carried out in future to confirm this trend. However, at the end of follow-up, only 9 10 317 34.6% were able to RTW and 31% were –administratively- censored, remaining disabled. 11 12 318 13 14 319 Strengths, limitations and needs for further research 15 For peer review only 16 17 320 The main strength of this study is the representativeness of the data and the generalizability of our results. 18 19 321 We included in the analysis all Belgian workers disabled due to cancer between 2007- 2011, excluding 20 21 322 civil servants and individuals for whom we detected coding errors. 22 23 323 24 25 324 In most work disability studies, survival analyses are used to estimate the time to an event of interest. The 26 27 325 end of work disability is, however, more complex than this and it may be caused by multiple factors. 28 29 30 326 Therefore, the use of competing risks analysis becomes appropriate to avoid over or under-estimating the 31 32 327 probability of experiencing each event [56]. This model is still rarely used in work disability studies and http://bmjopen.bmj.com/ 33 34 328 its use should be encouraged. 35 36 329 37 38 330 Regarding the objective of predicting disability, the two models showed their ability and effectiveness in 39 40 331 predicting the length and the reasons for ending work disability among Belgian cancer survivors. Our on September 29, 2021 by guest. Protected copyright. 41 42 43 332 second model presents original findings, using the survival rates to identify social inequalities. 44 45 333 Nevertheless, differences still exist among workers of the same age, gender, occupational class and cancer 46 47 334 type; this calls for a more complex model with information on treatment and symptoms [57] and on the 48 49 335 working environment [58]. This is feasible in the future, e.g. by linking data from Cancer Registries to 50 51 336 data on employment and socio-economic status. Results could be used to develop rehabilitation 52 53 337 programmes for cancer survivors similar to those that already exist in other countries [59-62]. 54 55 56 8 Since 2011, disabled workers can resume work (to an adequate extent) without prior authorisation from the health insurer's 57 occupational doctor. 58 59 Version 30 août 2016 60 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 14 of 36
1 2 3 338 4 5 339 6 While our paper focuses on work disability among cancer survivors in Belgium, it is important to realize 7 8 340 that the methods and principles used are generic and applicable to addressing work disability as whole. 9 10 341 Therefore, this report is also relevant to other conditions and SSS. This paper contributes towards closing 11 12 342 the gap on the transition among cancer survivors from long-term work disability to ability to RTW. 13 14 343 Linking these important results to predictions of cancer incidence9 should make it possible to plan cancer 15 For peer review only 16 344 rehabilitation needs and related sickness absence benefits. 17 18 19 20 21 345 Contributorship statement 22 346 23 24 347 Régine Kiasuwa Mbengi, MPH, PhD Student 25 26 348 Designed the study 27 349 Requested/Collected the data 28 350 Performed the analysis 29 351 Wrote the paper 30 352 31 32 353 Mioara Alina Nicolae, MSc,PhD http://bmjopen.bmj.com/ 33 354 Provided statistical support to build the models 34 355 Provided advices on the writing of the sections on methods and results 35 356 36 357 Prof. Els Goetghebeur, MSc,PhD 37 358 Provided statistical support to build the models 38 39 359 Provided advices on the writing of the sections on methods and results
40 360 on September 29, 2021 by guest. Protected copyright. 41 361 Renee Otter, MD, PhD 42 362 Provided clinical expertise support to build cancer categories 43 363 Substantially contributed to the writing 44 45 364 46 365 Katrien Mortelmans, MD, PhD 47 366 Substantially contributed to the writing 48 367 49 368 Sarah Missine, MSc, PhD 50 369 Provided advice and support to integrate the inequalities perspective in the paper and in the 51 52 370 statistical model 53 371 Supported the writing of the paper 54 55 9 56 Cancer incidence predictions up to 2025 (Belgian Cancer Registry): http://www.kankerregister.org/Tenyears. 57 58 59 Version 30 août 2016 60 14
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 15 of 36 BMJ Open
1 2 3 372 4 5 373 Marc Arbyn, MD 6 374 Substantially contributed to the preparation of the tables and figures and the abstract section 7 375 8 376 Prof. Catherine Bouland, MsC, PhD 9 377 As the co-superviser of the PhD of Miss Kiasuwa, Prof. C. Bouland substantially contributed to 10 378 11 the preparation of the study, the design and the revision of the paper 12 379 13 380 Prof. Christophe de Brouwer, MD, PhD 14 381 As the superviser of the PhD of Miss Kiasuwa, Prof. C. de Brouwer substantially contributed to 15 382 the preparation Forof the study, peer the design and review the revision of the paperonly 16 383 17 384 18 19 385 Funding and Competing interests 20 21 386 22 This study has been funded by the National Institute for Health and Disability Insurance. We 23 24 387 have read and understood BMJ policy on declaration of interests and declare that we have no 25 26 388 competing interests. 27 28 29 389 This article does not contain any studies with human participants or animals performed by any of 30 31 390 the authors. 32 http://bmjopen.bmj.com/ 33 391 34 35 36 392 Data sharing statement 37 38 393 39 The complete and anonym dataset used for this study can be made available from the
40 on September 29, 2021 by guest. Protected copyright. 41 394 corresponding author for researchers interested in comparative studies. This request would be 42 43 395 subject of approval from the Belgian National Institute for Health and Disability Insurance, from 44 45 46 396 the Scientific Institute of Public Health and the Université Libre de Bruxelles. 47 48 397 49 398 50 51 399 52 53 400 54 401 55 56 402 57 58 403 59 Version 30 août 2016 60 15
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 16 of 36
1 2 3 404 Reference List 4 405 5 6 406 1. Amir Z, Wilson K, Hennings J, Young A: The meaning of cancer: implications for family finances and 7 407 consequent impact on lifestyle, activities, roles and relationships. Psychooncology 2012, 21: 1167-1174. 8 9 408 2. Daniel C.Lustig: Causal Relationships Between Poverty and Disability. Rehabilitation Counseling Bulletin 10 409 2007, 50: 194-202. 11 12 410 3. Schuring M, Robroek SJ, Otten FW, Arts CH, Burdorf A: The effect of ill health and socioeconomic status 13 411 on labor force exit and re-employment: a prospective study with ten years follow-up in the 14 412 Netherlands. Scand J Work Environ Health 2013, 39: 134-143. 15 For peer review only 16 413 4. Reichard A., Stransky M., Phillips K., Drum C., Mc Clain M.: Does Type of Disability Matter to Public 17 414 Health Policy Practice? Californian Journal of Health Promotion 2015, 13: 25-36. 18 19 415 5. OECD. Sickness, Disability and Work. Breaking the Barriers. 2010. 20 416 Ref Type: Report 21 22 417 6. Luengo-Fernandez R, Leal J, Gray A, Sullivan R: Economic burden of cancer across the European Union: 23 418 a population-based cost analysis. Lancet Oncol 2013, 14: 1165-1174. 24 25 419 7. Betsch N: Kankerpatiënten en werkhervatting: rechten en plichten. Instituut Voor Arbeidsrecht, KU Leuven; 26 420 2013. 27 28 421 8. Desiron HA, Donceel P, de RA, Van HE: A conceptual-practice model for occupational therapy to 29 422 facilitate return to work in breast cancer patients. J Occup Rehabil 2013, 23: 516-526. 30 31 423 9. Desiron HA, Donceel P, Godderis L, Van HE, de RA: What is the value of occupational therapy in return http://bmjopen.bmj.com/ 32 424 to work for breast cancer patients? A qualitative inquiry among experts. Eur J Cancer Care (Engl ) 33 425 2015, 24: 267-280. 34 35 426 10. Desiron HA, Crutzen R, Godderis L, Van HE, de RA: Bridging Health Care and the Workplace: 36 427 Formulation of a Return-to-Work Intervention for Breast Cancer Patients Using an Intervention 37 428 Mapping Approach. J Occup Rehabil 2016. 38 39 429 11. Rommel W. Werken na kanker:welke problemen ervaren (ex-)patiënten die het werk hervatten? 2012. 40 430 Vlaams Liga Tegen Kanker. on September 29, 2021 by guest. Protected copyright. 41 431 Ref Type: Report 42 43 432 12. Tiedtke C, Donceel P, Knops L, Desiron H, Dierckx de CB, de RA: Supporting return-to-work in the face 44 433 of legislation: stakeholders' experiences with return-to-work after breast cancer in Belgium. J Occup 45 434 Rehabil 2012, 22: 241-251. 46 47 435 13. Tiedtke C, Dierckx de CB, Donceel P, de RA: Workplace support after breast cancer treatment: 48 436 recognition of vulnerability. Disabil Rehabil 2014, 1-7. 49 50 437 14. Alexanderson K, Norlund A: Swedish Council on Technology Assessment in Health Care (SBU). Chapter 51 438 12. Future need for research. Scand J Public Health Suppl 2004, 63: 256-258. 52 53 439 15. Alexanderson K, Hensing G: More and better research needed on sickness absence. Scand J Public Health 54 440 2004, 32: 321-323. 55 56 57 58 59 Version 30 août 2016 60 16
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 17 of 36 BMJ Open
1 2 3 441 16. Alexanderson K, Norlund A: Swedish Council on Technology Assessment in Health Care (SBU). Chapter 4 442 1. Aim, background, key concepts, regulations, and current statistics. Scand J Public Health Suppl 2004, 5 443 63: 12-30. 6 7 444 17. INAMI. Rapport annuel. Satistiques des indemnités. 2014. 8 445 Ref Type: Report 9 10 446 18. Coleman MP, Gatta G, Verdecchia A, Esteve J, Sant M, Storm H et al.: EUROCARE-3 summary: cancer 11 447 survival in Europe at the end of the 20th century. Ann Oncol 2003, 14 Suppl 5: v128-v149. 12 13 448 19. Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, Coebergh JW, Comber H et al.: Cancer incidence 14 449 and mortality patterns in Europe: estimates for 40 countries in 2012. Eur J Cancer 2013, 49: 1374-1403. 15 For peer review only 16 450 20. Pryce J, Munir F, Haslam C: Cancer survivorship and work: symptoms, supervisor response, co-worker 17 451 disclosure and work adjustment. J Occup Rehabil 2007, 17: 83-92. 18 19 452 21. Kennedy F HCMFPJ: Returning to work following cancer: a qualitative exploratory study into the 20 453 experience of returning to work following cancer. Eur J Cancer Care 2007, 16: 17-25. 21 22 454 22. Mehnert A: Employment and work-related issues in cancer survivors. Crit Rev Oncol Hematol 2011, 77: 23 455 109-130. 24 25 456 23. Mehnert A, de BA, Feuerstein M: Employment challenges for cancer survivors. Cancer 2013, 119 Suppl 26 457 11: 2151-2159. 27 28 458 24. Verbeek J, Spelten E, Kammeijer M, Sprangers M: Return to work of cancer survivors: a prospective 29 459 cohort study into the quality of rehabilitation by occupational physicians. Occup Environ Med 2003, 60: 30 460 352-357. 31 32 461 25. Dal Maso L, Guzzinati S, Buzzoni C, Capocaccia R, Serraino D, Caldarella A et al.: Long-term survival, http://bmjopen.bmj.com/ 33 462 prevalence, and cure of cancer: a population-based estimation for 818 902 Italian patients and 26 34 463 cancer types. Ann Oncol 2014, 25: 2251-2260. 35 36 464 26. Kiasuwa MR, Otter R, Mortelmans K, Arbyn M, Van OH, Bouland C et al.: Barriers and opportunities for 37 465 return-to-work of cancer survivors: time for action-rapid review and expert consultation. Syst Rev 38 466 2016, 5: 35. 39
40 467 27. Hewitt M, Rowland JH, Yancik R: Cancer survivors in the United States: age, health, and disability. J on September 29, 2021 by guest. Protected copyright. 41 468 Gerontol A Biol Sci Med Sci 2003, 58: 82-91. 42 43 469 28. Amir Z, Brocky J: Cancer survivorship and employment: epidemiology. Occup Med (Lond) 2009, 59: 44 470 373-377. 45 46 471 29. Collins C.G, Ottati A., Feuerstein M: Cancer Survivorship. In Handbook of Work Disability. Prevention and 47 472 Management. Springer edition. Edited by Loisel P, Anema J.R. New York: 2013:289-302. 48 49 473 30. Tiedtke C, de RA, Dierckx de CB, Christiaens MR, Donceel P: Experiences and concerns about 'returning 50 474 to work' for women breast cancer survivors: a literature review. Psychooncology 2010, 19: 677-683. 51 52 475 31. Tiedtke C, Dierckx de CB, de RA, Christiaens MR, Donceel P: Breast cancer treatment and work 53 476 disability: patient perspectives. Breast 2011, 20: 534-538. 54 55 477 32. van Muijen P., Duijts SF, van der Beek AJ, Anema JR: Prognostic factors of work disability in sick-listed 56 478 cancer survivors. J Cancer Surviv 2013, 7: 582-591. 57 58 59 Version 30 août 2016 60 17
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 18 of 36
1 2 3 479 33. Grunfeld EA, Cooper AF: A longitudinal qualitative study of the experience of working following 4 480 treatment for gynaecological cancer. Psychooncology 2012, 21: 82-89. 5 6 481 34. van Muijen P., Weevers NL, Snels IA, Duijts SF, Bruinvels DJ, Schellart AJ et al.: Predictors of return to 7 482 work and employment in cancer survivors: a systematic review. Eur J Cancer Care (Engl ) 2013, 22: 8 483 144-160. 9 10 484 35. Sterud T, Johannessen HA: Do work-related mechanical and psychosocial factors contribute to the social 11 485 gradient in long-term sick leave: a prospective study of the general working population in Norway. 12 486 Scand J Public Health 2014, 42: 329-334. 13 14 487 36. Piha K, Laaksonen M, Martikainen P, Rahkonen O, Lahelma E: Interrelationships between education, 15 488 occupationalFor class, income peer and sickness absence.review Eur J Public Health only 2010, 20: 276-280. 16 17 489 37. Melchior M., Krieger N, Kawachi I., et al.: Work Factors and Occupational Class Disaprities in Sickness 18 490 Absence: Findings From the GAZEL Cohort Study. American Journal of Public Health 2005, 95: 1206- 19 491 1212. 20 21 492 38. Melchior M, Goldberg M, Krieger N, Kawachi I, Menvielle G, Zins M et al.: Occupational class, 22 493 occupational mobility and cancer incidence among middle-aged men and women: a prospective study 23 494 of the French GAZEL cohort*. Cancer Causes Control 2005, 16: 515-524. 24 25 495 39. Melchior M, Krieger N, Kawachi I, Berkman LF, Niedhammer I, Goldberg M: Work factors and 26 496 occupational class disparities in sickness absence: findings from the GAZEL cohort study. Am J Public 27 497 Health 2005, 95: 1206-1212. 28 29 498 40. Holm LV, Hansen DG, Larsen PV, Johansen C, Vedsted P, Bergholdt SH et al.: Social inequality in cancer 30 499 rehabilitation: a population-based cohort study. Acta Oncol 2013, 52: 410-422. 31 32 500 41. Lindbohm ML, Kuosma E, Taskila T, Hietanen P, Carlsen K, Gudbergsson S et al.: Early retirement and http://bmjopen.bmj.com/ 33 501 non-employment after breast cancer. Psychooncology 2014, 23: 634-641. 34 35 502 42. Carlsen K, Hoybye MT, Dalton SO, Tjonneland A: Social inequality and incidence of and survival from 36 503 breast cancer in a population-based study in Denmark, 1994-2003. Eur J Cancer 2008, 44: 1996-2002. 37 38 504 43. Carlsen K, Dalton SO, Diderichsen F, Johansen C: Risk for unemployment of cancer survivors: A Danish 39 505 cohort study. Eur J Cancer 2008, 44: 1866-1874.
40 on September 29, 2021 by guest. Protected copyright. 41 506 44. Gerkens S, Merkur S: Belgium: Health system review. Health Syst Transit 2010, 12: 1-266, xxv. 42 43 507 45. Kiasuwa Mbengi R. La réinsertion socioprofessionnelle des patients atteints de cancer. Exposé des positions. 44 508 2014. 45 509 Ref Type: Online Source 46 47 510 46. Onkelinx L. Toespraak van de Minister van Volksgezondheid en Sociale Zaken op het Symposium Evaluatie 48 511 van het Kankerplan. 2012. 49 512 Ref Type: Report 50 51 513 47. Rijken M, Spreeuwenberg P, Schippers J, Groenewegen PP: The importance of illness duration, age at 52 514 diagnosis and the year of diagnosis for labour participation chances of people with chronic illness: 53 515 results of a nationwide panel-study in The Netherlands. BMC Public Health 2013, 13: 803. 54 55 516 48. de Rijk A.: Work Disability Theories: A Taxonomy for Researchers. In Handbook of Work Disability. 56 517 Prevention and Managment. Edited by Loisel P., Anema J.R. New York: Springer; 2013:475-499. 57 58 59 Version 30 août 2016 60 18
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 19 of 36 BMJ Open
1 2 3 518 49. Belgian Cancer Registry. Cancer Survival in Belgium. 2004-2008. 1-109. 2012. 4 519 Ref Type: Report 5 6 520 50. Regional variation in incidence for smoking and alcohol related cancers in Belgium. Edited by Kris Henau, 7 521 Elizabeth Van Eycken, Geert Silverman, Eero Pukkala. Cancer Epidemiology 39[1], 55-65. 2016. 8 522 Ref Type: Generic 9 10 523 51. Mehnert A, Koch U: Predictors of employment among cancer survivors after medical rehabilitation--a 11 524 prospective study. Scand J Work Environ Health 2013, 39: 76-87. 12 13 525 52. Ervasti J, Kivimaki M, Dray-Spira R, Head J, Goldberg M, Pentti J et al.: Socioeconomic gradient in work 14 526 disability in diabetes: evidence from three occupational cohorts. J Epidemiol Community Health 2016, 15 527 70: 125-131.For peer review only 16 17 528 53. Woods LM, Rachet B, Coleman MP: Origins of socio-economic inequalities in cancer survival: a review. 18 529 Ann Oncol 2006, 17: 5-19. 19 20 530 54. Holm LV, Hansen DG, Larsen PV, Johansen C, Vedsted P, Bergholdt SH et al.: Social inequality in cancer 21 531 rehabilitation: a population-based cohort study. Acta Oncol 2013, 52: 410-422. 22 23 532 55. Krokstad S, Johnsen R, Westin S: Social determinants of disability pension: a 10-year follow-up of 62 000 24 533 people in a Norwegian county population. Int J Epidemiol 2002, 31: 1183-1191. 25 26 534 56. Noordzij M, Leffondre K, van Stralen KJ, Zoccali C, Dekker FW, Jager KJ: When do we need competing 27 535 risks methods for survival analysis in nephrology? Nephrol Dial Transplant 2013, 28: 2670-2677. 28 29 536 57. Glimelius I, Ekberg S, Linderoth J, Jerkeman M, Chang ET, Neovius M et al.: Sick leave and disability 30 537 pension in Hodgkin lymphoma survivors by stage, treatment, and follow-up time-a population-based 31 538 comparative study. J Cancer Surviv 2015. 32 http://bmjopen.bmj.com/ 33 539 58. Bradley CJ, Bednarek HL: Employment patterns of long-term cancer survivors. Psychooncology 2002, 34 540 11: 188-198. 35 36 541 59. Scott DA Mills M, Black A, Cantwell M, Campbell A, Cardwell CR, Porter S et al.. Multidimensional 37 542 rehabilitation programmes for adult cancer survivors (Review). Cochrane.Database.Syst.Rev. [3]. 2016. 38 543 Ref Type: Generic 39
40 544 60. Mewes JC, Steuten LM, Groeneveld IF, de Boer AG, Frings-Dresen MH, IJzerman MJ et al.: Return-to- on September 29, 2021 by guest. Protected copyright. 41 545 work intervention for cancer survivors: budget impact and allocation of costs and returns in the 42 546 Netherlands and six major EU-countries. BMC Cancer 2015, 15: 899. 43 44 547 61. de Boer AG, Taskila TK, Tamminga SJ, Feuerstein M, Frings-Dresen MH, Verbeek JH: Interventions to 45 548 enhance return-to-work for cancer patients. Cochrane Database Syst Rev 2015, 9: CD007569. 46 47 549 62. Zaman AC, Bruinvels DJ, de Boer AG, Frings-Dresen MH: Supporting cancer patients with work-related 48 550 problems through an oncological occupational physician: a feasibility study. Eur J Cancer Care (Engl ) 49 551 2015. 50 552 51 553 52 53 54 55 56 57 58 59 Version 30 août 2016 60 19
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 20 of 36
1 2 3 Predicting End of Work Disability After Cancer. A 7-years Follow-up Population-based Cohort Study 4 Using Competing Risks Analysis. 5
6 a,c b b a d a 7 R. Kiasuwa Mbengi , M.A Nicolae , Els Goetghebeur , R. Otter , K. Mortlemans , S. Missinne , e c c 8 M. Arbyn , C. Bouland and C. de Brouwer 9 10 11 12 Group of diseases 2007 2008 2009 2010 2011 2012 2013 13 Mental heath 74,054(33%) 78,112(34%) 83,247(34%) 88,535(34%) 92,899(34%) 98,171(35%) 104,291(35%) 14 Muskuloskeletal and connective 58,032(26%) 60,595 (26%) 65,146(27%) 69,583(27%) 74,192(28%) 79,643(28%) 86,071(29%) 15 Circulatory diseases For19,372(9%) peer 19,216(8%) review 19,427(8%) 19,571 only(8%) 19,549(7%) 19,772(7%) 19,963(7%) 16 Traumatic injuries and poisoning 15,302(7%) 15,776(7%) 16,538(7%) 17,080(7%) 17,635(7%) 18,383(6%) 18,955(6%) 17 Tumours* 13,592(6%) 14,266(6%) 15,103(6%) 16,083(6%) 16,742(6%) 17,591(6%) 18,462(6%) 18 Others (13 other conditions) 43,332(19%) 44,188(19%) 45,748(19%) 47,083(18%) 48,482(18%) 49,981(18%) 51,666(17%) 19 TOTAL 223,684 232,153 245,209 257,935 269,499 283,541 299,408 20 (100%) 21 Table 1. Number of cause-specific disabled workers in Belgium. Top 5 Evolution 2007-2013. 22 Annual Report NIHDI, 2014. *including cancers and benign tumours 23 24 25 Belgian socially insured workers in work disability due to cancer on 1st January 2007 + 26 st 27 those who entered up to 31 December 2011 n=21.701 28 29 30 Exclusion of observations with entrance into 31 work disability before 01-Jan-2007, n= 6.098 32 http://bmjopen.bmj.com/ 33 34 35 Belgian socially insured workers in work 36 disability starting between 2007- 2011 37 n=15.603 38 39
40 Exclusion of records with inconsistencies: on September 29, 2021 by guest. Protected copyright.
41 42 In dates of start and end of disability 43 n=5 (end date of disability before start date) 44 45 Gender/cancer site, n=33 46 men with breast cancer, female, n=23 female with breast cancer, male, n=6 47 women with male genital organs cancer, n=2 48 men with ovary cancer, n=1 49 men with female genital organ cancer, n=1 50 51 duplicated records n=22 52
53 54 Belgian socially insured workers with work 55 disability due to cancer starting in 2007 – 56 2011 and no inconsistencies in records 57 n=15.543 58 59 60 1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 21 of 36 BMJ Open
1 2 3 Figure 1. Flowchart of the number of workers disabled because of cancer between 2007 and 2011 in Belgium. 4 5 GROUPS The 11 cancer groups by anatomical location 5-year 5-year Survival Frequency 6 relative relative category observed in 7 survival rate survival rate the data 8 men*(%) women*(%) 9 1 Other malignancies and undefined sites, CIS NA 1 247 10 Benign tumours 11 2 Head and neck: lip, oral cavity, nasal cavities, middle ear 50.0 57.0 medium 877 and accessory sinuses, pharynx, larynx 12 3 Digestive tract 22.8 22.7 low 2 400 13 14 Esophagus 257 15 Stomach For peer review only 218 16 Colon & rectum 1 479 17 Pancreas 209 18 Other malignant neoplasms of digestive organs and 237 19 peritoneum 4 Respiratory tract 14.6 19.5 low 1 417 20
21 Trachea and lung 1 404 22 mesothelioma 13 23 5 Haematological 1 660 24 25 Hodgkin Disease 86.1 85.0 high 263 26 Non-Hodgkin Disease 67.0 68.9 medium 711 27 Acute lymphoid leukemia and Lymphoid leukemia. Other 81.3 76.7 high 161 28 Myeloid leukemia and others 38.5 40.6 low 307 6 Breast 78.2 88.0 high 5 511 29 30 Brest female 5 494 31 Breast male 17 32 7 Female genital organs 821http://bmjopen.bmj.com/ 33 34 Corpus uterus - cerv. ut. 69.8 medium 273 35 Cervix uteri corp. ut. 79.6 high 147 36 Ovary ovary 54.1 medium 362 Others 37 8 Male genital organs 95.3 - high 486 38 39 Prostate 377
40 Testis 94 on September 29, 2021 by guest. Protected copyright. 41 Others 16 42 9 Urinary tract 388 43 Kidney 71.0 0.7 high 147 44 Bladder 56.6 49.2 medium 178 45 Others 63 10 Central nervous system (CNS) 22.7 25.8 low 709 46 11 Bone and connective tissue (sarcoma’s) 61.9 59.7 medium 47 Melanoma of the skin 86.2 91.0 high 48 Malignant neoplasms of skin other than melanoma NA 49 Thyroid and other endocrine glands 89.3 94.1 high 50 Other malignancies and undefined sites, invasive 51.5 39.1 medium/low 51 Tumours of uncertain and unspecified behavior NA 52 TOTAL 15 543 53 Table 2. The 11 cancer groups used for the analysis. 54 *Reference: Belgian Cancer Registry. Cancer Survival in Belgium, 2004-2008.
55
56 57 58 59 60 2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from BMJ Open Page 22 of 36
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 Figure 2. Kaplan-Meier estimator for the time in work disability, stratified by the year of entrance into work 24 disability. 25 26 27 28 29 30 31 32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 29, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from 4 4 ) )
(censored) (censored) Work disability disability Work 7.46 (17.76- 37.17) 37.17) (17.76- 7.46 27.75 (9.83- 45.68) (9.83- 45.68) 27.75 (7.11- 40.51) 23.81 (0.00- 48.43) 24.05 (0.00- 48.49) 21.33 41.29) (14.28- 27.78 (0.00- 41.44) 20.42 57.68) (31.57- 44.63 (0.00- 48.53) 19.96 (2.65- 32.13 17.39 11.36 (0.00- 40.58) (0.00- 40.58) 11.36 2 33.25) (23.11- 28.18 (0.00- 39.94) 19.49 (5.78- 36.86) 20.50 27.49) (10.29- 18.89 37.47) (16.68- 27.07 31.53) (12.07- 21.80 36.24) (25.95- 31.09 (8.60- 38.44) 23.51 (NA) 0.01 < 33.29) (16.47- 24.88 33.44) (21.40- 27.41 37.31) (26.13- 35.02 41.90) (28.15- 35.03 32.00) (10.23- 21.11
work work Ability to return-to- to Ability 41.09) (33.10- 37.09 24.54) (18.25- 21.40 28.55) (24.68- 26.61 (8.51- 11.76) 10.14 45.17) (40.18- 42.67 59.11) (49.24- 54.17 35.89) (29.19- 32.54 (35.76-45.28) 40.52 28.23) (19.41- 23.82 22.36) (16.18- 19.27 32.63) (25.40- 29.01 31.85) (29.62- 30.74 51.58) (43.80- 47.70 33.32) (28.96- 30.96 26.75) (24.39- 25.57 46.79) (41.41- 44.10 56.27) (51.82- 54.05 41.98) (36.06- 40.52 33.93) (27.55- 30.74 25.63) (20.60- 23.12 36.23) (32.02- 34.12 36.41) (33.20- 34.80 35.73) (32.41- 34.07 37.22) (32.54- 34.88 49.47) (44.04- 46.76 % (IC95%) % (IC95%) - - - Reason of ending work disability disability work ending of Reason Retirement Retirement (1.01 - 4.20) 2.61 (0.86 - 4.24) 2.55 (3.86 - 7.18) 5.52 (0.68 - 6.45) 3.57 (2.07 - 4.26) 3.16 (2.28 - 3.95) 3.12 (0.00 - 1.00) 0.47 (9.72 - 20.21) 14.96 (3.92 - 11.65) 7.78 (0.00 - 0.90) 0.42 (0.20 - 4.97) 2.58 (2.08 - 3.51) 2.80 (1.63 - 3.98) 2.80 (8.04 - 12.31) 10.25 (3.75 - 5.60) 4.68 (2.40 - 4.20) 3.30 (2.23 - 5.07) 3.65 33.63) (28.09- 30.86 (3.87 - 5.60) 4.73 (2.69 - 4.00) 3.35 (1.23 - 2.26) 1.75 (0.42 - 1.02) 0.72 (0.14 - 0.93) 0.54 http://bmjopen.bmj.com/
BMJ Open Death Death - 20.66) (10.68 15.67 - 61.91) (50.27 56.09 - 53.01)) (47.94 50.48 - 77.45) (72.42 74.93 - 29.20) (24.20 26.70 - 16.14) (13.77 14.96 - 47.45) (38.91 43.18 - 24.17) (16.77 20.47 - 53.05) (41.07 47.06 - 56.50) (48.54 52.52 - 52.37) (43.59 47.98 - 39.61) (36.96 38.28 -31.45) (28.57 30.01 - 40.40) (34.33 38.28 5017 (32.28%) 5017 (2.20%) 342 (34.63%) 5383 (30.89%) 4801 52.52) (49.21- 50.86 26.59) (24.46- 25.53 26.13) (22.19- 24.16 29.88) (26.89- 28.38 43.65) (40.53- 42.09 48.97) (43.08- 46.02 37.91) (34.61- 36.26 36.06) (32.80- 34.42 34.11) (30.82- 32.46 31.02) (27.72- 29.37 - 34.17) (29.01 31.59 on September 29, 2021 by guest. Protected copyright.
.30 [1.23-1.37] [1.23-1.37] .30 [3.24- 4.98 [1.41-1.93] 1.69 [1.18-1.41] 1.31 [0.69-0.83] 0.75 [1.66-2.08] 1.83 [1.95-2.30] 2.10 [1.37-1.82] 1.56 [1.52-2.24] 1.73 [1.32-2.16] 1.67 [1.18-1.89] 1.46 [1.03-1.52] 1.24 1.79 [1.63-1.97] [1.63-1.97] 1.79 [1.72-1.89] 1.79 1 (years) ( IC 95%) IC 95%) (years) ( in work disability disability work in Median time spent spent time Median 1.59 [1.52-1.66] [1.52-1.66] 1.59 sp median time (overall ent in disability) [1.12-1.26] 1.18 [1.84-2.05] 1.94 [1.415-1.69] 1.51 [1.637-1.88] 1.76 [1.580-1.79] 1.70 [0.797-1.06] 0.91 [1.780-2.048] 1.89 [1.717-1.960] 1.83 [1.621-1.889] 1.76 [1.410-1.670] 1.54 [0.413-0.487] 0.45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only n (%)n 703 (37%) (37%) 703 15543 15543 (4%) 614 (5.6%) 877 (15.4%) 2400 (9.1%) 1417 (10.7%) 1660 (35.%) 5511 (5.3%) 821 (3.1%) 486 (2.5%) 388 (4.6%) 709 (4.2%) 660 2125 (13%) (13%) 2125 individuals individuals TOTAL TOTAL (50%) 7715 5
(38%) 5874 (62%) 9669 (15.6%) 2421 (32.5%) 5052 (44.7%) 6946 ( 7.21%) 1121 (22.2%) 3454 (24.2%) 3760 (23.4%) 3630 (21.8%) 3388 (8.4 %) 1311
Table 3. DescriptionTable 3. of the characteristics of the individuals included in the cohort, mediantheir spenttime the in disability, rate of those one experiencing of the three competing events at the end of andfollow-up the rate of those remaining work disabled theat end of follow-up.
n=15,543 n=15,543 TOTAL n (%) n TOTAL Bone&Con (sarc.)/Skin/Thyroid Bone&Con Central nervous system (CNS) system nervous Central Urinary tract Urinary Male genital organs genital Male Female genital organs Female Breast Breast Haematological Haematological Respiratory tract Respiratory Digestive tract Digestive Head & Neck & Head CIS/Ben CIS/Ben Cancer site site Cancer Assisting spouse & & spouse Self-employed Assisting White collar White Blue collar collar Blue Occupational Class Occupational Women Women Men Men Gender Gender >=60 >=60 50-59 50-59 40-49 40-49 17-39 17-39 Age at entry entry Age at 2011 2010 2010 2009 2009 2008 2008 2007 2007 Year of entry Year entry of
Page 23 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 24 of 36 5 5
3c 3c
http://bmjopen.bmj.com/ BMJ Open on September 29, 2021 by guest. Protected copyright. 3e 3e 3b
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only
and occupational class group, d) gender, cancer c) b) age, a) per follow-up,end of stratified the by ability RTW of to incidence cumulative 3Non-parametric a-e. Figures 3d 3a e) year of entrance. entrance. of year e) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from 6 6
http://bmjopen.bmj.com/ BMJ Open on September 29, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only Figures 4 a-d. Box4 Figures a-d. plots of time to any of the by gender, three group,events, age year of entrance Youoccupationaland class. could add a box by plot survival category of the cancer.
Page 25 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from Page 26 of 36 7 7
http://bmjopen.bmj.com/ BMJ Open on September 29, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only ox plot of time to events, by cancer group by cancer to time events, of plot ox B Figure 5 e.Figure 5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2016-014094 on 17 February 2018. Downloaded from
8 8