Psychosocial working conditions and diabetes self-management at work: A qualitative study

Accepted Manuscript Psychosocial working conditions and diabetes self-management at work: a qualitative study Adrian Loerbroks, Xuan Quynh Nguyen, Patricia Vu-Eickmann, Michael Krichbaum, Bernhard Kulzer, Andrea Icks, Peter Angerer PII: DOI: Reference:

S0168-8227(17)31221-4 https://doi.org/10.1016/j.diabres.2018.03.023 DIAB 7280

To appear in:

Diabetes Research and Clinical Practice

Received Date: Revised Date: Accepted Date:

29 July 2017 2 March 2018 9 March 2018

Please cite this article as: A. Loerbroks, X. Quynh Nguyen, P. Vu-Eickmann, M. Krichbaum, B. Kulzer, A. Icks, P. Angerer, Psychosocial working conditions and diabetes self-management at work: a qualitative study, Diabetes Research and Clinical Practice (2018), doi: https://doi.org/10.1016/j.diabres.2018.03.023

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Psychosocial working conditions and diabetes self-management at work: a qualitative study

Adrian Loerbroksa,b, Xuan Quynh Nguyena, Patricia Vu-Eickmanna, Michael Krichbaumc,d, Bernhard Kulzerc,d,e, Andrea Icksf,g,h, Peter Angerera

a

Institute of Occupational, Social and Environmental Medicine, Centre for Health and

Society, Faculty of Medicine, University of Düsseldorf, Universitätsstraße 1, 40225 Düsseldorf, Germany b

Mannheim Institute of Public Health, Social and Preventive Medicine, Mannheim Medical

Faculty, Heidelberg University, Ludolf-Krehl-Straße 7-11, 68167, Mannheim, Germany c

Research Institute Diabetes Academy Mergentheim (FIDAM), Bad Mergentheim, Theodor

Klotzbücher Str. 12, 97980 Bad Mergentheim, Germany d

Diabetes Center Mergentheim, Theodor Klotzbücher Str. 12, 97980 Bad Mergentheim,

Germany e

Department of Clinical Psychology, University of Bamberg, Markusplatz 3, 96047 Bamberg,

Germany f

Institute for Health Services Research and Health Economics, Centre for Health and

Society, Faculty of Medicine, University Düsseldorf, Moorenstraße 5, 40225 Düsseldorf, Germany g

Institute for Health Services Research and Health Economics, German Diabetes Centre,

Auf'm Hennekamp 65, 40225 Düsseldorf, Germany h

German Centre for Diabetes Research, Ingolstädter Landstraβe 1, 85764 München-

Neuherberg, Germany

1

Corresponding author: Dr. Adrian Loerbroks Institute of Occupational, Social and Environmental Medicine Centre for Health and Society, Faculty of Medicine, University of Düsseldorf Universitätsstraße 1, 40225 Düsseldorf, Germany Telephone: +49 (0) 211 - 81 08032 Fax: +49 (0) 211 - 81 18586 Email: [email protected]

Conflicts of interest: None

2

ABSTRACT Aims: We conducted a qualitative study to expand our current understanding of the potential link between psychosocial working conditions and diabetes self-management at work. Methods: Thirty employed adults with diabetes mellitus living in Germany (n=19 with type 1, n=11 with type 2, 57% female, aged 24-64 years) were recruited. Using a topic guide, we carried out in-depth interviews in face-to-face contact or by telephone. Interviews were transcribed and content-analyzed using MaxQDA. Results: Psychosocial working conditions perceived to detrimentally affect self-management activities included, amongst others, a high workload, poor job control, unhygienic working environments, the requirement to work under high or fluctuating temperature, perceived social norms at the workplace, and the attitude to prioritize work-related demands as opposed to diabetes-related demands. The types of self-management activities considered to be adversely affected related to glucose monitoring, insulin injections, dietary control, the ability to recognize hypoglycemia and health care use. Conclusions: Various types of occupational psychosocial factors may determine diabetes self-management practices at the workplace. Quantitative studies are needed to confirm our observations. Subsequently, interventions could be developed and evaluated to improve opportunities to adequately engage into diabetes self-management at work.

Keywords: Diabetes mellitus, Qualitative study, Self-management, Workplace

3

1. INTRODUCTION Patients with diabetes need to carry out various self-management activities, whose implementation is often complex and time-consuming. Such activities relate to dietary control, physical exercise, monitoring or adjustment of glucose levels, and utilization of health care services. Effective self-management likely improves glycemic control [1] and thereby reduces the risk of diabetes-specific disability and complications [2]. Patients’ opportunities to appropriately implement recommended self-management activities may however be constrained in real life [1]. The occupational context may be particularly challenging in this respect, but its potential impact remains poorly understood and under-researched [3-5]. The limited earlier work suggests though that key occupational factors that may impair diabetes management include time pressure, the requirement to attend to unexpected tasks or events, disruption of usual working routines [3, 5], physically demanding tasks (i.e. complicating glycemic control) [3], and unavailability of suitable foods [5]. Also, unfavorable attitudes were identified, which comprised feelings of guilt when spending time on diabetes management while being at work [3, 5] or perceiving one’s diabetes entirely as “personal business” which should not interfere with job tasks [4]. As a consequence of these working conditions and attitudes, many individuals report to skip glucose monitoring, physician appointments and physical exercise sessions and report that their consumption of sugary foods may increase [5]. Also, glucose levels may deliberately be kept too high to prevent impairment of workability or hypoglycemia [4]. By contrast, work characteristics perceived as beneficial were, for instance, social support from colleagues and from other workers with diabetes [3]. While the above-mentioned qualitative studies have provided initial insights, research remains sparse. We therefore aimed to conduct a qualitative study to expand our current understanding of the potential link between psychosocial working conditions and diabetes self-management at work.

4

2. SUBJECTS, MATERIALS AND METHODS Participant recruitment We recruited 30 adults with diabetes mellitus type 1 (DMT1) or diabetes mellitus type 2 (DMT2), who were in employment. Participants were recruited through various pathways. These included a large diabetes clinic (Diabetes Center Mergentheim, Germany, which enrolled 17 patients), physicians (i.e., diabetologists, internists, and occupational physicians, n=6), diabetes support groups (n=6), and private contacts (n=1). We further aimed to recruit a minimum number of five participants for each sex x type of diabetes stratum (i.e., five women with DMT1, five men with DTM1, five women with DMT2 and five men with DMT2). Individuals who provided informed consent participated in a qualitative interview and completed a standardized questionnaire to collect background data. The questionnaire covered basic socio-demographic variables, occupational factors and health. Our study was approved by the Institutional Review Board of the Medical Faculty of the University of Düsseldorf, Germany.

Data collection The study team created a topic guide, which considered findings from prior qualitative research [4, 5], but seemed sufficiently broad to allow for new themes to emerge. The topic guide (see Table A in supplementary materials for details) started out with broad open-ended questions that aimed to make participants feel at ease to talk and to cognitively retrieve the full range of potential experiences related to work and diabetes self-management. The guide then proceeded to work situations that may impair self-management and the types of activities affected. The mode of interviewing (i.e., face-to-face [n=9] or by telephone [n=21]) was determined by each participant’s preferences. Throughout the process of data collection we conducted interim analyses to refine our topic guide and to gain an understanding of potential saturation and thus conclusion of data collection.

5

Data analysis Data were collected between June and August 2016. Interviews were audio-recorded and were all carried out by the same interviewer (XQN). XQN was trained prior to interviewing and received feedback by PVU and AL on the interview conduct during data collection based on recordings. Interviews were transcribed and content-analyzed [6] using the MaxQDA 12 software package. Briefly, we employed each primary question included in the topic guide as overarching category (deductive coding). These broad categories were further broken down into subcategories based on the findings that emerged from analyses of the transcripts (inductive coding). Initially, XQN coded all transcripts (i.e. systematic identification of relevant statements, assigning of relevant statements to existing or new codes) to derive an initial structure of categories. Next, PVU and AL each coded five different interviews and reviewed all transcripts as well as all categories and codes created by XNQ. For instance, PVU and AL checked whether they would assign highlighted statements to the same code, if additional statements could be assigned to a given code, if new codes or categories were needed, if the codes within a category seemed coherent, if codes were coherent with the category label and to what extent categories seemed to capture (dis)similar content. Analyses were discussed during consensus meetings (XQN, PVU and AL) and considered in a second coding round. Upon its completion, PVU and AL again reviewed all categories and codes, but there was only little further discussion among team members. Therefore, a subsequent third round of data analyses was carried out by XQN, which was considered final.

3. RESULTS Table 1 shows characteristics of the study participants. Briefly, the sex distribution was fairly balanced and most participants were in their mid-40s with coverage of a broad age range. Participants from diverse occupational contexts were included, i.e. individuals in both fulltime and part-time work, both employed and self-employed, with and without leadership responsibility, and with varying perceptions of their overall work stress. Two thirds reported to have been diagnosed with DMT1 and the remaining reported DMT2. Diabetes diagnoses had 6

been received two to 41 years ago (mean=15.0 standard deviation [SD]=10.5) and 43.3% of the participants reported at least one diabetes complication. Virtually all participants needed to carry out glucose measurements for their diabetes management. Two interviews needed to be excluded due to poor recording quality and thus 28 interviews were content-analyzed. The following findings emerged from the data:

Which work conditions impair diabetes self-management at work? High workload: According to study participants, a high workload represents a major barrier to effective diabetes self-management at work. Periods of varying work intensity, highly complex tasks, the need to respond to simultaneous requests from different individuals (e.g., customers, supervisors, colleagues), and shortage of staff are thought to induce a sense of time pressure, the requirement to perform multi-tasking, and overtime work and thereby restrict the duration and frequency of breaks (see statements WL1-WL3, Table 2). In particular frequent social contacts at work seemed to contribute to the perception of a high workload. Poor job control: Unpredictable or unscheduled events were also referred to as challenging in the context of diabetes. Such events may result in a temporary lack of control of work processes (see JC1, Table 2). Further, organizational structures may generally impair perceived job control and thereby restrict opportunities for self-management (JC2). Such challenging exposures are experienced, for instance, when meetings are scheduled on short notice (e.g., within hours), when meetings are postponed or ongoing meetings are unexpectedly extended. Additional problematic situations are when colleagues need to be substituted on short notice, when the duration of phone calls expands too much or when one is being called at a time scheduled for diabetes self-management (JC3, JC4). Also business trips and external appointments seem problematic, as they interfere with routines, which are necessary to reconcile the demands of one’s job with diabetes self-management (JC5). Further, inflexible work routines (JC6), assembly line work (JC7) and shift work (JC8, JC9) were perceived to inherently restrict job control and self-management opportunities. 7

According to participants, breaks are taken irregularly or are being skipped when job control is poor (JC2). Physical working environment: Except for the restricted time resources to engage in diabetes self-management at work, lack of a refuge to socially withdraw from work seems to represent a significant issue (e.g., no opportunities to inject insulin without disturbance) (PWE1, Table 3). Also, unhygienic working conditions (PWE2), large fluctuations of temperature (PWE3) and working under high temperatures (PWE4) are perceived to detrimentally affect selfmanagement. Perceived social working environment: Many participants expressed that their ability to overtly attend to their diabetes at work is limited. The reported barriers relate to the perception of implicit social norms with regard to the measurement of blood glucose and the injection of insulin in public contexts (SWE1, Table 3). Self-management impairment due to those perceived social norms may be particularly pronounced in jobs with frequent interpersonal contact, for instance, with children (SWE2), colleagues (SWE3-SWE5), customers, patients, or business partners (SWE4, SWE5, SWE1, SWE6). In those contexts, participants report to feel considerably impaired or unable to measure their blood glucose levels and to inject insulin. These situations are characterized by perceived discomfort, restricted freedom and a sense of being closely observed by others (SWE3, SWE5). In addition, some attribute their restricted opportunities to overtly attend to their diabetes at work to a perceived lack of understanding by colleagues and employers; specially, there may be poor understanding of the fact that short time windows are needed to adequately manage one’s diabetes at work (SWE7, SWE8).

What types of self-management activities are affected by working conditions? Many interviewees expressed that short time windows for self-management activities are lacking, e.g., when breaks cannot be taken as scheduled. As a result, self-management activities (e.g., related to glycemic control) are forgotten, deliberately omitted, postponed or carried out irregularly (SMA1-SMA5, see Table 4). As alluded to above, a lack of privacy, a 8

sense of being watched, and unhygienic conditions may also contribute to those adverse behavioral patterns. Further, working under time pressure (SMA6) and large fluctuations of temperature (SMA7) are felt to impair the perception of physical symptoms and of potential hypoglycemia. According to participants, unscheduled events disrupt one’s work and selfmanagement routines. As a result, the individually tailored sequencing of insulin injections and breaks for food consumption is disturbed (e.g., when insulin has been injected, but then one is asked to take over unexpected tasks) (SMA8, SMA9). Participants feel that shift work disrupts their sleep and eating cycles and thereby detrimentally affects blood glucose levels (SMA10). Further, external appointments are perceived to render glycemic control challenging, because the nutritional composition of some served meals and foods remains uncertain (SMA11) and because attractive meals challenge one’s dietary compliance (SMA12). It seems further problematic that time slots for diabetes-related physician appointments are often in conflict with working times. In participants’ views, one’s opportunities of attending physician appointments are greatly determined by the employer’s and colleagues’ understanding of the need for such appointments (SMA12). In summary, the above-mentioned working conditions may detrimentally affect self-management activities related to glucose monitoring, insulin injections, dietary control, the perception of hypoglycemia and health care use.

What types of attitudes towards self-management at work emerged? Some participants expressed that they did not feel socially restricted when attending to their diabetes and that they even viewed communication about their diabetes at work as vitally important (ATS1, ATS2, see Table 5). Advocacy for such overt dealings seemed to be largely determined by one’s personal positive experience of having received support and understanding by colleagues (ATS1, ATS3, ATS4). It was also expressed that diabetes selfmanagement is mainly one’s personal responsibility (ATS5). Nevertheless, participants emphasized that working processes and procedures need to provide them with opportunities to engage into adequate diabetes self-management (e.g., the permission to carry along 9

measurement devices, insulin pens and appropriate foods). Some stated that they prioritize their diabetes self-management when this is made difficult by high work demands (ATS6). Yet many others seemed to prioritize job demands at the cost of diabetes self-management (ATS7). Reasons for this decision were, for instance, i) that one feels to violate the norm of being a „reliable worker“ when attending to one’s diabetes, ii) that working conditions were simply perceived not to allow for any diabetes self-management (ATS8), or iii) fear of loosing one’s job (ATS9).

4. DISCUSSION 4.1. Summary of findings Our qualitative data suggest that diabetes self-management at the workplace is perceived to be detrimentally affected by factors related to a high workload (e.g., time pressure, complex tasks, handling of multiple requests), poor job control (e.g., unexpected events, meetings, business trips), the physical working environment (e.g., lack of a refuge for injections, unhygienic and hot workplaces), the perceived social working environment (e.g., perceived social norms at work, lack of understanding by others) as well as specific attitudes towards reconciling one’s diabetes with one’s job demands. Activities that were primarily affected according to study participants related to glycemic control, food consumption and attendance of physician appointments for diabetes.

4.2. Findings in light of earlier qualitative research While some of our findings are in line with those from the three prior qualitative studies [3-5], we also contribute novel findings. A high workload has also emerged as detrimental to diabetes self-management in a study from Ireland, carried out by Balfe et al. [5], among young employees with DMT1. In line with our observations, participants in that study specifically seemed to conceptualize a high workload in terms of work intensity, time pressure and overtime work [5]. Our study further adds two observations. Firstly, also the demand to handle simultaneous requests may adversely affect diabetes self-management 10

(e.g., in service-related professions). Secondly, individuals with diabetes may experience pronounced difficulties in managing their condition when they are employed in professions, in which any shortage of staff implies that the workload for the remaining staff further increases (e.g., in health care professions). With regard to aspects related to poor job control, our study confirms earlier reports that highlighted that non-routine work [3, 5], irregular working hours [5], expanding business meetings [5], and shift work [5] are perceived to detrimentally affect diabetes selfmanagement at work. However, in accordance with a prospective qualitative study from the US [4], our study also suggests that work procedures that are overly structured and inflexible may likewise be unfavorable, because they do not allow for short breaks to attend to one’s diabetes, e.g., in assembly line work [4]. Again, our study expands those available insights to the important role of staffing: it emerged from our data that the unexpected requirement to substitute colleagues will not only increase one’s workload, but will likewise reduce job control thereby limiting opportunities for diabetes self-management. Another relevant finding that saliently emerged from our data relates to the challenges associated with business trips. Such trips seem to not only disrupt behavioral self-management routines, but also often imply exposure to unknown or extraordinary foods and thereby complicate dietary control. Regarding aspects of the physical working environment with implications for diabetes selfmanagement, our participants referred to unhygienic working conditions as particularly challenging, which is in keeping with findings from a study from the UK [4]. Our study adds further important insights relating, firstly, to the lack of a refuge to safely check glucose levels and inject insulin. Secondly, working under high or fluctuating temperature was perceived as detrimental to the recognition of potential hypoglycemia. It seems likely that those aspects did not emerge from prior work, because individuals exposed to such working conditions (e.g., in industrial professions) were not included at all [5] or considerably underrepresented [4]. By contrast, our sample included, for instance, an industrial electrician, a warehouseman, and an industrial fitter.

11

With regard to aspects related to the perceived social working environment our study confirms the potential challenges inherent to norms related to the social appropriateness of overtly monitoring insulin levels and/or injecting insulin at work [4]. Also, just like in earlier research [3, 4], participants in our study appeared to experience the understanding and support from colleagues, managers and employers as limited. Possibly though, support resources are partly determined by the type of occupational branch. Social support may be more easily received by colleagues when one works in a profession related to health services or social services [3]. Just like in our study, there seems to be the consistent observation that in particular selfmanagement activities related to glycemic control are detrimentally affected at work (e.g., diet, glucose measurement and insulin injection) [3-5]. Also, as confirmed by our study, attendance of physician appointments seems to be challenging [4, 5]. Unlike in earlier studies [3, 5], physical activity has not emerged from our study as a key behavior that is pressured by unfavorable work conditions. This could be due to sample differences (e.g., age and professional branches) or due to the fact that our study participants did not conceptualize physical activity as a key self-management activity. In terms of selfmanagement activities, our study adds that the timely perception of potential hypoglycemia could be challenging, in particular when high physical activity levels or exposure to high temperatures are required at work. Finally, the experience of a conflict related to timecompeting demands of one’s work and diabetes was found in our and all prior studies [3-5] and often seems to translate into the habit of running higher than optimal glucose levels. Finally, while the nature and strength of the relationship between working conditions and selfmanagement may partly be disease-specific, some challenges may be experienced in the vast majority of chronic conditions. For instance, the finding that the utilization of health services (e.g., for regular check-ups or during episodes of disease exacerbation) is constrained by one’s working times and the ability to take time off has also been observed in workers with other conditions than diabetes [7, 8] (i.e. asthma and chronic kidney disease). Also, the finding that a lack of privacy at work may contribute to one’s decision not to take 12

medication as prescribed or as needed has been confirmed for other conditions that may require medication intake during working days [9].

4.3 Implications for practice Based on the available evidence it seems promising to pursue the development of interventions that improve the quality of working experiences among individuals with diabetes. Such interventions could address various types of aspects of the working environment. Some of these aspects seem straightforward and easy to communicate to employers (e.g., provision of a refuge for injections). However, the degree to which the modification of working conditions is feasible likely varies significantly across workplaces. For instance, high temperatures, unhygienic conditions and little job control may be inherently related to one’s tasks (e.g., construction work and assembly line work) and thus leave very little or no opportunity for modification. The awareness among employers, managers and colleagues of the implications of working with diabetes needs to be increased. This may be useful in reframing detrimental social norms regarding overt glucose management, in building social support resources, and in adjusting to some extent problematic working conditions (e.g., taking care to schedule realistic durations for meetings to ensure that individuals with diabetes can prepare accordingly). With regard to workers, psychological counseling may help to reflect, reframe and to better reconcile work demands and diabetes demands. Also, patient education programs may put a stronger emphasis on the communication about one’s diabetes at work and on strategies to build social support resources. In some professions, interventions related to “job crafting” [10] (i.e. the employeeinitiated redesign of work characteristics) are potentially useful to adjust job control or to increase social support resources. In addition, workers with diabetes could be supported by easy access to healthy foods whenever needed and by company initiatives that help to incorporate physical activity into sedentary jobs (e.g., lunchtime walking) [11]. Occupational physicians may also be key players, as they could act as advocates for workers with diabetes. Occupational physicians could, for instance, help supervisors and colleagues to 13

recognize the implications of working with diabetes, they could support potential job redesign to accommodate diabetes needs and could provide medical check-ups for diabetes to reduce time away from work due to routine physician appointments (e.g., HBA1c monitoring). It needs to be mentioned though that the widespread involvement of occupational physician will be challenging, at least in Germany, for small- and medium-sized companies who usually lack in-house occupational health services. Besides the above-mentioned approaches, legislative action could help reducing some of the conflicts between work demands and diabetes self-management. Examples may be the obligatory provision of social refuges at work to allow for diabetes self-management or to ensure that counseling services for workers with diabetes are provided and are easily accessible.

4.4 Implications for research Prior to the development and evaluation of interventions it seems that additional research is needed, especially an intermediate phase of quantitative (i.e. survey-based) research. As of today, very few such studies have been published [12, 13] and they have provided limited and inconsistent evidence. Those earlier quantitative studies have considered workload [12, 13] and job control, and examined diabetes self-management in terms of glucose management [12, 13] and dietary habits [12]. One of those studies [13] showed a link between a high workload and poorer glucose management, but this was not confirmed by another study [12]. Also, in the latter study, workload was not associated with dietary behavior. Likewise, poor job control was not associated with any of those self-management behaviors [12]. The inconsistency of the findings from qualitative and quantitative studies may stem from sample differences (e.g., smaller samples and more selective participation in qualitative studies), but may also reflect that subtle associations, as perceived by workers, may not necessarily translate into significant (or pronounced) associations in quantitative surveys (for instance, because adverse work conditions are experienced infrequently and therefore do not exert sustained effects). However, since evidence remains sparse, additional quantitative studies are required. Those future studies could further test various 14

additional hypotheses that can be deduced from our and other qualitative work [3-5] (e.g., “Individuals who work under poor hygienic conditions are less likely to monitor glucose levels” or

“Individuals who prioritize their work demands as opposed to diabetes self-

management are less likely to monitor glucose levels”). Also, to identify target areas for preventive action, quantitative studies are needed to determine the prevalence of determinants (i.e. both barriers and enabling factors) of self-management as experienced by workers with diabetes. Moreover, quantitative studies need to expand the scope of examined self-management behaviors from glucose management and dietary habits to physical activity and health care use. Furthermore, research may be expanded to clinical outcomes to examine, for instance, whether i) glycemic control is poorer among workers with diabetes exposed to adverse psychosocial working conditions as opposed to those with better working conditions and ii) to what extent such relationships are mediated by poor self-management [14, 15]. Finally, the current understanding of the occupational determinants of diabetes selfmanagement at work may be improved by studies that collect and analyze both qualitative and quantitative data in the same sample to reduce the likelihood that findings vary due to sample differences. This may be particularly promising when those data sources are analytically mixed (so-called mixed methods research). Such approaches remain underutilized in diabetes research [16].

4.5. Methodological considerations Strengths of our study include the recruitment process, which drew on various pathways (e.g., a large diabetes clinic, physicians, and diabetes support groups) and ensured reasonably balanced recruitment of individuals from both sexes and with both types of diabetes. This was done to maximize the likelihood of obtaining a sample with diverse diabetes self-management experiences at work. Accordingly, the characteristics of our samples showed large variation (Table 1) and it seems reasonable that a broad range of perspectives had emerged from our data. Further, in line with each participant’s preference 15

interviews could be carried out in face-to-face contact or by telephone. This approach was chosen to reduce potential selection due to demanding jobs. It has been suggested that faceto-face interviews and telephone interviews generally yield comparable insights [17]. To increase the likelihood that content analyses were carried out in a well-balanced manner, a team of three analysts worked with the data. This team was mixed in terms of gender, professional backgrounds and research experience. When interpreting the observations from qualitative research, one needs to bear in mind that - in contrast to quantitative or epidemiological studies - it is not the primary aim of qualitative studies to provide statistically generalizable data. Instead, the qualitative research methods applied in the current study aim to reveal as much as possible the full diversity of potential views and opinions that can possibly be held by members of a specific population [18]. As mentioned above, subsequent epidemiological studies are needed to produce generalizable quantifications of the relationships of working conditions with diabetes self-management practices and to estimate the prevalence of specific working conditions or self-management practices at work. Some limitations of our study also deserve mentioning. Firstly, we do not know to what extent the willingness to participate in our study may be related to the quality of diabetes selfmanagement at work. Secondly, only individuals in employment were eligible, which may imply that those who experience the most severe difficulties in managing their diabetes at work had possibly quit work and were not included (“healthy work selection bias”). Thirdly, while the generated codes and categories were reviewed by multiple analysts, only five transcripts were newly coded by all analysts. Possibly, the analytical process had been superior if multiple raters had coded the entire material. This approach, which is considered inefficient in terms of cost and effort [18], was not feasible though. Moreover, it has been suggested that it is sufficient when multiple researchers review selected proportions of the data along with the coding framework [18], which is in line with the approach adopted in the current study.

16

4.6. Conclusion Various types of occupational psychosocial factors may determine diabetes selfmanagement practices at the workplace. Quantitative studies are needed to confirm our observations and expand the evidence to clinical outcomes. Subsequently, interventions could be developed and evaluated to improve opportunities to adequately engage into diabetes self-management at work.

Acknowledgements: For their support related to participant recruitment we are indebted to the patient group „Diabetes Talk“, Düsseldorf (headed by Mrs. Katinka Driesen), the patient group at the German Diabetes Center Düsseldorf (headed by Mrs. Evelyn Bruns), Hausarztpraxis Vi & Weng (Düsseldorf), and Dr. Kaltheuner (Leverkusen). For her translation of quotes (Table 2) we are grateful to Ms. Ursula Goldberger, a certified translator from the Mannheim Institute of Public Health, Social and Preventive Medicine, Mannheim Medical Faculty, Heidelberg University, Mannheim, Germany.

Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

17

References 1. Chrvala CA, Sherr D, Lipman RD. Diabetes self-management education for adults with type 2 diabetes mellitus: A systematic review of the effect on glycemic control. Patient Educ Couns 2016;99:926-943. 2. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med 2008;359:1577-1589. 3. Bose J. Promoting successful diabetes management in the workplace. International Journal of Workplace Health Management 2013;6:205-226. 4. Ruston A, Smith A, Fernando B. Diabetes in the workplace - diabetic's perceptions and experiences of managing their disease at work: a qualitative study. BMC Public Health 2013;13:386. 5. Balfe M, Brugha R, Smith D, Sreenan S, Doyle F, Conroy R. Why do young adults with Type 1 diabetes find it difficult to manage diabetes in the workplace? Health Place 2014;26:180-187. 6. Schreier M Qualitative content analysis in practice. London: Sage; 2012. 7. Kolbe J, Vamos M, Fergusson W, Elkind G. Determinants of management errors in acute severe asthma. Thorax 1998;53:14-20. 8. McQuoid J, Welsh J, Strazdins L, Griffin AL, Banwell C. Integrating paid work and chronic illness in daily life: A space-time approach to understanding the challenges. Health Place 2015;34:83-91. 9. Zhao D, Cheung J, Smith L, Saini B. Exploring asthma in the workplace: A triangulation of perspectives from management, employees and people with asthma. J Asthma (in press) 10. Wrzesniewski A, Dutton J. Crafting a job: revisioning employees as active crafters of their work. Acad Manag Rev 2001;26:179-201. 11. Thogersen-Ntoumani C, Loughren E, Duda J, Fox KR. Step by step: The feasibility of a 16-week workplace lunchtime walking intervention for physically inactive employees. J Phys Act Health 2014;11:1354-1361.

18

12. Weijman I, Ros WJ, Rutten GE, Schaufeli WB, Schabracq MJ, Winnubst JA. The role of work-related and personal factors in diabetes self-management. Patient Educ Couns 2005;59:87-96. 13. Hakkarainen P, Moilanen L, Hanninen V, Heikkinen J, Rasanen K. Work-related diabetes distress among Finnish workers with type 1 diabetes: a national cross-sectional survey. J Occup Med Toxicol 2016;11:11. 14. Young J, Waclawski E, Young JA, Spencer J. Control of type 1 diabetes mellitus and shift work. Occup Med (Lond) 2013;63:70-72. 15. Annor FB, Roblin DW, Okosun IS, Goodman M. Work-related psychosocial stress and glycemic control among working adults with diabetes mellitus. Diabetes Metab Syndr 2015;9:85-90. 16. Hennink MM, Kaiser BN, Sekar S, Griswold EP, Ali MK. How are qualitative methods used in diabetes research? A 30-year systematic review. Glob Public Health 2017;12:200219. 17. Sturges J, Hanrahan K. Comparing telephone and face-to-face qualitative interviewing: a research note. Qualitative Research 2004;4:107-118. 18. Barbour RS. Checklists for improving rigour in qualitative research: a case of the tail wagging the dog? BMJ 2001;322:1115-1117. 19. Metzenthin P, Helfricht S, Loerbroks A, Terris DD, Haug HJ, Subramanian SV, Fischer JE. A one-item subjective work stress assessment tool is associated with cortisol secretion levels in critical care nurses. Prev Med 2009;48:462-466. 20. Eng A, Mannetje A, Pearce N, Douwes J. Work-related stress and asthma: results from a workforce survey in New Zealand. J Asthma 2011;48:783-789.

19

Table 1. Characteristics of the study population (n=30) Sociodemographics

Occupational data

Diabetes-related data

1

Female sex, n (%)

17 (56.7)

Age in years, mean (SD), min-max

44.7 (11.3), 24-64

Married, n (%)

15 (50.0)

High educational level1, n (%)

17 (56.7)

Years in current job, mean (SD), min-max

16.3 (13.0), 0-37

Working full-time (as opposed to part-time)

21 (70.0)

Employed (as opposed to self-employed)

26 (86.7)

Job with leadership responsibility

15 (51.7)

Work stress2 (1-10), mean (SD), min-max

6.4 (2.4), 1-10

Diabetes disclosed at work, n (%)

29 (96.7)

Type 1 diabetes present (as opposed to type 2), n (%)

19 (65.5)

Years since diabetes diagnosis, mean (SD), minmax

15.0 (10.5), 2-41

At least one diabetes complication 3, n (%)

13 (43.3)

Need to carry out glucose measurement, n (%)

28 (93.3)

High level of school education (i.e. “Abitur“ or ”Fachhochschulreife” versus lower degrees or no

formal degree) 2

We measured overall work stress by a numeric rating scale ranging from 0 (low stress) to 10 (high

stress). Similar single-item measures of overall work stress have been used earlier and were found to correlate with markers of physiological stress responses (i.e. cortisol) [19] and adverse health outcomes [20]. 3

Measured by self-reported hypertension, myocardial infarction, stroke, circulatory disorders at the

legs, retinopathy, polyneuropathy, and nephropathy

20

Table 2. Quotes of the participants related to workload and job control. High workload WL1: R: Well if there are 300 or 400 persons on site, [..] you don’t think about it. About measuring your blood sugar level. Then you ’re basically occupied with something else. WL2: R: Well, being pressed for time, I would agree. Particularly when a colleague was absent (,) one had, I (,) I don’t want to say twice the amount of work, but quite, quite some more. And that came on top, right? That creates pressure. And that’s more important. So the first, one’s first thought isn’t that the blood sugar level could be too low. WL3: I: Can you think of situations when you find it difficult to inject or measure? When you don’t get around to doing it? R: Being understaffed due to illness or vacation. A lot of client contacts. Yes, then it happens quickly that you forget about it. Poor job control JC1: R: Well, yes, when a device stopped working or there is a task that needs to be done immediately because of the operating pro cedure. And in this moment I can’t say: “Alright everyone, I can’t help you right now, I have to eat something first”, or something like this. It can sometimes happen that then I have to skip a meal or something like that. JC2: R: It varies quite a lot. That’s actually a problem for me. Sometimes there’s a break, sometimes there isn’t. And they don’t always occur at the same time. And I realize that doesn’t do me good. It can’t be changed. That’s how it is. I can’t influence it. When I sta rt a conversation I don’t know, will it take 15 minutes, 20 minutes or 45 minutes? JC3: R: The problem is that it’s possible that a colleague won’t show up, I drive to school, expect a calm afternoon, I have already injected insulin, and then I arrive at school and am told that I have to replace a teacher and teach class 3A and give sports lessons. JC4: It’s actually a general problem that when I have prepared for lunch and injected 20 minutes before having lunch and then I get a phone call. And that happens quite often. Then I just don’t have time to eat. JC5: R: [..] Prevention can become very, very difficult, especially during business trips, because the food is different and there is also a lot of food that is delicious, and I like to eat. JC6: There were appointments made by others I had to attend. And arriving on time was defined as arriving two minutes in advance. If you also have to do something else then you can’t split into two people, so you turn into a long-distance runner […] fulfilling two tasks at the same time, yes. JC7: R: in our production hall in the summer we have 43 – 45 degrees. And I have already found out that when I carry around the measuring device all the time, then the device is too hot. So I have to leave it the workshop. And if I had to walk to the workshop among the machines, well I’d say that I’d lose five to ten minutes each time. And, well, then our head of production would get angry. Because that’s ten minutes of unproductive machine time. JC8: I: Do you think that problems are more frequent during rotating shifts? R: Yes, they have become more frequent since they int roduced rotating shifts. I didn’t have as many problems before, because I could almost always eat at the same time. JC9: During late shift one naturally goes to bed later and gets up later in the morning. And therefore the process of adaption, especially during the first two or three days, that’s usually quite problematic when it comes to blood sugar levels. I = interviewer; R = respondent

21

Table 3. Quotes of the participants related to the physical and perceived social working environment Physical working environment PWE1: R: When only my colleague was in the room, I did in fact inject insulin at my desk, below the edge of the desktop. But when there were clients or other colleagues, then I would always have to go the rest room. You’re not alone there as well, because one (…) I think the cubicle itself is extremely unhygienic. You can’t even lay something down there. Around the washbasins it was better. It didn’t happen terribly often, but it still bothered me. And then I started to sometimes go to a store-room, closed the door and injected. PWE2: I: In what type of situations did you find it particularly difficult to take care of your diabetes? R: […] Especially during a second job […] as a meter-reader. […]. One had dirty hands. And one didn’t have the opportunity to wash one’s hands and didn’t want to prick one’s finger, right?. PWE3: R:(...) the feeling of having a low blood sugar level, when temperatures change, then you’re not always sure whether it’s the blood sugar level or the weather that makes you sweat. PWE4: R: in our production hall in the summer we have 43 – 45 degrees. And I have already found out that when I carry around the measuring device all the time, then the device is too hot. So I have to leave it the workshop. And if I had to walk to the workshop among the machines, well I’d say that I’d lose five to ten minutes each time. And, well, then our head of production would get angry. Because that’s ten minutes of unproductive machine time. Perceived social working environment SWE1: R: Well, basically you can’t expect everyone to have no problem at all when you inject while sitting at a table. That would be the case with external clients. So you have to make a choice if this is adequate or not. Sometimes you don’t want to take a risk. SWE2: R: I work at a special needs school. This means that I can’t take out my measuring devices and insulin pens everywhere. They just fascinate the children, I have to lock them away. SWE3: R: When I attend a meeting, then I don’t necessarily want to take out my blood sugar meter and prick my finger. And then everybody watches me. ID 1SWE4: R: When only my colleague was in the room, I did in fact inject insulin at my desk, below the edge of the desktop. But when there were clients or other colleagues, then I would always have to go the rest room. […] SWE5: Situations in which it is difficult, when I talk to a patient and notice: ‘ah, now it would be good to measure’. Then it’s more daunting to say: ‘I have to do this now’. Especially when they don’t know that I have diabetes. I: Did you have fewer opportunities during your previous jobs to measure your blood sugar levels or to inject insulin? B: No, I did do it (…), but without the same feeling of freedom, right? So I felt more restricted and like somebody is watching me. Because there were colleagues, doctors’ assistants, who sometimes watched me or so. And that was a bit, um (..) on the one hand you don’t want special treatment, but on the other hand it’s necessary to handle this disease. And here I don’t […] owe any explanations. SWE6: I have to (,) take care about eating, that’s not always easy. Because when it comes to appointments I have to consider patients’ needs. And then it becomes a bit difficult, if there is a request to visit a patient at home during lunchtime when I should actually have lunch. (…) I can’t do it in patients’ homes, right? I: Yes.(..) How do you arrange for having time for meals? R: I have to do it before leaving. SWE7: R: in our production hall in the summer we have 43 – 45 degrees. And I have already found out that when I carry around the measuring device all the time, then the device is too hot. So I have to leave it the workshop. And if I had to walk to the workshop among the machines, well I’d say that I’d lose five to ten minutes each time. And, well, then our head of production would get angry. Because that’s ten minutes of unproductive

22

machine time. SWE8: (...) But if one was measuring blood sugar levels all the time and one wouldn’t do it openly but at the (incomprehensible) or in the bathroom or and, then people would say ‘two minutes here, two minutes there’, you also spend ten minutes time just for taking care of diabetes. Yes, there simply isn’t (…) enough understanding. (...) I: Would you wish for people to have more knowledge and understanding about diabetes? R: (…) Of course. I just don’t know if people would understand, if one can really understand without having experienced it. I = interviewer; R = respondent

23

Table 4. Quotes of the participants related to the types of self-management activities which are affected by working conditions SMA1: R: When I am somehow immersed in my work, then I just don’t think about it. Well, then accordingly I don’t inject. Basically that’s no problem. However, it’s important to eat some carbohydrates in between to keep the blood sugar level stable. SMA2: Well, I usually only feel bad when my blood sugar level is low. And then I take the patient I’m caring for at the moment, who I ha ve just taken care of, (…) I finish what I’m doing and take him to the dining room and then I have a glass of juice and I just go o n. I don’t even measure or inject then, I do it sometime later. SMA3: R: Well, yes, when a device stopped working or there is a task that needs to be done immediately because of the operating pro cedure. And in this moment I can’t say: “Alright everyone, I can’t help you right now, I have to eat something first”, or something like this. It can sometimes happen that then I have to skip a meal or something like that. SMA4: I: How do you deal with regularly measuring your blood sugar level and injecting insulin during working hours? R: Well, I don’t do it regularly, I do it irregularly. Because I also eat at irregular times. SMA5: I: And how was it before during your working hours (...) R: I have practically never measured, never. Because I didn’t have the time, right? SMA6: R: Well, being pressed for time, I would agree. Particularly when a colleague was absent (,) one had, I (,) I don’t want to say twice the amount of work, but quite, quite some more. And that came on top, right? That creates pressure. And that’s more important. So the first, one’s first thought isn’t that the blood sugar level could be too low. SMA7: R:(...) the feeling of having a low blood sugar level, when temperatures change, then you’re not always sure whether it’s the blood sugar level or the weather that makes you sweat. SMA8: R: The problem is that it’s possible that a colleague won’t show up, I drive to school, expect a calm afternoon, I have already injected insulin, and then I arrive at school and am told that I have to replace a teacher and teach class 3A and give sports lessons. SMA9: It’s actually a general problem that when I have prepared for lunch and injected 20 minutes before having lunch and then I ge t a phone call. And that happens quite often. Then I just don’t have time to eat. SMA10: During late shift one naturally goes to bed later and gets up later in the morning. And therefore the process of adaption, especially during the first two or three days, that’s usually quite problematic when it comes to blood sugar levels. SMA11: Every now and then there’s a situation where there’s food that I don’t know how to estimate. When it’s difficult to say no. Let’s say a Japanese colleague invites everyone to a Japanese restaurant. They offer unusual things, fried, breaded, and in combination with Japanese noodles. I have no idea at all, because we don’t eat this here, how much can I eat? Then I‘m a bit more cautious. SMA12: R: It has already happened that it [the physician appointment] was scheduled during working hours. Then it was always a problem, well, (…) I don’t want to say that my boss doesn’t like it, but he sometimes says, my colleagues also do it: ‘Oh well, he has to see the doctor yet again’. And it’s (…) I don’t want to say that one’s excluded. But that’s how it is. (…),Yes, one (,) some of the colleagues address it in conversation. And, yes, it makes me feel quite uncomfortable. I = interviewer; R = respondent

24

Table 5. Quotes of the participants related their attitudes towards diabetes self-management at work ATS1: Somehow I am in the happy situation that I don’t feel ashamed to inject insulin immediately, for example, during lunch in the cafeteria, and my colleagues are very tolerant of it. ATS2: R: Diabetes is a part of me. And I am open about it. […] I don’t see any reason why I should keep it secret. In my opinion, keeping it secret would mostly do me harm. ATS3: Once you start keeping something secret then I think your psychological burden is so high that the whole thing becomes a problem that perhaps you wouldn’t have had in the first place. At least that’s my opinion, and therefore I’ve always taken it easy. And when others see that I take it easy then that gives them the chance to take it easy as well, so that they can continue to see me as a colleague just as they did before without thinking „She’s got diabetes“, and then I reached my goal. I think many others should behave in the same way, but perhaps they just don’t have the chance at their workplace. ATS4: Well I don’t consider my diabetes as limiting. And yes, I also have the impression that if you tell others openly that you get along well with it, then, well, it isn’t perceived as limiting. ATS5: R: He [a colleague] sometimes says: „You should go and measure your blood sugar level, you’re not looking well“. But in the end no one can help me because no one really knows enough about it. ATS6: If I think it’s necessary I stop doing whatever I do and take care of my health. ATS7: Because one must keep in mind that I still have to perform. I always tell myself - if I occasionally measure or eat half an hour later - one can rather put up with this than with the employer saying ‘sorry, but that’s not how it works’. Right? So you adapt somehow. ATS8: Well, I usually only feel bad when my blood sugar level is low. And then I take the patient I’m caring for at the moment, who I have just taken care of, (…) I finish what I’m doing and take him to the dining room and then I have a glass of juice and I just go on. I don’t even measure or inject then, I do it sometime later. ATS9: R: If you go for convalescent care or something like this, it would be very, very difficult to actually claim your rights in a small company. Not from a legal point of view, but socially, because you don’t want to lose your job in the long term. Not that one has threatened or something like this, but it’s just difficult. I = interviewer; R = respondent

25

Table A. Relevant sections of the initial topic guide Broad category

Primary questions

Introductory questions



Work-related selfmanagement

Barriers

Facilitators

Conclusion

You are employed and you have diabetes - how do you experience a typical working day?  As an individual with diabetes you may need to take care of many things for your diabetes, such as injecting insulin, measuring blood sugar, taking pills, deciding what and when to eat, regular visits to physicians, checking your feet. What exactly do you need to do for your diabetes?  You just told me about your diabetes-related tasks. To what extent can you carry out these tasks during working days?  What are situations at work (if there are any) that make it difficult to carry out those diabetes-related tasks?  What do these situations look like? [Try to understand the nature of barriers in detail]  Which diabetes-related tasks are particularly challenging to carry out at work – and why?  What do you do when it is difficult to carry out those diabetesrelated tasks at work? What things make it easier to deal with your diabetes at work? [Note to interviewer: probe after both personal and environmental factors, such as organizational structures, work routines etc.) What are important aspects that we did not address yet?

26

Highlights    

We examined the link of work conditions with diabetes self-management. Glycemic control, diet and physician consultations were primarily affected. Novel observations highlight, first, the role of multi-tasking and understaffing And second, challenges related to business trips and temperature at work.

27