Technical report: Sickness absence and health in the workplace: understanding employer behaviour and practice

Question 1

Statement of compliance

Accepted Answer

This research complies with the 3 pillars of the Code of Practice for Statistics: value, trustworthiness and quality.

Value of this research

the research provides a description of the health and wellbeing initiatives used by employers and contributes to the growing evidence base
findings from this report have informed the ongoing development of policy decisions relating to employers

Trustworthiness

this research was conducted, delivered and analysed impartially by Ipsos MORI, working to the Government Social Research code of practice
authors: Ipsos MORI – Trinh Tu, Kelly Maguire, Karl Ashworth and Sarah Tipping

Quality

the survey was carried out using established statistical methods
the research has been quality assured using Ipsos MORI’s internal quality checking processes
the report has been checked thoroughly by Employers, Health and Inclusive Employment (EHIE) analysts to ensure it meets the highest standards of analysis and drafting

Question 2

1.  Introduction

Accepted Answer

The ‘Sickness absence and health in the workplace’ survey and follow-up qualitative research was commissioned by the Employers, Health and Inclusive Employment (EHIE) team, a UK government unit, which brings together officials from the Department for Work and Pensions (DWP) and the Department of Health and Social Care (DHSC). EHIE leads the government’s strategy to support working-age disabled people and people with long-term health conditions to enter, and stay in, employment.

‘Improving Lives: The Future of Work, Health and Disability’^{[footnote 1]} outlined the role of employers in helping people with disabilities or health conditions stay, and thrive, in work, as well as to prevent unnecessary sickness absence, presenteeism and health-related job loss. Disabled people and people with long-term health conditions are at greater risk of falling out of work^{[footnote 2]}, and in 2019, the Work and Health Unit (WHU) launched a consultation seeking views on the different ways in which government and employers could take action to reduce ill-health-related job loss^{[footnote 3]}. This report builds on the 2011 ‘Health and well-being survey of employers’^{[footnote 4]} which was commissioned to provide evidence into a range of measures, including employers’ perceptions of the importance of work to health and health to work, the provision of health and wellbeing initiatives, and employers’ occupational sick pay (OSP) policies.

This report also helps to develop the limited evidence base in this area, as identified in a scoping review conducted by the University of Nottingham (see Appendix A), commissioned by EHIE as part of this research. In total, 2,564 interviews were achieved with organisations with 2 or more employees in Great Britain (GB), broken down by sample type as detailed in the table below (See Table 1.1).

Table 1.1: Size breakdown of overall sample

Size of employer	Number of interviews	Number of follow up interviews
Small (2 to 49)	1457	16
Medium (50 to 249)	584	8
Large (250+)	523	6

The survey was conducted at a ‘head office’ level with the most senior person with responsibility for personnel issues (or the owner or manager of the organisation if no such a person existed), where an employer had 2 or more employees. The survey was carried out by Computer Assisted Telephone Interviewing (CATI).

The research was designed to help explain the following research questions:

employers’ attitudes towards employee health and wellbeing in the workplace
employers’ understanding of their legal responsibilities
employer provisions of health and wellbeing initiatives
management of sickness absence and health conditions in the workplace, with a focus on retention and reintegration
employers’ provision of sick pay
employers’ use of occupational health (OH) services

Following the completion of the survey, EHIE commissioned Ipsos MORI to design additional qualitative work with a sample of 30 employers (for size breakdown see Table 1.1) to explore some of the survey findings in greater depth^{[footnote 5]}.

The research was carried out in accordance with the requirements of the international quality standard for Market Research, ISO 20252, and with the Ipsos MORI terms and conditions.

The following chapters of the Technical Report provide additional detail on each stage of the project.

Question 3

2.  Survey technical details

Accepted Answer

This section provides more technical detail about the survey design and development, covering sampling, fieldwork details, response rate, and data processing and weighting.

2.1. Survey and questionnaire development

The questionnaire was developed by Ipsos MORI and approved by EHIE. The questionnaire was tested with 9 cognitive interviews undertaken by Ipsos MORI between 3 and 23 April 2018. Interviews were around 45 minutes and conducted by telephone. They were recruited using a sample provided by Dun and Bradstreet (a commercial sample provider), with minimum quotas by size and sector.

A pilot survey consisting of 24 telephone interviews with eligible employers (from the Inter-Departmental Business Register (IDBR) sample frame, as used for the main survey) was then conducted between 14 and 18 May 2018. Small changes to the questionnaire were made based on feedback from the cognitive testing and pilot survey (see Appendix B for the final questionnaire used). Prior to fieldwork, sampled employers were sent an advance letter, which provided information about the survey and gave them the opportunity to opt-out. Telephone matching was also undertaken for leads without a valid telephone number.

2.2. Sampling

The sampling frame for this survey was sourced from the Office for National Statistics (ONS) IDBR. The findings are weighted by size and sector to be representative of GB employers. 92% of employers are small (2 to 49 employees), 6% medium (50 to 249) and 2% large (250+). Large employers were oversampled, making up 20% of the raw sample to allow for analysis within the size category.

Table 2.1: Count of employers by size band in Great Britain

Size band (number of employees)	Number of units	Percentage
Small (2 to 49)	1,200,780	96.6%
Medium (50 to 249)	34,835	2.8%
Large (250+)	7,510	0.6%
Total	1,243,125

Source: Business population estimates, 2018^{[footnote 6]}

Sample preparation and selection

Prior to the pilot and main stage sample selection, the sample was cleaned. This involved removing leads with invalid telephone numbers (such as overseas numbers and numbers with too many/few digits) and removing duplicate leads. Ipsos MORI also sent advance letters to the selected sample giving them an opportunity to opt out of the research.

For the main stage survey, a sample of 12,742 employers was selected. Not all of this sample was used in the main stage – in total, 12,210 leads were issued – and the sample that was used was released in batches. Each batch was stratified based on disproportionate size and sector targets (aimed at achieving enough sample in certain sectors and amongst large employers for subgroup analysis) as well as the relative unadjusted response rate of each size by sector cell in the previous batch.

Table 2.2: The sample of employers by industry sector (SIC07) and size

SIC07 category*	Small (2 to 49)	Medium (50 to 249)	Large (250+)	Total
(A) Agriculture, Forestry and Fishing	196	27	23	246
(BDE) Mining and Quarrying; Utilities, Waste Management and Remediation Activities	48	30	52	130
(C) Manufacturing	617	452	441	1510
(F) Construction	906	143	110	1159
(G) Wholesale and Retail Trade; repair of motor vehicles and motor cycles	1164	334	395	1893
(H) Transport and Storage	294	107	130	531
(I) Accommodation and Food Service Activities	798	222	179	1199
(J) Information and Communication	516	134	150	800
(K) Financial and Insurance Activities	184	154	165	503
(L) Real Estate Activities	187	27	87	301
(M) Professional, Scientific and Technical Activities	1,001	247	281	1529
(N) Administrative and Support Service Activities	693	276	394	1363
(O) Public Administration and Defence; compulsory social security	1		2	3
(P) Education	91	58	41	190
(Q) Human Health and Social Work Activities	387	251	150	788
(R) Arts, Entertainment and Recreation	117	53	108	278
(S) Other Service Activities	288	17	14	319
Total	7,488	2,532	2,722	12,742

* Two SIC codes were excluded from the sample design: (T) Activities of Households as Employers; undifferentiated goods-and-services-producing activities of households for own use and (U) Activities of extraterritorial organisations and bodies.

2.3. Fieldwork

A probability sample telephone survey was undertaken with 2,564 employers. Fieldwork took place between 4 June and 13 August 2018. With the probability sample nature of the survey, each lead was called a minimum of 12 times unless it had already achieved a final outcome, and an outcome was sought for every lead in the sample. The following respondents were screened out of the survey at the start as ineligible:

those with fewer than 2 employees
those who were not based in Great Britain

2.4. Fieldwork outcomes and response rates

The survey achieved an adjusted response rate of 43.7%. A breakdown of the sample outcomes is presented in Table 2.3.

Table 2.3: Sample outcomes

	Total sample	Valid sample*
Completed interviews	2,564	2,564
Ineligible	197
No reply	2,863
Refusals and abandoned interviews	2,904	2,697
Still dialling (no outcome)	652	605
Unusable numbers	3,030
Grand total	12,210	5,866
Eligibility^{[footnote 7]}	92.9%
Co-operation rate^{[footnote 8]}	46.9%
Response rate^{[footnote 9]}	43.7%

* Selected outcome codes × eligibility rate

Table 2.4: Response rate by employer size band

Size band (number of employees)	Interviews (N)	Response rate (%)
Small (2 to 49)	1,457	43.1%
Medium (50 to 249)	584	49.8%
Large (250+)	523	39.8%
Total	2,564	43.7%

2.4. Data processing and weighting

The data were weighted by size and sector to be representative of the eligible GB employer population, based on population profile figures from the IDBR. The data was anonymised, showing only fundamental employer characteristics. The ONS’ Business Population Estimates (BPEs), 2018^{[footnote 10]} placed the employer population with at least 2 employees at 1.2 million, whereas the IDBR estimates used by Ipsos MORI to design the sample and weight the data totalled 1.09 million. This discrepancy was due to:

timing: the BPEs and our sample were drawn at different points in March 2018. The estimates were published, but the IDBR data used by ONS was subsequently updated and used to determine our sample
unit type: the sample was run using reporting units (either a one-to-one relationship with an enterprise, whereby individual sites belonging to the same organisation are reported on as a whole, or a one-to-many, whereby individual sites belonging to one organisation are reported separately). Where there is a one-to-many relationship, reporting units are often split by geography or by activity. This means that the sample may include enterprises who are represented across multiple activities and locations
unregistered businesses: the IDBR only holds businesses that are registered for value-added tax (VAT) or Pay As You Earn (PAYE). The BPEs make an estimate for the number of these businesses who are not registered for VAT or PAYE and include them in the figures

Table 2.5: Sample profile weighted by employer unit and employee volume

Variable – size:

Categories	Unweighted	Weighted by employer unit	Weighted by employee volume
Small	1457	92%	37%
Medium	584	6%	18%
Large	523	2%	45%

Variable – sector:

Categories	Unweighted	Weighted by employer unit	Weighted by employee volume
Agriculture and Energy	107	4%	4%
Manufacturing	351	8%	12%
Construction	225	13%	7%
Distribution, Hotels and Restaurants	572	26%	30%
Transport and Communications	220	11%	10%
Financial, Professional and Administrative Services	657	26%	24%
Public Administration, Education and Health	283	6%	8%
Other Services	149	7%	5%

Variable – ease of recruiting staff:

Categories	Unweighted	Weighted by employer unit	Weighted by employee volume
Easy	733	27%	30%
Difficult	1,346	53%	50%

Variable – ease of retaining staff:

Categories	Unweighted	Weighted by employer unit	Weighted by employee volume
Easy	1,565	68%	55%
Difficult	479	15%	22%

Variable – employee representation or trade union:

Categories	Unweighted	Weighted by employer unit	Weighted by employee volume
Yes	355	5%	21%
No	2,169	94%	77%

Variable – decisions on daily work tasks:

Categories	Unweighted	Weighted by employer unit	Weighted by employee volume
Employees	208	12%	7%
Managers or supervisors	898	33%	35%
Both employees and managers	1,443	55%	57%

Question 4

3.  Advanced analysis

Accepted Answer

This chapter covers the details of the advanced statistical analysis used in the report, including the methodology, rationale, and interpretation.

3.1. Latent class analysis

What is latent class analysis?

Latent class analysis (LCA) is a statistical technique which classifies individuals into unobserved classes based on observed behaviours and/or characteristics. In this survey, it was used to analyse responses to the multi-coded questions, where respondents picked a number of relevant responses in answer to a question.

Where respondents choose common combinations of options they reveal a ‘latent class’ which distinguishes the respondents from one another based on the pattern of their chosen options. In essence, LCA is a way of looking at commonly reported combinations of responses to particular questions. These combinations are called ‘latent classes’. The approach calculates the probability that each response option belongs to an underlying latent class category. Therefore, we can predict the membership of the categories by determining the probability of a respondent falling into a particular latent class based on their chosen options. This can then be used to predict the frequency of responses to each permutation of the observed variables from the question.

A goodness of fit (GoF) statistic is then calculated by comparing the observed and predicted frequency of permutations across the response options. There is no absolute measure of the ‘correct’ number of categories in the best fitting latent class model. Because of this, it is necessary to compare the GoF for each model that is run. The number of categories may vary substantially, and it is then a case of repeating the analysis to indicate the appropriate number of categories to provide the best fitting models. Each question for which LCA was conducted is described in more detail below.

3.2. LCA: Workforce composition

Employers were asked approximately what proportion of their workforce was made up of the following categories^{[footnote 11]}:

managers or professionals (managers include directors and managers of internal departments or sections, and professional occupations usually require a degree or equivalent formal qualification)
technicians or skilled trades (these occupations require a substantial period of full-time training or further study)
semi- and unskilled occupations (these occupations involve mostly routine tasks. Most do not require formal educational qualifications)

LCA was undertaken to group employers into categories based on the predominant occupation within their organisation. This analysis enabled us to better understand the predominant mix of occupational levels within organisations, rather than looking at the proportions of each occupational level in isolation, across employers as a whole. Of the 2,564 respondents, 2,445 provided usable responses (employers who could not give proportions were excluded) and the LCA resulted in 4 categories. The categories identified were then utilised in the logistical regression analysis (see Section 4.2 in main report) to help determine the relationship between the skill level of the workforce and the sophistication of an employers’ health and wellbeing provisions (OH and OSP). Please note that all figures are weighted, unless specified:

mixed (37% of employers) – this class represented firms where the workforce was reasonably balanced by the different types of employee occupations, i.e. all 3 occupational statuses were present in moderately low levels
predominantly technicians and skilled trades (27%) – this class represented firms dominated by technicians and skilled trades workers, with moderately low proportions of professionals and managerial occupations, and low proportions of semi- and unskilled workers
predominantly semi- and unskilled occupations (20%) – this class was predominated by moderately high and high proportions of semi- and unskilled workers, respectively, and mostly low proportions of skilled or managerial workers
predominantly professional and managerial occupations (16%) – this class was primarily composed of high proportions of professional and managerial workers, with low proportions of technical and semi- or unskilled workers

Table 3.1: Occupational LCA by size of employer

	Total	Small	Medium	Large
Mixed	37%	37%	30%	58%
Technicians or skilled trades	27%	27%	28%	18%
Semi- and unskilled occupations	20%	19%	37%	18%
Professional and managerial occupations	17%	17%	5%	6%
Base	2,445	1,427	562	456

Base: All employers selected for LCA, excludes those who could not give an answer at B7: OCCUPATION (unweighted).

Table 3.2: Occupational LCA by sector

	Total	Agriculture and Energy	Manufacturing	Construction	Distribution, Hotels and Restaurants	Transport and Communications	Financial, Professional and Administrative Services	Public Administration, Education and Health	Other Services
Mixed	37%	40%	31%	45%	32%	33%	44%	27%	33%
Technicians or skilled trades	27%	15%	41%	39%	22%	36%	17%	40%	30%
Semi- and unskilled occupations	20%	27%	18%	5%	37%	11%	12%	17%	24%
Professional and managerial occupations	17%	18%	10%	11%	10%	20%	27%	16%	14%
Base	2,445	102	333	210	552	213	625	267	143

Base: All employers selected for LCA, excludes those who could not give an answer at B7: OCCUPATION (unweighted)

3.3. LCA: Provision of retention and reintegration measures

Employers were asked which measures, if any, they had used in the last 12 months to support employees with health problems to remain in-work or support them in returning to work, from the following^{[footnote 12]}:

meetings with employees
phased returns to work from sickness absence (returning to full duties and hours at work gradually, over a defined time period)
amending employee workload or job role (for example, reduced hours/days, extra breaks, or different duties)
workplace adjustments (such as different chairs or desks, building modifications, or other specialised equipment)
a job coach or personal assistant (for example, a sign-language interpreter for meetings)
additional external support or advice (for example, clinical support such as psychological therapy or physiotherapy, or another expert or specialist)
none of these

LCA was undertaken to group employers into categories based on the predominant combinations of wellbeing measures they provided. This analysis enabled us to better understand the predominant combinations of wellbeing offers, i.e. their ‘suite’ of wellbeing provision, rather than looking at the proportions of each measures provided in isolation. It also allowed us to look at these classes in relation to other employer characteristics and uncover key relationships (see Section 5.1 in the main report).

Overall, 688 employers (unweighted, 27%) said they gave no support of this type to their employees. These employers were excluded from the analysis. LCA was undertaken exploring solutions with between 2 and 6 latent classes, and the model chosen included 5 latent classes, as it had the best goodness of fit.

The 5 classes are as follows:

1. Latent class 1 (17%) – these employers provided the most comprehensive wellbeing packages, offering between 4 and 6 wellbeing measures. Employers in this latent class more commonly provided external support and job coaches or personal assistants, compared to other classes.
2. Latent class 2 (41%) – these employers provided between 2 and 4 wellbeing measures and these focused on meetings with employees, phased returns to work, amended workloads and workplace adjustments.
3. Latent class 3 (15%) – these employers provided one to 3 wellbeing measures, focused on phased returns, amended workloads and/or workplace adjustments. They were notable by the absence of meetings with employees.
4. Latent class 4 (8%) – these employers predominantly provided workplace adjustments, supplemented by one other service, with the exception of a phased return to work.
5. Latent class 5 (20%) – these employers relied largely on meetings, supplemented mostly by one of the other alternative measures, with the exception of a job coach or personal assistant.

Table 3.3: Number of retention and reintegration measures by latent class (raw figures, unweighted)

	Latent class 1	Latent class 2	Latent class 3	Latent class 4	Latent class 5
One measure	0	0	61	46	100
Two measures	0	48	78	6	109
Three measures	0	271	41	19	11
Four measures	65	404	5	3	0
Five measures	489	0	0	0	0
Six measures	120	0	0	0	0
Base	674	723	185	74	220

Base: All employers selected for LCA, excludes those who could not give an answer or did not provide any measures at G2: WELLBEING (unweighted)

Table 3.4: Retention and reintegration LCA by size of employer (weighted)

	Total	Small	Medium	Large
Latent class 1	17%	13%	29%	64%
Latent class 2	41%	41%	44%	32%
Latent class 3	15%	16%	13%	1%
Latent class 4	8%	9%	1%	*
Latent class 5	20%	21%	13%	2%
Base	1,876	829	538	509

Base: All employers selected for LCA, excludes those who could not give an answer or did not provide any measures at G2: WELLBEING (unweighted)

Table 3.5: Retention and reintegration LCA by sector

	Total	Agriculture and Energy	Manufacturing	Construction	Distribution, Hotels and Restaurants	Transport and Communications	Financial, Professional and Administrative Services	Public Administration, Education and Health	Other Services
Latent class 1	17%	12%	17%	8%	15%	11%	18%	31%	21%
Latent class 2	41%	42%	43%	40%	44%	41%	35%	46%	41%
Latent class 3	15%	22%	15%	15%	18%	18%	15%	8%	15%
Latent class 4	8%	2%	4%	7%	6%	9%	13%	2%	8%
Latent class 5	20%	22%	22%	30%	17%	21%	20%	14%	15%
Base	1,876	69*	292	125	396	168	479	244	103

Base: All employers selected for LCA, excludes those who could not give an answer at B7: OCCUPATION (unweighted).

* Caution should be taken with figures where the base size is less than 100.

3.4. Logistic regression analysis

What is logistic regression analysis?

Logistic regression is a type of statistical analysis which assesses the degree to which 2 or more variables are related to each other. It allows us to understand which factors (independent variables) have an impact on the variable you are trying to understand (dependent variable), whilst controlling for other variables. In this survey, 2 regressions were carried out to explore which employer behaviours and characteristics were independently associated with:

the provision of occupational health services
whether employers offered more than statutory sick pay to their employees

The aim of these regressions was to examine which employer characteristics had a bearing on, or resulted in greater likelihood of, employers providing occupational sick pay (OSP), and, separately, occupational health (OH) services. The regression approach complements exploration of bivariate tables because many of the independent variables have relationships with each other as well as with the dependent variable. The regression ‘controls’ for these correlations between independent variables, allowing for a better understanding of the degree of the relationship between each employer characteristic specifically and the provision of these services. We can, therefore, use the results of the regression model to understand which combinations of characteristics are associated with the various likelihoods of an employer providing them.

Many potential characteristics were available to explore their impact on the likelihood of increasing (or decreasing) occupational health and sick pay provision. The pool of characteristics was first created through choosing those that previous research and theory suggested were likely to be associated and also showed an association at a bivariate level. The pool was then separated into those characteristics which represented firmographic, attitudinal and behavioural dimensions. Separate regressions were run on each outcome by each of the 3 broad dimensions of characteristics (firmographics, attitudes and behaviours). Those firmographic, attitudinal and behavioural characteristics which were significant in the preliminary regressions were retained for inclusion into a ‘full’ model. Finally, a ‘reduced’ model was produced which excluded variables which were non-significant in the full model.

This staged approach was applied for 2 reasons. First, because many of the characteristics were correlated with each other (multicollinearity) and shared associations can hide some important associations, i.e. they become non-significant because the shared relationship is ascribed to other, more dominant characteristics. Second, the inclusions of many colinear characteristics into a single model can produce a poor quality regression output which is difficult to interpret.

The presence of high degrees of multicollinearity between many variables is problematic because the size (and statistical significance) of the estimated coefficients for a characteristic depend upon which other (colinear) variables are included in the model. Consequently, we thought it prudent to explore different versions of the model and to retain those characteristics from each of the firmographic, attitudinal and behavioural dimensions which appeared most robust in the preliminary models and combine them in the final model. In this way, we attempted first to reduce the impact of multicollinearity between characteristics measuring similar concepts (for example, firmographics, attitudes, behaviours). However, variables from different dimensions were also colinear, hence previously significant results become non-significant in the ‘full’ model.

We provide estimated odds ratios to identify the degree to which characteristics increased the likelihood of an employer offering occupational health (OH) and occupational sick pay (OSP).

3.5. Occupational Health (OH) regression

A regression was undertaken to explore the relationship between the provision of OH services and a range of employer characteristics. The explanatory characteristics included firmographic variables, along with other health and wellbeing measures:

employer characteristics: size, sector, age profile of the workforce, proportion of employees with long-term health conditions or disabilities in the workforce, staff turnover, and occupational mix
employer practices and perceptions: ease of meeting legal responsibilities, practices around managing and monitoring sickness absence, type of sick pay provided to employees on sickness absence, provision of health and wellbeing services, and employer expectations of when employees should return to work following a period of sickness absence

The explanatory characteristics were chosen by EHIE through consideration of the characteristics assessed to have a potential relationship with OH provision. As with the OSP regression (see Section 3.6), many of the above characteristics were related to one another and had both a direct and an indirect relationship with OH provision. However, many of the above characteristics were related to one another and had both a direct relationship and an indirect relationship with OH provision, i.e. indirect associations are the source of the multicollinearity, discussed above.

The primary focus of the analysis was to identify the characteristics that were most strongly associated with OH. Typically, these characteristics will have both direct and indirect links to OH provision.

Due to the bivariate tables showing strong links between employer size and the sophistication of their health and wellbeing provisions, 2 separate regressions were run: one for all employers, and another for micro and small employers only, to identify key relationships within size brackets i.e. control for size. In both cases, a full model was run initially, and then refined to produce a reduced model by removing the variables which were not significant^{[footnote 13]}, allowing us to focus on the key terms.

Table 3.6: OH regression – all cases, full model

	Co-efficient	Standard error	Statistic	p. value	Odds ratio
(Intercept)	-2.66	0.42	-6.38	0	0.07
Size: 50 to 249 employees	1.07	0.23	4.64	0	2.93
Size: 250+ employees	2.95	0.38	7.7	0	19.18
Staff turnover between 5% to 9%	-0.29	0.28	-1.04	0.3	0.75
Staff turnover between 10% to 14%	-0.2	0.33	-0.62	0.54	0.82
Staff turnover between 15% to 19%	-0.9	0.42	-2.16	0.03	0.41
Staff turnover 20% or more	-0.59	0.32	-1.87	0.06	0.55
Staff turnover do not know, refused or other	0.08	0.48	0.18	0.86	1.09
Occupational level^{[footnote 14]}: Low-skilled data missing	-0.26	0.43	-0.61	0.54	0.77
Occupational level: Low-skilled 21-50% of workforce	-0.16	0.2	-0.8	0.42	0.85
Occupational level: Low-skilled 50% or more of workforce	-0.49	0.22	-2.19	0.03	0.61
Sector: Construction	-0.51	0.2	-2.61	0.01	0.6
Sector: Public Administration, Education and Health	0.46	0.26	1.77	0.08	1.58
Organisation type: Public sector	0.47	0.41	1.14	0.25	1.6
Percentage of the workforce aged 50+: 50% or more	-0.47	0.22	-2.14	0.03	0.62
Employers with at least some employees with disabilities in the workforce	0.09	0.18	0.48	0.63	1.09
Health concerns: mainly reporting mental ill-health	-0.04	0.32	-0.12	0.91	0.96
Health concerns: mainly reporting musculoskeletal conditions	-0.06	0.34	-0.17	0.86	0.94
Health concerns: Do not know	0.55	1.12	0.49	0.62	1.73
Sick pay provision: Above Statutory Sick Pay (SSP)	1.1	0.32	3.44	0	3.01
Sick pay provision: SSP only	0.24	0.31	0.76	0.45	1.27
Sick pay provision: Do not know	0.29	0.54	0.53	0.6	1.33
Had instances of long-term sickness absence (LTSA)	-0.1	0.21	-0.46	0.65	0.91
Support for employees with health problems^{[footnote 15]}: 1 or 2 provisions	0.19	0.24	0.8	0.42	1.21
Support for employees with health problems: 3 or 4 provisions	0.5	0.23	2.16	0.03	1.64
Support for employees with health problems: 5 or 6 provisions	0.55	0.29	1.89	0.06	1.73
Measures to prevent ill-health or improve general health and wellbeing: Health and Safety training	0.8	0.27	2.91	0	2.22
Measures to prevent ill-health or improve general health and wellbeing: An Employee Assistance Programme	1.17	0.19	6.15	0	3.22
Barriers to supporting return to work (RTW) following LTSA: Lack of expertise or specialist support	-0.55	0.27	-2.06	0.04	0.57
RTW barriers: Lack of time or staff resource	0.35	0.25	1.43	0.15	1.42
RTW barriers: Lack of flexibility in the way work is organised	-0.09	0.26	-0.36	0.72	0.91
RTW barriers: Employee engagement	0.53	0.26	2.09	0.04	1.7
RTW barriers: Lack of capital to invest	-0.18	0.23	-0.8	0.43	0.83
RTW barriers: Lack of senior leader support	-0.07	0.38	-0.19	0.85	0.93

The full model showed a range of employer characteristics were associated with the propensity to provide OH. Characteristics which showed no association with OH provision were dropped from the full model as well as any terms which were not significant^{[footnote 16]} to produce a ‘reduced’ model (Table 3.7). The reduced model provides a more focused set of results.

Table 3.7: OH regression – all cases, reduced model

	Co-efficient	Standard error	Statistic	p. value	Odds ratio
(Intercept)	-2.62	0.41	-6.42	0	0.07
Size: 50 to 249 employees	1.1	0.22	5.03	0	2.99
Size: 250+ employees	2.94	0.37	7.96	0	18.96
Staff turnover between 5% to 9%	-0.3	0.28	-1.07	0.29	0.74
Staff turnover between 10% to 14%	-0.17	0.33	-0.53	0.59	0.84
Staff turnover between 15% to 19%	-0.92	0.42	-2.17	0.03	0.4
Staff turnover 20% or more	-0.57	0.3	-1.91	0.06	0.56
Staff turnover do not know, refused or other	0.09	0.49	0.19	0.85	1.1
Sector: Construction	-0.25	0.43	-0.58	0.56	0.78
Sector: Public Admin, Education and Health	0.5	0.25	2.02	0.04	1.65
Sick pay: Above SSP	1.14	0.32	3.58	0	3.13
Sick pay: SSP only	0.25	0.31	0.81	0.42	1.29
Sick pay: Do not know	0.3	0.55	0.55	0.59	1.35
Measures to prevent ill-health or improve general health and wellbeing: Health and Safety training	0.76	0.27	2.77	0.01	2.13
Measures to prevent ill-health or improve general health and wellbeing: An Employee Assistance Programme	1.17	0.19	6.26	0	3.23
Percentage of the workforce aged 50+: 50% or more	-0.45	0.22	-2.09	0.04	0.63
Support for employees with health problems^{[footnote 17]}: 1 or 2 provisions	0.2	0.23	0.87	0.39	1.23
Support for employees with health problems: 3 or 4 provisions	0.54	0.21	2.59	0.01	1.72
Support for employees with health problems: 5 or 6 provisions	0.58	0.28	2.1	0.04	1.79
Barriers to supporting RTW following LTSA: Lack of expertise or specialist support	-0.46	0.23	-2.05	0.04	0.63
RTW barriers: Employee engagement	0.52	0.24	2.18	0.03	1.68

Analysis of all employers provided a reasonable model in that around 21% of the pseudo-variance in OH receipt was explained by the model (both the full and final models). The pseudo variance provides a way of thinking about how well the model predicts the dependent variable, with a maximum value of 100%. Consequently, a value of around 20% suggests a reasonable improvement over guessing the mean value through use of the model characteristics. Similarly, if we directly calculate the accuracy with which the predicted response of OH provision by an employer corresponds to the observed score, the overall predictive accuracy is around 78%.

Employer size was, as expected, strongly associated with receipt of OH provision with large (250+) employers being around 19 times more likely to provide OH than small/micro employers. In addition, sector was also influential. Employers in the Construction sector were only three-quarters as likely as employers in ‘other’ sectors to provide OH; whereas employers in Public Administration, Education and Health were about 1.7 times more likely to provide OH than ‘other’ employers.

An employee turnover rate of 15% to 19% reduced the likelihood of OH provision by over one half (0.4) compared to having 0% staff turnover. The highest rates of employee turnover were not quite significantly associated with OH reductions but were in the same lowered direction. Provision of OSP increased the odds of OH provision substantially by around 3 times more. Additionally, provision of the following wellbeing measures was also associated with OH receipt:

Health and Safety training (2.13)
an Employee Assistance Programme (3.23)

Employers who provided more than 2 wellbeing services were over 1.7 times more likely to provide OH than were employers providing fewer wellbeing support services. Employers reporting a lack of expertise or specialist support were less likely to provide OH (0.6) but those who reported employee engagement in the process were 1.7 times more likely to provide OH.

The model results for small and micro employers were not as good as their counterparts for all employers; the full model accounted for 13% of OH pseudo-variance with the final model accounting for 11%. Predictive validity of the final model was 71%. Characteristics such as sector, OSP provision and well-being provision remain significant in the final model. However, the proportion of the workforce aged 50+ seems important in the full model for small/micro employers but is not significant in the reduced model.

3.6. Sick pay regression

A regression was undertaken to explore the relationship between occupational sick pay (OSP) provision and a range of employer characteristics. The explanatory characteristics (independent variables) covered a range of firmographic variables, included in a stepwise manner, as described above, along with other health and wellbeing measures:

employer characteristics: size, sector, age profile of the workforce, proportion of employees with long-term health conditions or disabilities in the workforce, staff turnover, and occupational mix
employer practices and perceptions: ease of meeting legal responsibilities, practices around managing and monitoring sickness absence, whether employers offered OH services, provision of health and wellbeing services, and employer expectations of when employees should return to work following a period of sickness absence

The explanatory characteristics were chosen by EHIE through consideration of the characteristics assessed to have a potential relationship with OSP provision in the bivariate tables. However, many of the above characteristics were related to one another and had both a direct relationship and an indirect relationship with OSP provision. An indirect relationship exists where one characteristic is associated with OSP through its association with another characteristic. For example, provision of OH services was strongly linked with provision of OSP services, but both practices were more common amongst large employers. These indirect relationships are the source of the multicollinearity, discussed above.

The primary focus of the analysis was to identify the characteristics that were most strongly associated with OSP. Typically, these characteristics will have both direct and indirect links to OSP provision.

Due to the bivariate tables showing strong links between employer size and the sophistication of their health and wellbeing provisions, 2 separate regressions were run: one for all employers, and another for micro and small employers only, to identify key relationships within size brackets i.e. control for size. In both cases, a full model was run initially, and then refined to produce a reduced model by removing the variables which were not significant^{[footnote 18]}, allowing us to focus on the key terms.

Table 3.8: OSP regression – all cases, full model

Variable	Co- efficient	Standard Error	t-statistic	Odds ratio
(Intercept)	-1.08	0.35	-3.13	0.34
Size: 50 to 249 employees	0.19	0.21	0.88	1.21
Size: 250+ employees	1.35	0.44	3.09	3.86
Organisation collects sickness absence data	-0.08	0.17	-0.45	0.93
Office-based sector	0.47	0.18	2.65	1.61
Mix of office-based and manual sector	0.17	0.18	0.91	1.18
Percentage of the workforce aged 50+: 50% or more	0.35	0.18	1.92	1.42
Employers with at least some employees with disabilities in the workforce	0.28	0.17	1.68	1.33
Occupational level^{[footnote 19]}: Low-skilled data missing	0.05	0.5	0.11	1.06
Occupational level: Low-skilled 21% to 50% of workforce	-0.06	0.18	-0.35	0.94
Occupational level: Low-skilled 50% or more of workforce	-0.31	0.2	-1.55	0.73
Staff turnover between 5% to 9%	0.44	0.24	1.87	1.56
Staff turnover between 10% to 14%	0.5	0.26	1.9	1.64
Staff turnover between 15% to 19%	0.07	0.42	0.16	1.07
Staff turnover 20% or more	-0.04	0.25	-0.17	0.96
Staff turnover do not know, refused or other	-0.86	0.48	-1.78	0.42
Employer finds it easy to meet their legal responsibilities	-0.02	0.19	-0.12	0.98
Employer finds it difficult to meet their legal responsibilities	-0.07	0.25	-0.27	0.93
Provides access to OH services	0.85	0.18	4.67	2.34
Employee expected to return to work (RTW) as soon as they can do most of their work	0.05	0.17	0.3	1.05
Employee expected to RTW as soon as they can do all of their work	-0.51	0.22	-2.29	0.6
Employee expected to RTW: Do not know	0.49	0.27	1.84	1.64
No instances of long-term sickness absence (LTSA)	-0.1	0.21	-0.47	0.9
Long-term sickness risks posed to organisation: having to pay sick pay	-0.26	0.29	-0.92	0.77
Used meetings with employees to support employees with health problems to remain in-work or support in returning to work	0.11	0.18	0.58	1.11
Measures to prevent ill-health or improve general health and wellbeing: Health and Safety training	-0.19	0.21	-0.93	0.82
Measures to prevent ill-health or improve general health and wellbeing: Health and wellbeing promotion programmes	0.33	0.18	1.83	1.39
Measures to prevent ill-health or improve general health and wellbeing: Interventions to prevent common health conditions becoming a problem	0.34	0.17	1.94	1.4
Measures to prevent ill-health or improve general health and wellbeing: Training for line managers	-0.37	0.18	-2.03	0.69
Measures to prevent ill-health or improve general health and wellbeing: An Employee Assistance Programme	0.26	0.2	1.31	1.3

Even though regressions can cope with correlations between independent variables, when there are many such correlations these can obscure some of the more important effects (as discussed in Section 3.4 on the impact of ‘high degrees of multicollinearity’).

The reduced model revealed some further associations with OSP which were not as clear in the full model (Table 3.9 below). For example, organisations with employees with a disability were 1.4 times more likely to provide OSP. Moreover, employers who provided health and wellbeing promotion programmes or who offered intervention programmes were around 1.5 times more likely to provide OSP than those who did not. Conversely, organisations who expected employees to return to work following sickness absence only when they were able to do all of their work were less likely to provide OSP.

Table 3.9: OSP regression – all cases, reduced model

Variable	Co-efficient	Standard Error	t-statistic	Odds ratio
(Intercept)	-1.33	0.17	-7.96	0.26
Size: 50 to 249 employees	0.26	0.2	1.32	1.3
Size: 250+ employees	1.35	0.38	3.61	3.87
Office-based sector	0.58	0.17	3.42	1.79
Mix of office-based and manual sector	0.17	0.18	0.96	1.19
Employers with at least some employees with disabilities in the workforce	0.35	0.16	2.25	1.42
Provides access to OH services	0.87	0.17	5.02	2.39
Employee expected to return to work (RTW) as soon as they can do most of their work	-0.01	0.17	-0.08	0.99
Employee expected to RTW as soon as they can do all of their work	-0.55	0.22	-2.51	0.57
Employee expected to RTW: Do not know	0.51	0.26	1.94	1.67
Measures to prevent ill-health or improve general health and wellbeing: Health and wellbeing promotion programmes	0.4	0.17	2.32	1.5
Measures to prevent ill-health or improve general health and wellbeing: Interventions to prevent common health conditions becoming a problem	0.37	0.17	2.15	1.44
Measures to prevent ill-health or improve general health and wellbeing: Training for line managers	-0.38	0.18	-2.17	0.68

It should be emphasised that the regression models captured only some of the main characteristics which distinguish between employer propensity to provide OSP. Other characteristics were not identified and measured in this study. One standard measure of fit is the amount of variation explained in the dependent variable by the independent variables in the regression model, which is indicated by a percentage figure (100% being a perfect model^{[footnote 20]}). With categorical dependent variables the variance explained cannot be measured directly, therefore it is necessary to use a ‘pseudo’ variance or adjusted R^2.

The final reduced model accounted for an estimated 12% of the (pseudo) variance. Dropping the non-significant variables, as expected, had little impact on model quality, which explained just an additional percentage (13%) of the pseudo variance. Another quality measure is predictive accuracy. This uses the predicted probabilities from the model to estimate the likelihood of providing OSP or not. Perfect accuracy means the predicted value matches the observed value. Predictive accuracy was 69%.

OSP regression: micro and small employer models

The models focusing on micro and small employers were less effective than the models analysing all employers. The final model accounted for 7% of the pseudo variance and the full model, 10%, i.e. these models were a poorer fit than were the models for all employers. Looking at how well the predicted values from the model compared to the actual response of OSP regression showed that overall predictive accuracy was 66%, i.e. two-thirds of the predictions were accurate.

The final model (Table 3.10) showed:

office-based micro and small employers were twice as likely to provide OSP
mall and micro employers with at least some employees with a disability were 1.54 times more likely to provide OSP
small and micro employers with older workers were 1.53 times more likely
there small and micro employers provided OH services, this increased the likelihood of providing OSP by 2.35
micro and small employers who provided health and wellbeing promotion programmes to improve employees’ physical activity or lifestyle were 1.56 times more likely to provide OSP

Table 3.10: OSP regression – small and micro employer model, full model

Variable	Co- efficient	Standard Error	t-statistic	Odds ratio
(Intercept)	-1.23	0.37	-3.3	0.29
Organisation collects sickness absence data	-0.1	0.18	-0.56	0.91
Office-based sector	0.51	0.19	2.66	1.66
Mix of office-based and manual sector	0.17	0.2	0.84	1.18
Percentage of the workforce aged 50+: 50% or more	0.4	0.19	2.11	1.49
Employers with at least some employees with disabilities in the workforce	0.32	0.18	1.77	1.38
Occupational level^{[footnote 21]}: Low-skilled data missing	0.06	0.61	0.11	1.07
Occupational level: Low-skilled 21% to 50% of workforce	-0.01	0.19	-0.06	0.99
Occupational level: Low-skilled 50% or more of workforce	-0.24	0.22	-1.06	0.79
Staff turnover between 5% to 9%	0.57	0.28	2	1.76
Staff turnover between 10% to 14%	0.51	0.29	1.73	1.67
Staff turnover between 15% to 19%	0.18	0.49	0.37	1.2
Staff turnover 20% or more	-0.05	0.27	-0.17	0.95
Staff turnover don’t know, refused or other	-1.09	0.61	-1.79	0.34
Employer finds it easy to meet their legal responsibilities	0.01	0.2	0.03	1.01
Employer finds it difficult to meet their legal responsibilities	-0.15	0.27	-0.55	0.86
Provides access to OH services	0.79	0.2	3.98	2.2
Employee expected to return to work (RTW) as soon as they can most of their work	0.09	0.19	0.46	1.09
Employee expected to RTW as soon as they can all of their work	-0.4	0.23	-1.74	0.67
Employee expected to RTW: Do not know	0.5	0.3	1.67	1.64
No instances of long-term sickness absence (LTSA)	-0.01	0.24	-0.05	0.99
Long-term sickness risks posed to organisation: having to pay sick pay	-0.35	0.36	-0.97	0.7
Used meetings with employees to support employees with health problems to remain in-work or support in returning to work	0.13	0.19	0.68	1.14
Measures to prevent ill-health or improve general health and wellbeing: Health and Safety training	-0.21	0.21	-0.97	0.81
Measures to prevent ill-health or improve general health and wellbeing: Health and wellbeing promotion programmes	0.35	0.2	1.78	1.42
Measures to prevent ill-health or improve general health and wellbeing: Interventions to prevent common health conditions becoming a problem	0.32	0.19	1.7	1.38
Measures to prevent ill-health or improve general health and wellbeing: Training for line managers	-0.3	0.2	-1.52	0.74
Measures to prevent ill-health or improve general health and wellbeing: An Employee Assistance Programme	0.22	0.22	0.98	1.24

Table 3.11: OSP regression – small and micro employer model, reduced model

Variable	Co- efficient	Standard Error	t-statistic	Odds ratio
(Intercept)	-1.59	0.16	-9.66	0.2
Office-based sector	0.68	0.18	3.79	1.98
Mix of office-based and manual sector	0.19	0.19	1	1.21
Percentage of the workforce aged 50+: 50% or more	0.43	0.17	2.45	1.53
Employers with at least some employees with disabilities in the workforce	0.43	0.16	2.67	1.54
Provides access to OH services	0.85	0.19	4.58	2.35
Measures to prevent ill-health or improve general health and wellbeing: Health and wellbeing promotion programmes	0.44	0.18	2.49	1.56

3.7. Cluster analysis

What is a cluster analysis?

Cluster analysis is a statistical technique used to identify independent sub-groups (or ‘clusters’) of individuals within a data set. In this instance, the technique was used to group employers according to their:

occupational health provision
sick pay provision
benefits offered to employees
support for employees with health problems
measures in place to prevent ill health
line manager’s role in managing sickness absence
use of information sources on retaining employees with health conditions

Employers were grouped together because they provided similar responses to other employers in their cluster, in response to the questions asked in the survey. They are therefore expected to have greater similarity in the levels of occupational health provision, provide similar levels of employee support, and hold similar views towards employee support within each cluster than between them.

The clusters were developed to improve our understanding of how the different aspects of employee support interact with each other and help us understand co-occurring measures (i.e. whether an employer providing X, is also more likely to provide Y). Importantly, the analysis also tells us how large each of the clusters are within the population.

The method

There are a range of techniques available to generate clusters and the choice of method depends on the data structure; a method that is optimal for grouping individuals according to their responses to a set of attitudinal scales will not be optimal for grouping individuals according to (categorical) demographic characteristics.

The information that had been collected about occupational health and employer support was held in the dataset as categorical variables, where numerical values represent specific characteristics. For the purposes of analysing the data these variables were recoded into binary responses, where an employer’s answer to each question response was coded as either present (=1) or absent (=0).

Table 3.12 contains the binary response variables that were used to generate the clusters. These codes were based on the question names as per the survey found in Annex B. To illustrate, employers were asked 2 questions:

1. H1: Does your organisation provide access to occupational health services for your employees?
2. H2: Which types of providers have you used for your occupational health services?).

This created 3 response options that were of interest to us:

1 = No OH or do not know if have OH services
2 = OH purchased privately on case by case basis, or do not know how OH is delivered
3 = OH delivered in house, through a public sector body, or privately on a long term contract

These options became 3 binary response questions, which were then entered into the cluster analysis.

Table 3.12: Binary response variables included in the cluster analysis

Variable	Definition
H1_H2_1	No (OH services provided), or do not know if they have OH services
H1_H2_2	OH purchased privately on case-by-case basis, or do not know how OH is delivered
H1_H2_3	OH delivered in-house, through a public sector body, or privately on a long-term contract
D3_1	Information sources if employer wanted to find out more information on how to retain an employee with a long-term health condition: Internet search
D3_2	Information sources on retaining employees with health conditions: Professional or personal networks
D3_3	Information sources: Legal sources
D3_4	Information sources: OH professional or provider
D3_5	Information sources: Human Resources (HR) team
D3_6	Information sources: Other options
D3_7	Information sources: Nowhere or do not know
E2_1	Additional benefits offered to employees: Flexible working regularly used by employees (for example, working from home, term-time working, compressed week, taking time off in lieu)
E2_2	Additional benefits offered to employees: Employer contribution to employee pensions (above statutory requirements of 3% contribution)
E2_3	Additional benefits offered to employees: More than 20 days paid annual leave (plus in addition to bank holidays)
E2_4	Additional benefits offered to employees: Enhanced maternity and paternity pay above statutory levels (weekly rate for first 6 weeks is 90% of the employee’s average weekly earnings, weekly rate for remaining weeks of 52 weeks is £140.98 or 90% of the employee’s average weekly earnings, whichever is lower)
E2_5	Additional benefits offered to employees: None of these, or do not know
F1_1	Nature of sick pay provision: No sick pay or do not know sick pay type
F1_2	Nature of sick pay provision: SSP only
F1_3	Nature of sick pay provision: Above SSP
G5_G6_1	Sickness absence management: Line managers take primary responsibility for managing (short and long-term) sickness absence and they receive training and/or tailored support (for example online support, care conference with HR) for handling sickness absence
G5_G6_2	Sickness absence management: LM responsible for managing sickness absence and not trained
G5_G6_3	Sickness absence management: LM does not take responsibility for managing sickness absence
G2_1	Measures used in the last 12 months to support employees with health problems to remain in-work or support in returning to work^{[footnote 22]}: Nothing
G2_2	Measures used in the last 12 months to support employees with health problems to remain in-work or support in returning to work: Maximum of 2 provisions
G2_3	Measures used in the last 12 months to support employees with health problems to remain in-work or support in returning to work: Those who offer 3 or 4 provisions
G2_4	Measures used in the last 12 months to support employees with health problems to remain in-work or support in returning to work: Those who offer 5 or more provisions
E1_1	Current provisions to prevent employee ill-health or improve the general health and wellbeing of the workforce: Health and safety training or guidance
E1_2	Current provisions to prevent employee ill-health or improve the general health and wellbeing of the workforce: Health and wellbeing promotion programmes to improve employees physical activity or lifestyle
E1_3	Current provisions to prevent employee ill-health or improve the general health and wellbeing of the workforce: Interventions to prevent common health conditions becoming a problem
E1_4	Current provisions to prevent employee ill-health or improve the general health and wellbeing of the workforce: Training for line managers on ways to improve employee health and wellbeing
E1_5	Current provisions to prevent employee ill-health or improve the general health and wellbeing of the workforce: An Employee Assistance Programme (EAP), staff welfare or counselling programme
E1_6	Current provisions to prevent employee ill-health or improve the general health and wellbeing of the workforce: We do not provide anything, or do not know
E1_7	Current provisions to prevent employee ill-health or improve the general health and wellbeing of the workforce: Other responses

Once the data had been recoded, the distribution of the variables were checked to ensure there were no variables with less than 5% response, since these would cause the clustering not to run^{[footnote 23]}. The clusters were generated using a technique called ‘hierarchical agglomerative clustering’ with the Jaccard distance measure and ‘within groups linkage^{[footnote 24]}’. These methods are recommended for binary data.

The Jaccard distance measure (also called the Jaccard similarity coefficient) compares the members of 2 different clusters and identifies what proportion of cases within the 2 clusters are similar, in that they share similar characteristics. For this analysis, this meant calculating what proportion of employees in 2 clusters gave the same responses to the questions in Table 3.12. A score was created, ranging from 0 to 100, that was calculated as the number of cluster members with the same characteristics divided by the total number of cluster members across the 2 clusters, multiplied by 100. This score reflects the proportion of cases across the 2 clusters that share similar characteristics. The higher the score, the more similar the 2 clusters are.

‘Within groups linkage’ means the distance between 2 clusters is calculated as the average distance between all pairs of cases within the cluster. This method is considered robust as it means the distance measure is less affected by outliers.

Hierarchical agglomerative clustering starts by treating each case in the sample as a single cluster. The algorithm identifies the 2 cases closest to each other (based on the Jaccard distance function) and links them together. Cases are repeatedly linked to the nearest case (or group of cases) until all cases are linked. Hierarchical clustering does not require the target number of clusters to be set in advance, as the clustering algorithm will continue to link cases until no further links are possible. The output can be used to determine what number of clusters are optimal. The results indicated that a 7-cluster result was optimal, as further splits lead to additional clusters containing very few sample members.

The 6-cluster result contained a larger, less distinct cluster (it combined the second and third cluster of the 7-cluster result) that was less useful in practical terms. The 7-cluster result split this into 2 distinct clusters that were different in terms of the occupational health services provided. Chi-square tests were used to check that there was significant level of variation in the occupational health and employer support measures used to create the clusters.

The analysis resulted in a set of 7 clusters that were distinct and contained sufficient sample numbers to profile. The distribution of the clusters is given in Table 3.13.

Table 3.13: 7-cluster solution

	Approximate number of employers in the GB population	Percentage of employers
Structured Support	136,158	12%
Intensive Support	64,258	6%
Informal Support	158,305	14%
Pragmatic Support	95,854	9%
Reactive Support	111,251	10%
Reluctant Support	207,305	19%
Minimal Support	320,222	29%

Figure 3.1: Cluster analysis variables

Measures in place to prevent ill health	Structured	Intensive	Informal	Pragmatic	Reactive	Reluctant	Minimal	Total
Health and safety training or guidance	9%	10%	27%	15%	54%	44%	39%	32%
Health and wellbeing promotion programmes	5%	2%	6%	3%	4%	7%	5%	5%
Interventions to prevent common health conditions becoming a problem	12%	18%	15%	8%	7%	10%	7%	10%
Training for line managers on ways to improve employee health and wellbeing	22%	10%	19%	39%	10%	16%	11%	17%
An EAP, staff welfare or counselling programme	44%	54%	11%	22%	6%	8%	5%	16%
We do not provide anything, or do not know	1%	0%	16%	7%	18%	13%	32%	17%

Occupational Health (OH) provision	Structured	Intensive	Informal	Pragmatic	Reactive	Reluctant	Minimal	Total
OH purchases privately on case by case basis, or do not know how OH is delivered	15%	95%	5%	3%	5%	3%	4%	11%
OH delivered in house, through a public sector body, or privately on a long term contract	62%	14%	4%	6%	7%	4%	2%	12%
No OH or do not know if have OH services	30%	1%	91%	91%	89%	93%	94%	79%
Has OH services	71%	99%	9%	9%	11%	7%	6%	21%

Perks offered to employees	Structured	Intensive	Informal	Pragmatic	Reactive	Reluctant	Minimal	Total
Regular flexible working	80%	82%	81%	76%	76%	62%	60%	70%
Pensions contributions above statutory requirements	78%	83%	64%	56%	53%	57%	45%	58%
More than 20 days paid annual leave	87%	91%	77%	73%	79%	74%	67%	75%
Enhanced maternity and paternity pay	41%	50%	33%	23%	25%	17%	18%	26%
None of these, or do not know	1%	1%	2%	3%	4%	7%	12%	6%

Sick pay provision	Structured	Intensive	Informal	Pragmatic	Reactive	Reluctant	Minimal	Total
No sick pay or do not know sick pay type	13%	4%	20%	5%	15%	7%	33%	18%
Statutory Sick Pay (SSP) only	24%	33%	0%	94%	69%	91%	57%	54%
Above SSP	63%	63%	80%	0%	17%	2%	10%	28%

Support for employees with health problems	Structured	Intensive	Informal	Pragmatic	Reactive	Reluctant	Minimal	Total
Nothing	32%	17%	42%	11%	1%	70%	81%	49%
Maximum of 2 provisions	17%	19%	18%	35%	98%	3%	6%	21%
Offered 3 or 4 provisions	27%	47%	35%	40%	0%	23%	12%	23%
Offered 5 or more provisions	24%	18%	5%	14%	1%	3%	1%	7%

Line managers’ role in managing sickness absence	Structured	Intensive	Informal	Pragmatic	Reactive	Reluctant	Minimal	Total
Line Manager (LM) responsible for managing sickness absence and receives training	44%	62%	34%	93%	1%	15%	5%	27%
LM responsible for managing sickness absence and not trained	8%	19%	61%	3%	56%	82%	5%	34%
LM does not take responsibility for managing sickness absence	48%	19%	5%	4%	42%	3%	89%	39%

Information sources on retaining employees with health conditions	Structured	Intensive	Informal	Pragmatic	Reactive	Reluctant	Minimal	Total
Internet search	34%	26%	59%	53%	50%	52%	44%	47%
Professional or personal networks	22%	30%	28%	29%	36%	23%	24%	26%
Legal sources	19%	15%	9%	13%	11%	6%	6%	10%
OH professional or provider	24%	26%	3%	5%	1%	10%	6%	9%
HR team	8%	16%	7%	6%	8%	4%	3%	6%
Nowhere or do not know	12%	7%	6%	5%	4%	13%	24%	13%

Question 5

4.  Qualitative follow-up research

Accepted Answer

This section outlines the aims and approach to the follow-up qualitative interviews.

4.1. Background and aims

The survey data was supplemented by qualitative research with employers who had consented to be recontacted following their participation in the survey. The qualitative research was designed to provide a greater understanding of employers in regard to health and wellbeing – what drives their decisions, what support they put in place for their employees, and why – as they play a key role in preventing unnecessary sickness absence, presenteeism and health-related job loss. The topic guide was developed with EHIE and is included in Annex C.

4.2. Recruitment and fieldwork

Ipsos MORI conducted follow-up depth interviews with 30 employers over the telephone, between 22 July and 9 August 2019. Employers were sampled from those who had consented to be recontacted following their participation in the survey.

A range of quotas were set to ensure the employers broadly reflected the employer population and had characteristics of interest to explore in more depth in the qualitative interviews (Table 4.1).

Table 4.1: Qualitative sampling matrix

Description	Characteristics	Quota	Total
Size	Micro (2 to 9) Small (10 to 49) Medium (50 to 249) Large (250+)	Min. 12 Min. 12	3 13 8 6
Organisation type	Private sector Charity or voluntary sector Government financed body	Mix and monitor	26 2 2
HR	Internal HR External HR (consultancy) None	Mix and monitor	24 4 2
Sickness absence	Long-term sickness absence in the last year Instances of recurring sickness absences	Min. 10 Min. 6	23 19
Employment type	Mainly 0 hours contracts or casual employment Mix of casual/permanent contracts No 0 hours contracts or casual employment	Min. 10 Min. 10 Min. 10	7 11 12
Use of occupational health services	Yes No	Min. 8 Min. 8	20 10
Sick Pay	SSP only SSP and OSP Neither	Min. 8 Min 8 -	11 16 3
Employees with a disability or long-term health problem	Yes	Min.15	17

4.3. Analysis

Our approach was divided into 2 key phases: data management and data analysis. At the data management stage, the research team explored and learnt about the data through regular data familiarisation sessions, held throughout the fieldwork period. The questions and hypotheses arising from these discussions were followed up by further investigation of the existing data, or explored further in the remaining interviews.

Following the data familiarisation sessions, the research team constructed an initial thematic code frame in Excel. This code frame covered the key research questions, driven by the survey data, the key areas in the discussion guide, as well as other important themes based on the research discussions.

Following data management, the team analysed the data using the thematic code frame. The code frame makes it easier to look across the data as a whole for links between cases, and to look for relevant factors that seem to explain different experiences.

Question 6

Appendix A: Scoping Review

Accepted Answer

Retention of employees with health conditions and returns to work after periods of long-term sickness absence: a scoping review

Louise Thomson, University of Nottingham

Overview

This report describes the results of a scoping review of the literature on workplace-based or employer-led practices to support the retention of employees with health conditions and returns to work after periods of long-term sickness absence, as well as barriers and enablers to action.

Research Context

The cross-cutting questions for this review are to identify: what works in supporting the retention of employees with health conditions and returns to work; what the current practice is and why; and what the supporting policy levers are. Within these overarching questions we also sought to identify literature which addressed the following 6 themes:

employer attitudes and perceptions towards sickness absence
sickness absence management practice and facilitation of return to work
employer-paid sick pay
Occupational Health (OH) and employers’ work ability assessments
workplace adjustments to aid returns to work
employer investments in expert-led advice for example, OH services

Methodology

A search strategy was developed to examine numerous databases and websites across both the academic and grey literatures. Relevant studies and sources of information were identified and the evidence mapped across the 6 themes and the 3 cross-cutting questions. All types of empirical study designs were included and the search applied to any employment setting. International studies were included if they are published in the English language and the search included studies from 2000 to present to ensure a focus on current practices. Long-term sickness absence was defined as over 4 weeks.

A narrative synthesis was conducted to summarise the volume of the current available evidence under the themes, the types and quality of evidence available, and identify the gaps in research.

Results

A total of 218 records were included in the final mapping; 119 peer-reviewed articles and 99 web-based reports. Journal articles came from studies based in a wide number of countries, including 21 in the UK. Table A.1 shows the number of peer-reviewed articles and web-based results identified through the search for each of the 6 themes and 3 cross-cutting review questions.

Table A.1: Number of articles by themes and review questions

1. Cross-cutting review question: What works

	Number of Journal Articles	Number of Web Results
1. Employer attitudes and perceptions	13	42
2. Sickness absence management practice	45	63
3. Employer-paid sick pay	0	3
4. Occupational Health and work ability assessments	4	0
5. Workplace adjustments	15	26
6. Employer investments in expert-led advice	10	7

2. Cross-cutting review question: current practice

	Number of Journal Articles	Number of Web Results
1. Employer attitudes and perceptions	42	47
2. Sickness absence management practice	43	53
3. Employer-paid sick pay	2	3
4. Occupational Health and work ability assessments	6	2
5. Workplace adjustments	17	25
6. Employer investments in expert-led advice	3	3

3. Cross-cutting review question: policy levers

	Number of Journal Articles	Number of Web Results
1. Employer attitudes and perceptions	7	22
2. Sickness absence management practice	9	30
3. Employer-paid sick pay	2	1
4. Occupational Health and work ability assessments	0	0
5. Workplace adjustments	2	16
6. Employer investments in expert-led advice	1	1

Conclusions

The review identified a greater volume of evidence relating to the ‘what works’ and ‘current practice and why’ review questions as compared to the ‘policy levers’ question. Across the 3 key questions, more records were found relating to ‘sickness absence management practice’ and ‘employer attitudes and perceptions’ as compared to the other 4 themes.

Although the review returned a considerable amount of evidence on the effectiveness of employer-led return to work interventions to reduce sickness absence, there was very little evidence on which components of the return to work interventions are the most important in achieving outcomes. Indeed, there was considerable variation in the outcome measures used across studies.

Evidence gaps also remain across the 3 key questions specifically in relation to the UK context; the vast majority of studies were from other European countries with varying health, welfare and insurance systems. As well as this, the review highlighted the small amount of Randomised Control Trial evidence in this area.

Overall, the scoping review has identified a need to undertake wider research on a number of key areas, including available policy levers and particularly within the UK experimental context.

Question 7

Appendix B: Questionnaire

Accepted Answer

Below is a copy of the questionnaire used for the mainstage survey.

Introduction

Read out to all.

Good morning/Good afternoon, my name is … and I’m calling from Ipsos MORI, the independent research organisation. Is this [Insert organisation name from sample]?

We are conducting a major research study with employers of all sizes across Great Britain for the government, to find out how employers are currently managing employees’ health and sickness absence, and what information and support they need to better address these issues in the future.

I would like to speak with [Insert name from sample] please?

Interviewer instruction: If sample name blank: I would like to speak with the most senior person here who has responsibility for personnel issues for [Name of organisation] in Great Britain? If you have more than one site, that person should be responsible for personnel issues across all sites where your HR policy applies.

Interviewer instruction: If no such role exists within the organisation, then please ask for the owner, managing director, or general manager.

You should have received a letter in the post about this research last week.

If necessary: Would you like me to email you a copy of this letter?

The survey takes around 20 minutes depending on your organisation’s circumstance. Is now a good time to find out your views?

Reassurances if necessary:

the survey is being conducted by the Work and Health Unit, a joint unit of the Department for Work and Pensions and Department of Health and Social Care
taking part is totally confidential and anonymous for all individuals and organisations. It will not be possible to identify you, your organisation or your employees from the published findings
findings will be published on the GOV.UK website in early 2019 to help employers like you, as well as the Government, to understand how best to support employee health and wellbeing
your organisation has been selected at random from records from a national database

Ask all.

Before we start, I just want to clarify that participation in the survey is voluntary and you can change your mind at any time. Are you happy to proceed with the interview?

If necessary: If you would like to read it beforehand, you can access the Privacy Notice.

Yes – continue with interview.
No – thank and close.

S1. Sndem

Interviewer: Does an email need to be sent?

Yes.
No.

Ask all.

A1a. Datasheet

I can send you a summary of key questions you may find helpful to know in advance, to speed up the interview. Would you like us to email you this and arrange a convenient time to call you back?

Yes – confirm size and send reassurance email/datasheet.
No – arrange call back or proceed to interview.
All other outcome codes:

a. 171 refused – confidentiality concerns
b. 172 refused – think survey is not genuine
c. 173 refused – company no name policy
d. 174 refused – soft refusal
e. 175 refused – hard refusal
f. 176 refused – survey length too long

Ask if A1a=1

A1b. Dtasht_size

I’d like to check the number of employees you have to send you the correct questions. Including yourself, how many employees do you currently have on the payroll in Great Britain]? Prompt to code

One employee – close.
2 to 49.
50 to 249.
250 or more.
Not sure.

Send short datasheet if Code 2, everyone else gets long version.

Ask if datasheet was sent.

I4: Datasheet_use

When we first contacted you, we emailed you some questions that would be asked in the survey. Do you have this to hand?

Reassure: It does not matter if you don’t have it to hand.

Yes.
No.

Section A: Employer characteristics

Read out to all.

First, I’d like to confirm some details about you and your organisation.

Ask all.

A1. Role

How would you describe your role?

Do not read out. Prompt to code.

Owner/Director
Senior management
HR/Office Manager
Occupational Therapist/Wellbeing professional
Company Secretary
Other (please specify)

Ask all.

A2. Org_type

Are you…?

Read out

Single code only.

Mainly seeking to make a profit (i.e. private sector).
A charity or voluntary sector organisation or a social enterprise
A local-government financed body
A central government financed body
Other please specify
Do not read out: Private-Public Partnership (PPP)
Do not read out: Don’t know

If code 1, text subsition “business” throughout. Codes 2 to 7: “organisation”

Ask all

A3. Site

Is the site at which you are based…?

Read out

Single code only

The only site; or
One of a number of sites within a larger organisation in Great Britain
Do not read out: Don’t know

Read out if >1 Site (Code 2): Throughout the survey I would like you to think about permanent employees in the whole of your organisation in Great Britain. This means thinking about all sites where your Human Resources policy applies, rather than thinking just about the site where you work.

Ask all

A4. N_employees

Including yourself, how many employees do you currently have on the payroll [IF MULTI-SITE: in Great Britain]? Please include both full-time and part-time permanent employees. Please do not include contractors or agency staff or other temporary staff.

Interview note: Permanent employees have an indefinite contract whereby they are employed by the company until such time as the employer or the employee no longer wish to work there.

Allow 2 to 400,000 Soft check 300,000

Type in number

(If 1 – then end interview and thank and close – Screen out)

Do not read out: Ref (Screen out)
Do not read out: do not know

Ask if do not know at A4

A5. B_employees

Prompt to range: Is it…

Read out. If do not know, prompt using sample information: [sameemployees]

Single code only

2 to 4 employees
5 to 9
10 to 49
50 to 99
100 to 249
250 to 499
500 to 999
1,000 to 4,999
5,000 to 9,999
10,000 or higher
Do not know – screen out

Dummy size band: combine band and numeric

Ask if sampsector not blank

A6. Sector_1

We have [Sampsector] as your main industry sector. Does this sound about right?

Single code only

Yes – correct.
No – not correct.
Do not read out: Do not know.

Ask if sampsector blank or sector_1=2 or 3

A7. Sector_2

What is the main activity of this business?

Prompt to code

Single code only

(A) Agriculture, Forestry and Fishing
(B), (D), (E) Mining and Quarrying; Utilities, Waste Management and Remediation Activities
(C) Manufacturing
(F) Construction
(G) Wholesale and Retail Trade; Repair of Motor Vehicles and Motorcycles
(H) Transportation and Storage
(I) Accommodation and Food Service Activities
(J) Information and Communications
(K) Financial and Insurance Activities
(L) Real Estate Activities
(M) Professional, Scientific and Technical Activities
(N) Administrative and Support Service Activities
(O) Public administration, defence and compulsory social security
(P) Education
(Q) Human Health and Social Work Activities
(R) Arts, Entertainment and Recreation
(S) Other Service Activities
Other please specify

Dummy sector: combine sample and response

Ask if private sector or voluntary sector

A8. N_Turnover

What was your approximate [If private sector: turnover] [If voluntary sector: income] for the last financial year?

Please refer to Q1 on datasheet

Allow 1 to 200,000,000. Soft check 90,000,000

Enter number in hundreds, thousands or millions.

Check: To confirm your [If private sector: turnover] [If voluntary sector: income] was [Insert amount in word]

Do not read out: Do not know
Do not read out: Refused

Range checks
Hundreds: 1 to 999
Thousands: 1 to 999
Millions: 1 to 200 Ask if do not know at N_Turnover

A9. B_Turnover

Would you say the [If private sector: turnover] [If voluntary sector: income] of your [business] [organisation] [If Code 2 at “Site”: in Great Britain] in the last financial year was…

Read out. If do not know, prompt using sample information: [Samturnrover]. Probe fully
Probe for best estimate before coding DK
Singlecode

Less than £1,000
£1,000 to less than £10,000
£10,000 to less than £50,000
£50,000 to less than £100,000
£100,000 to less than £500,000
£500,000 to less than £5 million
£5 million to less than £10 million
£10 million to less than £50 million
£50 million or more
Do not read out: Do not know

Dummy turnover band: Combine numeric and bands and sample

Ask all

A10. Outlook – question deleted

Ask all

A11. Spending

I am going to read out 6 possible spending or investment priorities. Please tell me how important they are to your [business] [organisation] in the next 12 months.

Please rate each using a scale of zero to 10, where zero means no importance and 10 is of extreme importance.

[If multi-site: Please think about spending or investment priorities across the [business] [organisation] as a whole in Great Britain].

Read out

Single code each 0 to 10. Allow do not know

Rotate list

a) Training and skills development of employees
b) New business, service or product development
c) [If org_type=1]: Focusing on existing core activities and brand strength
d) Recruitment of new employees
e) Improving employee health and well-being (for example, via pay, benefits and flexible working options). Add if necessary: Health and wellbeing refers to a positive state of mind and body, feeling safe and able to cope
f) Investment in infrastructure (for example, machinery, property, equipment)

Ask all

A12. Hlt_Concerns

Which of the following health concerns affect the most number of staff in your [business] [organisation]?

Read out. Rotate except other which should be last

Stress
Anxiety, depression or other common mental ill-health conditions
Musculoskeletal conditions, repetitive strains or injuries
Cardiovascular or heart problems
Respiratory problems
Visual problems
Physical injuries caused by workplace accidents
Other (please specify)
None of these

Ask if >1 selected at A12

A13. Hlt_Concerns_1

And which of these affects your staff the most?

CATI display answers chosen at A12. Allow one response only.

Section B: Workforce characteristics

DP note: If employees is less than 10, ask questions in numeric only. Do not allow percentages.

Ask all

B1. Full-time

You mentioned earlier that there are [Insert numeric] [If band: between insert band] permanent employees currently on your [business’] [organisation’s] payroll [If multisite: in Great Britain].

If number of employees is numeric: How many of these employees work full-time, that is 30 hours or more per week? If 10+ employees: You can answer in number or percentage.

If number of employees is banded: What percentage of these employees work full-time, that is 30 hours or more per week?

Please refer to Q2 on datasheet

CODE:
Number
Percentage

Dummy N_Full-time
Dummy N_Part-time
Dummy %Full-time
Dummy %Part-Time
Take mid-point for bands (for top band, compare with sample information to cap)

Add check: Readout: This means that [Dummy N_Part-time] [Dummy %_Part-time] of your permanent employees work part-time?

Correct
Not correct [Skip back]

Ask all

B2. Tempstaff

Does your [business] [organisation] employ staff on temporary contracts – either directly or via an agency?

Interviewer instruction: We are referring to the hiring of staff for a limited period of time to meet the needs of your business such as seasonal workers, self-employed contractors or temps.

Single code only

Yes
No
Do not know

Ask if N_Employees <10 or B_Employees=1 or 2

B3. Sml_Aged_50+

How many of your [Insert employees] are aged 50 years or older?

Enter number
Do not know

Ask if N_Employees>10 or B_Employees= 3 to 10

B4. Lg_Aged_50+

Approximately what proportion of your [Insert employees] employees are aged 50 years or older? Is it…?

Read out

Single code only

None
Less than a quarter
A quarter to half, or
More than half of your workforce
Do not read out: Do not know

Dummy aged_50+: calculcate proportion at small_aged_50+ to match codes at aged_50_, and combine into one variable (aged_50+)

Ask if N_Employees > 10 or B_Employees=3 to 10

B5. Disability_LG

Approximately what proportion of your employees have a disability or a long-term health condition?

Explain if required: This includes any physical or mental impairment that has a ‘substantial’ and ‘long-term’ negative effect on the employee’s ability to do normal daily activities.

None
Less than 10%
11% to 50%
51% or higher
Do not read out: Some do, but don’t know the proportion
Do not read out: Not to my knowledge

Ask if n_employees <10 or b_employees=1 or 2

B6: Disability_sml

How many of your employees have a disability or a long-term health condition?

Explain if required: this includes any physical or mental impairment that has a ‘substantial’ and ‘long-term’ negative effect on the employee’s ability to do normal daily activities.

Enter number
Allow dk

Dummy disability: calculate proportion at disability_sml to match codes at disability_lg, and combine into one variable (disability)

Ask all

B7: Occupation

I am going to read out 3 types of occupation:

1) managers and professionals
2) technicians or skilled trades
3) semi or unskilled occupations

Please tell me approximately what percentage of your [insert employees] are in each.

Please refer to q3 on datasheet

Read out. Prompt with examples on showcard a if required.
Must sum to total 100%. Allow dk/ref.

	Interviewer prompts [have this column on a showcard and not in the script]
(1) managers or professionals. Managers include directors and managers of internal departments/sections. Professional occupations usually require a degree or equivalent formal qualification:	Managers include directors and managers of internal departments/sections. Professional occupations usually require a degree or equivalent formal qualification: - programmers, software development and it professionals - accountants, solicitors, lawyers - economists, financial project managers - journalists, newspaper editors - engineering professionals - doctors, nurses, midwives, social workers
(2) technicians or skilled trades. These occupations require a substantial period of full-time training or further study.	Technicians include: - graphic designers, photographers, writers - estate agents, auctioneers - IT / science / engineering / lab technicians Skilled trades include: - electricians, plumbers - butchers, bakers, fishmongers - vehicle technicians, mechanics and electricians - IT engineers - florists - chefs
(3) semi-and unskilled occupations. These include occupations that involve mostly routine tasks. Most do not require formal educational qualifications.	- general admin/secretarial work - sales and customer service - caring work (nursery, ambulance staff, healthcare assistants) - drivers/machine operators - cleaners, packers, farm workers, postal workers - hairdressers

Ask if n_employees >10 or b_employees=3 to 10

B8: Lg_empturnover

In any typical year, excluding years with exceptionally good or bad results – what percentage of permanent employees would leave the [business] [organisation]?

Add if necessary: please include permanent staff only. Include staff who leave for whatever reason: termination, retirement, death, interagency transfers and resignation.

Interviewer note: if rate varies year-on-year by more than the amount in the codes below, ask for the most recent year.

Please refer to q3a on datasheet

Prompt to code
Single code only

Less than 5%
5% to 9%
10 to 14%
15% to 19%
20% or more
Do not know

Ask if n_employees <10 or b_employees=1 or 2

B9: Sml_empturnover

In any typical year, excluding years with exceptionally good or bad results – how many permanent employees would leave the [business] [organisation]?

Add if necessary: please include permanent staff only. Include staff who leave for whatever reason: termination, retirement, death, interagency transfers and resignation.

Interviewer note: if rate varies year-on-year by more than the amount in the codes below, ask for the most recent year.

Enter response
Do not know

Dummy empturnover: calculate proportion at sml_empturnover to match codes at lg_empturnover, and combine into one variable (empturnover)

Ask all

B10: Staff

Generally, how easy or difficult is it for your [business] [organisation]…

If multi-site: generally, across your [business] [organisation] as a whole, how difficult is it…

a) To find employees with the required skills to carry out your main function [i.e. Non-support/admin staff]?
b) To retain these employees?

Add if necessary: please think about the largest group of non-support staff that you employ

Read out. Reverse scale
Single code only

Very easy
Fairly easy
Neither easy or difficult
Fairly difficult
Very difficult
Do not read out: don’t know

Ask all

B11: Represent

Is there an employee representative group or trade union within your [business] [organisation]?

Do not read out
Single code only

Yes
No
Do not know

Ask all

B12: Pln_exc

Who normally decides on how employees prioritise their daily work tasks?

Add if necessary: please think about the extent to which staff in the largest occupational group have discretion over how they work.

Read out

Employees undertaking the task
Managers or supervisors
Both employees and managers or supervisors
Do not read out: don’t know

Ask all

B13: Pln_exc_2 – question deleted

Section C: cross-cutting attitude statements

Read out to all

I am now going to read out some statements to do with employee health and wellbeing. As far as possible, please give the views held by your organisation rather than your personal opinion. Throughout this survey, health and wellbeing refers to a positive state of mind and body, feeling safe and able to cope.

Ask all

C1: Attitudes

How much do you agree or disagree with the following statements?

Rotate statements

a) Employers have a responsibility to encourage employees to be physically and mentally healthy
b) Currently, sickness absence is a barrier to productivity in this organisation
c) The financial benefits of spending money on employee health and wellbeing outweigh the costs
d) There is a link between work and employees’ health and well being
e) The things that affect employees’ health and well-being are out of our control
f) We know what to do to improve our employees’ health and well-being at work
g) It is difficult for us to find time to do things to improve the health and well-being of our employees

Strongly agree
Tend to agree
Neither agree nor disagree
Tend to disagree
Strongly disagree
Do not read out: do not know

Ask all

C2: Invest

How important are the following factors in your [business’] [organisation’s] decision on whether to invest in employee health and wellbeing initiatives?

Please answer using a scale of zero to 10, where zero means no importance and 10 is of extreme importance. Add if necessary: health and wellbeing refers to a positive state of mind and body, feeling safe and able to cope. Read out. Single code each 0 to 10. Allow dk.

Rotate list.

a) Helping to satisfying legal obligations to do with health and wellbeing at work
b) Meeting expectations from employees or their representatives
c) Maintaining or increasing productivity through improved health or wellbeing
d) Maintaining the organisation’s reputation
e) Helping recruitment or retention of talent
f) Helping to minimise cost resulting from sickness absence

Section D: compliance, information, advice

Ask all

D1: Lgl_understand

Thinking about health and safety, disability, and sick leave, how well does your business understand the legal responsibilities around these issues?

Read out. Single code only.

Very well
Fairly well
Not very well
Not well at all
Do not read out: don’t know

Ask all

D2: Easy_lgl

And how easy or difficult is it for your [business] [organisation] to meet these legal responsibilities?

Read out. Single code only

Very easy
Fairly easy
Neither easy nor difficult
Fairly difficult
Very difficult
Do not read out: don’t know Ask all

D3: Info_oh

If your [business] [organisation] wanted to find out more information on how to retain an employee with a long-term health condition, where would you look for advice?

Do not prompt

Multicode ok

Occupational health/professional health provider
Internet search (for example, Government or acas website, occupational health specialist sites etc)
Professional/personal networks or contacts (for example, Trade body)
Legal sources
Other – specify
No-one/no-where
Do not know
HR team

Section E: health and wellbeing initiatives

Ask all

E1: Provision

Which, if any, of the following do you currently provide to prevent employee ill-health or improve the general health and wellbeing of your workforce?

Explain if required: please include any services offered, even if you only offer them to some employees.

Add if necessary: health and wellbeing refers to a positive state of mind and body, feeling safe and able to cope.

Code all that apply. Only read out examples if required.

Health and safety training or guidance (such as injury prevention and stress management).
Health and wellbeing promotion programmes to improve employees’ physical activity or lifestyle (for example, Healthy food choices, health advice or events, dedicated health and wellbeing section on the intranet, loans or discounts on bicycles, free or subsidised gym member ship).
Interventions to prevent common health conditions becoming a problem (for example, Free health checks, free vaccinations, smoking or weight loss support).
Training for line managers on ways to improve employee health and well-being.
An employee assistance programme (EAP), or staff welfare/counselling programme provided by an external organisation.
We provide something else (write in).
We do not currently provide anything.

Ask all

E2: Benefit

Which, if any, of the following does your [business] [organisation] offer your employees?

Read out. Code all that apply. Rotate list. Codes 5 and 6 single code only

Flexible working regularly used by employees (for example, Working from home, term time working, compressed week, taking time off in lieu).
Employer contribution to employee pensions (above statutory requirements of 3% contribution)
More than 20 days paid annual leave (plus in addition to bank holidays).
Enhanced maternity and paternity pay above statutory levels (weekly rate for first six weeks is 90% of the employee’s average weekly earnings, weekly rate for remaining weeks of 52 weeks is £140.98 or 90% of the employee’s average weekly earnings, whichever is lower).
Don’t know.
None of these.

Dummy one benefit = only one benefit selected

Dummy two_three benefits = 2 to 3 benefits selected

Dummy all benefits = all benefits selected

Section F: sick pay

Read out to all

I would now like to ask you about the pay that permanent employees receive when they are off work sick.

Ask all

F1: Sickpay

Which of these forms of sick pay do you provide to employees when they are off sick?

Add if necessary: please include even if you provide to some employees only.

Interviewer note: occupational sick pay is where an organisation chooses to provide a contractual sick pay that is more generous than the statutory minimum (ie. Statutory sick pay).

Read out

Single code only

Statutory sick pay
“occupational sick pay” scheme
Both
Neither
Do not know

Ask if offer OSP

F2: OSP

Do you offer occupational sick pay to all or some of your employees?

Single code only

All
Some
Do not know
Refused

Ask if some employees receive OSP

F3: OSP_criteria

What determines who receives occupational sick pay?

Read out

Multicode ok

Seniority of employee
Length of service
Based on the certain occupations or skills the [business][organisation] needs
The type of contract employees are on (for example, Permanent or temporary)
Something else - please specify
Do not know

Ask if pay OSP

F4: OSP_length

How long would you pay occupational sick pay to eligible employees in any one period of absence? Would you prefer to give your answer in terms of working days or working weeks?

If necessary: by occupational sick pay, i mean payment over and above statutory sick pay

Working days
Working weeks
Do not read out: don’t know how long payment is for [go to f7]
Do not read out: no fixed policy [go to f7]
Do not read out: refused [go to g1]

Ask if working days

F5: OSP_days

How many working days in total would you pay occupational sick pay for any one period of absence? If it varies for different employees, please tell me the number of days that applies to the majority of your workforce.

Please refer to q4 on datasheet

Minimum allowed: 1 day. Maximum allowed: 365 days. Soft check if <10 days or >130 days.

Enter number of working days
Indefinitely
Do not read out: refused

Ask if working weeks

F6: OSP_weeks

How many working weeks in total would you pay occupational sick pay for any one period of absence, where one working week is five working days? If this varies for different employees, please tell me the number of weeks that applies to the majority of your workforce.

Please refer to q4 on datasheet

Minimum allowed: 1 week. Maximum allowed: 52 weeks. Soft check if >26 weeks

Enter number of working weeks (cati convert to days – 1 week=5 working days)
Indefinitely
Do not read out: refused

Ask if pay OSP

F7: OSP_rate

Over the course of one period of sickness absence, does the rate at which you pay occupational sick pay to eligible employees reduce over time?

Please refer to q5 on datasheet

Yes
No
Do not read out: do not know
Do not read out: no fixed policy
Do not read out: refused

Ask if reduces overtime

F8: Avg_OSP_rate

What is the average rate of OSP as a percentage of usual wage? If payment varies for different employees, please tell me the percentage that applies to the majority of your workforce.

If necessary: please give me your best estimate

Minimum allowed: 1% maximum allowed: 100%

Enter %
Do not read out: pay a fixed sum rather than % of salary
Do not read out: do not know
Do not read out: refused

Section G: retention and reintegration

Ask all

G1: Support

In the last 12 months, have any of your employees requested support to accommodate their health condition or disability?

Do not read out. Single code only

Yes
No
Don’t know

Ask all

G2: Wellbeing

In the last 12 months, have you used any of the following to support employees with health problems to remain in-work or support in returning to work?

Read out. Only read out examples in brackets if needed

Multicode

Rotate list

Meetings with employees.
Phased returns to work from sickness absence (returning to full duties and hours at work gradually, over a defined time period).
Amending employee workload or job role (for example, Reduced hours/days, extra breaks, or different duties).
Workplace adjustments (such as different chairs or desks, building modifications, or other specialised equipment).
A job coach or personal assistant (for example, A sign-language interpreter for meetings).
Additional external support or advice (for example, Clinical support such as psychological therapy or physiotherapy, or another expert or specialist).
Do not read out: none of these

Ask all

G3: Proactive

Which of the following best describes your organisation?

Read out. Single code only

a. We take steps to identify and address employee health and well-being issues at the earliest possible opportunity.
b. We take action as and when employee health and well-being becomes a problem

Ask all

G4: Monitor

Do you collect sickness absence data?
Do not read out. Single code only

Yes
No

Ask all

G5: Lm_resp

Do line managers take primary responsibility for managing (short and long-term) sickness absence?

Do not read out. Single code only

Yes
No

Ask if line managers take responsibility

G6: Lm_train

Do line managers receive training and/or tailored support (for example online support, care conference with hr) for handling sickness absence?

Do not read out. Single code only

Yes
No

Ask all

G7: Flex

Does your [business][organisation] adapt policies for managing sickness depending on the employee? Do not read out. Single code only

Yes
No
Do not read out: in some circumstances/it depends
Do not read out: do not know

Ask all

G8: Challenge

Which, if any, of these barriers does your [business] [organisation] face in supporting employees on long-term sickness absences return to work?

Read out. Prompt to code. Multicode ok except code 9 which is single code only.

A lack of expertise or specialist support.
Lack of time or staff resource.
Lack of flexibility in the way work is organised.
Employee engagement in the process.
A lack of capital to invest in support.
A lack of support from senior leaders.
The benefits of investing in retaining an employee do not warrant the investment.
Other.
We do not face any barriers.

Ask if more than one challenge mentioned

G9: Challenge_1

And which of these is the greatest barrier?

Cati display options chosen at challenge

Cati combine single code responses at challenge with response at challenge_1

Ask all

G10: Return_when

Which of the following best describes your [business’] [organisation’s] approach: “an employee should return to work as soon as they can do…”

Read out 1 to 3.

…some of their work
…most of their work
…all of their work
Do not know

Ask all

G11: LTSA_4wk

In the last 12 months, have any of your employees had instances of long term sickness absence of 4 or more weeks?

Do not read out. Single code only

Yes
No

Ask if had long term sickness absence

G12: Return_4wk

Has your [business] [organisation] used any of the following to manage these employees’ returns to work after long term sickness absence?

Read out

Multi code ok

Rotate list

Regular meetings.
Develop return to work plans.
external, specialist support to manage the employee’s return.
Opportunities for employees to return to work in a flexible manner (for example, Phased returns, or reduced workload)
Independent assessment of employees work capacity (including oh assessment).
Something else please specify.
None of these.

Ask all who have reported long term sickness absences

G13: LTSA_risk

What are the main business risks or costs associated with long term sickness absence?

Multicode. Prompt to codes

Having to pay sick pay.
Additional cost/time arranging temporary cover/recruiting and training new staff.
Covering work within the organisation (additional pressure, readjusting work processes).
Keeping job open.
Uncertainty of return to work and planning around it (including reintegrating employees back into the business, time involved).
Low morale among rest of staff.
Legal risk resulting from employees who do not feel they have had appropriate support.
Reputational risk resulting from employees who do not feel they have had appropriate support
Impact on productivity or quality of work.
Missing client deadlines/dissatisfied clients.
Something else - specify.
Do not know.

G14: LTSA_main

Ask all who gave >1 answer at LTSA_risk

New question

Which of these is the greatest risk or cost to your organisation?

Section H: occupational health

Ask all

H1: Oh_provision

Does your [business] [organisation] provide access to occupational health services for your employees? Add if necessary: by occupational health services, i mean advisory and support services which help to maintain and promote employee health and wellbeing. Oh services support organisations to achieve these goals by providing direct support and advice to employees and managers, as well as support at the organisational level for example, to improve work environments and cultures

Do not read out. Single code only

Yes
No
Do not know

Ask if use occupational health services, else go to barriers

H2: Provider

Which types of providers have you used for your occupational health services?

Read out

Multi code ok

In-house providers, for example OH specialist on site
Private contractor – purchase access on case-by-case basis
Private contractor – long-term contract
Public sector bodies (for example, NHS health at work service)
Do not read out: don’t know

Ask if use occupational health services

H3: Reason

What are your reasons for providing occupational health services for staff?

Do not read out

Prompt to code

Fulfilling a legal obligation.
Meeting expectations from employees or their representatives.
Maintaining or increasing productivity.
Maintaining the organisation’s reputation.
Helping recruitment or retention.
Helping to minimise sickness absence/improve employee health and wellbeing.
Helping to minimise cost resulting from sickness absence.
Other please specify.
Do not know.

Ask if more than one enabler chosen

H4: Mainreason

And which is the main reason?

Cati combine single code responses at ex: reason with response at ex: mainreason

Ask if do not provide occupational health services

H5: Barriers

What has prevented your [business] [organisation] from providing occupational health services for staff?

Do not read out

Prompt to code

Cost/too expensive.
Unable to get funding for it.
Too complicated/too much administration involved.
Too few cases to justify the expense.
No employee demand/employees not disclosing they are in need of OH.
Doesn’t help solve the issues that the organisation faces.
Lack of knowledge – what services to buy, who to buy services from/who are good suppliers.
Lack of time to investigate.
Negative experience of oh services in the past.
Lack of awareness or support amongst senior management/no ‘champion’ to take forward.
General make-up of the workforce doesn’t make it worthwhile, for example Mainly part-time or temporary staff, high levels of staff turnover.
Not a priority for this organisation.
Other please specify.
Do not know.

Ask if more than one barrier chosen

H6: Mainbarrier

And which is the main reason?

Cati combine single code responses at barriers with response at mainbarrier Ask if use occupational health services

H7: OH_use

Earlier you mentioned that your [business][organisation] has occupational health services. How is your [business][organisation] making use of these services?

Read out. Multi code ok. Prompt to code.

To manage an employee’s sickness absence.
To prevent and remove health risks arising in the workplace.
To ensure your [business][organisation] meets its statutory responsibilities.
To help maintain a healthy workforce.
To provide screening and health surveillance services.
To give independent and professional diagnosis, prognosis and advice about staff unable to work due to health problems.
Other (specify).
Do not know.

Ask all

H8: Policy

Which of the following policies, if any, do you use mostly to manage an employees’ sickness absence from work? Read out. Multicode ok

Capability policy
Sickness absence management policy
Disciplinary policy
Wellbeing at work policy
Other (please specify)
We do not have a specific policy

Section I: closing questions

Ask all

I1: Recontact

Thank you very much for taking part in this survey. The joint department of health and social care and department for work and pensions’ work and health unit is planning to carry out further research on this topic in the next six months.

Would you be willing to be contacted by ipsos mori about participating in this research? This would be a more in-depth discussion with you on similar topics as we have discussed today. You do not have to commit to anything now, just indicate a willingness to be contacted again.

Single code only

Yes confirm name, email and telephone number
No

Ask all

I2: – question deleted

Ask all

I3: Feedback – question deleted

Thank and close.

Question 8

Appendix C: Topic guide

Accepted Answer

Below is a copy of the guide used in the qualitative interviews with employers.

1. Introduction (2 to 3 mins)

Orientates interviewee, prepares them to take part in the interview.

Outlines the ‘rules’ of the interview (including those we are required to tell them about under Market Research Society (MRS) and General Data Protection Regulation (GDPR) guidelines).

Thank participant for taking part. Introduce self, and explain nature of interview: informal conversation; gather all opinions; all opinions valid. Interviews should take around 45 minutes.

Introduce research and topic – the Work and Health Unit (a joint unit between the Department for Work and Pensions and Department of Health and Social Care) has commissioned Ipsos MORI to conduct research with employers to understand their reasons and practices regarding a range of health and wellbeing measures for their employees. They have been asked to take part in this interview because they took part in a survey on a similar topic over the summer.

Role of Ipsos MORI – Independent research organisation (i.e. independent of government), we adhere to the MRS Code of Conduct.

Confidentiality – reassure all responses anonymous and that identifiable information about them will not be passed on to anyone, including back to WHU or any other government department.

Consent – check that they are happy to take part in the interview and understand their participation is voluntary (they can withdraw at any time).

Ask for permission to digitally record – transcribe for quotes, not detailed attribution. Only non-identifiable information will be passed back to WHU.

Any questions before we begin?

2. Context (10 mins)

Provides contextual background information about the participant and their organisation.

To start off with, I’m going to ask a few broad questions about your role and the nature of your organisation.

Can you tell me a bit about what the organisation does?

sector (confirm)
organisational type (confirm)
size (confirm)
financial turnover (confirm)
length of time in operation
single site v. multiple site
briefly explore the nature of the work and range within the business (for example, manual v. office based) and the workforce that you employ (for example, skills level, contract, staff turnover)

And can you talk me through what your role involves?

specific HR role / senior management / owner or director / Occupational Therapist or Wellbeing professional
explore their responsibilities/experience/qualifications in relation to employee health and well-being
length of time at the organisation

Can you talk me through how the organisation makes decisions about staff health support, programmes, schemes or policies?

Moderator: By ‘health support, programmes, schemes and policies’ we mean all things to do with employee health (both physical and mental health), including the management of ill-health (including long-term and re-occurring health conditions), LTSA, retention of employees and disability.

who is involved in making these decisions? (for example, director, senior management, owner). Probe on specific health based role roles: HR, external professionals. If no internal HR resource: How do you work with your external HR consultants to make these decisions?

Can you tell me what support, programmes, schemes or policies you have in place, if any, to support the health of your employees? And when they were put in place / reviewed?

Moderator, probe for:

sickness Absence (including short and long term)
sick pay
Return to Work (including a phased return to work)
reasonable adjustments or additional needs
Occupational Health
are these standalone policies or integrated into other policies?
are there other processes that go above and beyond this?

[Moderator: Take note of these as they will be brought up later]

How important would you say supporting employee health is to your organisation?

why is it important?
how important is investing in employee health and wellbeing compared to other business priorities? Why?
how important is it to prevent people falling out of work (due to ill health or disability?

What measures does your organisation put in place to support employees’ health and wellbeing?

probe on: health and safety guidance and/or training, health and wellbeing programmes, interventions to prevent health conditions, an Employee Assistance Programme, OH services, other benefits, regular catch-ups, directing employees to healthy lifestyle support, phased returns to work, amending job role, workplace adjustments
why do you offer these? Probe: Is this a legal requirement, good for reputation, good for staff, for preventative reasons?
what, if anything does your organisation do to try and prevent health issues?

3. Managing sickness absence (15 mins)

Collect information on the employee sickness journey. Covering sickness absence (short-term and long-term), returning to work and adjustments.

Next, we are going to talk through what happens when one of your employees become ill, or requires additional support, and how this is dealt with within your company.

What do you think your roles and responsibilities are when supporting employees with sick leave and/or health conditions?

what are your legal responsibilities? Where does the role of the NHS fit in? Probe: funding, treatment. When does your role end?
what do you believe is your moral duty?
what about when it’s non-work related? What if there’s an impact on the employees’ productivity?
where do you think your responsibilities lie in terms of getting staff back to work or keeping them in work if their sickness absence is not work related?

What steps do you take when you become aware that an employee’s health is impacting on their ability to carry out their role?

at what point do you take these steps?
note to moderator: probe on timescales, if and how they use of OH, who refers OH. (If not already covered) Also probe on polices that support this, including disciplinary or capability policies, and dismissal

Do Line Managers receive any training to help with identifying and managing this?

prompt to include: nature of condition, severity, disclosure by employee
is the absence/health condition handled at a team level or is it referred broader in the organisation?

Where would you go for further advice to support employees with ill health? Why?

what further advice or information do you feel would be helpful to help you best support employees? Probe: does this include legal / expert advice?
(If employer has OH) Would your OH provider be your first point of call? If not why?

How long does your organisation consider to be a ‘short-term’ sickness absence? And how long does your organisation consider to be a ‘long-term’ sickness absence?

Moderator note (Do not read out to participant): A short-term sickness absence is an absence from work lasting less than 4 weeks. A long-term sickness absence is an absence from work lasting 4 or more weeks. Generally, for both, employers request some form of medical evidence (such as a fit note from a GP) for sickness absences beyond the ‘self-certification period’. Employees can self-certify for a period of 7 calendar days.

What are the costs, if any, associated with sickness absence?

prompt if only mentioning direct costs: These could be direct (cover costs) and indirect costs (stress on other staff members)
how are the costs different depending on whether the absence is short or long term?
what impact do these have on your organisation?
how do you respond to these costs? What are the strategies to minimise these costs associated with LTSA? How are these costs absorbed? (Prompts: support employees to return, dismissal, phased returns)
how long do you keep a job open? At what point do you consider dismissal? Is there a point at which you can no longer support the employee?

Could you tell me about the policies you use or (if you don’t have policies) things you do when managing staff sickness absences?

when and how are the policies used?
are they used consistently for short-term SA, long-term SA and LTHC? If so/if not why?
- probe: Are differences based on contract type (FT, PT, casual, 0hr), skill level, length of service, seniority? Why?
- are the policies adjusted to support employees with a disability? If so how? If not why not?
how confident do you feel in managing sickness absences?
- how is this impacted by who deals with this – for example, do line managers get training on this. Probe: how well can line managers manage these absences?
- what do they have in place and what do line managers do in practice?
- is this different for long term vs. short term sickness absences or recurring sickness absences?
how do you prevent a short-term absence becoming a long-term absence?

4. Return to Work (5 mins)

Now, I would like to focus on the processes that are in place in terms of enabling employees to return to work.

Moderator ask if they have a Return to Work policy: You mentioned earlier that you had a return to work policy, how is it applied in practice?

Moderator ask if they do not have a Return to Work policy: You mentioned earlier that you do not have a formal return to work policy, so how do you manage an employee returning to work?

how do you handle an employees’ return to work after LTSA? What methods/tools do you use? How? (Probe: phased returns, frequent contact, OH, external support)
- who is involved in this? (employee, HR, line manager, GP, OH) Who is responsible? At what point are they involved? When does action start to be taken?
- if a phased return was used, how has it worked? Is there any additional support that would be useful?
in your experience, do you feel that all managers or staff manage RtW in the same way? If no, what differences are there?
what are the challenges associated with helping an employee to return to work? Both in terms of the company and the employee. (Moderator: probe on finances, resources, knowledge, employer confidence and employee experience)
- is there anything that can get in the way of how SA is handled in practice?
- do you have any experiences of an unsuccessful RTW? (for example, LTSA resulting in no return or dismissal, or a LTHC that resulted in employee not being retained)
what costs and resources are there associated with facilitating a RtW following a LTSA?
how does the RtW process vary depending on the nature of the condition?
what do other employers do (both good and bad) when it comes to RtW?
- what further advice or information do you feel would be helpful to help you best support employees? Is more expertise needed internally?
- are you aware of any types of support on offer / information that helps keep employees in work? If yes, what sources do you use to inform these decisions?

5. Adjustments (5 mins)

Provide information on how adjustments are managed and when they are made.

We are now going to briefly focus on any adjustments that have been made to accommodate an employee’s needs or what would happen if this was required.

Can I ask, have you had experiences of in work adjustments? Please provide examples relating to non-permanent and long-term health conditions or disability

Moderator note: this section applies to non-permanent, long term health conditions or disability (not just those who have been absent). Please also note that there are many types of adjustments i.e. physical, adjustments to workload and hours.

at what point do you find out an employee needs an adjustment? Does the timing of this cause any problems in terms of finding out too late or the timescales required to put the adjustments in place? What conversations are had, and who are these between? Is this on a formal or more informal basis?
who is involved in assessing, deciding and implementing adjustments? (employee, HR, line manager, GP, OH)
how do you decide what adjustments are needed to support the employee and whether they can be accommodated?
how are adjustments paid for and by who? And who decides this? Probe: Is there a formal decision made? How is it recorded? Listen out for ‘Access to Work’
what types of adjustments are you able to make for an employee? Moderator probe on; physical, adjustments to workload, hours, different job roles and responsibilities, additional support, nature of condition
what adjustments would you consider ‘appropriate? Are there any you feel that would be inappropriate?
how do you monitor adjustments?
how are adjustments phased out? For example, someone returning to full time hours, after working part time
what are the challenges in accommodating adjustments?
are you aware of what other types of employers do (both good and bad) when it comes to adjustments? Probe: Mental health, physical health
what information or support would you like as an employer around adjustments? (if any)

Have you heard of the Equality Act 2010? What does it include?

what does this Act mean for you as an employer?
what do you think your responsibilities are?
does the Act have an impact on practice?
what do you have in place to support existing or new employees with a disability or long-term health problem?

6. Sick Pay (5 mins)

Collects information on the offer of OSP and reasons for offering this.

Next, we will touch on the offer of sick pay.

Do you offer Occupational Sick Pay (OSP) or is it only the Statutory Sick Pay (SSP) which is offered?

Note for moderator: note if participant mentions welfare or benefits

If above SSP (i.e. OSP) is offered

why is this? What factors were involved in this decision?
how does OSP work within your company? At what rate is it set? How long is it paid for? (Note, the rate should be as a percentage)
what happens when OSP ends?
- probe if required: falling to SSP, back to work, end of employment, dismissal, claim ESA or other welfare/benefits, use of insurance of some kind?
who can access OSP? What factors make an employee eligible? (does it depend on: cost, the employee, the condition)?
how does the payment of OSP effect RTW, if at all?
- how does it affect employees’ incentive to RTW?
- how does it affect your incentives?

If only SSP is offered

why is this? What factors were involved in this decision?
- how do you feel about SSP being a legal minimum set by the government? Impact of rules changing?
- probe: Cost implications? Haven’t thought about it?
have you offered OSP in the past? How do you feel about it? Why might other employers offer it?
how does SSP effect RTW, if at all??
what happens when statutory sick pay ends?
do employees come back too early? If so why do you think this is?

(If no form of sick pay is offered) Why is this?

7. Use of Occupational Health (5 mins)

Explores the use of OH and reasons for offering, as well as gaining an understanding of why employers don’t offer OH and the impact that this has.

Why do you invest in the health of employees?

how is the level of support for employees determined? Is support provided to employees in a routine (and/or indiscriminate) or case by case (and/or discriminate) way? Full-time vs. part-time employees vs. 0 hour contracts?
- in terms of treatment, where are the lines for funding it?
does it make economic sense to invest in employees staying in, or returning to, work? What does this depend on?
what do your employees expect of you? (Probes: covering legal minimum, for work not to make them sick)

Employers who use OH

We will now touch upon Occupation Health Services. By Occupational Health services, I mean advisory and support services by qualified professional to help employers to maintain and promote employee health and wellbeing.

What support and treatment do you offer as part of OH?

explore nature of OH contract – is it a permanent or ad hoc contract? What does the payment structure look like? (i.e. monthly payment then reduced treatment costs, lump sum to provide unlimited access, pay when needed)
why have you chosen this contract and payment structure?
- how do you justify the cost of a permanent or ad hoc OH contract? How do you determine whether it is value for money or not?
(if their OH provider is a public-sector body) What type of public sector bodies are used for OH?
if OH identifies that an employee needs treatment, how is this dealt with?
- in which cases will you offer to pay for treatment (does it depend on: cost, the employee, the condition)?
- where’s the line between your responsibility and the responsibility of the NHS?
- probe for examples if this has been the case in the past
is OSP considered to be part of your OH offer and / or wider health offer for employers?
what would help you prioritise OH more in your organisation? Probe on finances, incentives, more information, other

Employers who do not use OH

We will now touch upon Occupation Health Services. I understand that you do not use OH services as an organisation, is that right?

Can you describe to me what you understand by the term ‘Occupational Health’ services? Interviewer: Reassure participant that ‘do not know’ is a valid response

what kinds of things do you think employers might use OH services for?

Why do you not offer OH?

do you think it would be useful for your company? If not, in what circumstance would it be useful? In what situations?
is it something you have considered using in the past? What factors were considered? (does it depend on: cost, the employee, the condition)?
if cost related, if you had more money, do you think you would invest in OH services? Would OH be a priority?
where do you go for advice/support on this type of matter? For example, health visitors / nurses
if an employee needs treatment, how is this dealt with? Probe for examples if this has been the case in the past – when were they able to return to work? Were any adjustments made? Who made these decisions? Would you provide OH if the need arose?
do you have any policies in place that outline how to deal with this situation?

8. Summing up (2 to 3 mins)

Brings the conversation to a close, and allows participants time to mention anything that has not already been covered.

We are coming to the end of the interview, but I have a few final questions before we finish.

Is there anything else you would like to mention that we haven’t had the opportunity to discuss?

Thank and reiterate confidentiality. Incentives = a ‘thank you’ from Ipsos MORI for their time and contribution. £50 donation to a charity of their choosing.

Explain next steps for the research and close – report will be published in Autumn 2019.

DWP and DHSC, ‘Improving Lives: The Future of Work, Health and Disability’, 2017. ↩
DWP and DHSC, ‘Health in the workplace – patterns of sickness absence, employer support and employment retention’, 2019. ↩
DWP and DHSC, ‘Health is everyone’s business: proposals to reduce ill health-related job loss’, 2019. ↩
GfK NOP Social Research, ‘Health and well-being at work: a survey of employers’, 2011. ↩
A more detailed explanation of the qualitative aims and method is included in Chapter 4. ↩
Read Business population estimates 2018. ↩
Completes / completes + ineligible ↩
Completes / completes + refusals and abandoned interviews ↩
Completes / valid total ↩
Please see: Business population estimates 2018. ↩
For the full question wording and example classifications, please refer to B7: OCCUPATION in Appendix B. ↩
For the full question wording and example classifications, please refer to G2: WELLBEING in Appendix B. ↩
Variables significant at P<0.1 in the full model were included in initial runs for the final model. However, variables significant at P<0.05 only were included in the final model. ↩
Please note that these variables are based on B7_3 in the questionnaire, as the regression was carried out before the LCA on B7: OCCUPATION. ↩
Survey response options included:

1. Meetings with employees.
2. Phased returns to work from sickness absence (returning to full duties and hours at work gradually, over a defined time period).
3. Amending employee workload or job role (for example, reduced hours/days, extra breaks, or different duties).
4. Workplace adjustments (such as different chairs or desks, building modifications, or other specialised equipment).
5. A job coach or personal assistant (for example, a sign-language interpreter for meetings).
6. Additional external support or advice (for example, clinical support such as psychological therapy or physiotherapy, or another expert or specialist).
7. None of these or do not know. ↩
Variables significant at P<0.1 in the full model were included in initial runs for the final model. However, variables significant at P<0.05 only were included in the final model. ↩
Survey response options included:

8. Meetings with employees.
9. Phased returns to work from sickness absence (returning to full duties and hours at work gradually, over a defined time period).
10. Amending employee workload or job role (for example, reduced hours/days, extra breaks, or different duties).
11. Workplace adjustments (such as different chairs or desks, building modifications, or other specialised equipment).
12. A job coach or personal assistant (for example, a sign-language interpreter for meetings).
13. Additional external support or advice (for example, clinical support such as psychological therapy or physiotherapy, or another expert or specialist).
14. None of these or do not know. ↩
Variables significant at P<0.1 in the full model were included in initial runs for the final model. However, variables significant at P<0.05 only were included in the final model. ↩
Please note that these variables are based on B7_3 in the questionnaire, as the regression was carried out before the LCA on B7: OCCUPATION. ↩
In principle, the maximum percentage of variance which can be explained is 100 per cent. However, in practice, it is not usual to expect models to aspire to this value, except perhaps in time-series situations, where lagged variables of the outcome are included in the model. ↩
Please note that these variables are based on B7_3 in the questionnaire, as the regression was carried out before the LCA on B7: OCCUPATION. ↩
Survey options included: Meetings with employees; Phased returns to work from sickness absence (returning to full duties and hours at work gradually, over a defined time period); Amending employee workload or job role (for example, reduced hours/days, extra breaks, or different duties); Workplace adjustments (such as different chairs or desks, building modifications, or other specialised equipment); A job coach or personal assistant (for example, a sign-language interpreter for meetings); Additional external support or advice (for example, clinical support such as psychological therapy or physiotherapy, or another expert or specialist); and Do not know or none of these. ↩
If only a few people give a positive response to a binary variable, the analysis may provide unstable solutions. If there are a number of such variables in the analysis, the algorithm may not be able to complete at all. ↩
Run using SPSS v25. ↩

Cookies on GOV.UK

Applies to England, Scotland and Wales

Statement of compliance

Value of this research

Trustworthiness

Quality

1. Introduction

Table 1.1: Size breakdown of overall sample

2. Survey technical details

2.1. Survey and questionnaire development

2.2. Sampling

Table 2.1: Count of employers by size band in Great Britain

Sample preparation and selection

Table 2.2: The sample of employers by industry sector (SIC07) and size

2.3. Fieldwork

2.4. Fieldwork outcomes and response rates

Table 2.3: Sample outcomes

Table 2.4: Response rate by employer size band

2.4. Data processing and weighting

Table 2.5: Sample profile weighted by employer unit and employee volume

Variable – size:

Variable – sector:

Variable – ease of recruiting staff:

Variable – ease of retaining staff:

Variable – employee representation or trade union:

Variable – decisions on daily work tasks:

3. Advanced analysis

3.1. Latent class analysis

What is latent class analysis?

3.2. LCA: Workforce composition

Table 3.1: Occupational LCA by size of employer

Table 3.2: Occupational LCA by sector

3.3. LCA: Provision of retention and reintegration measures

Table 3.3: Number of retention and reintegration measures by latent class (raw figures, unweighted)

Table 3.4: Retention and reintegration LCA by size of employer (weighted)

Table 3.5: Retention and reintegration LCA by sector

3.4. Logistic regression analysis

What is logistic regression analysis?

3.5. Occupational Health (OH) regression

Table 3.6: OH regression – all cases, full model

Table 3.7: OH regression – all cases, reduced model

3.6. Sick pay regression

Table 3.8: OSP regression – all cases, full model

Table 3.9: OSP regression – all cases, reduced model

OSP regression: micro and small employer models

Table 3.10: OSP regression – small and micro employer model, full model

Table 3.11: OSP regression – small and micro employer model, reduced model

3.7. Cluster analysis

What is a cluster analysis?

The method

Table 3.12: Binary response variables included in the cluster analysis

Table 3.13: 7-cluster solution

Figure 3.1: Cluster analysis variables

4. Qualitative follow-up research

4.1. Background and aims

4.2. Recruitment and fieldwork

Table 4.1: Qualitative sampling matrix

4.3. Analysis

Appendix A: Scoping Review

Overview

Research Context

Methodology

Results

Table A.1: Number of articles by themes and review questions

1. Cross-cutting review question: What works

2. Cross-cutting review question: current practice

3. Cross-cutting review question: policy levers

Conclusions

Appendix B: Questionnaire

Introduction

S0. Consent

S1. Sndem

A1a. Datasheet

A1b. Dtasht_size

I4: Datasheet_use

Section A: Employer characteristics

A1. Role

A2. Org_type

A3. Site

A4. N_employees