The WHAA model below (Figure 1) demonstrates the flow between the various ROI components. Ultimately all “returns” will be measured by the three parameters on the right - productivity, accident/Injury and illness/absence. These can be impacted upon directly (
solid lines) or via intermediary pathways such as fatigue/energy, morale/engagement, culture and attraction/retention (dotted lines).
The measurement of the intermediary components is therefore unnecessary as their impact can be determined through the measurement of the three “downstream” measures, all of which (fortunately) are more amenable to quantification in financial terms.
In other words, measuring productivity will pick up the benefits that flow from improvements in health, fatigue/energy, morale and engagement, obviating the need to measure these directly.
The bottom line is that the entire value of the “return” in “return on Investment” can be determined through the measurement of three parameters – Productivity, Injury/accident and illness/absence. Indeed, we would question the efficacy of a program that resulted in improved employee morale and engagement without an appreciable change in the downstream metrics. In business terms this represents a “failure to capitalise” on a benefit.
So we’ve reduced the number of required parameters for effective ROI calculation to three. Let’s take a closer look at these.
Whereas absenteeism is “not at work – not working”, presenteeism is often defined and “at work – not working” (or in a more humorous vein; “appearing not to be absent”). It is a huge issue. In a modern economy, it stands at around 20% - one fifth of all the time spent at work by an employee does not generate an outcome for the organisation.
Whilst a detailed Fatigue & Energy Attraction/Retention Corporate Culture Morale & Engagement Employee Health discussion of presenteeism is beyond the scope of this document, readers are referred to Mitchell & Bates (2011)8 for a more comprehensive review on the topic. Even though research shows that companies with successful health and productivity programs have superior business outcomes and financial returns, methods to measure productivity differ and a validated method to monetize costs associated with lost work time, especially presenteeism, is still lacking.
Measuring productivity/presenteeism comes down to a choice of methodology. If a relevant output measure is available (i.e. number of claim forms processed, tonnes of coal mined or time to install insulation in a roof), productivity can be measured directly. In reality, very few jobs are amenable to such an approach. Indeed, without accounting for the cost of errors/omissions, physical/psychological stress/burnout and/or quality of service delivered, such measurements lack veracity. If a direct measure of productivity is chosen, the ROI component relating to productivity is simply the increased value of production/output over the cost of program delivery.
As stated, very few jobs or work environments are candidates for direct measurement, and productivity must therefore be measured indirectly. There are two generally accepted methods for indirect measurement; a) Self-assessed, and b) Assumptive.
Self-Assessment of Productivity/Presenteeism
Whist there is understandably some resistance to using “self-report” survey tools to assess employee productivity (“how hard to you work?” … “very hard indeed sir!”), these concerns should be tempered by the fact that the tools currently in use have been developed, deployed and monitored with the highest levels of academic rigour. The two most commonly used tools; The Harvard/WHO Health and Productivity Questionnaire (HPQ)9 and the Work Limitations Questionnaire (WLQ)10 have an impeccable pedigree which attests to their reliability and validity. Notwithstanding the credibility that goes with the WHO and Harvard brands, collectively these survey tools have been;
- Validated against 300,000 face to face interviews
- Used as the survey tools of choice in over 2,500 published studies covering tens of millions of employees in a wide range of business/industry sectors
- Validated in all industrialised economies, and more recently in a number of Asian and Middle Eastern countries11
These tools have been used by many Australian WPH providers and WHAA members. A recently published study which deployed the HPQ to assess presenteeism in the Australian Health Care and Education sectors showed that the cost of presenteeism ($AU 2006) was $8,338 and $8,092 per worker per annum respectively.
Although the tools themselves are quite lengthy to deploy in full, a presenteeism score can be easily derived from 3-4 simple questions added to an existing Health Risk Assessment (HRA) or staff survey tool.
Details for HPQ use can be found at:
http://www.hcp.med.harvard.edu/hpq/info.php The WLQ questionnaire is available from Debra Lerner, The Health Institute, Tufts-New England Medical Center, Boston MA, USA. WLQ@tufts-nemc.org.
The Assumptive Method The simplest and most widely used technique to measure the productivity impact of health status change is to use data from exiting studies and assume that the published relationship holds true for the employee population being monitored.
Over the past decade, evidence published/presented from groups such as Medibank Private, Australian Health Management (AHM) and Good Health Solutions/Alere has demonstrated that;
- High Risk employees (5+ Risks) are at work but not productive 32.7% of the time compared to Low Risk Employees (0-2 Risks) who are not productive 14.5% of the time13 . The productivity difference between healthy and unhealthy employees is therefore 18.2% or 45 working days per annum
- High Risk employees average 5.1 hours/month absence versus 2.4 hours/month for Low Risk.11 This amounts to 32.4 hours (over 4 days) days per annum.
- In a more recent study14 , healthy employees average 1-2 sick days per annum versus 18 days for those in the lowest health and well-being category
- This study also showed that the most unhealthy employees are productive for only about 49 hours out of each month compared to around 140 hours/month for the most healthy
- Poor health can account for an average 5% loss in productivity across the entire Australian workforce with the unhealthiest group reporting a 13% drop in productivity (using internationally validated criteria). About half this is related to chronic conditions such as headaches, hay fever and neck/back pain, whilst half can be accounted for by lifestyle factors such as inactivity, smoking, obesity etc.
The bottom line is that unhealthy employees represent an unmanaged risk to their employers. The literature is quite definitive – this risk can be mitigated to the economic advantage of the organisation, and given that the Average Australian employee carries 4 risk factors, these results are alarming.
Using this information, the productivity loss per risk factor can be determined and applied to changes in risk status pursuant to the deployment of a workplace health program. For example, using the first bullet point above, the productivity loss per risk factor can be calculated at around 3.6%. A program that reduces risk factor prevalence at a worksite from say 4 risks to 3.5 risks per employee would, under this assumption, result in performance improvements of 1.8% per employee. Using 2,000 working hours per annum as a rough estimate of full time worker hours, this accounts for 36 hours of additional work – roughly one week. At the average Australian wage ($75K in 2014) this amounts to around $1,440 per annum of benefit [it is significantly more if we consider the “value” of an employee as their contribution to the profit of the organisation – The Commonwealth Bank of Australia.
(CBA) employs 52,000 people and generated a profit of $8.7B in 2014/15 - $167K per employee per annum. 1.8% of this figure would put the benefit of the program at around $3K per employee, and that is before consideration of illness/absence and accident/injury effects].
Illness and Absence
Most organisations have basic payroll data on illness related (unscheduled) leave/absence. Unfortunately this is rarely detailed enough for serious analytics (half of all Australian public service departments were unable to provide the National Audit Office with absence data for their annual performance audit16).
Payroll data is often sufficient to determine gross trends such as seasonal variations due to colds/flu. “Flu season” generally causes a spike in absence from a baseline of around 0.8 days per month in the warmer months to 1.3 days per month during the colder months, a trend evident even on the most basic of systems. Although it seems like a formality to monitor the impact of an intervention such as an influenza vaccination program on unscheduled absence by comparing data year on year, annual variations in the virulence of infections often swamp the vaccination effect.
To overcome this problem, and provide the capacity to monitor the impact of a flu vaccination program, employers need to move from a longitudinal trend model (where year to year fluctuations in the severity of the flu mask program impact) to a model which compares absence data of vaccinated vs unvaccinated employees in a given flu season.
The most significant factor hindering analysis of illness related absence is therefore the inability of HR/Payroll systems to provide the appropriate data – rarely does a payroll system contain data on the vaccination status of employees.
This problem is exacerbated when looking at other factors which impact absence. Few Australian organisations are brave enough to countenance the idea of a HR/payroll system which houses data on employee blood pressure, BMI, Stress levels, smoking, alcohol and exercise habits (notwithstanding privacy considerations), yet the data would prove invaluable in assessing the absence contribution of health and lifestyle to ROI.
One must also consider that the “cost of absence” varies widely across professions, and not necessarily in a manner that relates to remuneration level. Teachers, train and bus drivers must be replaced, many other types of employees play “catchup” in the days following illness related absence. These dynamics alter the ROI assumptions significantly. Of the 373,000 teachers in Australia, over 15,000 are absent daily at a replacement cost of $350 per teacher per day resulting in over $1B in the cost of casual replacements each year (plus admin costs!). Alter absence by just 1% and there’s a direct saving of $10M. This “direct cost of replacement” is not evident with bank tellers, builders or biomedical engineers.
As with productivity/presenteeism, a direct measurement approach may not be possible (but is certainly the preferred option if available), so the “assumptive” model may be the only available option (although the HPQ and the WLQ do have survey questions on selfreported absence as well as self-reported productivity).
Assumptive Method for Absence ROI
If we assume that the target employee population responds to health and lifestyle issues in a manner consistent with the academic literature, likely absence rates can be related to risk factor prevalence data. This analysis can be performed in relation to specific risks (see table 1) where each factor is accorded its own weighting17, or, more simply, the aggregate number of risk factors per employee can be tallied and absence calculations made (table 2).
Sample Calculation – Company ABC has 1000 employees with an average income of $65,000/year. Post HRA data shows “risky” alcohol and tobacco use has halved since the previous HRA as a result of various interventions – Alcohol from 10% to 5% of employees and smoking from 16% to 8%.
Absence Cost Pre = 10% of 1000 employees x 5% x $65K = $325,000 Absence Cost Post = 5% of 1000 employees x 5% x $65K = $162,500
Value of Benefit = $162,500 Smoking Effect Absence Cost Pre = 16% of 1000 employees x 2.84% x $65K = $295,360 Absence Cost Post = 8% of 1000 employees x 2.84% x $65K = $147,680 Value of Benefit = $147,680
If the cost of delivering these programs was $30pp for the alcohol education program and $200 per smoker for the 160 smokers, net cost is $30K and $32K for the alcohol and smoking programs respectively. This gives on ROI of 5.4:1 for the alcohol program ($5.40 return for each dollar invested) and 4.2:1 for the smoking program.
Note that these results represent the ROI contribution from Absence only, and only two risk factors were considered. Applying the presenteeism calculations from column 3 of Table 1 yields significantly better ROI outcomes, mainly because the presenteeism effect of smoking is almost double its absence impact (alcohol has a similar impact on both absenteeism and presenteeism).
Also note the disparity between the absenteeism/presenteeism effects related to each condition. Asthma effects absence four times more that it affects presenteeism, whereas physical activity affects presenteeism sixteen times more that it impacts absence.
The Simpler Method
Without going through the formidable task of calculation ROI by individual risk, one can apply a more generalised analytical technique which simply assumes absence and productivity impacts can be aggregated according to number of risk factors and number of medical conditions18 (see Table 2 below).
Accidents, Injuries and Workcover claims
In WHAA’s Best Practice Guidelines we promote the integration of Workplace Health and Occupational Health &Safety (OH&S). The reasons for doing so are compelling;
- Over 500,000 work-related injuries occurred in Australia in 2013/14, with 117,815 of these being serious19. Serious claims are 10 times more likely in physically demanding work roles than in professional/technical/finance/insurance roles.
- The top 4 “mechanisms of injury” for serious claims are related to either muscular stress or falls. Together these account for around 43% of claims.
The 5th most common cause is “Mental Stress”, although the unit cost of mental stress claims is 2-3 times that of the top 4. Clearly, safety practices and training for manual handing are vitally important for prevention of musculo-skeletal injuries and falls, however closer scrutiny of the data reveals many upstream issues which exacerbate the problem and provide potential early intervention opportunities. Many of these point to the physical and psychological health of the worker in ways which are outside the scope of traditional OH&S paradigm, but firmly within that of Workplace Health.
Consider for example that claims are twice as prevalent, seven times more costly and result in 13 times more lost work days in those with high BMI’s.21 It is immediately evident that a risk mitigation strategy should encompass worker anthropometrics. This is the domain of workplace health, rather than OH&S, and points to an integrated solution as the best way forward.
Similar data exists for smoking and work related injury (38% higher risk in smokers22), and chronic conditions such as hypertension and diabetes (workers with 2 or more chronic conditions have 2½ times longer injury related absences than healthy workers - 9.3 days pa vs. 3.7 days pa) .
Sleep disorders and associated fatigue represent possibly the single largest contributor to the economic burden of the nation in terms of workplace accidents and productivity loss. In their 2006 publication “The Economic Cost of Sleep Disorders”24 Access Economics calculate the cost of sleep related workplace injuries at almost $2Billion, with an additional $1.2Billion in productivity losses.
Smoking cessation, chronic disease management, weight management, fatigue, sleep hygiene and stress management have a massive impact on OH&S outcomes, but are generally outside the scope of OH&S interventions, hence the imperative for an integrated solution.
The unit cost of a workplace Injury is, on average, an order of magnitude above the presenteeism/absenteeism costs of an employee. The data of Burton et al25 showed the average productivity loss per risk factor to be around 2.4%, translating to around $2,000 per risk factor for the average Australian worker. Given the average worker harbours 4 risk factors, the net loss related to health risk and lifestyle is around $8K per employee per annum. The average cost of a Workcover claim for “slips, trips and falls” is ten times this26 (see Table 3 below), and for mental stress, thirty times. A WPH strategy that has even a small impact on such claims will yield significant ROI.
It is clear that the confluence of OH&S, personal health/anthropometric and behavioural factors provides the input parameters for issues such as “risk of falling” (not to mention concentration/fatigue issues and numerous other factors). This being the case, an integrated approach combining elements of OH&S and return to work processes with Wellness is most likely to deliver the best outcome. The inclusion of an ROI component which relates employee health to accidents, injuries, claims and premiums would strengthen the case for an integrated approach and gain much needed support from senior managers.
Due to compliance factors related to OH&S, this is generally where the best “cost” data sits. The problem however, as with absence and productivity, is one of data linking. If a WPH program benefit is to be demonstrated, linking health risk and lifestyle data to accident, injury and claims data is imperative. Unfortunately it is rarely performed due to issues such as data linking cost/complexity and privacy issues which mitigate against this approach.
To do this effectively requires that an organisation obtain the following;
- Who participated in the employee health program? (enables unit cost to be determined and also ensures program ROI is not inflated by any improvements in employee health which occur outside the program )
- What health outcomes were achieved? (allows focus on the cohort of participants for whom health improvement are likely to result in changes to the parameters being investigated)
- Was there a difference in the accident/injury/claims in those employees that demonstrated health improvements vs those that didn’t and non-participants? If so, what was the magnitude of the change?
Analysing this data, with the associated injury and claims/premium data, will readily allow ROI to be determined. The benefit value is aggregation of cost saved through reduction in number and cost of claims (together with savings in admin and avoidance of premium increases). The investment cost is the cost to deliver the weight management program.
Comcare/Workcover recognise the multifactorial aetiology of workplace accidents/injuries. The model below (from WorkCover Qld27) outlines how personal and work related factors influence worker health and wellbeing and the performance of a business.
A word of caution is pertinent when attempting to extrapolate from data of North American origin (which constitutes over 80% of the global data set). US employee health programs generally include a lifestyle component and a disease management component, the latter inclusion due to the fact the US companies are directly responsible for the health care costs of employees. Two RAND Corporation publications from 201328 and 201429 are illuminating in terms of the components of ROI attributable to lifestyle and disease management. The overall ROI was $1.50—that is, a return of $1.50 for every dollar that the employer invested in the program, but the returns for the individual components differ strikingly: $3.80 for disease management but only $0.50 for lifestyle management for every dollar invested. As a result, ROI data for US based studies is significantly inflated compared to the Australian context (Australian employees fund health care cost indirectly through the Medicare Levy, and one could argue that and ROI calculation should reflect cost savings through the national medical scheme).
Factors Affecting ROI
One of the most critical elements impacting program ROI in participation rate. One might expect that the relationship between participation and ROI in linear, but due to the fact that the least healthy employees are also the least likely to attend/participate in WPH programs, increasing participation rates draws more higher risk people into the program. It is these people that generate the best ROI if their health risks are managed effectively.
As the graph below demonstrates, increasing participation can double or triple program ROI30. This underscores the need to support WPH with a comprehensive promotional strategy which instils a strong wellness culture and a particular focus on high risk employees.
Numerous other factors contribute to Program ROI. These include;
- Risk factor prevalence of employees (A “zero risk” population has little room for improvement)
- Targeted program promotion and content – adapting the program to employee needs/risks
- Sound behavioural foundations – Most interventions require employees to change risky behaviours. Basing a program around a sound theoretical model such as the Health Belief Model, the Transtheoretical Model, the Theory of Reasoned Action, a Self-Efficacy model or Emotional Intelligence (or a combination of these) will tend to yield better results than an ad-hoc approach.
- Services delivered via different mediums – A self-managed program delivered online or through an “app” may deliver a similar outcome to multiple 1 on 1 counselling sessions at a fraction of the cost.
- Use of effective IT in program administration (Bookings, reporting etc.)
ROI is Improving over Time
The move to a results-oriented framework has underpinned the improved outcomes which are now delivered by workplace health programs. Programs have evolved over recent years to embrace a strong Human Capital Management framework which is highly personalised, more results oriented, has higher participation and more robust evaluation criteria.
Health portals, online gamification, wearable devices, social networking, self-service health kiosks, health “apps” on smartphones/tablets, rewards programs, motivational messaging and a move towards managing/preventing chronic disease through technology are all changing the landscape in terms of program content and delivery of workplace health programs.
One of the critical issues in relation to technology and Workplace Health is the ability to automate processes and enhance efficiencies when deploying programs. This includes functions such as booking, data collection, analytics and reporting. Costly when deployed manually, but cost effective in the context of a good IT solution.
Without good IT infrastructure, efficiency and scalability are compromised, and inadequate reporting threatens program sustainability if management are unable to measure impact over time.
The landscape is changing significantly in Australia with the 2014 Buck Survey31 showing telehealth related activities comprised 4 of the 5 fastest growing areas of workplace health in Australia and NZ.
“How workplaces engage, educate and motivate change and healthier behaviours within their people is also evolving. While more traditional work health methods such as wellness seminars and screening programs remain popular, many workplaces are looking to take advantage of developments in technology which increase the scale, accessibility and affordability of programs. Digital health platforms including wellness portals and mobiles, as well as telehealth programs are all expected to increase in use amongst work health strategies in workplaces during 2015/16” The 2015 Bupa Survey
Online health portals have the capacity to automate numerous processes. In the absence of a multifaceted online solution, many of these processes would not occur, or be slow, manual and prohibitively expensive, absorbing much of the employee health budget in administrative costs rather than effective service delivery. The impact on ROI is devastating.
As economic cycles ebb and flow in the Australian business sector, inevitable periods of economic hardship arise which result in scrutiny of discretionary expenditure for cost saving purposes. Lack of hard data has, in the past, caused the cessation of many programs which probably delivered great results, but reporting and measurement to demonstrate this were lacking. Essentially, ROI is a mechanism to make WPH recession proof.
The data presented in this WHAA White Paper demonstrate what a powerful tool ROI can be. Ultimately many organisations will continue to support WPH under the assumption that it makes good business sense for a host of tangible and intangible reasons, regardless of whether they take the time and energy to measure outcomes. Should they choose to however, the information provided here will assist in the formulation of a potent, evidence based business case in favour of organisational support for employee health.
1 Marlo, K. and Serxner, S. (2015), Beyond ROI: Building employee health & wellness value of investment. National Business Group on Health/Optum - White Paper. 2 Shepherd, R. Medicine and Science in Sports and Exercise (1982) 3 Pelletier, K. Journal of Occupational & Environmental Medicine: A Review and Analysis of the Clinical and Costeffectiveness Studies of Comprehensive Health Promotion and Disease Management Programs at the Worksite: Update VIII 2008 to 2010; November 2011 - Volume 53 - Issue 11 - p 1310–1331 4 Edington, D, Alyssa B. Schultz, (2008) "The total value of health: a review of literature", International Journal of Workplace Health Management, Vol. 1: pp.8 - 19 5 Steven G. Aldana (2001) Financial Impact of Health Promotion Programs: A Comprehensive Review of the Literature. American Journal of Health Promotion: May/June 2001, Vol. 15, No. 5, pp. 296-320. 6 Chapman, L., American Journal of Health Promotion, July/August, pp 1-14, 2005. 7 Baxter, S. et. al., American Journal of Health Promotion; Vol.8, #6, (2014)
8 Mitchell, R.J. (2011). Measuring Health-Related Productivity Loss. Pop. Health Management Apr; 14(2): 93–98. 9 Kessler, R.C., et. al. (2003). The World Health Organization Health and Work Performance Questionnaire (HPQ). Journal of Occupational and Environmental Medicine, 45 (2), 156-174. 10 Lerner D, et. al. (2001), The Work Limitations Questionnaire; Medical Care. 2001 Jan;39(1):72-85. 11 Pournik, O. et al. Reliability and validity of Persian version of World Health Organization health and work performance questionnaire in Iranian health care workers. Int J Occup Environ Med. 2012 Jan;3(1):33-8. 12 Scuffham PA, Vecchio N, Whiteford HA (2014). Exploring the validity of HPQ-based presenteeism measures to estimate productivity losses in the health and education sectors. Med. Decision Making. 34(1):127-37.
13Musich, S. et.al. (2006): Am. J. Health Promotion, 20(5); 353-363. 14 Medibank Private/Vie Life Report, 2007 15 Lang, G.J. (2007), GHS Health, Absence and Productivity Survey, IHPM Conference.
16 Audit Report No 52 (2002/3). Absence Management in the Australian Public Service. Aust. Nat. Audit Office.
17 Goetzel, R., et al. (2004). Health, absence, disability, and presenteeism cost estimates of certain physical and mental health conditions affecting U.S. employers. J. Occ. Env. Med., 46, 398-412.
18 Mitchell, R.J. (2011). Measuring Health-Related Productivity Loss. Pop. Health Management Apr; 14(2): 93–98. 19 Key Workplace Health and Safety Statistics Australia (2015), Safe Work Australia 20 Australian Workers Compensation Statistics, 2012-13, Safe Work Australia
21 Ostbye, T. et.al. Arch Intern Med. (2007) Apr 23;167(8):766-73. 22Dong, Z.S., Wantg, Z. and Largay, J.A. (2015) 23 Casimirri, E., et.al., (2014) International J. Occupational Medicine and Environmental Health. 27(3): 343–354 24 Hillman, D.R., et. al. Sleep (2006); 29(3): 299-305 25 Burton WN, Chen CY, Conti DJ, Schultz AB, Pransky G, Edington DW. The association of health risks with on-the job productivity, J Occupational and Environmental Medicine (2005), 47(8):769-77. 26 Safe Work Australia (2012). The Cost of Work-related Injury and Illness for Australian Employers, Workers and the Community.
27 See https://www.worksafe.qld.gov.au/injury-prevention-safety/health-and-wellbeing-at-work/impacts
28 Mattke S, et al, Workplace Wellness Programs Study: Final Report, Santa Monica, Calif. RaND Corporation, RR254-DOl (available at www.rand.org/t/ RR254), 2013. 29 Caloyeras JP, et al, “Managing Manifest Diseases, But Not Health Risks, Saved PepsiCo Money Over Seven Years,” Health Affairs, Vol. 33, No.1, 2014, pp. 124–131. 30 Chapman, L.S. (2007) Proof Positive. An Analysis of the Cost Effectiveness of Worksite Wellness. Seattle, WA: Chapman Institute
31 Buck Consultants (2014). Working Well: A Global Survey of Health Promotion and Workplace Wellness Strategies. 32 Workplace Health in Australia – Bupa Benchmark Survey 2015. Available at; http://www.bupawellness.com.au/index.php