by Scott MacStravic
July 2, 2007 at 11:17 am
· Filed under Population health management
There have always been significant challenges in measuring the results achieved through PHM efforts, particularly with employee populations where “results” can include a host of gains over and above reductions in healthcare/insurance costs. These challenges include both determining what changes have occurred in whatever success measures are selected as consequences of interest, and gathering evidence that demonstrates, or at least strongly indicates that the consequences actually occurred because of PHM interventions.
This is enormously more complicated than tracking the results of sickness care interventions, where there is a targeted acute or chronic condition treated by a specific medical/surgical intervention. But even that is complicated enough, depending on what success measures are considered worth measuring. A the New York Hospital for Special Surgery, for example, surgeons thought that being pain-free was the key success outcome that patients wanted. But, when asked, patients identified restoration of ability to do household chores, dress themselves, engage in favorite activities, etc. as equally important. Had their employers been asked, productivity and job performance would likely have been added. [“Patients and Doctors Often Differ on What Constitutes Successful Surgery” Strategic Health Care Marketing 20:3 March 2003 p.12]
The most obvious approach to measuring success would be to identify whatever success metrics are important to the providers and customers of PHM efforts, particularly the return on investment (ROI) achieved, and measure those. But in order to document success, a “causal chain” of effects must also be identified in order to have solid grounds for accepting that success was achieved because of the PHM investment, rather than through unrelated reasons. And this causal chain has many links.
In sickness care, it is usually necessary only to identify the diagnosed condition(s) identified as the problem, and the prescribed treatment(s) employed, then track changes in the problem, while attributing them to the treatments. There is a rule about “therapeutic specificity” that calls for attributing to the intervention only those changes shown by evidence-based medicine (EBM) studies to be responsible for such changes.
But in PHM, there is often neither a single problem nor a single intervention, particularly when aiming for results such as worker productivity and performance. In the simplest case, workers with diabetes may be identified as having low levels of adherence to their blood sugar monitoring and control responsibilities, for example. The PHM intervention might choose as its success metrics first improvements in adherence, and second improvements in control of blood sugar, or HbA1c metrics.
But diabetes, in addition to being a chronic disease, is also a major risk factor for heart, kidney, and eye diseases, as well as circulatory problems, and amputation of limbs that result therefrom. Moreover, diabetes patients are typically burdened by high blood pressure and cholesterol, overweight/obesity, depression and other “co-morbidities”. The best DM programs aimed at diabetes patients would tend to include far more than adherence to medications, to include diet and exercise improvements, and self-management efforts relative to co-morbidities as well.
As a consequence, success metrics related to diabetes DM may include first measures reflecting specific changes in behavior by patients, and often by their personal physician as well, with respect to monitoring of blood sugar, pressure, cholesterol, etc. Improvements in diet, exercise, and other healthy activities would commonly be tracked. And these may be causes of other improvements, in weight, fitness measures, energy levels, self-confidence, etc.
While these added measures complicate the evaluation process, they also create the foundation for proving or at least making a strong case that the PHM intervention caused the ultimate value and ROI success measures, such as reduced absences and improved productivity, and reduced sickness care costs (health costs for managing diabetes would be expected to increase due to improved medications adherence). Reductions in disability wage replacement, overtime and temporary worker replacement would also be tracked as related results.
By tracking the chain of events, from DM participation by patients to changes in their behavior to changes in their objective clinical metrics to changes in absences/productivity/performance to changes in labor costs, quality performance, customer satisfaction, and operating margin — enables making the case that PHM interventions, and the resulting changes in individuals, have produced changes in the organization’s performance.
Measuring all the “links” in the chain of causes will necessarily add to the costs of the DM or other PHM intervention. But it will also provide a comprehensive basis for managing the intervention in order to improve its performance over time. By checking levels of employee enrollment, participation in and completion of PHM interventions, then changes in their behavior, in objective physiological metrics, then in sickness care use, absence, productivity, quality, turnover, etc., employers will be able to identify where gaps occur, and focus their efforts accordingly.
The success measure most often missing from PHM evaluations is the consequences that employee participation, behavior changes, and objective physiological improvements have on their own overall health and life quality. It is usually these positive consequences that employees look for when investing their own time, effort, and often money as well, in the PHM effort. Their anticipation of potential positive effects, their perception of those that have occurred, and confidence in continued improvement are often the major reason for their continued PHM participation or adherence to healthier behaviors.
Moreover, once significant and valued improvements in their health and lives are perceived by employees, and credited to their engagement in PHM initiatives, it should become less necessary or even useful to offer them incentives and rewards in order to get them to participate therein. To the extent that this happens, employers can reduce their costs and increase their ROI from PHM investments. Increasing success over time is the normal or typical pattern of PHM returns anyway, and reducing costs will add to this pattern and employer performance outcomes.
Moreover, once significant health/life quality effects occur and are perceived by employees, this has a good chance of moving them past the challenge of overcoming productivity and performance impairment related to poor health or motivation, and moving them toward what has been called “positive presenteeism”, where they improve their performance beyond what has been previously deemed “normal” levels. This will also tend to improve employee retention and save other labor costs in that manner as well.
While outsource providers of PHM services are normally chosen for conducting interventions, due to privacy/confidentiality of employee and patient data, for example, the employer is usually the only one who can track many of the success measures involved. Measuring PHM success will necessarily be a partnership effort requiring the cooperation of employees, providers, and the employer, but without a comprehensive measurement effort, there is far less likelihood that the full potential value of PHM, to both employers and employees, will be reached. And that full potential should, in most cases, more than cover the added costs of comprehensive success management. Moreover, it is the only way that employers can even know if they are achieving the full potential.
Permalink
[…] Nile Virus Link to Article diabetes Measuring Population Health Management Results » Posted at […]
Measuring ROI of Population Health Management (PHM) Programs
Marathon Health,Inc. is a vendor in the emerging market of onsite PHM providers leveraging onsite clincians to serve as personal health coaches to effectuate risk reduction and disease management in addition to providing onsite acute care. The key to success of PHM prograams in engagement, and Marathon Health’s onsite clinicians are averaging 70% engagement rates compared to 25% with telephonic coaching.
Marathon Health divides its return on investment methodology into two sections:
1. Hard dollar savings
2. Soft dollar savings
Hard Dollar Savings
This category includes three items: onsite primary care, reduced utilization of physician and hospital services and disability/absenteeism.
1. Primary care delivered onsite by the Marathon Health clinician(s) and incurred but not reported (IBNR) for the purpose of generating a claim requiring provider reimbursement. Marathon Health uses the carrier specific or zip code specific fee schedule for relevant CPT-4 codes in calculating IBNR. Companies that have established on-site health clinics found improved quality and cost savings. Employers typically saved 20% to 25% of their previous health care costs. Illnesses is caught sooner and absenteeism is reduced since the on-site visits took about 15 minutes and off-site visits took close to three hours. Employee satisfaction and overall corporate image also improved. Marathon Health estimates approximately $175,000-$250,000 in savings per 1,000 employees per year.
2. Reduced utilization of physician and hospital services for illnesses associated with preventable claims. The percentage of medical expenses attributable to lifestyle claimants ranges between 50% and 70% of total medical payments. Using evidenced based savings achieved through population health management, Marathon Health estimates that approximately 25% of all claims can be prevented. These claims include the following ICD-9 codes which the medical literature shows are directly related to high blood pressure, hyperlipidemia, obesity, smoking, stress, alcohol and substance abuse and other lifestyle conditions : 430-438 (Cerebrovascular disorders including occlusion of cerebral arteries, cerebral hemorrhage, transient ischemia and other cerebrovascular diseases); 401-405 (Hypertensive disease including essential hypertension, hypertensive heart disease, hypertensive kidney disease, and secondary hypertension); 410-417 (Ischemic Heart Disease including heart attack, angina pectoris and other forms of chronic heart disease); 250-251 (Diabetes); 490-496 (Chronic Obstructive Pulmonary Disease including asthma, bronchitis, and emphysema); 290-319 (Mental disorders including substance abuse, anxiety disorders, acute reaction to stress, depression and psychogenic disorders); 720-724 (Musculoskeletal Disorders including carpal tunnel syndrome, knee, neck and back pain, and sprains, strains and tears of tendons, ligaments and muscle); 531-537 (Disorders of the Digestive System including dyspepsia, gastritis, ulcers, acid reflux, diarrhea and bowel disorders). Marathon Health has conducted claims analysis on over 20 businesses and found the range of preventable claims falling into the above categories to be 20%-40% of all claims with an average of 25%.
The Kaiser Health Education Research Trust 2007 Survey of Employer Sponsored Health Benefits reports an average monthly premium of $373 for all health plans. Taking this amount and using the Marathon Health methodology described above we can estimate the average costs and savings per 1,000 employees as follows:
Employees 1,000
Cost per EE per month $373
Total annual medical $4,476,000
Percent preventable 25%
Amount preventable $1,119,000
Total Saved $464,385
Total Saved as % of all claims 10%
Goetzel and colleagues (2005) conducted a meta-analysis of 44 studies investigating the financial impact and return on investment from population health management programs. Interventions for asthma, diabetes, heart disease and depression were included. While results across 44 studies were predictably mixed, many showed positive results. Four studies for programs managing multiple risks had intervention program expenses averaging $230/yr while savings averaged $1,563/yr.
A program called The Contract for Health and Wellness at GlaxoSmithKline employed population health management techniques to reduce health care costs and improve productivity. Stave et al. (2003) reported that the annual savings associated with the program after five years were $613 per EE/Yr.
There are many examples in the literature documenting a positive ROI for worksite population health management programs. An exhaustive compendium can be found in the Cost Benefit Analysis and Report, Health Management Research Center, University of Michigan, 2006.
Verification of projected savings is can be completed by a time series (program onset vs. 15 months post program onset and repeated each year) matched cohort analysis of claims data for preventable claims and/or applying the value of excess cost associated with risk factors, and any reduction of risk factor prevalence in the managed population, as described by Edington (op. cit.).
3. Reduced absenteeism and disability associated with illness and injury. Marathon Health uses the Work Loss Data Institute’s Official Disability Guidelines to assess work day lost days due to illness. For example, the Work Loss Institute’s national database identifies an average of 14 days away from work per year for employees diagnosed with anxiety (ICD-9 codes 300.0-300.02). Marathon Health’s ROI model, and its experience, is that the onsite health coach identifies those employees, through either data mining claims, the screening process or uptake during an acute care visit, and that by working with those employee, the average number of days away can be reduced by 20%. The ODG identifies 3.84 days lost due to illness and disability per worker per year. Marathon Health’s ROI model uses $450/day as an average cost per missed day, translating to a total of $1.7M absenteeism and disability cost per 1,000 workers. Edington (2006) in the Cost Benefit Analysis and Report (op. cit.) documents a savings range of 25%-66% from absenteeism and disability given case management of sick and injured employees. This includes studies by Strakal (1996), Green-McKenzie (1998), and Atcheson (2001). Marathon Health uses a savings capture rate of 20% in its ROI model. This translates to a total savings of $345 per employee per year ($1.7M x 20% = $345,000/yr).
Verification of these savings can be completed by a time series (program onset vs. 15 months post program onset and repeated each year) matched cohort analysis of disability and absenteeism data (if available) and/or using ODG averages based on the number of employees.
Soft Dollar Savings
This category includes two items: presenteeism and turnover. Neither of these items is included in the ROI analysis used to guarantee performance. However, they are significant benefits of population health management and should therefore be understood.
1. Presenteeism is a phenomenon receiving increased attention from employers concerned about performance and productivity. First cousin to absenteeism, presenteeism is defined as an employee being at work but functioning somewhere below 100% capacity. Presenteeism is a complex phenomenon generated by a compound of at least six interactive elements including burnout, physical health, mental health, work distractions, life distractions and entitlement. Cigna Health Care reports an average percent presenteeism score of 9%. Scores range from a low of 3% (healthy) to a high of 58% (high risk and unmanaged chronic disease). These values indicate functioning, on average, at 91% of capacity with some workers functioning at 97% capacity and other functioning as low as 42% of capacity. The average presenteeism payroll cost is estimated at $5, 875 per employee per year. In calculating the value of reduced presenteeism Marathon Health estimates movement an employer’s employee population between high, medium and low risk categories over time. Marathon Health’s experience is that up to 70% of medium and high risk employees engage with the health coach and of that 50% improve by at least one risk category resulting in an estimated average savings from reduced presenteeism of $564 per employee per year.
2. Turnover involves the cost associated with recruiting and training new employees. The estimated cost for employee turnover ranges from $3,500 to $25,000 per employee. The annual turnover rate for wellness program participants of the Canada Life Assurance Company of Toronto was 1.8%, compared to the company-wide average of 18%. British Columbia Hydroelectric’s wellness program participants had an annual turnover rate of 3.5%, compared to a company-wide average of 10.3%. The healthier and happier the work force, the less a company has to spend on hiring and training new personnel. Marathon Health estimates an average savings from reduced turnover costs of $429 per employee per year.
Summary of Average Savings per Employee
Savings ROI
Onsite Primary Care $175 0.5
Reduced medical claims $464 1.4
Disability and absenteeism $345 1.0
Sub-Total Hard Dollar Savings $984 3.0
Presenteeism $564 1.7
Turnover $429 1.3
Sub-Total Soft Dollar Savings $993 3.0
Total Annual Savings per Employee $1,977
Total Annual Cost per Employee $333 5.9
Mr. Demers received his master’s degree in public health from the University of California, Northridge Graduate School of Public Health in 1981. He is director of strategic planning and product development at Marathon Health, Inc., based in Colchester, Vermont. He has worked in health care for 25 years.
HTML-Tags:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>
