Health Benefits

10 Steps to Assess Vendor Programs That Promise to Lower Medical Costs

CFOs should carefully assess their vendor choices to decrease the cost curve and improve the quality of care for employees.
Jeff Levin-Scherz, MDJanuary 19, 2023
10 Steps to Assess Vendor Programs That Promise to Lower Medical Costs
Photo: Getty Images

With rising costs and a potential economic slowdown, employers are under enormous pressure to control spending on employee health benefits. Employers have been accustomed to a decade of medical trend rates under 5% annually, but the increase in total medical costs will likely exceed 6% in 2023, and climb higher in 2024.


  Jeff Levin-Scherz

In response, medical carriers, pharmacy benefit managers, and third-party vendors are offering a multitude of programs that they say will lower medical costs. The volume of programs and the abundance of statistics put forward to demonstrate effectiveness can be overwhelming. Here are key considerations for chief financial officers as their companies sift through offerings to identify which will benefit their employees while effectively addressing medical inflation.

1. Offer Programs That Target Members With Truly Modifiable Costs  

Five percent of adults in employer-sponsored health plans represent 52% of costs, and the top 1% represents 23% of costs. On the other hand, the bottom 50% of the population represents a mere 3% of total medical costs. Therefore, employers should focus on programs that carefully target plan members who are likely to have high costs which interventions can plausibly decrease. The choice of target conditions is important, too. Programs aimed at those with prevalent chronic conditions like diabetes that can be modified will reap the most savings for employers.

2. Vendors Should Identify Members Who Will Benefit From Their Program

Medical carriers and vendors should offer tailored identification programs that don’t overlook eligible members but avoid identifying members unlikely to benefit from the intervention. For instance, an intensive diabetes management program that gave out connected glucometers to diabetics not on insulin was unable to demonstrate improvement in diabetes measures, since so many participants were in good control even before the intervention.

Effective programs generally use a combination of medical and pharmacy claims data to identify potential participants; pharmacy data is especially valuable because data on who has filled prescriptions is available immediately. Identification programs should be inclusive and avoid algorithms that incorporate racial or other biases.

3. Have a Realistic Plan to Engage Members

Programs to improve care and lower medical costs only work if members enroll and actively participate. Nonetheless, many well-intentioned programs make little difference because it’s difficult to gain member participation.

A center of excellence (COE) program could direct members to orthopedists with the best outcomes, the fewest complications, and the lowest total cost of care for hip or knee replacements. But the program will only improve quality and lower cost if members actually use those providers. Member engagement depends on effective employer and vendor communication as well as infrastructure to make it easy for qualified members to enroll. Vendors and employers can deploy insights from behavioral economics to increase member engagement. Some programs, like COEs, work better when enrollment is mandatory or when the financial incentive is large.

4. Read the Fine Print of Studies That Promise Cost Savings

The ideal evidence that a program would lower medical costs comes from randomized, double-blind studies. But these studies are complex, expensive, and challenging to perform, so the best evidence available will often be observational studies. These should answer the questions “Is it likely that the program itself lowered costs?” and “Is the population studied similar to our membership?”

Some health plans or vendors produce white papers with validation from a third party, but employers should read these critically as the third party is being paid by the vendor and may be biased toward positive results. Claims of cost savings from internal data can be suggestive, but data can be cherrypicked and companies often make evaluation errors.

Data on “return on investment” should account for vendor costs or the costs of substitute care, should not assume an unrealistically high trend, and should do appropriate discounting to recognize the time value of money. Employers should also be cautious interpreting program performance data that comes from small samples or short observation periods.

5. Examine How the Intervention Will Integrate With Other Employer Programs 

Many employers already offer various health care programs, and adding overlapping programs increases costs, confuses employees, and decreases engagement. In an era where some employers work with 10 or more separate vendors, companies must encourage lines of communication across vendors to provide the best member experience.

Those with serious illnesses often have valuable relationships with providers in their local community, and new programs should support these relationships. For example, programs that adjust medications should be designed to optimize communication, coordination, and collaboration with local providers. The best vendors integrate with existing programs and in-person providers to offer a seamless experience for members.

6. Establish That the Vendor has the Resources to Effectively Implement the Program

Vendors require substantial capital to invest in program design, information technology, and hiring and training staff. Many of these companies lose money for the initial portion of each new contract, so employers should determine that the vendor has adequate financial and managerial resources to endure.

Companies that provide services for only a handful of small clients are less likely to succeed in a new installation for a large employer, and companies that have had success selling seasons might struggle to implement many new clients simultaneously. Assessing financial stability is of increasing importance as venture capital firms have decreased available funding for companies that are still in a high-growth, money-losing phase.

7. Consider What Channel to Use to Purchase Each Program

Employers and health plan sponsors can often purchase programs directly from a vendor, from their medical carrier, or from their pharmacy benefit manager (PBM). Purchasing vendor programs through the medical carrier or PBM can ease the burden of contracting. The carrier or PBM has already been vetted for information security and financial stability and can offer standard terms. The carrier or PBM can also vouch for the vendor’s financial and technical capabilities, and purchasing vendor programs through an intermediary can expedite appropriate data sharing.

However, larger purchasers can gain more program customization and might be able to negotiate preferred terms directly with the vendor. Further, contracting through a carrier or PBM can make it more difficult to change that carrier or PBM for financial or performance reasons. 

8. Agree On Evaluation Methodology Before Implementation

Agreeing about how a program will be evaluated in advance allows the vendor to collect actionable, targeted data, and guards against relying on selective data which can be misleading. Vendors should report agreed-upon metrics regularly, and a third-party review of a vendor’s evaluation methodology can increase the integrity of the assessment. Employers may need to integrate their claims data with vendor activities to do a proper evaluation of return on investment.

Beware of biases frequently seen in the evaluation of vendor programs, including:

  • Selection bias, where those getting the intervention were more likely to succeed even without intervention. For instance, individuals who voluntarily participate in a program are often more motivated to manage their condition than those who refuse.

  • Regression to the mean, where a program enrolls very sick people, and takes credit for their subsequent lower costs which would have been achieved without the program.

  • Survival bias, where companies report only the result of those who complete a program, who are more likely to succeed than those who dropped out.  

9. Incorporate Performance Guarantees Into Vendor Contracts

Vendors offer performance guarantees to demonstrate that they believe in their program, and employers who obtain performance guarantees are more likely to get good service from vendors. Well-designed performance guarantees around cost savings usually require evaluation of the claims experience of the entire population to overcome selection bias; truly effective programs should lower costs for the entire population. Vendor “book of business” outcome statistics might be appropriate for employers with populations too small to measure credibly. Member satisfaction can be measured with surveys, and net promoter score is especially good at identifying world-class vendors.

Structural performance guarantees, such as staff hired and technology in place on time, and process performance guarantees, such as the number of enrollees and continued engagement, help focus vendor attention. If the program elements are not in place or if members don’t enroll, the program cannot lower costs. Performance guarantees should be straightforward to reconcile; complex performance guarantees can distract vendor attention from providing their intervention.

10. Eliminate Poorly Performing Programs

Many employers and health plan purchasers are already investing in multiple programs to address high-risk members. But these programs cannot improve health or save money if they have few enrollees or limited engagement. Terminating or reconfiguring underperforming programs can save substantial dollars, as well as lessen the burden on staff to monitor contracts and coordinate across vendors.

Look for programs with low utilization, or where members leave quickly or fail to “graduate.” Many programs are paid on a “per employee per month” basis, and companies can be unpleasantly surprised when they calculate the cost per member who actively participates in the program. Companies can also be surprised that some programs continue to assess fees long after a member stops participating. Increasingly, the costs of clinical interventions are paid through medical claims, and companies should incorporate all related claims costs when they assess the value of the program.

Savings from eliminating ineffective programs can be repurposed to support more effective programs or to lower the overall cost of care.

The coming years of high medical inflation will increase the pressure on health care purchasers to effectively control costs and could worsen the health status of employees and their families. Companies that carefully assess their vendor choices can optimize their portfolio of programs to improve the quality of care delivered, decrease the cost curve, and offer a better care experience for employees.

Jeff Levin-Scherz, MD, MBA, is a managing director and population health leader of the North American health and benefits practice at WTW. He is an assistant professor at the Harvard TH Chan School of Public Health. Follow him on Twitter.