In KPI Ninja’s “School District Claims Analysis” comparing claims costs under the Nextera plan and the competing fee for service (Choice) plan, the “Analyst” overlooked two major differences between the plans in how the “School District” pays “Claims“.
- Nextera members pay post-deductible coinsurance at a 20% rate and the district pays an 80% share. But for the exact same claim by a member of the competing fee for service plan (Choice member), the split is 10% and 90%.
- While both cohorts have the same $2,000 deductible, in theory, only the Choice plan members have access to an employer-paid $750 health reimbursement account that provides first dollar claims coverage, delaying the onset of the deductible and effectively reducing it to $1,250 dollar.
When two claims for exactly the same service rendered can draw different employer payments for Nextera members and Choice members, that difference payment has nothing whatever to do with Nextera’s effectiveness. Yet, the different effective rates at which claims are paid obviously have substantial effects on the total claims amounts for each group. Accordingly, a large part of any difference between the totals for the two groups is the result of SVVSD benefit design, not the result of anything Nextera does that reduces costs.
To accurately reflect only the savings attributable to Nextera, it is necessary to normalize the district’s average payment rate between the two populations. KPI Ninja did not see the need to do this.
Our method for doing this was to estimate and compare the actuarial values of the medical coverage in the two plans using the publicly available CMS actuarial value calculator developed for the Affordable Care Act’s individual market coverage.[**] We arrived at a downward adjustment from Choice plan total employer costs given by a factor of 0.905.
Correcting just this one oversight by KPI Ninja makes a difference of $311 in the overall savings claims, deflating Nextera’s brag by over one-third. See the computations here.
Every one of over two dozen claim cost comparisons in the KPI Ninja report needs this same adjustment (plus others discussed both in other posts at this blog and in the Nextera Manuscript that can be viewed through a menu item above).
A happy by-product of understanding how claims are divided between the district and its employees is that it essentially resolves KPI Ninja’s concerns about not having been provided employee payment data. Once we have put employer payments for the two groups on a normalized scale, the details of how the employer and employee divided the costs of particular claims among themselves is of little or no value in assessing Nextera’s aggregate contribution to overall savings. In other words employee cost-sharing issues need not impair our ability to evaluate Nextera’s performance.
[**] The paired plans have AV of 78.0% (Nextera plan) and 86.2% (Choice plan) , giving a ratio of 0.905. That ratio represents a conservative adjustment of the employer payments reported by KPI Ninja for two reasons. First, the computed AVs include both medical and pharmaceutical claims. For pharmaceutical claims, however, the cost sharing is identical, so pharmaceutical claims (about one-fifth of all claims) play no role in generating the difference. All of the 10.5% overall AV difference is generated from the difference between the paired medical claims, which must therefore be appreciably larger than 10.5%. Also, the study data included claims from two plan years, 2018 and 2019. The 80:20 coinsurance split for Nextera patients applied to 2018. For the 2019 calendar year, for Nextera patients, the district replaced coinsurance with copayments for several selected services, including e.g., $200 for advanced imaging and ED visits and $60 for specialist visits. See SVVSD 2019 benefits guide. We determined the actuarial value of both the 2018 Nextera plan and the 2019 Nextera plan. For 2019, the Nextera plan had an ever lower actuarial value, i.e, the net effect of the 2019 changes was to decrease overall employer payments for Nextera members.
DPC Referee 