The Two-Plan Health Cost Teardown, Two SBCs to a Yearly Number

You have two health plans in front of you - two job offers, or your current plan against a new one - and you need to know which one costs you less over a year before you sign. This guide is for candidates at the offer stage who are tired of guides that stop at "build a table and request the SBC." I carry two real plans from their Summary of Benefits and Coverage all the way through premiums, deductible, coinsurance, the out-of-pocket maximum, and employer HSA money, to three usage-weighted annual numbers and one decision.

This picks up where a whole-offer decision framework leaves off. That framework treats the health plan as a single line item. It is not a line item. It is a figure you have to compute, and the cheaper plan on premium is routinely the more expensive plan on total cost.

What the SBC gives you and what it hides

The Summary of Benefits and Coverage (SBC) is the standardized document an issuer must provide for any private plan, and it is built so you can compare costs and coverage apples-to-apples between plans based on price, benefits, and features. It gives you the cost-sharing structure. It does not give you the premium.

Here is what you read off it, plan by plan:

• Deductible - how much you spend before the plan starts paying, except for free preventive services. Listed for individual and family.
• Out-of-pocket maximum - the most you spend for covered services in a year, after which the plan pays 100 percent.
• Coinsurance - the percentage you pay after the deductible, per service.
• Copayments - the flat amounts for common events: an office visit, an MRI or CAT scan, a hospital stay, and prescription drug tiers.
• Network rules - in-network and out-of-network provider information, and whether out-of-network care is covered at all.

What the SBC deliberately leaves out is the premium. Information about the cost of the plan is provided separately, because premiums vary by income, age, and tobacco use. So one of your first moves is to pull the per-paycheck worker premium from HR and never look for it on the SBC.

The reason this matters: HealthCare.gov's documented procedure is to compare estimated total yearly costs for each plan, not just the premium. Total cost is premium plus deductible plus out-of-pocket costs plus any copayments and coinsurance. The deductible and coinsurance are deferred premium - money you only pay if you use care. Which means the ranking depends on how much care you use, and that is the whole problem this guide solves.

The worked example: Plan A versus Plan B

Let me carry two concrete plans through to the end so you can follow along on your own case. I will keep the numbers round and the structure real.

Plan A - a PPO. Wide network, out-of-network covered, moderate-to-high premium. Say the worker premium is $120 per paycheck, 26 pay periods, so $3,120 a year. Individual deductible $1,000, coinsurance 20 percent after the deductible, out-of-pocket maximum $5,000. No employer HSA money, because a PPO at a standard deductible is not HSA-eligible.

Plan B - an HDHP. A high-deductible health plan is a cost structure, not a network type; most HDHPs are themselves either PPOs or HMOs. Say the worker premium is $55 per paycheck, so $1,430 a year. Individual deductible $2,500, coinsurance 20 percent, out-of-pocket maximum $5,000. It is HSA-eligible and the employer seeds $1,500 into the HSA.

On premium alone, Plan B wins by a mile: $1,430 against $3,120. That is the trap in Failure Mode 1. The premium rank and the total-cost rank often disagree, and I am about to show the ranks cross.

Netting out the employer money first

Before any usage, subtract employer account dollars. Plan A gets nothing. Plan B gets a $1,500 HSA seed. The employer HSA seed is the cleanest subtraction you can make, because the IRS limit is a combined cap on every dollar entering your HSA, yours and your employer's. For 2026 that cap is $4,400 self-only and $8,750 family, so a $1,500 seed sits well inside it and is simply money you did not have to spend.

• Plan A net premium: $3,120 - $0 = $3,120
• Plan B net premium: $1,430 - $1,500 = -$70

Yes, Plan B's net premium goes slightly negative, because the employer is handing you more in HSA dollars than you pay in premium. That is not a trick; it is why an HDHP that looks expensive can flip once the seed is counted (Failure Mode 3). One rule: count the seed the same way on both plans. Do not model the full tax shield on one and a flat seed on the other, or you will double-count (Failure Mode 6). Pick a method and apply it identically.

Building low, expected, and high usage

Three scenarios, each a different volume of care, each run through the same machinery: deductible first, then coinsurance, capped at the out-of-pocket maximum. Build the scenarios from your own prior-year claims, not national averages (Failure Mode 7). The averages below are only to sanity-check that your assumptions are not wild.

Low usage - preventive only plus one or two sick visits. Preventive is free. Say $300 of billed care, all below the deductible, so you pay the full $300 on either plan.

Expected usage - roughly three office visits plus one or two maintenance prescriptions. This matches the population anchors: the US office-visit rate is about 3.2 per person per year, and roughly 50 percent of Americans took a prescription in the last 30 days. Say $3,000 of billed care.

• Plan A (deductible $1,000): you pay the first $1,000, then 20 percent of the remaining $2,000 = $400. Cost sharing $1,400.
• Plan B (deductible $2,500): you pay the first $2,500, then 20 percent of the remaining $500 = $100. Cost sharing $2,600.

High usage - a hospitalization or surgery that drives spend past the out-of-pocket maximum. Both plans cap at $5,000, so your cost sharing is exactly the OOP max on each: $5,000. This is the point readers over-fear. The out-of-pocket max bounds your downside; once you hit it the plan pays 100 percent, so the high case is computable, not open-ended. Just remember the cap excludes premiums, balance billing, non-network costs, and non-covered services (Failure Mode 5), so keep those outside this number.

The out-of-pocket max turns the scariest scenario into the easiest one to compute: net premium plus the cap.

Carrying each scenario to an annual number

Add net annual premium to each scenario's cost sharing. That gives the two-plan by three-scenario grid that is the entire point of the exercise.

| Scenario | Plan A (PPO) | Plan B (HDHP) | Cheaper | | Low | $3,120 + $300 = $3,420 | -$70 + $300 = $230 | Plan B | | Expected | $3,120 + $1,400 = $4,520 | -$70 + $2,600 = $2,530 | Plan B | | High | $3,120 + $5,000 = $8,120 | -$70 + $5,000 = $4,930 | Plan B |

In this worked example Plan B wins all three, because the $1,500 seed is large relative to the premium gap and the out-of-pocket maxes match. That is a real outcome, not a rigged one - but do not generalize it. Change the seed to $500, or make Plan B's OOP max $8,500 instead of $5,000, and the high-case winner can flip to Plan A. The grid is the deliverable precisely because it exposes where the crossover sits for your numbers.

The procedure, start to finish

Here is the full run as a sequence you can do with your hands. Budget about two hours across all eight steps.

Network type is a hidden cost gate

Before you trust any grid, confirm your providers are actually covered, because an uncovered provider is an unbounded cost the grid never shows. The plan type tells you the rules.

| Plan type | Network | Out-of-network | Referral needed | | PPO | Wide | Covered | No | | HMO | Limited | Not covered | Yes, via PCP | | EPO | Limited | Not covered | No | | HDHP | Varies (PPO or HMO) | Depends on base | Depends on base |

The failure here is quiet (Failure Mode 4). On an HMO or EPO you generally pay all costs out-of-pocket for out-of-network or unreferred care unless it is a true emergency. That spending does not count toward the out-of-pocket max either, so it blows past your high case entirely. The check is mechanical: list your must-have providers and confirm each one's status on each plan before you rank.

There is also an optionality lesson. PPOs are the most common employer plan - 46 percent of enrollments - and the most expensive. PPO optionality is the most-paid-for, least-used feature. An EPO covers in-network care only, like an HMO, but needs no PCP or referrals, like a PPO. If every provider you need is in-network, an EPO captures most of the PPO savings without the referral friction. You only pay the PPO premium if you genuinely need out-of-network reach.

How this comparison goes wrong

Most bad plan decisions come from one of seven predictable errors. Each has a tell and a check.

• Comparing premiums only. The "cheaper" plan has the low premium but a deductible that erases the gap in a high-use year. Check: always carry to the high scenario, capped at the OOP max.
• Reading premium off the SBC. It is not there; the cost of the plan is provided separately. Check: confirm the per-paycheck figure with HR.
• Ignoring employer HSA or FSA money. An HDHP looks expensive until a $1,500 seed flips it. Check: subtract employer dollars before ranking.
• Assuming out-of-network care counts. On HMO and EPO it generally does not. Check: verify each provider's network status per plan.
• Forgetting what the OOP max excludes. It excludes premiums, balance billing, non-network costs, and non-covered services. Check: keep those outside the capped figure.
• Double-counting the HSA benefit. Counting both the employer seed and the full tax shield overstates savings. Check: pick one method, apply it to both plans identically.
• Using national averages as your inputs. Averaging to three visits hides a chronic condition needing a monthly specialist. Check: use your own prior-year claims, not the CDC rate.

The single most common of these is comparing premiums only. Your own numbers for scale: the 2025 average worker contribution was $1,440 for single coverage and $6,850 for family, against total premiums of $9,325 and $26,993. The employer carries most of the premium, which is exactly why a premium-only comparison tells you so little about your own out-of-pocket exposure.

| Coverage | Total premium | Worker share $ | Worker share % | | Single | $9,325 | $1,440 | 16% | | Family | $26,993 | $6,850 | 26% |

Getting the numbers a plan document won't give you

Two inputs routinely stall this teardown: the exact employer HSA seed, and whether your specific providers are in-network. Neither is always printed in the materials a candidate receives pre-signature. The fix is to ask a human at the company. For a US offer that is easy. In Refolk's index of professional profiles there are 16,034 US benefits and total-rewards contacts - Benefits Manager, Benefits Specialist, and Total Rewards Manager titles - any of whom could send you an SBC or confirm the seed. The same index holds only 383 equivalent UK contacts, roughly 42 times fewer, so the "ask a human" step is far easier for US offers than cross-border ones.

| Segment | Count | Vs US total | | US benefits/total-rewards contacts | 16,034 | 100% | | US Director/VP benefits leaders | 1,049 | 6.5% | | UK benefits/total-rewards contacts | 383 | 2.4% of US |

The most useful person to reach is often not the recruiter. A total rewards specialist can confirm the 2026 employer HSA seed and any FSA match directly, which closes the one gap that most changes your net-premium line. Refolk turns a plain-language ask into that contact list, so you can request the SBC and the seed figure from someone who actually owns the plan rather than guessing from a benefits brochure.

If you are weighing a current plan against a new one and already have last year's explanation-of-benefits statements, you have the strongest scenario inputs there are: your actual visits, prescriptions, and claims. Use them. The CDC rate of 3.2 visits per person per year and the figure that 12.8 percent of Americans take five or more medications are only there to tell you whether your own assumptions are plausible.

Before you call the decision final

Run this check before you sign. It catches the errors that quietly reverse a ranking.

Keeping the comparison current

The limits in this teardown move on a schedule, so re-check them rather than memorizing a value. The IRS resets HSA and FSA limits each year: for 2026 the HSA cap is $4,400 self-only and $8,750 family, and the health FSA limit is $3,400 with a $680 carryover. The ACA out-of-pocket cap also moves - $10,600 individual and $21,200 family for 2026, rising to $12,000 and $24,000 the following year. When you re-run this next open enrollment, pull the current-year limits first, because an employer seed that was inside the cap one year can crowd your own contributions the next.

| Account | 2025 | 2026 | Change | | HSA self-only | $4,300 | $4,400 | +$100 | | HSA family | $8,550 | $8,750 | +$200 | | Health FSA | $3,300 | $3,400 | +$100 | | Dependent-care FSA | $5,000 | $7,500 | +$2,500 |

One honest limit of this method: sources disagree on how to value the HSA. Some model only the employer seed as a flat subtraction; others add the tax shield on your own contributions. I subtract the seed flatly because it is unambiguous and applies identically to both plans. If your marginal tax rate is high and you will max the HSA, the tax shield is real money - just model it on both plans or not at all. The decision you write down should name which method you used, so that next year's you can reproduce the result and see exactly where the crossover moved.