How do you compare health insurance plans side by side?

Pull the Summary of Benefits and Coverage for every plan so they are read from the same document, write down a fixed picture of expected care so the inputs do not move between plans, run a quiet-year, moderate-year and bad-year cost scenario for each, then check network, formulary, referral rules and plan-year dates. The cost scenarios narrow the field and those four columns usually separate whatever is left.

Is the plan with the lowest premium the cheapest plan?

Only in a year with almost no care. Premiums are certain and everything else is conditional, which is why premium dominates comparisons and why it misleads. Running premium × 12 alongside premium × 12 plus the full out-of-pocket maximum shows the range a plan puts you in, and a low-premium plan often has the wider range.

What is a Summary of Benefits and Coverage?

A short standardised document every US health plan has to publish, with the same headings in the same order for every insurer. That standardisation is what makes it the practical basis for a side-by-side comparison — a benefits portal summary is built on top of it and tends to omit the details that separate similar plans.

How many plans is it worth comparing?

Two or three is usually where the method stays useful. Beyond that the effort per plan drops and the comparison tends to get shallower rather than more thorough, so narrowing on one hard requirement first — a named clinician being in network, for instance — generally beats running a thin comparison across six.

Does this apply to car or home insurance too?

No. This page is about health plans, which are compared on premium, deductible, cost sharing, an out-of-pocket ceiling, a provider network and a drug formulary. Auto, home, life and disability policies are structured around different terms entirely, and the method here does not transfer to them.

How to Compare Health Insurance Plans: A Side-by-Side Method That Ends in a Decision

Why plan comparison stalls

A comparison that never ends usually has one of three problems, and none of them are solved by adding another column.

The plans are being read from different documents. One from a glossy benefits portal summary, one from a PDF, one from memory of last year. Those are not the same source and they do not carry the same detail.
The inputs move between plans. Plan A gets compared against an optimistic year and Plan B against a pessimistic one, so the arithmetic is measuring the mood, not the plan.
There is no definition of winning. Without deciding in advance whether the goal is the lowest likely spend, the lowest worst case, or keeping a particular clinician, every column looks equally important and the choice never resolves.

The fix for all three is order: same document, fixed inputs, stated goal — and only then arithmetic.

Step 1 — Get the same document for every plan

In the US, every health plan has to publish a Summary of Benefits and Coverage (SBC) in a standardised format, which is what makes it the one document that compares cleanly across insurers. A few pages, the same headings, the same order, every time.

A benefits-portal summary card is a marketing surface built on top of that document, and it is where the details that separate two similar plans tend to go missing. Gathering the SBC for every plan under consideration — not for two of the three — is the step that makes everything after it comparable.

Worth having open alongside it: the plan's formulary (its drug tier list) and its provider directory, both of which are usually separate documents and neither of which appears in the SBC.

Step 2 — Fix the inputs before comparing anything

The comparison is between plans, so everything that is not a plan has to be held constant. Writing this down once, before opening any arithmetic, is what stops the numbers from drifting:

Who is covered. One person, a couple, a family — and whether a family deductible is aggregate or embedded, which changes the maths considerably for a family where one person uses most of the care.
Regular prescriptions. The actual list, by name, including anything taken seasonally.
Clinicians and facilities that matter. Named — a specific physician, a specific hospital — not "my doctor".
Care already expected this year. A planned procedure, a pregnancy, ongoing therapy, regular specialist visits.
Employer contributions. Any HSA or HRA contribution attached to a specific plan is real money and is routinely left out of the comparison.

That list is the test case. Every plan gets run against the same one.

Step 3 — Run three cost scenarios per plan

One number per plan hides the shape of the plan. Three numbers show it:

Quiet year. Premium × 12, plus routine care that happens reliably — a few visits, regular prescriptions at their tier cost.
Moderate year. The quiet year, plus one unplanned thing: an injury, a short course of specialist care, a scan.
Bad year. Premium × 12, plus the full out-of-pocket maximum. This is the worst realistic case for covered in-network care, and it is the number that says whether a plan is survivable rather than merely cheap.

Then subtract any employer HSA or HRA contribution from that plan's totals, since it offsets real spending.

The pattern that usually emerges: a low-premium, high-deductible plan wins the quiet year and loses the bad one, and a richer plan does the reverse. Seeing the spread between a plan's best and worst year is more informative than any single figure — a narrow spread is predictability, a wide one is a bet.

If the term arithmetic itself is the unfamiliar part, what each of those four numbers actually does is worth reading first.

Step 4 — The four columns that decide it anyway

Plans that look close on cost are usually separated here, and none of these show up in a premium comparison.

Network

A lower premium sometimes buys a narrower network. The check that matters is each named clinician and the nearest hospital against this specific plan, not against the insurer — a single insurer can run several networks, and being "in network with the company" is not the same thing.

Formulary

Drug coverage runs on its own tier list and sometimes its own deductible. Looking up each regular prescription on each plan's formulary is where two otherwise identical plans often turn out to differ by a large amount, and it is also where a prescription can turn out to need prior authorisation or step therapy under one plan and not another. Anything that surfaces there splits into two different questions for two different people: whether an alternative exists at all is a clinical question for the prescriber or a pharmacist, and whether a given plan will cover it is a coverage question for the insurer or a licensed broker. Neither question is answerable from a spreadsheet.

Referrals and prior authorisation

Some plans require a primary care referral before a specialist visit is covered. For someone already seeing specialists, that is a change to how care actually happens, not just a cost line.

Plan year

Deductibles and out-of-pocket maximums reset on the plan year, which is not always January. Switching mid-year can restart an accumulator that was nearly satisfied — an expensive surprise that no cost scenario catches unless the dates are checked.

Step 5 — Make the call, and write down why

A comparison ends when a sentence can be written: this plan, because the goal was the lowest bad year and it costs $1,400 less there, and both named clinicians are in network.

Writing that sentence down does two things. It forces the goal to have been chosen rather than assumed, and it leaves next year's version of this task a starting point instead of a blank page — along with a note of anything that was uncertain, which is the list worth taking to the insurer or a licensed broker before the deadline.

Questions worth resolving before enrolment closes:

Is the family deductible aggregate or embedded?
Which services are covered before the deductible is met?
Is there a separate out-of-network out-of-pocket maximum, and does out-of-network spending count toward anything?
Is each named clinician and facility in network for this specific plan?
What tier is each regular prescription on, and is there a separate drug deductible or a prior-authorisation requirement?
Does a referral gate specialist visits?
When does the plan year start, and when do the accumulators reset?
What does the employer contribute to an HSA or HRA on each plan?

Want the side-by-side done for you? The Insurance Comparison Wizard is a prompt pack that walks two or three plans through this same structure one at a time — the three cost scenarios, the network and formulary columns, and a comparison table you can look at and actually choose from. $8.99 · instant download →

Common questions

How do you compare health insurance plans side by side?: Pull the Summary of Benefits and Coverage for every plan so they are read from the same document, write down a fixed picture of expected care so the inputs do not move between plans, run a quiet-year, moderate-year and bad-year cost scenario for each, then check network, formulary, referral rules and plan-year dates. The cost scenarios narrow the field and those four columns usually separate whatever is left.
Is the plan with the lowest premium the cheapest plan?: Only in a year with almost no care. Premiums are certain and everything else is conditional, which is why premium dominates comparisons and why it misleads. Running premium × 12 alongside premium × 12 plus the full out-of-pocket maximum shows the range a plan puts you in, and a low-premium plan often has the wider range.
What is a Summary of Benefits and Coverage?: A short standardised document every US health plan has to publish, with the same headings in the same order for every insurer. That standardisation is what makes it the practical basis for a side-by-side comparison — a benefits portal summary is built on top of it and tends to omit the details that separate similar plans.
How many plans is it worth comparing?: Two or three is usually where the method stays useful. Beyond that the effort per plan drops and the comparison tends to get shallower rather than more thorough, so narrowing on one hard requirement first — a named clinician being in network, for instance — generally beats running a thin comparison across six.
Does this apply to car or home insurance too?: No. This page is about health plans, which are compared on premium, deductible, cost sharing, an out-of-pocket ceiling, a provider network and a drug formulary. Auto, home, life and disability policies are structured around different terms entirely, and the method here does not transfer to them.

General educational information about how health plan comparison works. This is not medical advice, not legal or tax advice, and not individualised insurance advice, and nothing here is a recommendation of any plan, insurer or product. Following a method on this page is not a substitute for professional advice, and it does not make an outcome correct — coverage questions belong to the insurer, the plan administrator or a licensed insurance broker, who are the only parties able to speak to what a specific plan covers. A pharmacist can speak to medication questions, not to what a plan will pay. This page covers health plans only — auto, home, life and disability insurance are compared on entirely different terms and none of this transfers to them. The Summary of Benefits and Coverage for a specific plan is the authority, not this page; plan documents differ and change every year.