What a Plan Must Cover With No Cost Sharing

Answer first: the list is not written by the plan, and not by Congress either

Federal law requires most health plans and insurers to cover a defined set of preventive services and to charge nothing for them. The statute, 42 U.S.C. 300gg-13, says a plan and an issuer "shall, at a minimum provide coverage for and shall not impose any cost sharing requirements" for four categories of item and service. The implementing rule, 45 CFR 147.130, repeats that and names the three prohibited charges explicitly: "any cost-sharing requirements (such as a copayment, coinsurance, or a deductible)."

What makes this provision unusual is that the law does not itself name the services. It points at outside bodies and adopts whatever those bodies currently recommend. The four categories are: evidence-based items or services rated A or B in the current recommendations of the United States Preventive Services Task Force with respect to the individual involved; immunizations for routine use recommended by the Advisory Committee on Immunization Practices of the Centers for Disease Control and Prevention; for infants, children and adolescents, evidence-informed preventive care and screenings in comprehensive guidelines supported by the Health Resources and Services Administration; and, with respect to women, additional preventive care and screenings in HRSA-supported comprehensive guidelines. A fifth category in the regulation covers qualifying coronavirus preventive services.

So the boundary of "free" is set by expert panels, and it moves when those panels move. A service is inside the obligation because a panel rated it, not because a legislature listed it or a plan agreed to it. That is the single fact that explains most of the rule's behavior, including the timing rules described below.

A recommendation does not take effect the day it is issued

The statute directs the Secretary to establish a minimum interval between the date a recommendation or guideline is issued and the plan year in which coverage becomes required, and it sets a floor: "The interval described in paragraph (1) shall not be less than 1 year."

The regulation implements that. A plan or issuer must provide the coverage for plan years — in the individual market, policy years — that begin on or after September 23, 2010, or, if later, for plan years that begin on or after the date that is one year after the date the recommendation or guideline is issued.

The practical effect is a lag that can run well past twelve months. A recommendation issued in the middle of a calendar year does not attach to a plan year that begins less than a year later; it attaches to the next one that begins on or after the anniversary. Two people whose plan years start in different months can therefore face the same new recommendation on different dates, with neither plan doing anything wrong.

The rule also protects against a mid-year removal. Where a plan is required to cover an item or service on the first day of a plan or policy year, it must provide that coverage through the last day of the year "even if the recommendation or guideline changes or is no longer described in paragraph (a)(1) of this section" during the year.

The office visit is a separate question from the service

The most common surprise in this area is a bill arriving after a visit a person understood to be free. The regulation addresses that directly, and the answer turns on billing mechanics rather than on the nature of the service.

Where a covered preventive item or service is billed separately, or tracked as individual encounter data separately, from an office visit, the plan may impose cost sharing on the office visit. Where it is not billed separately and the primary purpose of the visit is the delivery of that item or service, the plan may not impose cost sharing on the visit. Where it is not billed separately and the primary purpose of the visit is not the delivery of that item or service, the plan may impose cost sharing on the visit.

The rule's own worked examples make the distinction concrete. In one, a person is screened for cholesterol abnormalities during a visit and the provider bills separately for the visit and for the laboratory work; the plan may not charge for the separately billed screening, but may charge for the office visit. In another, a person visits a provider to discuss recurring abdominal pain and receives a blood pressure screening during that visit, billed as an office visit; because the primary purpose of the visit was not to deliver the screening, the plan may impose a cost-sharing requirement for the office visit charge.

There is a further limit that catches people even more often. The regulation states that a plan may impose cost-sharing requirements "for a treatment not described in paragraph (a)(1) of this section, even if the treatment results from an item or service described in paragraph (a)(1)." Its example: a person screened for cholesterol is diagnosed with hyperlipidemia and prescribed a course of treatment not itself on the list. Because the treatment is not on the list, the plan is not prohibited from imposing cost sharing on it. The screening being free says nothing about what follows the screening.

Network rules survive, with one exception in the text

Nothing in the section requires a plan or issuer that has a network of providers to provide benefits for a listed preventive service delivered by an out-of-network provider, and nothing precludes it from imposing cost sharing on such a service delivered out of network. The zero-cost-sharing obligation is, as a general matter, a rule about in-network care.

The regulation carves out one situation. If a plan or issuer "does not have in its network a provider who can provide an item or service described in paragraph (a)(1)," the plan or issuer must cover the item or service when performed by an out-of-network provider and may not impose cost sharing on it. A separate paragraph requires coverage of qualifying coronavirus preventive services regardless of network status, with out-of-network reimbursement in an amount that is reasonable as determined in comparison to prevailing market rates.

The two rules together mean that a network directory is not a side issue for preventive coverage. Whether a network contains a provider who can deliver a specific listed service is the fact that decides whether the carve-out applies at all.

What the plan still gets to decide

Two provisions preserve substantial plan discretion inside an obligation that reads, at first, as absolute.

The first is medical management. Nothing prevents a plan or issuer "from using reasonable medical management techniques to determine the frequency, method, treatment, or setting for an item or service described in paragraph (a)(1) of this section to the extent not specified in the relevant recommendation or guideline." The qualifier does the work: where the underlying recommendation specifies a frequency, method, treatment or setting, the plan follows it; where the recommendation is silent, the plan may rely on the relevant clinical evidence base and established reasonable medical management techniques to fill the gap. Coverage without cost sharing is therefore not the same as coverage without conditions.

The second is the boundary of the list itself. The regulation states that nothing in the section prohibits a plan or issuer from covering items and services in addition to those recommended by the Task Force, the Advisory Committee or the HRSA guidelines, "or from denying coverage for items and services that are not recommended" by them. The statute says the same thing in its own words. A service outside the list gets no protection from this provision, whatever its clinical merit — the plan's ordinary coverage terms govern it.

One further limit applies to the whole section: the coverage requirement in paragraph (a)(1) is stated as subject to three other sections of the same part, and the section as a whole sits alongside § 147.140, which governs grandfathered plans. Whether a particular plan is grandfathered is a fact published in that plan's own documents.

Reading a real bill against the rule

The rule supplies a short set of questions that a bill either answers or does not. Was the item or service on one of the four referenced lists at the start of this plan year — not now, and not when it was recommended, but at the start of the plan year. Was it billed separately from the office visit. If it was not billed separately, what was the primary purpose of the visit. Was the provider in network, and if not, does the plan have any in-network provider who could have delivered that specific service. And is the charge for the preventive service itself or for a treatment that followed it.

Those five questions are not a claim about who is right in any particular dispute. They are the structure the regulation actually uses, which is why a charge that looks like a violation often is not one, and why a charge that looks routine sometimes is.

It is worth being precise about what this provision is and is not. It is a floor on cost sharing for a defined, externally maintained list. It is not a general rule that prevention is free, it is not a guarantee that any particular condition is screened for, and it does not reach the treatment that a screening leads to. Those are governed by the plan's own benefit terms, which are documents a plan is separately required to make available.