Top

In-Network vs Out-of-Network

In-network vs out-of-network describes whether a provider has a contracted rate and terms with a health plan or bills as a non-contracted provider. Network status drives allowed amounts, patient cost share, authorization rules, and how claims adjudicate.

Kathryn Thompson
Reviewed by Kathryn Thompson · Updated September 2026

What it means

What in-network vs out-of-network means

Health plans build a network of providers who sign contracts that set:

  • Allowed amounts and fee schedules
  • Patient cost share (copays, coinsurance, deductibles)
  • Authorization and referral rules
  • Billing rules and where claims are sent

An in-network provider is contracted with the plan under a specific tax ID and NPI combination, sometimes also tied to a site address and specific service types. The plan treats services from that provider as covered under in-network benefits.

An out-of-network provider has no contract for that plan and tax ID / NPI combination. The plan may:

  • Pay a lower allowed amount based on "usual and customary" or some OON schedule
  • Apply a separate, higher OON deductible and coinsurance
  • Not cover OON services at all, especially for HMO or Medicaid managed care
  • Allow balance billing to the patient, subject to state and federal surprise-billing rules

For RCM, network status is not a clinical question. It is a contracting and benefits question that directly changes how many dollars you actually collect and how much goes to patient responsibility.

Why network status matters operationally

Network status affects dollars and denials before you ever hit "submit claim." If you schedule and admit as if in-network, but the plan treats the provider as out-of-network, common outcomes are:

  • CO-50 or CO-96 non-covered denials if the plan has no OON benefit
  • PR-204, N130, or MA130 remark codes pointing back to plan rules or network status
  • Very low allowed amounts and huge patient balances that are practically uncollectible

For in-network services, you are trading rate for predictability. You know the contracted allowed amount, you can quote estimates, and you can set clean pre-service collection workflows. For out-of-network, you must assume:

  • Higher authorization risk and more frequent clinical review
  • Higher patient friction when large balances hit the statement
  • Longer AR cycles, more disputes, more refunds and adjustments

From a leadership view, network status drives net collection rate and writeoffs. A few large out-of-network residential or PHP cases billed and collected as if in-network can distort metrics for months. Getting network status wrong by the time of admission is a five-figure mistake in higher-acuity behavioral programs.

How to read and use network status in daily work

Network status needs to be nailed down during benefits verification, not when the first ERA comes back.

When you call or check a portal, you should confirm very specifically:

  • Tax ID and NPI: "Is tax ID X with NPI Y in network for behavioral health for this member"
  • Location: "Is this site address considered in network"
  • Level of care: "Are residential / PHP / IOP / outpatient therapy in network under this contract"
  • Provider type: "Is Dr. Smith, NPI Z, in network as a psychiatrist / psychologist / LCSW"

Document the answer in your eligibility note, including date, call reference, and any nuance like "facility OON, but psychiatrist in network" or "facility in network for PHP and IOP, OON for RTC." Supanote or any structured note system is the right place to standardize this so scheduling, utilization review, and billing all see the same story.

On the back end, read the EOB/ERA for hints that the plan treated the claim as OON even though you expected in-network:

  • Allowed amount is far below your expected contracted rate
  • Remark codes N130 or MA130, or plan language referencing OON benefits
  • Patient responsibility showing OON deductible or coinsurance buckets

When you see that mismatch, someone should check enrollment and contracting status and decide quickly: correct the claim data, appeal as misrouted network status, or reclassify the case as out-of-network with updated financial counseling to the patient.

Common mistakes

  • Assuming that a facility contract makes all providers in-network, so you schedule a psychiatrist visit under the facility NPI, but the psychiatrist is not credentialed with the plan and the claim pays out-of-network with a PR-204 and N130 remark.
  • Checking network status only at the group or facility tax ID level and not by level of care, so residential (H0019) is actually out-of-network even though PHP (H0035) is in-network, and you discover the problem when the first month of RTC claims deny as non-covered with CO-50.
  • Relying on eligibility cards or app screenshots that show the payer logo but not the behavioral carve-out vendor, so you bill to the medical plan as if in-network and get CO-16 or CO-18 denials until you redirect to the separate BH administrator.
  • Assuming an authorization turns an out-of-network provider into in-network, so you quote in-network benefits to the family, then ERAs show OON allowed amounts and high patient balances that you cannot ethically or practically collect.
  • Failing to update network status when a provider leaves or joins the group, so old NPI mappings ride along in your practice management system and new claims price incorrectly for weeks, confusing AR reporting and leading to avoidable appeals.

Why it matters in behavioral health

Behavioral health is where network status gets messy quickly. Many plans carve behavioral benefits out to a separate vendor, so a provider can be in-network on the medical side and out-of-network on the behavioral side. If your team does not explicitly ask about the behavioral administrator and network status during benefits verification, you can lose weeks to CO-16 and CO-96 denials while you rebill to the correct entity.

Levels of care also matter more in behavioral health. A payer may contract for outpatient therapy and med management in-network, but keep residential, PHP, or IOP narrow or entirely out-of-network. One patient can have in-network coverage for weekly therapy but out-of-network treatment for a 30-day RTC or PHP stay under the same plan. If you miss that nuance, your organization can absorb tens of thousands in writeoffs or create surprise bills that damage referral relationships.

Concurrent authorization often interacts with network status. For an out-of-network residential or PHP stay, plans may require more frequent clinical reviews, shorter initial auth periods, and tighter medical necessity criteria. Even if the case is approved, any units beyond the last approved date of service will deny, often with CO-197 plus policy remarks, and out-of-network status gives the payer more leverage to cut days.

State Medicaid and Medicaid MCOs can be even more restrictive. Some Medicaid managed care plans simply do not cover out-of-network residential or IOP unless there is no in-network provider available. Others require a single-case agreement or special authorization tier. If your census includes Medicaid-funded patients from multiple states, you need clear rules on what you will accept out-of-network and when you require payers to treat you as in-network through an ad hoc agreement.

How AI can help with In-Network vs Out-of-Network

For in-network vs out-of-network work, AI agents can handle the grind of checking portals, reading plan PDFs, and comparing claim outcomes to expected network status. An AI benefits-verification agent can pull back whether the tax ID, NPI, location, and level of care appear in-network and populate a structured eligibility note, so your humans spend their time on the messy cases instead of retyping portal screens.

Supabill uses agents to store payer-specific network rules and to cross-check ERAs against expected status. A claims-scrubbing agent can flag when a claim priced as out-of-network even though your contract table shows in-network, prompting a human to check enrollment or appeal for reprocessing. A denials agent can read every 835, cluster CO-50, CO-96, PR-204, N130, and MA130 by network cause, and surface patterns to your contracting and utilization teams. Humans still need to negotiate single-case agreements, decide when to accept OON cases, and have financial conversations with families. AI can point to the problem, but it cannot own those strategic and ethical decisions.

FAQ

What is the practical difference between in-network and out-of-network for behavioral health services?

For behavioral health, in-network means the provider has a contract with the health plan or behavioral health administrator for specific service types, such as outpatient therapy, IOP, PHP, or residential. The plan uses contracted rates and in-network copays and deductibles. Out-of-network means no contract, so the plan may pay less, apply higher deductibles, or deny coverage completely, especially for residential or long-stay programs. From an RCM perspective, in-network cases tend to have more predictable allowed amounts and better collection rates, while out-of-network cases require tighter financial counseling, stronger authorization management, and more tolerance for writeoffs. HealthCare.gov outlines the general difference between in-network and out-of-network benefits in its glossary of insurance terms. Source

Does getting prior authorization make an out-of-network provider count as in-network?

No. Prior authorization and network status are separate. Prior authorization means the plan agreed that the service is medically necessary under its policies. Network status comes from a contract between the provider and the plan. Some plans may approve an out-of-network provider when no in-network provider is available, or may issue a single-case agreement that pays at a negotiated rate, but the claim is still treated under out-of-network coverage rules unless the plan explicitly agrees to in-network treatment. CMS and Medicaid guidance both stress that authorization is not a guarantee of payment, which includes not guaranteeing in-network pricing. Source

How do I confirm if a specific behavioral health level of care is in-network?

You cannot assume that all behavioral health levels of care share the same network status. During benefits verification, you should ask the payer or behavioral health vendor directly whether each relevant level of care is in-network for the facility and tax ID, for example: outpatient therapy, medication management, IOP, PHP, and residential or detox. Portals sometimes show only generic mental health benefits and omit network detail by level of care. The safest approach is to document the payer's answer and, when possible, obtain a reference number. This is especially important for residential and PHP care, where some plans carve these services out to special networks or exclude them entirely. Source

How does in-network vs out-of-network work for Medicaid and Medicaid managed care plans?

Many state Medicaid programs use managed care organizations that build their own networks. For non-emergency behavioral health, these plans often require members to use in-network providers and may not cover out-of-network treatment, particularly for residential or long-term services, unless there is no in-network provider available or the state has special arrangements. Some states carve out behavioral health to separate entities that maintain their own networks. You should verify network status with the specific Medicaid MCO or behavioral health administrator, not just the state Medicaid agency, and document any special rules for out-of-network coverage or single-case agreements. Medicaid.gov's managed care resources describe how states and plans handle networks and coverage rules. Source

What should be documented in the chart or RCM note about network status for a behavioral health admission?

At a minimum, your note should capture which entity is in-network or out-of-network (facility, rendering providers), which behavioral health administrator applies, the exact level(s) of care that are in-network, the member ID and plan type, the date and method of verification, the representative or reference number if obtained, and any special conditions such as single-case agreements or network exceptions. A structured tool such as Supanote helps standardize these fields so scheduling, utilization review, and billing all work from the same network story. Good documentation limits disputes with families, supports appeals when claims price incorrectly, and reduces the risk of misclassifying a long episode as in-network when the plan never agreed to that. Source

Sources

AI agents that run your billing.

The first agentic RCM that actually works.

Book a live demo