Skip to content

Safe Harbor: Unnecessary with Dangerous Consequences


Multiple bills have been introduced in Congress that create a “safe harbor” for physicians by prohibiting claims from being denied “due solely to the use of an unspecified or inaccurate [ICD-10] subcode” (H.R.2247) or “for errors, mistakes or malfunctions” relating to submissions of ICD-10 codes (H.R. 2652). Supposedly, the goal is to ease physicians’ transition to ICD-10 and reduce their financial risk as they become accustomed to submitting ICD-10 codes. Even though these bills are intended to respond to physicians’ concerns, the safe harbor appears to apply to all providers including hospitals. These bills are predicated on the assumption that the coding of ICD-10 diagnoses directly impacts physician payments and that ICD-10 coding will be a burden because required ICD-10 detail is not readily known or available in the medical record. Both of these assumptions are false.

ICD-10 diagnoses do not directly impact physician payment

The amount paid to physicians is not impacted by the assignment of ICD-10 diagnosis codes. The amount paid is determined by the reported CPT codes. Furthermore, there is no solid evidence to support the contention that physician practices will experience a significant level of claim denials as a result of the ICD-10 transition. In fact, the denial rate due to ICD-10 coding errors in the most recent CMS end-to-end testing was only two percent. Therefore, the assumption that there will be substantial ICD-10 related financial disruptions for physicians is false.

ICD-10 diagnosis coding will not be a burden for physicians

Although there continues to be assertions that the clinical detail in ICD-10 will be a burden for physicians, no actual examples of what constitutes burdensome clinical detail in ICD-10 are ever provided. Much of the expanded detail in ICD-10 is very basic information that should be readily available in the medical record. For example, almost half the increase in the number of codes in ICD-10 is simply due to the reporting of laterality (side of the body affected by the condition). The right/left body part distinction should always be specified in any well-documented medical record. Indeed, laterality is already a subcode modifier in the CPT coding system developed by the AMA and used by physicians to report procedures. CPT codes are very detailed and yet CPT coding is not considered a burden. As with CPT, no physician will use all of the ICD-10 codes, but will only use the subset of codes applicable to his/her specialty and patient population. For situations in which the detail for complete ICD-10 coding is not known, the hierarchical structure of ICD-10 already contains non-specific codes. No payer, including Medicare, has suggested these nonspecific codes would not be allowed on ICD-10 claims.

Far-reaching, negative consequences

Beyond being based on false assumptions, creating a “safe harbor” has far-reaching, negative consequences for the healthcare delivery system:

  1. Safe harbor undermines determination of coverage, medical necessity and quality of care

The ICD-10 diagnosis codes are used for coverage determination, medical necessity and some quality of care measures. Under the safe harbor provisions, physicians would be exempt from providing the data necessary for Medicare to determine if the services being delivered are covered under Medicare guidelines, meet medical necessity standards and are of high quality. Thus, a safe harbor has no impact on physician payment, but would severely restrict Medicare’s ability to determine coverage, medical necessity and quality of care.

  1. Safe harbor ignores Medicare’s fiduciary responsibility to ensure proper payment

Submission of “unspecified or inaccurate” ICD-10 codes will prevent Medicare from accurately determining medical coverage and medical necessity. By allowing ICD-10 coding errors to go through the payment processing system unchecked, medical necessity and coverage would not be validated, leading to payments for medically unnecessary or uncovered services. For example, a diagnosis of unspecified disorder of prostate would not justify removal or destruction of the prostate. Under the proposed safe harbor, Medicare would be required to pay the claim for the removal of the prostate even though there was no diagnosis such as malignant neoplasm of the prostate that would justify the procedure.

  1. Safe harbor raises serious fraud and abuse concerns

Allowing “mistakes” and “errors” to be reported without fear of audit raises serious fraud and abuse concerns. Even intentional errors would be included in the safe harbor provisions. If a broken finger was reported as a broken neck, Medicare would have no choice but to accept that the services delivered were covered and met medical necessity standards. Thus, even when there was a strong indication of potential fraud or the intent to purposefully bill incorrectly by deliberately reporting incorrect ICD-10 codes, a claim could not be denied. In fact, a physician could potentially report the same ICD-10 diagnosis code on claims for all of his patients, without regard to the patients’ actual medical conditions. This is analogous to allowing a tax form to be submitted with erroneous or conflicting information that does not support a refund and prohibiting the IRS from validating the information before issuing a refund check.

The recent release of Medicare payment data showed that in 2013 five physicians received more than $10 million in Medicare payments and 3,900 physicians received more than $1 million in Medicare payments. With payments of this magnitude, is it fiscally responsible to essentially suspend determination of coverage, medical necessity and quality of care?

  1. Safe harbor encourages incomplete documentation—a quality of care issue

Assigning the most accurate code based on the medical record documentation has long been an obligation for all providers – whether the coding system is ICD-9-CM, CPT, HCPCS level II, or ICD-10-CM. The vast majority of clinical detail in ICD-10 was included at the request of medical specialty groups because it was viewed as critical for understanding patient care and outcomes. Failure to document essential information in the medical record is in itself a quality of care issue.

  1. Safe harbor would lead to widespread system disruptions

It is not clear whether the safe harbor would apply to just Medicare or to all payers. If the safe harbor applies only to Medicare, then other payers may or may not adopt a similar safe harbor. Basic system functions such as coordination of benefits between payers would be compromised due to an inconsistent reporting of diagnosis codes. Also, the safe harbor would actually increase the burden on physicians, as they would still have to be able to submit accurate codes to private payers. If the safe harbor applied to all providers, then payment systems that rely heavily on diagnosis codes such as DRGs could experience significant shifts in payment levels.

  1. Safe harbor is just another delay by a different name

The original proposal to implement ICD-10 included a three-year period for physicians to get ready for the transition. The two delays in the implementation date have extended the time period for preparation to five years. And yet some still claim not to be ready. The safe harbor proposals essentially amount to yet another delay. As long as physicians aren’t expected to document and code as specifically as possible, many of them will continue to delay doing so. It can be predicted with virtual certainty that at the end of the safe harbor, there will be new demands for a reprieve from ICD-10 because there hasn’t been enough time to prepare.

There is substantial evidence that physicians can be prepared with a very modest amount of effort, and there are ample free or very low-cost resources available from many sources to assist physicians who want to be ready. The endless rumors about another delay or safe harbor provisions postpone physicians from making the effort to get ready.


The entire healthcare industry has made the investment to be ready. Taxpayers have a right to expect that all providers are reporting accurate information to ensure that government healthcare funds are being used wisely. For a physician who does not perform uncovered or medically unnecessary services, the only impact of ICD-10 diagnosis coding will be on quality of care measures. Thus, the sole effect of a safe harbor will be to compromise the ability of Medicare to monitor quality of care. Such a result is completely contrary to a move from paying for volume to paying for value in the Sustainable Growth Rate (SGR) reform legislation (H.R. 2). A safe harbor will have no impact on physician payments and will only undermine Medicare’s ability to measure quality of care and validate coverage and medical necessity.

Leave a Reply