Four years ago I was a veteran internist when I suddenly became a fresh faced, idealistic, brand-new physician advisor.  I have been on a steep learning incline ever since and it has been quite the ride!  No doubt my hair is grayer and there are many days that I get off the phone with a grumpy colleague and find myself muttering, “Bless your heart.”  (Anyone who lives in the South knows this latter phrase has little to do with religion and more to do with a passive–aggressive suspicion that the colleague is vexing, obstinate, and lacking in knowledge.  Of course, I keep that under my hat!)

On my journey, however I have come to greatly appreciate one particular member of our hospital team - our coders. 

A Special Article published online by the New England Journal of Medicine on February 28, 2016, "Readmissions, Observation, and the Hospital Readmissions Reduction Program", looked at observation and readmission rates for Medicare patients to find out if a reduction in readmissions for conditions that could trigger readmission penalties under the Medicare Readmission Reduction Program (MRRP) indicated that hospitals were gaming the system by using observation to avoid readmissions. The researchers used risk-adjusted rates of readmission and observation within 30 days of an index discharge between 2007 and 2015. They reviewed 7,175,558 admissions for one of the original three index conditions - pneumonia, congestive heart failure, and acute myocardial infarction - and compared this to 45,495,870 similar stays for nontargeted conditions.

Readmission for both targeted and nontargeted conditions had been pretty stable prior to the Affordable Care Act (ACA) but started to decline in 2010 after passage of the ACA, which created the MRRP. The study found that "from 2007 to 2015, risk-adjusted rates of readmission for targeted conditions declined from 21.5% to 17.8%, and rates for nontargeted conditions declined from 15.3% to 13.1%" (The rate of decline was greater for targeted than for nontargeted conditions, suggesting that the attention being paid to preventing avoidable readmissions and the risk of penalties under the MRRP contributed to improved hospital discharge practices and an across the board reduction in readmissions.) But the researchers found that observation rates, which had been increasing steadily since the beginning of the study period in 2007 for both targeted and nontargeted conditions continued to rise, though the increase in observation stays was not correlated with the reduction in readmissions.

The Conditions of Participation (COPs) make hospitals responsible for verifying that the documentation present in the clinical record supports the medical necessity for inpatient care. In order to bill for Medicare Part A services, the need for hospital care must be substantiated throughout the course of treatment in the hospital. This burden usually falls upon the utilization management process established for each hospital system.

In reality, all healthcare workers who participate in the patient’s care share responsibility for the efficient and timely utilization of hospital resources. In some facilities, the responsibility for the length of stay management falls on an individual physician leader such as the CMO, VPMA, Chief of Staff, UM chair, or the physician advisor. Other healthcare institutions delegate that function or oversight to non-physician leaders such as the COO, CNO or to specific sub-committees or task forces.

After two and a half years, the two-midnight rule should not continue to be giving us as much trouble as it does. When explained properly, it really boils down to a simple two-step process.

First, the Centers for Medicare & Medicaid Services (CMS) only expects to pay for patients who require hospital care; hence, we should only expect payment for those patients who require hospital care. Second, CMS only wants hospitals to admit as inpatients those who have an expectation of a total of two midnights of care or meet one of the few exceptions. If the patient does not require hospital care, we should not expect payment from CMS (but do reserve the option to charge the patient).

An online commentary on the Forbes/ Pharma and Healthcare website concerning Donald Trump’s response to a question on “Sixty Minutes” concerning an alternative to Obamacare revealed a clue about why drug prices are so high in this country.

In the critical article, calling Trump’s plan a delusion, the author, Matthew Harper, said Mr. Trump seemed to suggest that he thought he could reduce health care costs by negotiating lower prices with the health care industry without reducing choices. (In all fairness, Harper said that Hillary Clinton shares the same “delusion”.)

A recent question raised on the rac-relief.com news group was whether case management screening of observation cases is required by Medicare regulations. The answer, to be precise, is, “yes and no”. Here’s why:

Medicare Conditions of Participation (42CFR 482.30) require that a participating hospital have a Utilization Review Plan and “the UR plan must provide for review for Medicare and Medicaid patients with respect to the medical necessity of admissions to the institution…” Since patients who receive observation services are outpatients, by definition they have not been “admitted” to the hospital and don’t fall under this requirement for “admissions”.

CMS apparently thinks that hospitals have more inpatients as a result of the 2-midnight rule. At issue are the CMS assumptions and actuarial calculations that led CMS to conclude that there would be 40,000 more annual net inpatient admissions with the 2-midnight rule. To account for that, CMS implemented a 0.2% market basket reduction on IPPS payments in FY 2014. Recently, a federal judge issued a ruling on multiple lawsuits that were combined into the Shands Jacksonville Medical Center, et al. v. Burwell case. The judge found significant deficiencies in the notice and comment process, since CMS only shared certain assumptions in the 2014 Inpatient prospective Payment System (IPPS) final rule and offered little explanation for those assumptions. The judge remanded the case back to CMS to better explain the rationale for the assumptions, but stopped short of vacating the rate cut outright.

Here is CMS's explanation: CMS estimated the number of encounters that would switch from outpatient to inpatient by examining claims with observation or a major procedure that spanned at least 2 midnights. Appendix B of the CMS release lists what was considered a major procedure - cases such as breast biopsies, cystoscopies, AICDs, and cataracts. There are two reasons why CMS says they underestimated the number of new inpatient encounters. First, unbilled (yet reported) observation stays were not included in the initial analyses. Second, CMS's original estimates used the start of observation service as the start of the 2-midnight clock. Had they used the start of outpatient care preceding observation, it would have further increased the conversions to inpatient. Summing the two effects together, CMS estimated an extra 170,000 inpatient cases CMS should have but did not account for. CMS did in fact change the start of the 2-midnight clock between the proposed 2-midnight rule and the final 2-midnight rule to allow for all outpatient care. So it wasn't exactly an oversight in their initial calculation (which CMS appears to have forgotten), but it was genuinely a change in favor of providers.

The state of Maryland has been an anomaly in the Medicare hospital payment world. The Centers of Medicare and Medicaid Services (CMS) gave Maryland hospitals a waiver; they have not been covered by the Inpatient Prospective Payment System (IPPS) and are not paid based on Diagnosis Related Groups (DRGs). Since January 01, 2014 Maryland hospitals have been engaged with CMS in a new five-year demonstration project that builds on the Medicare waiver.

Under the new waiver program:

In the FY 2016 Outpatient Prospective Payment System (OPPS) final rule, CMS has opened the door to physicians admitting Medicare patients to a hospital when they believe that inpatient admission is required due to “rare and unusual” circumstances even though they don’t expect a stay lasing at least 2 midnights. 

This policy would seem to fly in the face of the 2-midnight rule, which advises physicians to admit patients to the hospital when they expect medically necessary hospital care to include two consecutive midnights and treat them as outpatients (generally with observation services) when this is not the case. 

On August 6, 2015, President Obama signed into law the Notice Act, Public Law 114–42. This law creates, one year from signing, a "Medicare requirement for hospital notification of observation status." While initially incorrectly identifying "observation" as a status (observation is a hospital service provided to patients in outpatient status), the law quickly gets its status ducks in a row and stipulates hospitals notify of their status:

each individual who receives observation services as an outpatient at such hospital or critical access hospital for more than 24 hours, to provide to such individual not later than 36 hours after the time such individual begins receiving such services (or, if sooner, upon release) —

On October 1, 2015, amid much concern and angst following several delays, ICD-10 was launched in the United States. Remarkably, the launch occurred 20 years after England, and shortly after the rest of the world adopted ICD-10 for coding use. So why after over 30 years of using ICD-9 did the United States need a newer version? Well, the short answer is simple: we ran out of codes. The 5 character ICD-9 coding system was full, without any available open codes for the myriad of new diseases, conditions, and procedures that have been realized over the last 3 decades; for example, there is no code for Ebola in ICD-9.

Historically, the International Statistical Classification of Diseases and Related Health Problems (ICD) dates back to the late 1800’s when the French physician J. Bertillon introduced the Bertillon Classification of Causes of Death in 1893. From 1900 to the early 1970’s, the ICD coding system had been revised approximately every 10 years. The ninth rendition, ICD-9 was written in the late 1970s and adopted by most countries during the 1980s. The coding system has been used as the standard documentation tool for epidemiology, health management and clinical purposes, to monitor the incidence and prevalence of diseases and for other health matters.

There was a recent thread on the RAC Relief Google group about Medicare Advantage managed care plans and the problems providers are encountering with the various plans. As I read this thread, it occurred to me that this is really part of a larger discussion on payer contracting in general.

Effective contracting is a team sport that requires input from the whole revenue cycle team (see figure below). Once a payer contract is in effect, there is very little recourse when care management (CM), physician advisors, billers and other members of the team discover issues with that contract. Time spent in getting input, studying, clarifying and negotiating contract terms prior to signing will pay off big on the back-end with fewer denials and more clean claims.

The largest five health insurance companies appear poised to become the big three. Anthem Healthcare and Cigna have announced an agreement in principle to merge. Aetna and Humana have done the same. Both will face scrutiny and regulatory hurdles prior to government approval of their mergers. United Healthcare would remain the third largest non-government provider of health insurance.

The simple utterance of these names in healthcare organizations cause physician advisors and finance departments to grimace. These companies are already known for their egregious denials and appeals processes. They make up rules for healthcare organizations to follow as they see fit and appear to install speed bumps to the hospital’s case management process whenever possible.

The Office of Inspector General (OIG) in May of 2012 published audit findings of "Coding Trends of Medicare Evaluation and Management Services". What they found was that according to U.S. census data, from 2001 to 2010 the population age 65 and over had grown from 35,291,000 to 40,268,000 - a 14 percent increase - but at the same time, Medicare payments for evaluation and management (E/M) services increased by 48 percent. They also found that from 2001 to 2010 physicians increased their billing of higher-level E/M codes for all types of E/M services. Billing for the two highest-level outpatient E/M codes (99214 and 99215) increased by 17 percent and the highest inpatient codes (99232 and 99233) increased by 6 and 9 percent respectively; the highest ED code (99285) rose by 21 percent.

As a result of these findings, review of the billing for the evaluation and management CPT code 99214 was placed into the OIG work plan to review for potential fraud or abuse. But what the OIG failed to consider was the "EMR Effect". It is generally believed that prior to implementation of electronic medical records (EMRs), physicians only documented about 40 percent of what they actually did. When an EMR was implemented, however documentation of more material occurred and the EMRs began to suggest the proper E/M level for the visit. Therefore, by 2010 there may not have been a "code creep" due to more aggressive coding but rather a more accurate representation of actual care delivery.

Some hospitals use McKesson’s InterQual admission criteria; some use MCG (formerly Milliman). Some managed care plans use InterQual and some use MCG. The Centers for Medicare and Medicaid Services (CMS) has said it and its contractors may refer to either InterQual or MCG both they don’t recognize either as the deciding factor in establishing payment. And of course there’s the two-midnight rule for Medicare. Some managed care plans have said they use both InterQual and the two midnight rule, as did the recent announcement by the Quality Improvement Organizations (QIOs) that will be doing admission reviews, but they have not explained how they can be used in conjunction. And then there are the arbitrary payment decisions arising from managed care plans that don’t seem to follow any guidance except what’s best for the plan.

Amidst all this confusion the physician is expected to apply his or her clinical judgment, document the decision-making process, and accurately determine billing status. No wonder they roll their eyes, raise their hands, and tell the utilization nurse, “Tell me what to write.” Aye, but thar’s the rub: the landscape is so confusing that even when they know the proper status, it’s nearly impossible for them to explain it to physicians in the little time they have with their attention.

Penned from 8-year old author Virginia O’Hanlon in September 1897, this now famed letter to the editor of the NY Sun Times that questioned the reality of Santa Claus can be extended to all present-day “Virginia’s” who question the reality of ICD-10.  “Is it going to happen…?”  “Will there be a delay…?”  “Will our systems be ready for the flip from I-9 to I-10…?”

As we sit here on the eve of the transition, it is clear that all last minute delay tactics have been expired and we will move to full adoption of ICD-10 on October 01, 2015.  This first “ICD” (International List of Causes of Death in 1893) has changed multiple times over the years and for the US this last version has been a long time coming.  Our first go-live was delayed in 2013 because everyone realized that no one was ready.  The second delay occurred on April Fool’s Day in 2014 because the CMS reply to “are you ready” left everyone with a realization that they were not even close.  And now, the third go-live this year comes about after a series of full-cycle testing by CMS shows that they can indeed receive and process “dummy” ICD-10 claims with certainty across the spectrum of healthcare.  We all share the hope that the results reported thus far are true after 10-01-15 and that it will extend to the commercial payers (though commercial payers have not had to report testing).  Only time will tell….

"Our first goal is for 30% of all Medicare provider payments to be in alternative payment models that are tied to how well providers care for their patients, instead of how much care they provide - and to do it by 2016. Our goal would then be to get to 50% by 2018." - HHS Secretary Sylvia Burwell, HHS Blog, Jan 26, 2015.

One of the earlier steps along this road was actually the Inpatient Prospective Payment System (IPPS) itself, as it bundled payments for inpatient care episodes into Diagnosis Related Groups (DRGs). The tremendous complexity of the DRG system, however probably encouraged as much documentation and coding proliferation as it did efficient care.
 
The CMS Innovation Center created by the Affordable Care Act has been piloting Accountable Care Organizations (ACOs) through the Pioneer ACO model. The premise is that the value of care can be improved by incentivizing providers to decrease cost through sharing in savings (or losses) while maintaining quality. Much of what was learned from Pioneer ACOs was incorporated into the Medicare Shared Savings Program, which contains most of the Medicare ACOs today. So far very few ACOs have elected the 2-sided model (where they can earn or lose money, based on performance), selecting instead the 1-sided model where savings but not losses can be shared with the ACO. The safer 1-sided option was originally planned for only the first agreement period, and that's the model almost every system that developed an ACO chose. Even so, participation nationwide has been such that only about one tenth of the Medicare population is now covered under an ACO. Faced with the prospect of losing the vast majority of 1-sided ACOs that were not expected to renew for a second agreement under a 2-sided model, CMS proposed a new track to sweeten the deal, more favorable benchmark calculations, and agreed to allow a second term under the 1-sided model.
 
Separately, the CMS Innovation Center continues to trial various models to shift more risk and reward to ACOs through the Next Generation ACO. This newer ACO model allows “first dollar” shared savings or losses, rather than utilizing a minimum savings rate (MSR) or minimum loss rate (MLR) before the ACO feels any financial impact. The maximum percentage of savings or losses shared with the ACO is higher as well, potentially up to 100%. Though ACO models have previously been based on a FFS payment system or variations thereof, the Next Generation ACO model contains an option for “full capitation” for an ACO, where the ACO receives a per member per month (PBPM) payment and all providers are paid out of that amount.

Denials of coverage for services provided have become standard parts of doing business. The necessary appeals tax the resources of providers mightily. In fact, the skeptic in me assumes that the various governmental contractors  and commercial payors have a  certain part of their recovery predictions based on the fact that many providers lack the resources to respond or are intimidated.

Denials management 101 is based on having the best “up front” process possible.  That means an engaged medical staff that understands  documentation is no longer just a daily note for colleagues.  They must truly be documenting with the expectation that the record will be critically audited and must support the services provided.  The services must be ordered AND they must be necessary.  And do not make the mistake that any auditor will be agreeable to reading minds or reading between lines….not a chance.  Auditors have zero incentive to give any provider the benefit of the doubt when it comes to documentation.  Effective denial management also requires point of entrance status review and a motivated, well-educated, well supported Utilization Review staff who recognize that their jobs are as much about education as anything else. 

Many U.S. hospitals are facing financial stresses that can lead to forced sale, merger, or even bankruptcy. Many of those that are better off see shrinking margins and face financial pressures, too. They are all under the gun to reduce the cost of providing care. They know this, and smart CEOs and CFOs are working diligently to improve the efficiency of their operations and the management of their money, but there is one part of the cost equation that is beyond their control – and that is the behavior of their medical staff.

It has been said that the physician’s pen (or keyboard in the era of electronic medical records) is the hospital’s most expensive piece of equipment. The physician determines when a patient is admitted and whether the stay will be inpatient or observation. This decision has huge financial implications for the hospital because the payment for inpatient care is generally far more than for observation and admission errors are costly to fix – whether it be for the salaries of case managers, physician advisors, and appeals coordinators, the high cost of outsourcing these critical functions, or lost revenue due to denials and delays in receiving payment as the appeals process drags along. In addition, the physician orders the tests and treatments "on the hospital’s dime" and the attending determines when consultants are to be called and who they will be. Will they be the most cost-efficient consultants available or the attendings golf buddies? Will they be playing "pay back" for previous referrals? One would hope not.