Given its complexity, medical coding often seems like a “necessary evil” for health care organizations. Sure, coding is something that must be done before providers can be paid, but the data generated in the process can also be leveraged to improve business processes and patient care.

Some call it “health care analytics” and others call it “data mining.” Either way, it’s about scouring the vast quantity of information locked within medical codes, electronic medical records, laboratory results, billing systems, and other places to discover trends and opportunities.

Essentially, you can combine data from various IT sources to form a large data warehouse, which can then be queried. Let’s look at two simplistic examples that show the promise of health care analytics.

Patient referral analysis: Maximizing efficiency and revenue
An endocrinology patient may end up at an endocrinologist by a variety of paths: self-referred, referred by a primary care physician at the same medical center, referred by a nephrologist at the same medical center, referred by a primary care physician at a different medical center, and on and on.

An analytical software tool could examine (using ICD-9-CM and CPT codes, and billing information) the treatment pattern of each type of patient. For example, a retrospective analysis of the past two years of endocrinology patients might show that self-referred patients — those who simply call up and make an appointment — had minor ailments that could’ve been treated by their primary care physician. Such “one and done” patients take appointment slots away from — and generate less revenue than — patients who truly need long-term endocrinology care.

Lesson learned: The endocrinology department can use these findings to better prioritize how it fills its appointment slots, perhaps by more closely screening self-referred patients to determine whether they really need to see an endocrinologist.

Outcomes research: Improving quality of care
There’s no doubt that the “art” of medicine is aggressively moving toward the “science” of medicine. Federal research dollars are pouring into areas like comparative effectiveness research, where researchers analyze outcomes over time to determine which treatments are statistically better than others for different patient populations.

For instance, from the aforementioned data warehouse, a medical center could select all male patients between the ages of 25 and 44 diagnosed with headaches in the past year. Presumably, different patients will be prescribed different treatments, such as over-the-counter medication (ibuprofen, acetaminophen, etc.), massage, or acupuncture.

This group of patients could then be analyzed over time. In the six months following initial treatment, perhaps the data show that patients prescribed acupuncture sessions are statistically less likely (controlling, of course, for other variables) to return to the physician with continued headaches.

Lesson learned: If acupuncture is statistically more effective, the organization could build a clinical decision support tool to alert physicians at the point-of-care that their headache patient might benefit most from acupuncture.

Robust analytics are the future
Medical coding is here to stay, but it shouldn’t be viewed solely as a means to get paid for services rendered. Rather, health care organizations ought to look at medical coding as one of several crucial sources of data that can be combined and analyzed to help optimize future revenue, improve patient care, and more.

Products such as DataScout, recently released by our friends at CodeRyte, are leading the way in the growing health care analytics software market. pmFAQtory stands ready to help your health care organization implement and make use of these valuable tools.

What secrets are locked within all the data you’re generating each day? Get in touch with pmFAQtory and let one of our health care consulting experts show you the way.

In the past few weeks, we’ve looked at several code sets (ICD, ICD-CM, HCPCS, and CPT) that are significant in the medical coding arena. Now, let’s explore computer-assisted coding (CAC) applications and how they fit within the health care information technology environment.

As you know, code sets form the foundation of medical coding. If you don’t know which code represents a specific medical diagnosis or procedure, you can look it up in the appropriate code set — much like you’d look up an unfamiliar word in the dictionary.

Simple enough — if you are looking up a code or two. But medical centers may see hundreds, or even thousands, of patients each day. Every diagnosis made — and every procedure performed — by a clinician must be coded using one of the above code sets. (That is, if the medical center wants to be paid for services provided!)

Even if we assume that to some degree the Pareto principle is in effect — that the majority of cases seen at a medical center can be coded using a reasonably small number of codes — we’re still talking about human coders becoming intimately familiar with hundreds of codes (out of the 15,000 or so codes in ICD-9-CM).

So it’s no surprise that health care organizations are seeking out companies like pmFAQtory to help them implement CAC applications such as CodeRyte. These software tools do the initial assigning of codes, meaning you need fewer human coders — and those coders only have to “audit” the software’s decisions, and perhaps dissect and code the really complex cases.

Within a health care organization’s information technology landscape, CAC applications sit among many other software systems. Various systems feed information into the CAC application; once processed by CAC software, data is fed into more systems.

These other systems include:

• A phone/dictation system into which physicians dictate their patient notes


• A transcription application to convert audio dictations into text


• A routing interface to bring transcribed notes together with relevant patient information, such as a patient number, demographic data, and insurance information


• A billing system to match procedures with charges and generate a bill for third-party payers and/or the patient


• An electronic medical record to house each patient’s individual health record

The CAC application generally sits between the routing interface and billing system. As you might guess, there are various checks and balances, and lines of communication, between these systems to ensure data integrity. For example, if a human coder is checking the output from a CAC application and realizes more specificity about a diagnosis is needed from the physician, there’s an avenue for him or her to connect with that doctor.

When managing a CAC software implementation, pmFAQtory’s job — among other things — is to take a holistic look at the organization’s IT environment, determine the interrelationships among the myriad systems, and ultimately ensure that all necessary data flows to and from the new CAC software.

How can pmFAQtory add value to your health care IT project? Contact us.

Diagnosis-related groups (DRGs) are one ingredient in medical coding’s “alphabet soup” that we’ve yet to cover. They facilitate Medicare’s inpatient prospective payment system, under which the U.S. government reimburses hospitals for treating Medicare patients.

The Medicare DRG system (MS-DRG) is maintained by the Centers for Medicare and Medicaid Services (CMS) and its latest revision, version 27, has more than 750 “groups.”

Taking a step back, we should note that DRGs are themselves classified into 25 major diagnostic categories (MDCs) based on organ system (eye, respiratory, etc.). After assignment into an MDC, the patient is matched to a DRG.

By sorting patients into DRGs, the government can reimburse providers on a “per episode” basis, as all patients in a particular DRG should theoretically use similar amounts of hospital resources and therefore cost the hospital about the same amount to treat. For any particular patient, the DRG system is meant to encourage hospitals to avoid unnecessary services and procedures, since their reimbursement amount is set at a fixed level.

Prior to the 1983 implementation of the DRG system, hospitals were simply reimbursed for costs they incurred serving Medicare patients. Like everywhere else, incentives matter in health care. When providers are paid more when they do more, they tend to do more. From 1967 to 1983, Medicare enrollment rose by barely 50 percent, but spending ballooned from $3 billion to $37 billion.

Under the DRG system today, each Medicare inpatient is evaluated based on his or her principal diagnosis, surgical procedure(s), age, sex, coexisting conditions, and status at discharge. Each DRG has a weight assigned to it, which represents how resource-intensive a particular condition (on average) is for the hospital.

For example, in fiscal year 2010, viral meningitis without complications or comorbidities (MS-DRG 076) has a weight of 0.8336. Contrast that with a lung transplant (MS-DRG 007), which is weighted at 9.4543. Clearly, a patient needing the latter costs the hospital more than someone treated for the former. The initial payment amount is computed by multiplying this weight by the payment rate specific to that hospital.

After this initial computation, the payment amount is adjusted by a number of additional factors:

• Wage conditions: Payments in high-wage areas of the U.S. are greater than those in low-wage areas.


• Share of low-income patients: Hospitals that serve a disproportionate number of low-income patients receive add-on payments.


• Indirect medical education: Approved teaching hospitals receive an add-on payment based on the ratio of medical residents to hospital beds.


• New technologies: Cases that use certain technologies receive add-on payments.


• Outlier cases: Cases with unusually high overall costs are eligible for additional payments.

So where does medical coding fit in?

Coding must be done before you can even start talking about DRGs. Diagnoses and procedures — more accurately, diagnosis and procedure codes — are two inputs for the DRG “grouper” algorithm that sorts patients into DRGs.

And as we’ve discussed before, improperly coded or undercoded cases can negatively impact how a private insurance company reimburses a provider for services rendered; so too can it affect DRG classification. When the assigned medical codes understate the severity of a patient’s condition — and therefore assign the patient into a lower-paying DRG — providers get underreimbursed.

Capturing every legitimate diagnosis and procedure maximizes reimbursement, regardless of whether it’s so the government can assign a Medicare patient to a DRG or so a private insurer can reimburse using a fee-for-service model.

Good medical coding is good business.

Stop leaving money on the table. pmFAQtory can help you optimize your coding and billing efforts. Get in touch with us today.

As we’ve discussed, medical coding is the process of assigning standardized numbers to diagnoses/diseases, causes of death, and inpatient and outpatient medical procedures. The primary reason for medical coding is to ensure consistent classification and billing, as it enables physicians, medical centers, and third-party payers to “talk” in the same language.

So where do these codes come from? There are four major “code sets” in the medical coding world, each with a different use:

• International Statistical Classification of Diseases and Related Health Problems (ICD), maintained by the World Health Organization


• International Statistical Classification of Diseases and Related Health Problems, Clinical Modification (ICD-CM), maintained by the Centers for Medicare and Medicaid Services and the National Center for Health Statistics


• Healthcare Common Procedure Coding System (HCPCS), maintained by the Centers for Medicare and Medicaid Services


• Current Procedural Terminology (CPT), maintained by the American Medical Association

Here’s how they fit together:

ICD-9: The ninth revision of the ICD code set, ICD-9 was used to classify mortality (death) in the U.S. until Jan. 1, 1999, and is now obsolete (replaced by ICD-10).

ICD-9-CM: The ninth revision of the ICD code set with “clinical modifications,” ICD-9-CM is used today in the U.S. to classify morbidity (diagnoses/diseases) and inpatient medical procedures. It consists of three volumes: volume one (tabular listing of diagnosis codes), volume two (index of diagnosis codes), and volume three (procedure codes). An ICD-9-CM code has between three and five characters, such as 560, 553.3, or 560.81.

ICD-10: The 10th revision of the ICD code set, ICD-10 replaced ICD-9 for classifying mortality in the U.S. as of Jan. 1, 1999.

ICD-10-CM: The 10th revision of the ICD code set with “clinical modifications,” ICD-10-CM is an expanded and revised version of ICD-9-CM, and will be put into use in the U.S. as of Oct. 1, 2013, to classify morbidity. An ICD-10-CM code has between three and seven digits; examples are M05.339 and S26.020D.

When you hear rumblings that medical centers are anxious about planned ICD changes, it’s this forthcoming implementation of ICD-10-CM across the U.S. that’s causing the angst.

HCPCS: A two-level code set used to classify outpatient medical procedures. Level one consists of CPT codes (see below), while level two classifies non-physician services and supplies, such as ambulance transportation and medical equipment.

CPT: Serves as HCPCS level one and is used in the U.S. today to classify all outpatient medical and surgical procedures (ICD-9-CM volume three is used to classify inpatient procedures). A CPT code has five digits, such as 39520 or 00756.

A subset of CPT codes, called evaluation and management (E&M) codes, are used to classify non-surgical physician visits/consultations.

Put another way, in the U.S. today …

• ICD-9-CM volumes one and two are used to classify morbidity


• ICD-9-CM volume three is used to classify inpatient hospital procedures


• ICD-10 is used to classify mortality


• HCPCS level one (CPT) is used to classify outpatient procedures


• HCPCS level two is used to classify medical equipment, supplies, and drugs

The intersection of medical coding and health care information technology — pmFAQtory’s field — is pretty clear-cut. For humans, sorting through various code sets and tens of thousands of individual medical codes, while at the same time ensuring compliance with payer regulations, is about as much fun as doing taxes by hand.

That’s why numerous technological solutions have surfaced to help medical centers more efficiently assign codes and ensure accuracy. pmFAQtory specializes in guiding health care organizations as they implement these systems, which sit within a complex health IT landscape. We’ll examine that landscape in more detail in the next few weeks.

Ever feel like your doctor speaks in code? To be sure, there’s often a communication gap between patients and physicians. But that’s a topic for another day.

The medical coding we’re talking about here is the type that exists in health care information technology — the realm in which pmFAQtory works. This kind of coding deals with numbers; specifically, it’s about assigning numbers to represent medical diagnoses and procedures. Medical coding means standardization and simplification. This concept is simple enough, but as we’ll see, the reasons for medical coding extend far beyond convenience.

Medical coding began in the mid-19th century as a way to more consistently classify causes of death. Today, coding is used to classify not only mortality (deaths), but also morbidity (diseases), and inpatient and outpatient medical procedures.

Two major “code sets” you may have heard about are International Statistical Classification of Diseases and Related Health Problems (ICD) and Current Procedural Terminology (CPT). We’ll dig deeper into each of these in the coming weeks.

So what implications does medical coding have for those working in the health care industry?

Medical centers (hospital, clinics, etc.) rely on accurate coding to get paid — literally. Only by properly coding (in terms of both accuracy and specificity) diagnoses and procedures can medical centers be reimbursed by third-party payers such as insurance companies or the government (Medicare at the federal level; Medicaid at the state level).

Third-party payers use medical codes to determine whether — and how much — to pay for a given procedure. They won’t pay for procedures performed without a supporting diagnosis. If a procedure is miscoded or not coded to the appropriate level of specificity, the payer may delay or deny reimbursement. These types of errors trickle down to patients, who are on the hook for charges not paid by their insurance company.

Biomedical researchers — such as epidemiologists — rely on medical coding when studying health outcomes over time. Instead of searching huge batches of medical records for a specific term or phrase — the use of which can vary throughout the U.S. or the world — researchers can search for the applicable code(s), which means better and faster research.

And that’s the tip of the iceberg, as they say. Next time, we’ll start exploring ICD and CPT, and their association with health care IT.

At pmFAQtory, we’ve long known that computer-assisted coding is a powerful strategic tool that health care providers can use to maximize revenue and decrease expenses. That’s why we focus on helping health care organizations implement CodeRyte, the most sophisticated and accurate computer-assisted coding system in the world.

An article in Health Management Technology earlier this year even referred to computer-assisted coding as “the secret weapon” (“Computer-assisted coding: the secret weapon” by Mark Morsch).

Let me explain. Following a patient visit, a physician verbally dictates notes about the patient’s symptoms and diagnosis, and about any procedures performed during the visit. These notes are then transcribed by hand or voice-recognition software. But a provider can’t send all that raw information directly to a patient or third-party payer (e.g., an insurance company) as a bill.

Before a bill can be generated, someone — or something — must parse those notes and catalog “what happened” during the visit using a standard set of several thousand numeric “codes.” This process of coding, believe it or not, is still done solely by humans at many clinics and hospitals. Can you imagine the inefficiency?

Enter computer-assisted coding software. Think of these applications like filters. In go the rough physician’s notes, out come of the clean numeric codes. In some cases, a human may double-check the software’s accuracy, but oftentimes these codes can be sent straight into the provider’s billing system, after which a bill is created and sent out.

The Health Management Technology article specifically mentions productivity, accuracy, consistency, transparency, and compliance as benefits of computer-assisted coding. For health care providers, this adds up to better efficiency and money saved.

Let’s take a look at some real numbers. In 2009, pmFAQtory project-managed the implementation of CodeRyte’s CodeAssist application for a large health care organization in the Midwest.

We finished that CodeRyte project in July 2009. Since then, the organization has reduced their data entry staff by six full-time equivalents (FTEs) and their coding staff by one FTE. They’re also saving 1.8 million sheets of paper per year. (We’d tell you how many trees that is, but there’s no simple calculation!)

Today, pmFAQtory is busy implementing CodeRyte at the University of Chicago Medical Center, and expect them to realize similar — if not greater — benefits.

Are you in the “secret weapon” market? Get in touch.