Since MS-DRGs get used for all kinds of things beyond their intended use, which is prospective payment for Medicare recipients, one of the most commonly discussed differences between ICD-9 and ICD-10 is in the obstetrics DRGs. If you ever have seen, heard, or read even one presentation about ICD-10 codes, you have probably been told that one major difference is, “ICD-9 obstetrics codes are classified by whether the patient delivered during the encounter, and ICD-10 codes are classified by trimester of the pregnancy.”
Excellent bit of ICD-10 intel, but it is about three rungs up the ladder of abstraction for our purposes. To understand how this impacts MS-DRGs, I intend to plant both feet on the rhetorical ground. Let’s take the simplest of examples—a pregnant woman is admitted to the hospital in active labor. The obstetrician documents that the patient is full term at 38 weeks, and the pregnancy has been uneventful except for higher-than-normal weight gain. After several hours of labor, the patient delivers a healthy infant, with manual assistance from the delivery team and no other interventions such as episiotomy. The ICD-9 record for this encounter typically looks like this, with only the diagnosis code on the record and no procedure code: 646.11 Edema or excessive weight gain in pregnancy, without mention of hypertension, delivered, with or without mention of antepartum complication.
There exists an ICD-9 procedure code for manually assisted delivery: 73.59 Other manually assisted delivery. Some facilities put the code on the record and some don’t. Two reasons can explain why, one from the point of view of coding, one from the point of view of coded data. Take your pick.
- Because the diagnosis code by itself contains adequate delivery information, and there were no additional procedures beyond the manual assistance given to aid the delivery
- Because payers require only the diagnosis code to determine that a vaginal delivery occurred , and the facility doesn’t use the code for internal tracking
An ICD-9 code with only 646.11 on the record is assigned to MS-DRG 774 Vaginal delivery with complicating diagnoses. The translated ICD-10 record containing only O26.03 Excessive weight gain in pregnancy, third trimester would not be assigned to DRG 774, but assigned instead to DRG 781 Other antepartum diagnoses with medical complications. Why?
ICD-10 MS-DRGs do not have the luxury of using the diagnosis code to infer a vaginal delivery in the absence of any other obstetric procedure code. Below are the two codes side by side highlighting that difference in the way the two systems classify obstetrics patients. The ICD-9 MS-DRGs used the available information in the diagnosis code to infer the delivery; the ICD-10 MS-DRGs cannot.
ICD-9-CM : 646.11 Edema or excessive weight gain in pregnancy, without mention of hypertension, delivered, with or without mention of antepartum complication
ICD-10-CM: O26.03 Excessive weight gain in pregnancy, third trimester
So, the ICD-10 MS-DRG definitions for the vaginal delivery DRGs 774-775 requires an ICD-10 procedure code specifying a manually assisted delivery, episiotomy, obstetric suture repair or the like. This list is available in the ICD-10 MS-DRGs definitions online on the CMS website.
Someone trying to estimate ICD-10 impact on their organization might use software or write their own programs that use the GEMs to create ICD-10 records from their database of ICD-9 records. And for every one of those obstetrics records where a vaginal delivery occurred but no delivery procedure was recorded, the ICD-10 MS-DRG will be predicted to shift from the vaginal delivery DRGs 774-775 to the antepartum DRGs 781-782.
The remedy is ridiculously simple. To correct the impact analysis, change the record conversion process to ensure that all ICD-9 records that indicate delivery in the diagnosis code but do not have an ICD-9 delivery procedure code are converted to ICD-10 records containing a delivery procedure. The ICD-10 corollary to the ICD-9 manually assisted delivery code is 10E0XZZ Delivery of Products of Conception, External Approach. You could add this one code to all such records and declare victory. No more DRG shift. As for real coding for real money under ICD-10, rest assured that the HIM department is aware of this significant difference in the obstetrics ICD-10 codes and will work to ensure that delivery procedures are correctly and completely coded in ICD-10.
Since that was practically too easy, how about we tackle the type of shift itself. As the optometrist would say in his preternaturally calm voice, Is it one or three? And if you hesitate—one or three?—what I mean is this example could be classified as either the third type of shift (an important difference in the classification itself) or the first type of shift (an automated translation is flawed and declares a DRG shift, when a correctly coded ICD-10 record works just fine with no shift).
I think of it as the first type, since MS-DRGs have taken account of the difference in the DRG logic itself, so the MS-DRG shift is only apparent, not real. But that is just one opinion. And it just goes to show you, even an attempt to name three measly things by what type of thing they are will expose the essential challenge of classification—the universe of things is infinitely divisible, and working to find the balance between detail and usefulness is what it’s all about.
Rhonda Butler is a Senior Clinical Research Analyst with 3M Health Information Systems.