This column's July and August editions dispelled the myth that the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10) will impose a host of new requirements on physicians, forcing them to dramatically change documentation practices. ICD-10 does, however, present some new documentation opportunities, which we will explore this month and next month.
Ever since Medicare Severity Diagnosis-Related Groups (MS-DRGs) were introduced in 2007, there have been pervasive physician documentation deficiencies nationwide, with a negative impact on revenue, quality, outcome, and pay-for-performance measures. These will no doubt continue to some degree under ICD-10, but ICD-10 is not the culprit. There are only a few pertinent changes in documentation specificity using ICD-10 that were not already an issue with ICD-9.
Depression or major depressive disorder (MDD) is a significant comorbid condition for many inpatients, complicating their management. For depression to be classified as a significant comorbid condition under ICD-10, it should be specified as “major” or as MDD and designated as mild, moderate, or severe. Also, for all cases of depression or MDD, clinicians should indicate whether this is the first episode or a recurrent episode. In addition, it's important to document whether the condition is “in remission” or “not in remission” during any particular admission. “In remission” means the depression/MDD is stable without symptoms on or off medication and does not require any intervention such as medication adjustment or psych consult.
Many inpatients meet the authoritative guideline criteria for the diagnosis of sepsis. It has always been clinically challenging to understand and stay current with these diagnostic standards. ICD-10 added another small obstacle by removing “infectious” systemic inflammatory response syndrome (SIRS), or SIRS due to an underlying infection, as sufficient documentation for assigning the sepsis codes. However, since sepsis is defined clinically as SIRS due to infection, it should be a simple matter for clinicians in the habit of describing “SIRS due to an infection” to document the word “sepsis” for these patients. Also, the term “urosepsis” has no corresponding ICD-10 code, requiring the documentation of “sepsis.” If only “urosepsis” is used, the clinician will be required to clarify whether it was sepsis due to UTI or UTI without sepsis.
Respiratory failure may be clinically classified as hypoxemic or hypercapnic or sometimes a combination of both. The distinction is clinically important because they may have different causes, mechanisms, and management implications, but it has no impact on revenue or quality and performance measures. ICD-10 allows the specific coding of hypoxemic or hypercapnic respiratory failure (or both), and clinicians are encouraged to document this distinction when a patient has respiratory failure.
The diagnosis of acute hypercapnic respiratory failure nearly always depends on measurement of arterial blood gases showing an elevated Pco2 associated with a pH below normal (<7.35). In chronic respiratory failure, the pH is normal due to renal compensation by retention of bicarbonate. The term “respiratory acidosis” should not be used unless the condition is also identified as acute hypercapnic respiratory failure. The terms mean exactly the same thing, but respiratory acidosis is assigned a code that is not equivalent to, nor reflective of, acute hypercapnic respiratory failure.
The opportunity to document the important diagnosis of acute hypoxemic respiratory failure is frequently missed, primarily because clinicians fail to recognize the diagnostic standard of a Po2 less than 60 mm Hg or an equivalent oxygen saturation less than 91%, even measured by pulse oximetry (Spo2) on room air. In addition, clinicians frequently fail to calculate a Po2/Fio2 (P/F) ratio for patients when measurements are done while receiving supplemental oxygen. A P/F ratio less than 300 confirms acute hypoxemic respiratory failure, except in patients who have chronic respiratory failure requiring continuous home oxygen.
Patients who require continuous home oxygen are expected to have a Po2 less than 60 mm Hg or an Spo2 less than 91% on room air, so these hypoxemic criteria cannot be used when measured on room air. On the other hand, such patients are expected to have a Po2 greater than 60 mm Hg and an Spo2 of 91% or above while receiving their usual supplemental oxygen flow rate (or higher). In this situation the hypoxemic criteria can be applied when measured on their usual home oxygen or at a higher rate. Do not use the P/F ratio for any patients on continuous home oxygen, since it will always be less than 300.
ICD-10 provides for the coding of great detail for a cerebrovascular accident (CVA), including embolic versus thrombotic, hemorrhagic versus nonhemorrhagic, laterality, and specific artery affected. Fortunately, there is a code that can be used when this specificity is not available in the record, and again revenue, quality, and performance measures are not affected. To allow the efficient capture of such valuable coded data, when the clinician diagnoses simply CVA or “stroke” but does not include these other details, the coder is allowed to pick it up from other sources like CT, MRI, magnetic resonance angiography, transcranial Doppler, and angiography.
There are a few ICD-10 changes related to ischemic heart disease. Angina or unstable angina is now included in a combination code, requiring specification of the presence or absence of coronary artery disease (CAD). If CAD is not documented in the record, a default code for angina not associated with CAD will be assigned. If CAD is mentioned anywhere, the condition will be coded as “with CAD.”
Coding of myocardial infarction (MI) has some interesting provisions under ICD-10. An MI is classified as acute for all encounters occurring during a period of 4 weeks following its occurrence (not the discharge date). After that it is coded as an “old” MI (code I25.2). ICD-10 makes a provision for what is called a “subsequent” MI: a new MI (even in the same location) occurring within 4 weeks of a prior MI. A code must be assigned for both the new and the prior MI. Clinicians should state when the prior MI occurred—if not the exact date or time frame, then whether it was more or less than 4 weeks before.
Location of an MI (such as anterior, lateral, posterior, etc.) or the artery involved can also be identified by ICD-10 codes, but this specificity has no impact on revenue or quality and performance reporting. The unspecified code for documentation of simply “MI” is I21.3 (STEMI, unspecified site). While stating the location is encouraged, the coder is allowed to pick up this information from other sources such as electrocardiogram or catheterization reports.
For inpatient cardiac catheterization procedures to be assigned an ICD-10 procedural code for the hospital's claim, the contrast used must be specified as high or low osmolality. Alternatively, the coder can use the brand name of the contrast, if documented, to make a determination. If only 1 type (usually low osmolality) or brand is used at a facility, this can be communicated to coders to avoid unnecessary requests to clarify this necessary information.
In summary, ICD-10 offers some interesting and important new documentation opportunities, but most of these will not affect physician and hospital revenue or quality and performance measures. If the fundamental diagnosis is clear, coders will often be able to pick up more specific details from other medical record sources in many situations. By being mindful of the documentation opportunities discussed above, physicians can add clarity to their medical record documentation and offer a more informed reply to any documentation clarification requests from coders and documentation specialists.