Guest post: The problem of ‘critical values’

By Howard Rodenberg, MD, MPH, CCDS for ACDIS CDI Blog

You would think that actually defining something would make it easier to recognize. That statement assumes, of course, that you can actually define what you’re trying to see. (I’m reminded of Justice Potter Stewart’s quote on obscenity, “I shall not today attempt further to define the kinds of material I understand to be embraced […] but I know it when I see it.”)

The Fourth Universal Definition of Myocardial Infarction (MI) was released in 2018. This latest move towards a conclusive definition of MI emphasizes the presence of a characteristic rise and fall of cardiac troponins in the setting of clinical ischemia as definitive evidence for acute infarction. At least one troponin value must be above the 99^th percentile value for the test to be diagnostic of acute myocardial damage in the proper clinical scenario.

This is likely not news for anyone in the CDI profession. What might get your attention, however, is that the values reported as high, low, or critical in your local laboratory may not reflect the 99^th percentile requirement.

Let’s start the discussion by talking about critical values. Lab reports usually offer a “normal” range which distinguishes high and low values, and a “critical value” that demands immediate clinician attention. The critical value for any particular lab test is only what the lab and the clinicians who review the test say it is. It’s a purely subjective number. Is a sodium of 155 a critical value? 156? 160? It all depends on what the folks in charge of the lab determine it to be.

(Personally, I prefer the term “panic value” over “critical,” if only because I can have so much fun with it. One particularly crabby night in the ER, I got a 4 a.m. call from the lab about a “panic value.” Just because I was in a foul mood, I let loose with a primal scream of terror, sort of like the noise made by a prairie dog when it gets swept up by the falcon. Following fifteen seconds of silence, a timid voice asked, “What was that?” I replied that, “You told me to panic, so I did.” The lab didn’t call me anymore. Ever.)

Defining the 99^th percentile value is more definitive. The statistical definition is straightforward. A significant troponin value (in the context of the diagnosis of acute MI [AMI]) means that the measured value must be greater than all but 1% of all troponin values measured in normal sample population. That’s fair enough. But it’s important to note that the 99^th percentile value is usually not defined by your hospital laboratory. Instead, it’s most often defined by the manufacturer of the particular troponin test your laboratory uses, and the statistical data supporting that cutoff value is often found in the manuals and guidelines that come with the assay. (These documents often provide information about “normal” and “intermediate” sample distributions as well.)

The trick here comes in interpreting the meaning of the laboratory result. One might presume that since the 99^th percentile value of troponin meets the threshold for the laboratory diagnosis of AMI, that cut-off number might represent your panic value.

In reality, however, it’s not so simple. Your laboratory’s reported ranges for normal, intermediate, and critical values may not parallel the statistical parameters set forth by the manufacturer. Within most institutions, normal values and reference ranges for reporting are determined by pathologists, clinical chemists, and other in-house experts. The reporting ranges will be aligned when they’re established with reference to the manufacturer’s assay. But if there’s been a switch in the vendor and no corresponding change in the institutional reference ranges, or the ranges of normal and abnormal were determined through literature review and clinical opinion independent of the assay used, there’s room for problems. A test where the reported critical value exceeds the 99^th percentile raises the possibility of missed diagnoses such as Type 2 MI with significant impacts on patient care. Ranges where the value of most concern is actually below the 99^th percentile target promotes over-triage and excess utilization of healthcare resources.

Making sure your lab values correspond to clinical definitions is especially important when we consider that within the fiscal year 2022 Inpatient Prospective Payment System (IPPS) proposed rule, there is a provision that would establish the diagnosis of non-ischemic myocardial injury as CC within the MS-DRG scheme. This aligns clinically with the Fourth Universal Definition of MI concepts of acute and chronic myocardial injury, the latter of which is often due to a longstanding, non-ischemic condition such as heart failure or chronic kidney disease. The presence of an elevated troponin due to non-ischemic myocardial injury often results in additional monitoring, assessment, and care to exclude acute myocardial injury as the source of the laboratory finding.

The ability to use the diagnosis of non-ischemic myocardial injury as a CC best reflects the realities of clinical care and reinforces the need to ensure your institution’s laboratory parameters for reporting parallel those of the assay’s manufacturer. Doing so will help ensure query opportunities are not missed and ensure accurate capture of the complexity of care.