What you don’t know might hurt you…the potential negative impact of CDI programs on quality outcomes data

October 12th, 2015 / By Cheryl Manchenton, RN

We’ve all heard the phrase…”what you don’t know won’t hurt you.” That might be true in some settings, but in the world of documentation improvement this is definitely not the case.

Let’s look at a few commonly queried diagnoses and their impact on quality profiles. The first one is acute blood loss anemia (ABLA). Certainly this can be a diagnosis present on admission (POA), but many times it is a diagnosis clarified in the postoperative setting. And heaven forbid the provider document dilutional anemia even though it might actually be the case! On the plus side, this may increase reimbursement or impact severity of illness. On the negative side, ABLA not present on admission is a potentially preventable complication (PPC). Several states are now requiring hospitals to report their PPC rates with decreased reimbursement to those with higher than acceptable rates. Also recently reported, a commercial payer is now profiling their orthopedic surgeons based on acute blood loss anemia (as well as other complications) and placing them into two tiers. Those in the lower tier have high rates of ABLA and other complications.

So what do we do? Not query and receive decreased reimbursement? Query with a stain on the provider’s scorecard? How about honest clarifications when we can see the care change or the patient be affected by anemia? And we have to be honest with our orthopedic surgeons so that they don’t document it on every case just to avoid a query.

Another commonly queried diagnosis is encephalopathy. Obviously, capture of this diagnosis impacts reimbursement or reflects a sicker patient even if not present on admission. But yet again, encephalopathy is problematic. Not only does its presence increase the risk of external audits, it is also a PPC when not POA. What to do? Clarify only when you see increased care, investigation (I do not mean a CT on admission) or true impact. For example, altered mental status due to a UTI in the elderly is common and usually 24 hours of antibiotics tends to resolve it. The patient’s stay is not extended in any way and, commonly, it does not require extra nursing intervention, pharmaceutical management or further workup. So how can an institution justify capturing increased reimbursement that they did not “earn” and at the same time place the case at risk and increase the PPC rate?

Ok one more: acute kidney injury (AKI). You guessed it…it is a PPC when not present on admission. If it is POA, the provider should have noticed the increased creatinine and have been making adjustments to meds, holding diuretics or other offending medications or gently hydrating the patient. Is it really AKI if the CDI specialist is the only one noticing it? And be careful in patients with known kidney disease. The KDIGO criteria lists the second criteria as an increase of 1.5 times the baseline in a patient with known renal disease. So that increase of 0.3 in creatinine should not be used in a patient with chronic kidney disease.

One last thought. When is the last time you queried POA status (when clinically appropriate) of diagnoses such as UTI or acute MI when there was no impact on reimbursement? Yup, both conditions not present on admissions are PPCs. Until CDI programs focus more on complete and accurate documentation, programs may be doing harm when it was not intended. What you don’t know is hurting your hospital’s quality profile.

Cheryl Manchenton is a Senior Inpatient Consultant and Project Manager for 3M Health Information Systems.


Today’s value-based care models require a different kind of quality measure. Learn which quality measures lead to improved patient outcomes.