Data Corruption

One of the more exciting collaborations happening in our pharmacy is the opportunity to coordinate with a commercial healthcare insurance company,  managing our mutual patients with the goal of improving therapy outcomes and reducing overall healthcare costs. The insurance company has provided us with new tools, giving us access to some of the detailed statistics they collect. The insurer is also very interested in several disease states. The insurer provides us specific patients they believe would benefit from additional workups. One of the targets is to identify patients that would benefit from initiation of a medium or high-intensity statin.

We have been working on this project for quite some time and have succeeded in initiating statin therapy in numerous patients were it would be beneficial. We identified these patients both using our own data as well as the information the insurance company provides us. Recently we encountered an examples of information in the health record kept by the insurance company that reminds us that we need to critically evaluate all information we receive, no matter it’s origin.

In this example of Tales from the Counter, we look at a patient the commercial insurer forwarded to us as a potential candidate for the initiation of statin therapy. Our patient is a 53 year-old female taking lisinopril / hydrochlorothiazide for essential hypertension. She reports she previously tried a statin, but was unsure when/why it was stopped. She denies a present medical history of diabetes or smoking.

Labs from the primary care provider included:
BP 135/89
TC 251
TG 191
HDL 61
LDL 152
Our clinical staff calculated a ASCVD score of 2.9%
The patient denies any history of diabetes or heart disease, and their 10 year risk for a cardiovascular event is low. In other words, or staff found no reason the patient should be on a statin. Yet the insurer had flagged her as a candidate for moderate-high intensity statin therapy.
This is were things get interesting. Our staff contacted the insurer to determine the basis of their inclusion of this patient on their list. After investigating, the insurer found that the patient had an annual eye exam with an opthamalogist and the billing information submitted for that visit included a secondary coding for type two diabetes w/o complications (ICD-10 of E11.9).
There was no other supporting clinical information in the patient’s clinical records, and it was unclear if the inclusion of the diagnosis in the ophthalmologists claim submission was an error, a misunderstood question answered by the patient, or an patient-reported diagnosis that the opthamalogist simply noted. The interesting part is that the inclusion of this diagnosis in a bill submitted to the insurance resulted the patient moved into a different risk grouping.
When diagnosing a patient, it is important that the assessment is reproducible. For hypertension, multiple measured blood pressures have to be elevated before the patient is diagnosed. Diabetes is the same way, with a single high blood glucose reading not being enough to confirm a diagnosis. But once information is in the healthcare record, things are presumed to be accurate, and this is the take-home lesson. It is important for all healthcare professionals and insurance providers to look deeper than just a diagnosis on the chart. Be sure the diagnosis is accurate before treating the patient. Always question the accuracy of the information you are looking at.  Don’t assume that someone else has confirmed something. Take the time to be sure yourself. Make every Encounter Count!

 

Published by

Michael Deninger

Mike graduated from the University of Iowa with a BS in Pharmacy in 1991 and completed his Ph.D. in 1998. He has over 20 years of practice experience, over half of which is as a pharmacy owner. Areas of expertise also include technology in practice, including integration with data sources.

Discover more from The Thriving Pharmacist

Subscribe now to keep reading and get access to the full archive.

Continue reading