Emergency Care: CDI in the ED Part II

Amid the chaos of the ED, order is needed.

EDITOR’S NOTE: This is the final installment in a two-part series on clinical documentation integrity (CDI) in the emergency department (ED).

It is my opinion that investing CDI resources into the neglected stepchild that is the hospital system ED is a worthwhile endeavor.

Getting diagnoses captured early and repeatedly creates performance improvement. Even more importantly, improved documentation improves patient care, and that is the overarching goal of all healthcare institutions.

There are three main concerns when discussing ED CDI:

1. When and how to do it?
2. Independent or combined with case management (CM)?
3. Which conditions to focus on?

The answer to “concurrent or retrospective” is …yes. If you can station personnel in the ED for part of the day, you can perform concurrent querying, education, and training for the times when no one is present. Expect it to take longer for the ED and CDI folks to get to know one another; EPs work shifts, there are often part-timers or locum tenens, and their schedules may not be as predictable or regular as that of a hospitalist working seven days in a row.

Opportunities for querying are likely to be numerous and rapid-fire in the ED setting, depending on the census. CDI specialists (CDISs) should log their queries to be able to analyze their return on investment. Expect a high yield.

In an ED with multiple healthcare providers (HCPs), education about a specific patient and a specific topic represents an opportunity to teach multiple providers at the same time. This may involve HCPs tasked with the same patient (resident and attending), like when rounding on the floor, or there may be an exemplary case so significant that the CDIS expands the education to include others on duty who are not caring for that patient. EPs, like other HCPs, prefer case-based practical education to theory. However, there may be occasion to hold impromptu education sessions during lulls in the patient flow.

Emergency physicians (EPs) and other emergency HCPs, such as emergency residents, physician assistants, and nurse practitioners, are great at generalization. If you teach them to specify the type of fracture of a patient’s ankle and why, they will carry the practice on to other fractures. If you teach them the terminology of “functional quadriplegia” at 2:30 p.m. on a Tuesday, they will use it on an appropriate patient at 4 a.m. the following Saturday night shift.

Retrospective queries may be a little more challenging and require some system jerry-rigging. First, EPs have to be taught what a query is and what to do about it. There must be some mechanism to create an addendum or append their record, because the patient will no longer be in the ED or under their care. In my system, I created a “Documentation Clarification Note,” which made it palatable to generate post-discharge documentation.

If you use an interfaced query system, the EP needs to build in time to address any query that pops up, either during the shift or between shifts. EPs often don’t use the system email, so you may need to find a way to let them know they have a query. Most EDs have some mechanism of communication; you have to know that there is a potluck dinner on Thanksgiving and who to tell what you are bringing! The CDIS may need to write a note and put it in the EP’s mailbox, saying “please check the electronic query system for a query.”

In an ED that has residents or mid-level providers, CDI needs to decide if the query goes to them or the supervising/attending physician. If it’s the latter, you might be able to catch the HCP between draft documentation and finalization, meaning no formal addendum is required.

I worked in an ED that had a case manager present from 8 a.m. to 11 p.m. When we put in for a bed, she would warn us if the patient did not meet inpatient criteria and might recommend observation status. This advice was invaluable, and it informed future status decisions, as well as prompted beefing up the documentation to demonstrate medical necessity. I could envision cross-training personnel in CDI and medical necessity constraints specifically for the ED, because it is a confined domain and most EDs don’t have enough cases to keep someone busy doing one or the other exclusively. There could be a limited CDI repertoire of conditions, because this is not the sole opportunity to capture diagnoses, and there are fewer documents to scrutinize. I do not support doing this cross-training in the inpatient setting, however.

If the institution/system already has a case manager cross-covering the ED with limited availability, then CDI should assign ED coverage to a CDIS. Performing ED CDI remotely is probably suboptimal. Initially, I would recommend intensive, in-person CDI coverage, especially for the educational aspect. After the program matures, remote CDI with occasional visits might be adequate.

EM has some expected CDI features:

• Excellent documentation should be done on all patients. Patients in observation status convert to inpatient not infrequently. You don’t want to miss important diagnoses because “they aren’t being admitted.”
• Specificity is a little less imperative. In the ED, we often do not know certain specifics; they need to be teased out upstairs over the next few days. Get the acuity in place and the hospitalist can provide the details. Hierarchical condition categories (HCC) specificity may not always be possible, but if the ED CDIS is knowledgeable about HCCs as well as comorbid conditions and major comorbid conditions (CCs/MCCs), everyone benefits.
• EPs have had, “don’t use ‘rule out’” hammered into their heads. You will need to teach them about uncertain diagnoses. Tell them to marry the sign/symptom with the suspected diagnosis (e.g., “chest pain, rule out MI.”) Caution them that the differential diagnosis they have been taught to provide to demonstrate complexity of medical decision-making is not sufficient.
• Don’t describe, ascribe. It isn’t a sodium of 128, it is “hyponatremia.”
• Linkage! Use words/phrases such as “secondary to, due to, as a result of, from” liberally. Describe your thought process.
• Be consistent. Don’t admit a patient for “sepsis” and describe her as “non-toxic.” Patients can’t be “alert and oriented” and comatose at the same time.
• Think about what is impacting the diagnosis and document all the relevant comorbidities. There should be multiple diagnoses. This bolsters your complex medical decision-making for your level of service.
• Think about present-on-admission (POA) conditions and be sure to document them if present. Check the patient’s skin for decubiti; consider sepsis, central-line associated bloodstream infections, UTIs (don’t ask for CAUTI – if the patient comes in with a UTI related to a Foley catheter, it is not a “CAUTI,” although it is a T83.511A, infection due to indwelling urethral catheter.)
• Trauma is rife with opportunity, but it is very fast-paced and high-stakes. It may be hard to document concurrently. Consider attending trauma conferences (many systems have combined trauma/ED morbidity and mortality conferences to learn from bad outcomes) and giving CDI pearls.
• Being stationed in the ED may also afford the opportunity to educate and guide consultant or admitting physicians. Any providers who venture into the ED are fair game for the CDIS.
• There are predictable conditions/diagnoses the CDIS should be on the lookout for:
  • Respiratory failure (acute and chronic)
  • Sepsis and concomitant sepsis-related acute organ dysfunction
  • Gram-negative or aspiration pneumonia
  • Heart failure (may not get systolic or diastolic specificity, but consider in “fluid overload,” “pulmonary edema”)
  • Type 2 MI
  • Functional quadriplegia
  • Encephalopathy (don’t forget cases of “superimposed on dementia”)
  • Acute kidney injury
  • Shock, hypotension
  • Acute blood loss anemia
  • Acidosis
  • Liver dysfunction, including coagulopathy
  • 32 Hemorrhagic disorder due to extrinsic circulating anticoagulants (don’t need this exact verbiage – need linkage of coagulopathy with anticoagulation, and likely worsening or complication of bleeding from the coagulopathy)
  • Severe protein calorie malnutrition (If you have to stand at the door to make sure you see the skin between the ribs moving because they look like death, you can make this diagnosis in the ED)

Without leading, make it as easy as you can for the EP to understand what you are asking and what is at stake. They are juggling multiple balls and don’t need a lot of extra work. If you want them to call it “encephalopathy” instead of “altered mental status,” they will be delighted to save two words. Help them make macros or acronym expansions in the electronic health record (EHR).

Set up a system to give them feedback. You may want to keep statistics or make a dashboard. Everyone learns from feedback.

If you can get your providers to understand that the goal is to make the patient look as sick in the medical record as he or she looks in real life, you will pave the way for them to accept the quirks of CDI documentation to get them credit for caring for that sick patient. Emergency physicians are adaptable. Starting a CDI project in the ED will yield great results.

Give that neglected stepchild the attention he deserves and you will be amply rewarded!