Six Steps to Prevent Readmission Denials

Prevention of readmission denials improves case mix index (CMI), reimbursement, and quality metrics.

As a consultant, I am like the FBI – if you invite me in, I will likely take your case. This often leads to varied and challenging projects.

Often, I am engaged to assess opportunity in provider documentation, clinical documentation integrity (CDI) competence, or a focused condition, like sepsis or postprocedural respiratory failure.

Recently, I did a project for which I was evaluating 30-day readmissions to see if I felt something done or an action not taken on the index admission had resulted in the second admission. If that were the case, the second admission would get denied and bundled into the DRG payment for the first admission.

I believe that the system as it is intended should work; no one wants to see fraud and abuse from the provider side, but we also want denials to be justified and not gratuitous. It is very enlightening to work on the other side of the aisle and empowering to be part of the process.

I approved approximately 95 percent of the cases. My final determination was to approve, deny, or have someone else from the payor side weigh in because I found the case questionable or equivocal, but I also made a notation as to whether I felt the second admission was related or unrelated to the index admission, or whether it was causal. The majority of cases were related, which meant that the disease process that elicited both admissions was the same.

There were common themes that brought encounters to my attention:

• Repeat visits for the same malady in patients with bad disease processes, like recurrent exacerbations of chronic obstructive pulmonary disease (COPD) with acute-on-chronic hypoxic and/or hypercapnic respiratory failure, or chronic liver failure, or sickle cell disease. Heart failure and atrial fibrillation were also frequent offenders causing readmissions.
• Repeat visits for the same condition because the patient was noncompliant with discharge instructions, like if they didn’t fill their prescription or take their diuretic. There were also clinical scenarios in which the patient/family declined a recommended investigation or treatment – this cannot be counted against the hospital/provider.
• Complications from the treatment from the first admission sometimes caused the second admission, but I only recommended denial if it was predictable and avoidable. For instance, if a patient was put on an appropriate heart failure regimen that caused the patient to get lightheaded and fall, it isn’t the fault of the hospital/provider that the patient had an adverse effect from a properly dosed medication.
• There were completely unrelated visits, like a COPD exacerbation followed by a hypoglycemic attack due to diabetes medications, or a bout of COVID-19 contracted weeks after the index admission (as opposed to contracting it in the hospital).
• Planned readmissions don’t count unless there was no reason to discharge the patient without the procedure just to incur a second admission.
• It was less likely that I found causality if there was a long time between admissions.

Sometimes, there was causality without denial. If it was deemed appropriate to reinitiate anticoagulation, wherein the risks of being off anticoagulation were felt to outweigh the risks of being on it, and then the patient had a gastrointestinal hemorrhage, I did not deny the case. However, I also ran into the converse: the anticoagulation should have been discontinued (e.g., a diagnosis of acute deep vein thrombosis was erroneously made because the providers didn’t recognize that the condition was chronic by reviewing old records) and the risk of bleeding was high. This resulted in medical denial.

There were certain things I looked for in the records. Obviously, they are being judged on how they were documented in the encounter, because we are not there contemporaneously. The providers may have done absolutely everything right, but “if they didn’t document it (or I couldn’t infer it), it wasn’t done.” Some questions I asked myself included the following:

• Was the patient discharged on appropriate medications? For instance, if a patient had recurrent infections, were they discharged on prophylactic antibiotics, like methenamine for recurrent urinary tract infections? Was the dosage of diuretic appropriate after an admission for heart failure?
• Did the providers consider the right diagnoses from the clinical indicators and do the correct work-up and treatment?
• If there were risky treatments/medications, did the documentation reflect that the consultants had done the appropriate risk assessment and rendered a decision?
• Was it documented that the patient had improved sufficiently and was ready for discharge? Or did it seem as though the patient had been prematurely released? Copying and pasting was often the enemy of this indicator.
• Were the appropriate follow-up appointments made/recommended at an appropriate interval?
• If the discharge disposition seemed suboptimal, was there documentation as to why (e.g., a patient who seems like they should be discharged to a skilled nursing facility/SNF, but went home because the family declined alternate placement)?

I counted on the discharge summary to give me enough data to make my determination that the discharge instructions were clinically appropriate. If the discharge summary was poorly crafted, that information might be absent. My conclusion might be unfavorable (to the hospital/provider) not on the basis of bad medicine, but substandard documentation. Therefore, here are my recommendations:

1. Be sure to document an accurate disposition condition.
2. Have the narrative tell the story of the encounter. Don’t waste inordinate real estate on the admission history and short the hospital course. Detail key decisions and the thought process (e.g., “the gastroenterologist approved restarting anticoagulation for persistent atrial fibrillation since there was no active bleeding on EGD.”).
3. Have the diagnoses correspond to the work-up and clinical indicators. Best practice is to include acuity, severity, and specificity. Include the social determinants of health (SDoH) and diagnoses of noncompliance, when applicable.
4. Make sure the medications are clear and comprehensive – what was discontinued, what was initiated, were there changes in dosage? Why were changes made?
5. Make sure the disposition location is reasonable or explained (e.g., “it was recommended to send patient to a SNF, but family declined. Discharged home in fair condition with home health care daily”). Be sure the patient has the ability and information to access the services they need.
6. Follow-up appointments must be with the appropriate caregivers, and in an appropriate timeframe. There must be clear instructions as to reasons to seek further urgent or emergent medical attention. Discharge instructions regarding the clinical conditions should be given and recorded.

The discharge summary can make or break you on many levels, including readmissions. At the risk of sounding like a broken record, providers should practice excellent medicine, document what they did, have it coded accurately, and let the quality metrics and reimbursement fall where they belong.

For More Information: six steps to prevent readmission denials