Congestive Heart Failure In A Community Hospital: Do We Document The Pertinent Findings And Interpret The Data Correctly? Will Improved Documentation Or Interpretation Lead To Improved Classification?

Congestive Heart Failure in a Community Hospital: Do We Document the Pertinent Findings and Interpret the Data Correctly? Will Improved Documentation or Interpretation Lead to Improved Classification?

Meeting: Hospital Medicine 2011, May 10-13, Dallas, Texas.

Abstract Number: 42

Category:

Sub-Category:

Keywords:

Background:

Congestive heart failure (CHF) can be a difficult diagnosis, requiring the physician to integrate a symptom complex with physical and x‐ray findings. The introduction of brain natriuretic peptide (BNP) when used correctly has improved our ability to differentiate patients with dyspnea, but when used incorrectly may lead to mis‐classification. We tested the following hypotheses: (1) for some patients admitted with a diagnosis of CHF, there is inadequate documentation to support the diagnosis; (2) for some patients admitted with CHF, an alternative diagnosis may be more likely; (3) an elevated BNP will make CHF a default diagnosis without supporting evidence.

Methods:

Retrospective chart review was performed on 100 charts, with CHF as a primary diagnosis. The charts were reviewed for predetermined signs, symptoms, BNP, and radiological finding to support the diagnosis of CHF. Patients were stratified based on a modified version of the Framingham criteria (FC) for CHF. We reviewed the charts’ documentation, findings, or treatment supporting an alternative diagnosis to CHF.

Results:

Seventy‐one percent of our patients were Hispanic, with a median age of 73 years. We found documentation of an S3 in 6%, paroxysmal nocturnal dyspnea in 10%, jugular venous distention in 19%, dyspnea on exertion in 79%, weight gain in 15%, crackles in 53%, and pitting edema in 55%. Based on this documentation and the FC, CHF could be classified as high/definite in 33%, probable in 22%, and unlikely in 6%. CHF was mis‐classified in 39%: as chronic kidney disease volume overload in 9%, as pneumonia in 7%, as chronic obstructive pulmonary disease in 3%, and as other in 20%. Seventy‐nine percent of the patients had an elevated BNP, but only 33% had sufficient supporting documentation for CHF; 7% of patients did not have a BNP drawn.

Conclusions:

(1) Fifty‐five percent of our patients had good to strong documentation to support the diagnosis of CHF, whereas in 39% an alternative diagnosis was more likely. (2) BNP was obtained in 93%. It was used in part to make a diagnosis, but commonly without supporting findings/documentation. It is unclear if some of these patients had CHF but with inadequate documentation, whereas for others the diagnosis may be based only on an elevated BNP. We are concerned that the use of BNP may have resulted in premature closure, leading to misdiagnosis/misclassification. We believe our findings are not unique to our hospital, and with CHF a high‐stakes diagnosis, monitored by the Centers for Medicare/Medicaid Services for hospital statistics. The magnitude of this problem warrants the development of a guideline‐based rating systems for the diagnosis of CHF to improve our documentation to support this diagnosis.

Disclosures: