LVSD is an impairment of left ventricular contractile performance. An ejection fraction (EF) is an index of left ventricular systolic function (LVSF) and reflects the proportion of blood ejected during each ventricular contraction compared with the total ventricular filling volume.

• Results from an in-hospital LVSF assessment test performed during the hospitalization that may have been filed into the chart after discharge should still be used.

A.  Methodology:

• Use of reports:
  • If EF/LVSF is present in both the Impression/Conclusion section of a report and in other sections of that report, the EF/LVSF from the Impression/Conclusion section takes priority.
  • If both final and preliminary reports from the same test are present, the final report takes priority.
• Consider test reports and findings from a test outside of the report that are not labeled “preliminary” as final reports/findings.

1. If only one in-hospital test performed:
   1. Use report.
   2. If no report (final or preliminary) or no EF/LVSF findings noted in report, use other sources (e.g., progress notes) that clearly reference that test. Disregard findings labeled preliminary when final findings are noted.
   3. If no EF/LVSF results from the in-hospital test are documented anywhere, skip to step 3a below.
2. If two or more in-hospital tests performed:
   1. Use report from the most recent test* (the test done closest to the time of discharge).
   2. If no report (final or preliminary) from most recent test or no EF/LVSF findings noted in most recent test report, use report from the second most recent test*. Continue working backwards (if > 2 tests) and use EF/LVSF from the most recent test that has a report with EF/LVSF findings.
   3. If no report (final or preliminary) from any in-hospital test, or no EF/LVSF findings noted in any in-hospital test report, use other sources (e.g., progress notes) that clearly reference the most recent test*. Disregard findings identified as preliminary when final findings are noted.
   4. If no EF/LVSF findings from most recent test in other (non-report) sources, use other sources (e.g., progress notes) that clearly reference the second most recent test*. Continue working backwards (if > 2 tests) and use EF/LVSF from the most recent test that has EF/LVSF findings noted in other (non-report) sources. Disregard findings identified as preliminary when final findings are noted.
   5. If no EF/LVSF results from any in-hospital test are documented anywhere, skip to step 3a below.
      
      *If you cannot determine between two in-hospital tests which was performed closest to the time of discharge, use BOTH tests:
      1.   Use both reports.
      2.   If no reports from either test, use other sources (e.g., progress notes) that clearly reference either or both tests. Disregard findings identified as preliminary when final findings are noted.
      3.   If no EF/LVSF results from either in-hospital test documented anywhere, skip to step 3a below.
3. If no EF/LVSF results from any in-hospital test documented, in-hospital test not done, or documentation is not clear that one was done (e.g., echo ordered but no documentation to support that it was ever done):
   1. Assume notations of EF/LVSF where timeframe is not specified or determinable are from assessments done prior to arrival.
   2. Review sources and use most recent EF/LVSF from prior to arrival (closest to the time of arrival).
   3. If most recent EF/ LVSF is available in the form of a report and also clearly referenced in another source (e.g., progress notes), the report takes priority.
   4. If unable to determine the most recent EF/LVSF from prior to arrival:
      1)  Compile all EFs/LVSFs where documentation is not sufficient to determine whether or not they are the most recent.
      -If there is a pre-arrival test report with EF/LVSF findings which may be the most recent EF/LVSF, remember to apply the principles outlined above in “Use of reports.”
      -In a case where there is both a test report and notation of EF/LVSF results from that same test clearly referenced in another source, EF/LVSF findings from the report take priority.
      2)  Select “Yes” if there is an EF less than 40% or an LVSF consistent with moderate or severe systolic dysfunction. E.g., “Echo done last month showed EF 25%” per H&P, “EF 45%” per progress note (test/timing not specified) – select “Yes.”

Apply the following priority order in cases of conflicting documentation within ANY ONE STEP in Methodology above (except 3d), where there are two or more different descriptions of EF/LVSF:

1. Use lowest calculated EF. Assume EF is calculated if not documented as estimated per #2 below (e.g., “EF 30%”).
   • If calculated EF < 40%, select “Yes.” If calculated EF ≥ 40%, select “No.”
2. Use lowest estimated EF. Estimated EFs often use descriptors such as “about” or “approximate.” They may be described as visually determined (e.g., “EF appears to be 35%,” “Visually EF is 45%”), or documented either as a range (use mid-point) or less than or greater than a given number (e.g., “EF < 40%”).
   • If estimated EF < 40%, select “Yes.” If estimated EF ≥ 40%, select “No.”
3. Use narrative description with severity specified. Select “Yes” if description consistent with term from Inclusion list A. If description with severity specified is NOT consistent with term from Inclusion List A (e.g., normal, mild, preserved), select “No.”
4. Use narrative description without severity specified. Select “Yes” if description consistent with term from Inclusion list B. Otherwise, select “No.”

• Consultation notes
• Emergency department record
• History and physical
• Progress notes
• Discharge summary
• Diagnostic test reports
• Procedure notes