AHLTA E&M calculator HPI change
Several of you have noted changes with 838U21 with relation to the E&M Calculator’s calculation of HPI and ROS. This change is due to a change in the calculator made by Medicomp (several Medicomp releases ago, but first surfaced by us with the recent ICD9CM code updates). It is relevant to all builds with this recent 3rd QTR 2007 update, including 838u21 and 3.3u1.
I’ve been reviewing these changes (in the application and with Medicomp) and am starting to get a better understanding of what it is doing. Essentially they did a radical overhaul of the HPI and ROS calculation after a thorough review by CMS, and in collaboration with multiple university medical centers. While this brings the calculator more directly in line with the CMS guidelines and auditor requirements, it is a significant change to what everyone has learned and has been taught.
; HPI is now much more conservative and ROS is now somewhat more liberal.
; For many encounters, this trade-off will be a wash in the end-result
; It is now irrelevant how many HPI terms are selected. This old term-
counting method in HPI was ruled invalid by CMS auditors.
Following are some details:
The HPI is now calculated completely differently and rather conservatively- ONLY the attributes (location, quality, severity, duration, timing, context, modifying factors, and associated signs and symptoms) assigned in the Medcin knowledgebase for a given Medcin term are counted in the HPI calculation.
0-3 attributes = level 1
4 or more attributes = level 2.
No more counting of terms, it just counts attributes.
Important Points and Caveats:
; More than 1 term with a given attribute still only counts as a
o Example: headache & neck pain- both have the “location” attribute,
and thus only count as 1 attribute when documented together
; HPI free-text terms are counted as 1 attribute (each) in AHLTA
o HPI Free-text terms are officially “HPI”, “Preliminary
Background HPI”, “Supplemental HPI”
o “Subjective” and “Encounter Background Information” do NOT count
o Tricks of uncertain use (maybe in AIM forms??):
; Put a “2” in the “value” field for an HPI free text term
and it will make the HPI calculate to level 2
; Put any combination of the letters LQSDTCMA (first letter
of each attribute) and it will count those attributes in
; Free text on a symptom or other term doesn’t count at all
o This is where most of the “attributes” (the story) are often
included as free text
; Terms with out attributes (e.g. “nausea”) get counted only as the
“associated signs and symptoms” attribute
o Regardless of how many there are
o Thus current behavior of checking of symptoms will virtually
always get this attribute counted.
; Entering Duration or Onset will qualify as the “Duration” attribute
; The “Modifiers” drop-down selections do NOT get counted as an
attribute (though they should- usually as “Severity”- and this has
been reported to Medicomp to fix).
; Supposedly, each finding in the encounter that has a prefix of
“reviewed PMH for” gets counted (as an “attribute” equivalent), for
some reason, but I haven’t been able to confirm this in AHLTA.
In practice, to get 4 attributes takes a bit of work:
; “Location” gets counted for most, but not all symptoms (e.g.
headache and neck pain, but not nausea and fever), ; An additional symptom should count as the “associated signs and
; “Duration” will usually require using the Duration or Onset widget
; Most other attributes (severity, timing, context, modifying factors)
will typically require drilling down the tree
; Making sure you use the right free-text anchor term will get you an
(unspecified) attribute count
--> Obviously, with any amount of free text detail that includes the key
attributes, the user will likely need to just click the 2 level next to HPI
in Disposition to get appropriate credit.
Details Pane Notes:
; The HPI Details pane ONLY shows the FIRST item that counts towards a
given attribute (I find this highly confusing, but that’s what they
; The HPI Details pane “Count” column has been rendered almost
; The calculator (under the hood) can now report the associated
attributes, which is what is now relevant, but this isn’t turned on
in AHLTA (wasn’t documented or anticipated)
The ROS calculation has been made more liberal in that ANY HPI symptoms can ALSO count in the ROS calculation.
; Thus it is no longer necessary to put ANYTHING in the ROS section to be
able to get some ROS credit.
This change will often, but not always, offset the more conservative HPI calculation
Please see additional details in the “AHLTA E&M Calculator Business Rules.doc”, updated today (attached). This document is based mostly on Medicomp’s documentation with additional AHLTA-specific information.