(This document should be read in conjunction with the complete LOINC Users' Guide, available as LOINMAN1 in either WordPerfect (.WP6 extension) or MS Word (.DOC extension) from http://www.mcis.duke.edu/standards/termcode/loinc.htm )
Release Notes
The first release of Clinical LOINC was made available via Internet on December 3, 1996. It contains 1,492 observations and 37 fields. The first 35 fields are identical to the fields available in Laboratory LOINC; fields 36 and 37 (ANSWERLST2 and ANSWERLST3) provide an overflow for ANSWERLIST values that exceed the 254 characters allowable in Foxpro dBASE and Microsoft Excel files.
4.0 Clinical observations and measures
4.1 Introduction
We have applied the same general approach to common clinical observation measures as to laboratory observation measures. We use the same six major parts of the name, some with subparts. In our first release we have addressed the types of observations listed in Table 1.
Table 1
Subjects covered to date in clinical LOINC
|
Blood pressure (systolic, diastolic, and mean)
|
| Heart
rate (and character of the pulse wave)
|
| Respiratory
rate
|
| Critical
care measures
|
| Cardiac
output, resistance, stroke work, ejection fraction, etc.
|
| Body
weight (and measures used to estimate ideal body weight)
|
| Body
height
|
| Body
temperature
|
| Circumference
of chest, thigh, legs
|
| Intake
and output
|
| Major
headings of history and physical
|
| Major
headings of discharge summary
|
| Major
headings in operative note
|
| Electrocardiographic
measures
|
We assign LOINC code numbers to the clinical LOINC variables from the same sequence as we assign laboratory LOINC variables; that is, we have one unique sequence of numbers that includes both laboratory and clinical variables.
For most of the measures we include specific observations for summary data, e.g., minimum and maximum systolic blood pressure per hour and per shift, and separate measurements for the various estimated, reported and measured values. (E.g., a patient's report of his or her body weight is a different variable from a measured result or the physician's estimate.) We also provide varying degrees of pre-coordination for the observation, the body site at which it was obtained, and the method. E.g., a cardiac output based on the Fick method is distinguished from a cardiac output computed from 2D cardiac echo data.
Physiologic measures are often monitored continuously over time, and the instrument reports summary "statistics" over that reporting period. For vital signs these can include minimum, maximum, and mean value over a time period. For intake and output the total is the summary statistic usually reported. When we address measures taken over time, we usually include 1 hour, 8 hour, 10 hour, 12 hour, and 24 hour summaries. Durations of 8, 10, and 12 hours are included to cover the varying lengths of work shifts within and across institutions, and the summary statistic is usually chosen to coincide with shift duration.
The parts of clinical measurement names are the same as for laboratory measures. The fourth part, the system, usually identifies an organ system or a particular part of the anatomy. For a measure of systolic left ventricular pressure, the system would be "CARDIAC VENTRICLE.LEFT". In contrast to laboratory tests, where the component is usually some chemical entity, for clinical measurements the component usually identifies the specific aspect of a property that is measured. For example, the property type might be pressure. Then the component would identify the pressure measured as intravascular diastolic. In general the component is used to distinguish the various points or ranges, or inflections of a physiologic tracing, and to define precisely which of a number of possible dimensions of length or area are being measured in imaging.
Laboratory measures tend to be more regular than clinical measures. The system is usually a specimen and the component a chemical or molecular moiety, so in some sense the component is an attribute of a specimen. For most clinical measurements, the component is also an attribute of a patient or an organ system within a patient. However, attributes of non-patient entities are also often of interest in the case of clinical measurements. For example, we might want to know the class of instrument used to obtain the measurement: what was the vendor model number of the instrument used to measure the cardiac output; what was its institutional inventory number? Infection control might want the latter reported in order to track potential nosocomial infections. When attributes of an instrument or device are being reported, the system is the name of instrument. The same is true when we report characteristics of tubes used to move fluid in and out of body cavities. For example, we might want to report the size and type of a nasogastric tube. Variables are assigned for reporting such information.
To accommodate the special dimensions of clinical observations we have introduced new options for the kind of property, timing aspect, system, precision, and methods. Most of the new kinds of property are what you might expect from the new kinds of dimensions being measured (e.g., resistance, voltage, work per beat). However we have also introduced two important new properties:
ANAT is a special case of PRID (or taxon) which applies to anatomic sites only.
IMP (impression) is a diagnostic statement, always an interpretation or abstraction of some other observation (a series of test results, an image, a total patient), and almost always generated by a professional. (We could also consider the EKG cart's automated diagnoses as impressions.) Impressions are used in laboratory medicine as well as clinical medicine, so you will see them appearing there as well.
We have added a few more concepts to the timing aspect and have identified a new subpart for it as well. The subpart allows us to identify the statistical summary we are reporting over an interval of time. For instance, a report of the maximum systolic blood pressure over an hour of observation would be identified as 1H^MAX, the minimum over an hour would be 1H^MIN, and the mean would be 1H^MEAN. (In laboratory measures the mean is the default value for time). To date we have defined MIN, MAX, MEAN, FIRST, and LAST. (ENCTR^FIRST and ENCTR^MAX will be used to report certain values from emergency department and hospital visits for some purposes, which we have not yet released.)
We have also made a change to the fourth subpart of the first part of the name, (described in Section 2.1.4 of the LOINC Users' Guide). This is the part of the name that in the past has distinguished measures made on the patient from those made on the control, a blood product unit, a fetus, or other entities that are related to the patient but whose values are not the same for a given measure as that of the patient. While working on fetal ultrasound measures (not yet released) it became apparent that this aspect is really a part of the system definition, so we have moved it to the second subpart of the system. Using this rule, we now distinguish a head circumference of the patient versus a fetus by:
CIRCUMFERENCE:LEN:PT:HEAD^PATIENT:QN
and another measurement
CIRCUMFERENCE:LEN:PT:HEAD^FETUS:QN
The same would apply to distinctions between blood product units and controls vis-a-vis the patient.
With clinical terms we almost always have two ways of reporting. Using the first, we can report an observation by reporting a number of atomic variables which together fully describe the observation. For example, we have the following atomic observations for circumference measures. These variables let us deal with all of the odd ball kinds of circumferences for which we have not yet defined a pre-coordinated term
CIRCUMFERENCE:LEN:PT:BODYPART.XXX:QN
|
The actual measure of some circumference
|
CIRCUMFERENCE SITE:ANAT:PT:*:QL
|
Identifies the body part measured (specifies the system)
|
CIRCUMFERENCE METHOD:TYPE:PT:BODY PART.XXX:*
|
Identifies the measuring technique used to obtain the circumference (answers =
tape measure, derived, imaging)
|
We also provide pre-coordinated terms that combine some of the atomic variables into one LOINC code. For example we have
8279-2 CIRCUMFERENCE.AT NIPPLE LINE:LEN:PT:CHEST:QN
and
8290-2 CIRCUMFERENCE^INSPIRATION:LEN:PT:CHEST:QN
which provide more specificity and permit the key components of the measure to be expressed as one variable as is the convention in many clinical systems. We call these pre-coordinated codes "molecular" variables.
Within the LOINC database molecular variables will vary with respect to how many atomic components are aggregated. As is true in some laboratory areas, methods often not included as part of a name, nor are they always reported. The commonest molecular aggregation is between functional measure and a particular site of measurement. (E.g., the many different intravascular sites for blood pressure measurements.) But in some cases the molecular
variables represent combinations of specific measures and particular methods (e.g., the cardiac output measures). Please note that most molecular variables could also be accompanied by one or more atomic measures to provide special information about the measure, e.g., special circumstances of the measure, or the vendor model number or institutional inventory number of the measuring instrument.
We have introduced one more convention in this release. When we have a variable that really reports what would have been contained in the name in a fully pre-coordinated term, will have an asterisk placed in the part that will be reported as a value. For example, a variable that is used to report the anatomic site as an atomic variable, would have an asterisk (*) in the system part of the name. The variable used to report the method of a particular measure would have a asterisk (*) in the method part of the name.
This is a work in progress. We may not have included all of the necessary atomic variables for each of the primitive measures of interest. A few terms may have escaped our conventions during editing. In the next months we will be adding variables to the LOINC data base for cardiac echo and obstetrical ultrasounds, as well as for the variables in the Data Elements for Emergency Department Systems (DEEDS) emergency visit reporting data base. Variables for respiratory therapy and ventilator management are a high priority so that we can complete the terms needed for typical inpatient data reporting.