Patient Display

---

Fig 10. Patient Display

The Patient Display Window (Fig 10, right) appears when an item in the Patient List Window is double-clicked and it shows data for individual patients and tests. Patient identification information is shown at the upper left, data for the individual events is shown at the lower left, and a graph of the data with reference range(s) shown in gray is at the lower right. Physician and location information is taken from the most recent event. The Data List (lower left) contains the date, time, X-axis value in days, and result for each event. For tests that are drawn in a peak/trough manner, a peak, trough, or random indicator is shown in the rightmost column. If a rule set exists that matches the test, the name of the rule set will be used above the value list and graph. If there is no matching rule set, the test code from the data file will be used in place of the name.

If the patient list has been scanned using rules, the names of any rules that were triggered by the data will be shown in the Flags/Comments List at the upper right (the names of the individual rules are specified in the Rule Builder window and are displayed in the Rule Set window). The list is arranged chronologically and includes the date and time of the first event that completed the pattern specified by the rule and thus caused it to be triggered. The number of events that the pattern persists is shown after the time of the first event. If the pattern is flagged for only one event, "1" will be shown; if "3" is shown, that means the pattern persisted for two additional events after the first one. Persistence of patterns is discussed more below. If the patient list has not been scanned, the Flags/Comments List shows any comments (e.g., a history, sample instructions or diagnosis) associated with specimens.

The Patient Display Window contains a Button Bar at the upper right for management of the display and analysis:

The four Plot Buttons on the left manipulate the graph. The leftmost button regraphs selected data points (or all data points if none are selected) in a linear plot. The second button regraphs selected or all data points on a semilog plot; the third button creates a semilog plot with a linear least squares curve fit and half-life calculation (for negative slopes). The fourth button saves a copy of the graph (at currently displayed size) to disk as a PICT (Mac) or BMP (Win) file for import into presentation programs, etc. The saved file has the size and shape of the graph at the time it is saved, so larger or smaller graphs can be saved by resizing the graph to the desired size and shape before saving.

The "disclosure triangle" just below the leftmost button hides the Flag/Comment List and expands the graph to fill the right side of the window. Clicking the triangle again shrinks the graph and re-displays the list.

or The Notepad Button displays a text editing window for entering text to be saved with a particular patient. If text has already been saved for a patient, the Notepad will show "writing" and clicking it will display the current text for editing or deletion. Text is added or changed in a patient's record by entering it in the window and clicking the "Update" button in the window (clicking "Delete" deletes all added text). Added text is saved in patient archives. If the note window is left open when moving between patients using the Arrow Buttons (below), the text will update appropriately.

or The Flag/Comment Button toggles the Flag/Comment List between displaying rule flags and sample comments. It is disabled if no information is available. The Flag/Comment Button is also disabled when the graph is expanded and the Flag/Comment List is hidden using the disclosure triangle (see above).

or The Checkmark Button marks a patient for printing, list display or archiving. Clicking the button causes the patient to be checked in the Patient List Window. When patients are checked, the button has a depressed appearance. Clicking the button again "unchecks" the patient.

The Arrow Buttons are used to move to the next (right button) or previous (left button) patient in the Patient List Window. Display settings are maintained when moving between patients using the Arrow Buttons.

The Interactive Graph (shown at the right in somewhat compressed form) allows direct selection of data points from the graph. If the mouse button is held down with the pointer over the graph, cursors will appear that can be dragged to select a subset of the displayed data points (the corresponding lines of data will be selected in the data list on the left of the Patient Display window). Clicking one of the plot buttons in the button bar will replot just the selected data points and the graph will re-dimension automatically to fit them. Alternatively, one or more events can be selected directly from the data list using the mouse and shift/command/control keys as is usual for the operating system, and then replotted by clicking a plot button. This can be used for discontinuous selection (eliminating certain points from the plot). If no points are selected (i.e., by clicking in the white space under the lines of the data list, or anywhere on the Display window background) the plot buttons will plot all points.

When the Display Window opens, data is plotted by default on a rectilinear graph with data points connected. The plot buttons in the button bar can be used to select semilog plotting or curve fitting on a semilog plot (note that semilog plots do not contain a "0" value for the Y axis, so result values of zero are not plotted on them). If the semilog/fit button (3rd from left, see above) is clicked, all selected data points are plotted on a semilog plot, the regression line for the selected points is drawn, and the calculated half-life (from the slope of the regression line) is displayed. This is most useful in evaluating drug elimination rates but also may be helpful in assessing other processes that may adhere to first-order kinetics. Changes to the graph are saved within a work session and printed reports match the appearance of the graph in the Patient Display window.

Double-clicking on a result line on the Data List displays a marker on the graph indicating the corresponding data point. It also highlights the rule or specimen comment in the Flag/Comments List at the upper right of the Display window that corresponds to the clicked specimen. If a subset of data points is displayed on the graph, a marker is displayed only if the double-clicked data point is on the graph. Similarly, double-clicking a specimen comment in the Flag/Comment List selects the corresponding data line in the Data List and draws a marker to the point on the graph if the point is currently displayed on the graph.

Double-clicking a rule (pattern) name in Flag/Comment list produces a bit more complex reponse. If the pattern is not persistent (if only one event is flagged and "1" is shown after the date and time on the list), then double-clicking a rule or specimen comment in the Flag/Comment List selects the events in the Data List corresponding to the pattern, highlights the data point on the graph in red and draws a marker on the graph indicating the data point. In addition, the preceding data points that contributed to the flagged pattern are shown in dark red (see example at right). Note in the example that the rule for "Decreasing values" is highlighted. The corresponding rule in this case happened to be "3 values decreasing at more than 1.0 per day with the last value less than 1.0." The three highlighted points on the graph satisfy this rule. The marker on the graph is drawn to the point at which the rule was fully satisfied and this point is highlighted in bright red. The two preceding points that contributed to the pattern are shown in dark red. All three of the data points involved in the pattern are selected in the Data List when the line in the Flag/Comment List is double-clicked. The subsequent points do not continue to satisfy the rule, so they are not included in the pattern.

LabScanner supports the concept of persistent patterns, in which subsequent points after an initial pattern is identified continue to meet the criteria for that pattern. In the Flag/Comment List at the right, the selected pattern has a persistence value of "3," indicating that two additional points after the first flagged point continue to meet the criteria for this rule. When this line is double-clicked, all the points in the initial pattern plus the subsequent two points are selected in the Data List. On the graph (see right), the first point completely satisfying the rule is highlighted in red and indicated with a marker as above. The preceding points contributing to the pattern are highlighted in dark red, also as above. In addition, the subsequent points that continue to meet the pattern criteria are highlighted in red and the last point is indicated with a second marker. Persistent patterns thus have a start point (the first marker), an end point (the second marker), an event duration (the number of events between the two markers, inclusive) and a temporal duration (the time difference between the two markers). These values are tabulated and reported in the statistical summary of a data set which is discussed later.

Markers and highlighting can be removed from the graph by clicking on one of the plot buttons to refresh the display.

Peak-trough drug level values are displayed as closed (peak) and open (trough) circles on the graph. They are correspondingly marked as "P," "T," or "R" (for random) at the right side of the Data List. Both peak and trough reference ranges are shown, if entered into the corresponding rule set. By default, peak data points are connected to other peak data points and troughs to troughs. Unlabeled (random) values are plotted as gray points without connecting lines (see the point at 17-18 days on the graph above). Because samples are not generally drawn at every dosing interval, connecting peaks and troughs to each other would not make physiologic sense. Connecting peaks and troughs separately allows a rapid assessment of the relationship between peaks and troughs and any correspondence between changes in either. Note that semilog curve fits include any points that are selected at the time of the fit (peak, trough or random), so analyses that include peak, trough and random values can be carried out by selecting the desired data points directly and clicking the "curve fit" button.