Variance Trends Metric Overview

Variance Trends 

The Variance Trends metric dashboard is available at sites with nursing care area coverage. The dashboard highlights trends in documentation accuracy across departments, medications, and nursing team members. A variance is defined as any patient-medication combination identified in the audit process with incomplete or inaccurate documentation. ControlCheck users can use the application's assessment of priority to narrow down areas of greater risk and focus investigative energies.

Access the Dashboard

Navigate to the Analytics tab in the main dashboard menu and select User Analytics in the dropdown menu. This will forward you to the IRIS Dashboard where you can then select the Variance Trends Metric link to view its dashboard. 

Filters

The Variance Trends metric dashboard allows for the following filters:

• Start Date - allows you to select a start date for your data
• End Date - allows you to select an end date for your data
• Status
  • Closed - Unable to Reconcile - Items marked as discrepancies in the audit review process 
  • Closed - Reconciled - Items marked as resolved without further issue in the audit review process
  • Open > Than 3 Days - Items in the audit table's 'Open' queue for more than three days
• Standard Deviation
  • None; +1, +2, +3, +4 standard deviations from the norm

The default date range for all dashboards is the previous 30 days. Selecting a standard deviation filter will filter the results to show only data that is equal to or greater than the threshold selected.

Understand the Key

Priority is indicated by the following colors: Red-High Risk, Yellow-Medium Risk, and Green-Low Risk. Priority is also assessed based on a proprietary outlier priority assessment - in simple terms, by incorporating information about both the variance rate and the total variance counts. More specifically, the bars are ranked highest to lowest variance count, modulated by a function of the total number of medication events.

• Higher percentages of higher overall counts will be ranked higher in the order.
• A high variance - even as high as 100% - in a low-volume area may rank below a lower variance rate in a higher-volume area.

Most facilities should expect to see 30 - 40% of total measured data assessed as High and Medium Risk, but there is no requirement that results will include all three priority rankings. Based on the overall variability of a facility's data from the facility's norm, you may have a cluster of Medium Risk events with no High-Risk events, or a cluster of High-Risk events with no Medium Risk events. Some facilities may find all metrics assessed as Low Risk, even with some variance rates within the data.

The overall count of medication events for a specific metric - department, medication, individual is indicated by the length of the bar. Those medication events with no issues (those that are not variances) are indicated by the grey portion of the bar.

The variance rate is displayed as a percentage and reflects the number of variances compared to the total number of medication events.

Interact with the Dashboard

Users should first select a date range and any additional desired filters to reflect the nature of the investigative question or to limit the scope.

Based on the applied parameters, the Variances by Department bar display will be updated. By default, the results will be filtered to show only the top 25 departments; the full list may be viewed by deselecting the View Top 25 filter. The results are ranked by our proprietary outlier priority assessment (see definition above). The X-axis reflects the total number of medication events.

Users can hover over any department bar to view the summary statistics pop-up window. Once a specific department has been selected, the department's view will collapse, the Investigate Data option will activate in the pop-up window, and the medication and user detail views will populate.

By default, the Drug Detail and Nurse Detail results will be filtered to show only the top 25 entries in each; the full list may be viewed by de-selecting the View Top 25 filter. The Drug Detail results are grouped by generic medication.

Investigate the Data

To view the specific, detailed variances at the department level, first select the specific department for analysis. Once a department is selected, the pop-up window with the department's summary statistics will appear. Select the Investigate Data link in the pop-up to view all of the variances in that department; the results will load in a new browser tab. If a department has no variances, the Investigate Data link does not display.

To view the specific, detailed variances at the medication level, first select the specific department for analysis, then hover over the medication for analysis. On hover, the pop-up window with the medication's summary statistics will appear. Actively click on the medication's bar to open the detailed investigative data; the results will load in a new browser tab. If a medication has no variances, clicking on the bar will not have any results.

To view the specific, detailed variances at the individual user level, first select the specific department for analysis, then hover over the user for analysis. On hover, the pop-up window with the user's summary statistics will appear. Actively click on the user's bar to open the detailed investigative data; the results will load in a new browser tab. If a user has no variances, clicking on the bar will not have any result.

A Note on the Data & Methods

A single variance with activity from multiple nursing team members will appear in each nursing team member's metrics, but only once in a medication or department's counts. A single variance with activity in multiple departments will appear in each department metric, but only once in a medication or nursing team member's counts.

Documentation errors are excluded from the variance counts in this dashboard. Documentation errors are defined as automatically closed event summaries whose variance is not due to missing or incorrect medication amounts but instead due to a clinician's specific misstep in medication selection: the same medication, different concentrations.

Some facilities may also have documentation errors related to the selection of outpatient versus inpatient identifiers. These are also excluded from the variance counts in this dashboard.

Open variances created within the preceding three days are excluded from the variance counts in this dashboard.

Any variances connected to unmapped identifiers (medication, user), regardless of status or state, are excluded from the variance counts in this dashboard. Once the mapping is completed and the related data is processed, the resulting variances will be included in any relevant metrics.