Getting started on a registry for patients with depression

This document explains how to use standardized queries in your EMR to start building chronic disease registries. The instructions below focus on Depression. Depression affects about 5 percent of the Canadian population aged 15 years and over during any given year, and affects 12% of Canadians at some point in their lifetime [Statistics Canada’s 2012 CCHS]. The Conference Board of Canada said in a report released in September 2016 that depression costs the Canadian economy at least $32.3 billion annually in lost productivity. Despite the magnitude, burden of illness, and social and economic costs associated with depression, there continues to be societal stigma and self-stigma that discourage individuals and families from seeking treatment and support. While there are real barriers, FHTs can provide access to different treatment options through their interprofessional teams. In order for a FHT program to be efficient and effective, the right patients need to be enrolled. Normally programs require the doctor to remember to enroll the appropriate patient into a program. However, that process for a handful of patients will be slow and least effective. Instead, FHTs can use a Depression Search tool, developed by the AFHTO Algorithm Project team, to identify the patients in their FHT that have depression. Our depression search tool has been developed using the CPCSSN case definition and the input from subject-matter experts at Hamilton FHT and St Michaels’ Hospital. The tool has been tested, revised, and validated using the eHealth Centre of Excellence EMR environment. The depression search tool is comprehensive, does not require any data cleaning prior to use, and is plug-and-play. At final testing, the tool achieved 96% sensitivity and 62% positive predictive value (PPV); in simple terms, this means that if the search tool identifies 100 patients, 62 patients will actually have depression. This is a starting point for a manual validation of a depression registry. Query criteria (click to see larger view)            

The Depression query is intended for teams that do not yet have a reliable list of patients with depression and don’t have the time or resources to start from scratch in reviewing all their patients to generate such a list.  Right now, it is also only for teams with PSS or Accuro although work is continuing to expand the standardized query to OSCAR and Nightingale. The following steps will help your team use the query to generate a list of CHF patients, starting from your EMR.  

Step 1. Estimate how many patients you think this will affect.  Multiply the number of patients your team serves by 0.05 (the average rate of depression in Ontario) to get a rough idea of how many of your patients likely have depression.  If you still think this is a big enough group of patients for you to generate a registry for, carry on to step 2.

Step 2. Import the query into your EMR.  Right now, you can only do this if you have either Telus PSS or QHR Technologies Accuro EMRs.  You will likely need the help of your QIDSS, IT staff or other person who usually works with your EMR to do this.

We are in the process of creating similar queries for OSCAR. Contact improve@afhto.ca for more information.

Step 3.  Run the query in your EMR.  Again, you might need the help of your QIDSS, IT staff or other person who usually runs queries in your EMR. Running the query will produce a list of patients with depression. The list will not be perfect – probably 38% of the patients identified by the query will NOT have depression. The query gets you STARTED in building the depression registry but doesn’t do the whole job for you.

Step 4. Find the patients who might not have depression. Review the list of patients generated by the query to separate out those patients that are clearly already coded as having depression. What’s left will the list of patients who MIGHT have depression based on other data in the EMR besides formal coding.

Step 5. Prepare your physicians to review the list  Subdivide the list of possible patients with depression into separate, shorter lists for each physician. Work with your physicians to find out if they would prefer a list on paper or electronically and how they might like it sorted (i.e. by name or most recent visit or some other parameter).

Step 6.  Invite each physician to review their list of patients.  They know their patients best and can likely quickly confirm which ones do or do not have depression, even though that information might not be easy for others to find in the EMR.

Step 7.  Clean up your EMR data.  Add depression codes to the EMR for each patient that the physician confirms as having depression. This so-called “data cleaning” work is a great job for a student.  AFHTO has created a toolkit to assist members in recruiting and using students for data clean-up. Click here for the toolkit.

Step 8.  Re-run the query. After you have corrected the EMR, re-run the query to generate a list of patients with depression. This is your new depression patient registry. Going forward, you can run the query anytime you need to generate a list of patients with depression.  You can use the list to invite patients to a depression program, track progress with outcomes on these patients or any other purpose.   Once identified, you can recruit patients to your depression program and improve patient prognosis, management, and overall care. Here are some example outcome measures to apply for this program:

  • % of patients who show an improvement in PHQ-9 score
  • % of patients who show improvement on CES-D
  • % of patients hospitalized
  • % of patients with action plans
  • % of patients self-identifying as satisfied after a group session

See this space for more on resources and contacts in supporting teams set up care programs for depression. This query was produced by and for QIDSS with assistance from eHealth Centre of Excellence in support of all AFHTO members.  If you have any questions, please contact improve@afhto.ca.