Outcomes Framework - Time Period

Time Period is used for Effectiveness Metrics to select the time at which you would like to measure outcomes.

The Time Period designated refers to the second point in time which will be compared to the baseline/first measurement. By default, the template will define the timeframe as the third month of Treatment Episode. If the patient had more than one measurement in the designated time period, the system will automatically utilize the most recent/last measurement that falls within that time period. In the templates, the system will use the most recent/last measurement within the third month of Treatment Episode. Measurements completed in the second or fourth month of Treatment Episode would not be considered. By default, the template selects month 3 of Treatment Episode as the point in time for the last measurement for a patient to be in the Time Period. You may want to change this depending on your modality of Treatment Episode and the specifics of your care.


In order to change it, scroll down to the Time Period section within Configure Your Metric pictured below. Please note that this option will only appear if a template is selected or all fields are manually filled in and only for Effectiveness Metrics.

Mirah supports three Time Period options, which are described below.

Last measurement recorded

By selecting the Last measurement recorded option, all Treatment Episodes with at least two data points will be included, no matter how long into care they were measured.

This can be useful in a variety of situations:

  • When collection is sporadic.
  • When Treatment Episodes may go on for a long time.
  • When you want to see the final outcome, rather than freezing the Treatment Episode at a given point in time.

Each Treatment Episode is only included once, at the latest time point we have data for.

If you select this option, it is recommended to pair it with an Inclusion Criteria for ‘days between Treatment Episode start and selected measurement’ to ensure that you are only measuring outcomes where the patient has spent enough time in care to see a result.


Last measurement in month of Treatment Episode

By selecting the Last measurement in month ___ of Treatment Episode option and setting an appropriate month, you can freeze outcomes at the latest measurement taken in a particular month of care. 

The month of Treatment Episode is defined as follows:

  • Assuming the patient enrolled on e.g. April 15th.
  • Month 1 is considered the first ‘month’ of days, e.g from April 15th to June 14th
  • Month 2 is considered the second full month of care, e.g. from June 15th to July 14th.
  • Etc.

If there are multiple measurements in the month, only the last one will be used.

Measurement by month is the most common option:

  • It is not always possible to measure people in a specific week, so measuring in a month gives multiple opportunities for the patient to give data that will be valid.
  • It allows you to time-bound the analysis in a way that choosing the ‘Last measurement recorded’’ option does not.

Last measurement in week of Treatment Episode

By selecting the Last measurement in week ___ of Treatment Episode option and setting an appropriate week, you can freeze outcomes at the latest measurement taken in a particular week of care. This is the same functionality as Last measurement in month from above but in weeks instead of months.

The week of Treatment Episode is defined as follows:

  • Assuming the patient enrolled on e.g. Monday April 15th.
  • Week 1 is considered the first 7 days, e.g from April 15th to April 22nd
  • Week 2 is considered the second full week of care, e.g. from April 22nd to April 29th.
  • Etc.

If there are multiple measurements in a week, only the last one will be used.

It can be difficult to get patients to measure exactly in a seven day window, so Mirah’s general recommendation is to use Last Measurement in month where feasible to do so.



Adjusting Time Period

By adjusting the Time Period, you change which patients are included in the In time period number in the Statistics section. It specifies at what point outcomes are measured from the date of start of care. For example, you want to understand measure outcomes 5 weeks after the Treatment Episode started. You would select ‘Last measurement in week __ of Treatment Episode’ and type ‘5’ in the space. Only patients measured in the 5th week of Treatment Episode would be included in the In time period number. No outcomes outside of this window will be evaluated. The first day of the Treatment Episode is the first day of Month 1 and Week 1. Note that if there are multiple measures during that time period, only the latest time will be used. 


In general keep in mind that:

  • The shorter the period, the closer the outcomes will be from the start of care and more patients will have completed assessments. Therefore, there will be more results.
  • As you shorten the period, there is less time for Treatment Episodes to be effective, and the lower your overall success metric will be.

Aim for a middle point where most of your patients are measured, but the metric is far enough into care that patients should have a response to Treatment Episodes.


Here are some experiments to understand how the Time Period works:

  • Select Last measurement in month  ___  of Treatment Episode. Decrease the number using the arrows  down to ‘1’. Now increase the month using the arrows until In time period is 0. Notice how the In time period number decreases because generally, fewer patients are measured the further away from the start of care you go.
  • Try selecting Last measurement in week  ____ of Treatment Episode and adjust the number with the arrows in the same way as you did with months. It’s likely that no matter what week number you choose, the In time period number will be smaller than before because a week covers a smaller period of time than a month does.
  • If you select Last Measurement, you’ll notice that In time period is the same as Treated. This makes sense, because you’re taking the last available measure, which is always present if a patient completed a measure. Notice the effect it has on the Successful number. You are now including patients who may only have been in care for a couple of weeks.

If an unexpectedly few number of patients are In time period, check either of the following:

  • Check that the Time Period aligns with the time you administer a measure in your care timeline. For example, if you administer PHQ-9 on Week 12 of care, no results should exist for PHQ-9 outcomes in Week 10 of care.
  • Consider whether you need to put extra steps in place to encourage patients to fill out the measures at the appropriate time. If patients aren’t regularly taking measures, you may have issues getting enough data.