Skip to main content
Creating a new NHS England: Health Education England, NHS Digital and NHS England have merged. More about the merger.

Publication, Part of

Psychological Therapies: reports on the use of IAPT services, England September 2020 Final including report on the EA in IAPT pilot

Official statistics, Experimental statistics

Data resubmission

Please note that this publication was updated on 16th December 2021 to include revised data for some providers. This is following a one off resubmission exercise that took place in August 2021 and covered the data period September 2020 to May 2021 inclusive.

16 December 2021 09:30 AM

Page contents


Outcomes in IAPT are measured in terms of three measures:

  • recovery,
  • reliable improvement,
  • and reliable recovery.


Recovery in IAPT is measured in terms of ‘caseness’ – a term which means a referral has severe enough symptoms of anxiety or depression to be regarded as a clinical case. A referral has moved to recovery if they were defined as a clinical case at the start of their treatment (‘at caseness’) and not as a clinical case at the end of their treatment, measured by scores from questionnaires tailored to their specific condition.

The Government target is that 50% of eligible referrals to IAPT services should move to recovery.4

51.6% of eligible referrals moved to recovery

Calculating Recovery rates

In September 2020, the calculation to calculate the recovery rate is performed as below:

Count_Recovery / (Count_FinishedCourseTreatment - Count_NotAtCaseness) * 100

26,077 / ( 53,888 - 3,374  ) * 100 = 51.6%


Sub-national recovery rates are published in the Monthly Activity Data File as column ‘Percentage_Recovery’.

Reliable improvement

A referral has shown reliable improvement if there is a significant improvement in their condition following a course of treatment, measured by the difference between their first and last scores on questionnaires tailored to their specific condition.

68% of referrals finishing a course of treatment showed reliable improvement

Calculating Improvement rates

In September 2020, the calculation is performed as below:

Count_Improvement / Count_FinishedCourseTreatment *100 

36,663 / 53,888 * 100 = 68%


Sub-national reliable improvement rates are published in the Monthly Activity Data File as column ‘Percentage_Improvement’.

Reliable recovery

A referral has reliably recovered if they meet the criteria for both the recovery and reliable improvement measures. That is, they have moved from being a clinical case at the start of treatment to not being a clinical case at the end of treatment, and there has also been a significant improvement in their condition.

49.1% of referrals reliably recovered

Calculating reliable recovery rates

In September 2020, this calculation is performed as follows:

Count_ReliableRecovery / (Count_FinishedCourseTreatment - Count_NotAtCaseness) * 100

24,789 / (53,888 - 3,374) * 100 = 49.1%


Sub-national reliable recovery rates are published in the Monthly Activity Data File as column ‘Percentage_ReliableRecovery’.

The chart below compares recovery, reliable improvement, and reliable recovery rates across a period of twelve months.

Consistently, a higher proportion show reliable improvement than move to recovery; this is because reliable improvement only looks at the scale of change, and not whether the referral has moved below the clinical caseness threshold.

Reliable recovery, which requires both recovery and reliable improvement, is the most stringent measure and therefore has the lowest rate.

Each quarter, more detailed data are published about recovery, reliable improvement and reliable recovery. The most recent quarterly data, Quarter 1 2020/21, can be found at:

For an explanation of the terms used and further information about how measures are calculated in IAPT see the 'Guide to IAPT data and publications' at

Last edited: 15 December 2021 2:33 pm