Skip to main content

Improving Access to Psychological Therapies self-referral process research

Read our Improving Access to Psychological Therapies (IAPT) self-referral process user research report, published in April 2019.

Scope of discovery

This short discovery explored the user experience of IAPT services and identified user needs related to:

  • workflow and information needs when managing and processing self-referrals
  • parts of the process that work well and parts that are challenging

Who we interviewed

To understand how IAPT staff manage self-referrals, we conducted seven field trips. Each consisted of a short interview with a staff member, followed by observation of the admin processes used when receiving self-referrals.

One-to-one, face-to-face interviews and observation sessions (1.5 to 2 hours) took place in:

  • Doncaster IAPT
  • North Staffordshire Wellbeing
  • ThinkAction (London, Surrey and Kent)
  • Norfolk & Suffolk Wellbeing
  • Talking Therapies Berkshire
  • Leeds IAPT
  • Time to Talk Worthing

Executive summary

In November 2018 the NHS Digital mental health team ran usability tests of IAPT self-referral forms with 12 people suffering from mild/moderate anxiety or depression. The research showed that current forms contain serious usability issues which create significant barriers to successful online self referral (User research report. IAPT discovery. December 2018).

At the same time the team conducted phone interviews with 9 IAPT services to gather insights into how services are working with self-referrals. We uncovered disparity across services, but also commonality in shared challenges and issues (Research with IAPTs final report. December 2018).

In March 2019, the team carried out research consisting of 7 onsite visits to services around England. The aim was to gain understanding of the processes, difficulties and needs of IAPT services when working with self-referrals.

The research reinforced the outputs of the previous research, finding a wide variation between IAPTs in their understanding of the purpose of the online self-referral form and what information is needed at initial contact.

Forms used by the services show a wide variation in design and content. They range from a short “registration” form that collects contact details and just enough information to determine eligibility to a full 10-screen online assessment which includes seven measures, including PHQ9 and GAD7.

Individual IAPT services are hampered by lack in-house design expertise on how to make online referral easy. Most do not involve their users in the design of their online self-referral processes.

When online referrals are received, software commonly used within IAPTs is not fit for purpose and requires time-consuming workarounds by admin staff.

The team saw some evidence that offering users the opportunity to book their own assessment appointments helps to increase online referrals and reduce DNA rates.


Insight 1

There is little agreement between IAPTs as to the role of the online self-referral form which results in a wide variation in form design and content.

Online self-referral forms used by the services we visited range from a short “registration” form that collects contact details and just enough info to determine eligibility to a full 10-screen online assessment which includes 7 measures, including PHQ9 and GAD7.

Findings and other resources:

1. Role of the online self-referral form

2. Form design

3. Measures – how and when

Diagram 1 – Information gathering pyramid

Diagram 2 – Services information requirements at referral

“Prompt and clear routes into the service with no over-complicated referral processes or opt-in systems will support engagement.” - The IAPT Manual 2018.

Insight 2

Five of six IAPT services that have an online self-referral form had not researched user needs when designing their form or collected data on form completion rate.

Five services are using forms supplied by either Mayden or PCMIS. These have been modified by each service according to their internal processes and feedback from clinicians. 

Services have not trialled the forms with users and form changes tend to be driven by internal users.

ThinkAction worked with a third party supplier to make their referral form more user friendly and their own processes more efficient. Prior to this work more than 50% of people starting the self-referral form did not complete it.


1. Role of online self-referral form
2. Form designs

“In Surrey we were getting 50 per cent of our referrals online and it’s now 90 per cent. Our hypothesis is that, we’ve not got more people coming to the service. What we’ve got is more people completing the shorter referral form.”

Insight 3

IAPT services lack in-house design expertise to make online referral easy. Some IAPTs are outsourcing the design of online referrals and bookings to third parties. This costs money and risks replicating work across the programme.

IAPT services who have problems in managing and processing self-referrals lack in-house expertise in how to design better journeys. There appears to be little sharing of experience between IAPT services facing similar issues.

One service’s response to not being able to keep up with phone referrals was to outsource improvements to their online self-referrals to an independent design agency. 

2. Form design
6. Self-booking systems 

“Key features of a better performing IAPT service [include] simple and direct access that is not hindered by a complex patient opt in or confirmation system.” - The IAPT Manual 2018

Insight 4

Patient record software used within IAPTs is not fit for purpose and requires time-consuming workarounds by admin staff.

All of the IAPT services we visited had challenges with their software, including those that used IAPT-specific software.  

For example, not all information from Mayden self-referral forms is pulled through to IAPTUS. In 2 services, we saw admin staff work around this by cutting and pasting referral info into a Word document then manually copying it into IAPTUS. This is time consuming and also increases the risk of human error. Staff report finding this time-consuming.

"Whoever comes up with it isn’t the person that uses it. If initially it had gone out to an admin team - is it user friendly? – and then they’d taken feedback it would save everybody a lot of time.” - Senior administrator, Berkshire

9. Software workarounds

Insight 5

There is some evidence that allowing people to book and rearrange assessment appointments online helps to increase self-referrals, and also to reduce DNAs, empty slots and phone work.

ThinkAction saw a significant increase in online self-referrals after introducing online booking for assessments. People are automatically taken to a self-booking form to make their own appointment which can be cancelled up to 4 hours before and rescheduled. Since adding self booking the service’s DNA rates have fallen from over 30% to below 20% for the first appointment. Online self-referrals have increased from 50% to 90%. The service has also seen a reduction in phone work and has fewer empty treatment slots. 

Three other IAPTs we visited – Leeds, Doncaster and Norfolk – were investigating online booking systems.


5. Phone “cat and mouse”
6. Online assessment booking systems

“Having a live booking system has brought us more efficiency. It would be good to be able to send a letter out and say go online and choose an appointment.” 


Role of online self-referral form

Between IAPT services there is wide variation in the understanding of the purpose of the online self-referral form. 6 of the 7 services visited use an online form for self-referral. These forms vary from a short registration form to a full online self assessment and triage.

Leeds IAPT asks 6 eligibility questions followed by a full 10-screen online assessment, including GAD7 and PHQ9.

Worthing, Berkshire and Norfolk use single-screen forms to collect demographic info, contact details, ascertain eligibility and get a brief description of symptoms. They also seek info about long-term conditions.

ThinkAction and North Staffs assess eligibility and collect minimum details required for contact only.

Doncaster does not use an online form.

Related findings

Role of the online self-referral form; Form design; Measures – how and when; Diagram 1 – Information gathering pyramid; Diagram 2 – Services information requirements at referral.

Form design

The design of the online self-referral form varies widely and is influenced by different understandings of the purpose of the form and local changes to service. Most IAPTs have either made changes or are keen to make changes to their original online form.

Changes made or intended include:

  • reducing form to minimum name, contact details and eligibility criteria
  • offering people new options eg: video or short-notice appointments
  • adding crisis info and a risk tickbox
  • linking to a self-booking form

Related findings

Role of the online self-referral form.

Measures (GAD7, PHQ9 etc) – how and when

There is wide variation in how and when services ask people to complete measures prior to treatment. 

Measures were:

  • included in online self-referral forms
  • sent out by post or email
  • completed in the waiting room before assessment
  • done during phone or face-to-face assessments

Some services felt strongly that the first set of measures should be completed during a phone or face-to-face assessment as the start of the therapeutic process. 

One service included the measures in their online self-assessment form but these were not processed until 8 weeks after they were completed.

“If you talk to someone you can frame the questions, but on an email you can’t do that.”

“On an email you haven’t put a context around them. People don’t really understand what they’re for. They don’t speak the language.”

Related findings

Identifying people at risk

Inappropriate referrals - eligibility

Some IAPT services receive high numbers of inappropriate online self-referrals. This is often related to the design of their online form and leads to a high volume of phone work for staff.

Some IAPT services receive a high volume of referrals that don’t match their eligibility criteria. 60 per cent of online self-referrals received by 1 service cannot be processed without follow-up phone calls because:

  • people request appointment times outside the service’s working hours
  • they are out of area
  • people are not suitable for the service because of risk or complexity

Some GP referrals are also inappropriate. 

“The first thing you are trying to ascertain is whether people live in the right area. You don’t want to go through all the demographic details if they are out of area.”

Related findings

Role of online form, cat and mouse, form design.

Phone cat and mouse

Phoning and not getting through to people is a time-consuming pain point for most IAPTs we met. As well as managing phone referrals, phone calls are made to check those not found on Spine, to check info on referral forms, to fill slots and to make an assessment appointment time. 

Phoning to check online self-referrals is a big problem. 

Leeds and North Staffs - 60% of online self-referrals/assessments need to be phoned to check info.

ThinkAction shortened their form and initiated online booking for people to make (and change) appointments. This has reduced need for phone calls. 

Norfolk - dealing with volume of phone referrals is a struggle and they hope more online self-referrals and online booking will reduce volume. 

Berkshire believe their extended opening hours (8am-8pm) helps to reduce need for phone cat and mouse.

Doncaster spend a lot of phone time filling slots - “a complete pain”.

“You can phone a list of 10 people and you might only get one. You’ve used all that manpower in making those repeat phone calls.” 

“I never pick up calls from a withheld number either, especially during working hours.”

Related findings

Inappropriate referrals – eligibility; Self booking systems.

Online assessment booking systems

Online systems that let people book and change assessment appointments encourage more online self-referrals and reduce the number of phone call, empty slots and DNAs. 

ThinkAction initiated online assessment booking for people to make (and change) appointments. Combined with a shorter form the number of phone calls has reduced significantly and DNAs have reduced by 10 to 15%.

Norfolk are implementing online self booking and hope to increase online self-referrals and free up phone lines for people who need them. PWPs are keen as they bear the brunt of clients disgruntled by wait for phone call to book assessment.

Doncaster are considering an online booking solution for booking assessments and appointments but the system they are investigating needs to be tailored to their needs.

Related findings

Do not attend (DNA) rates and influences.

Priority groups

Priority groups of people such as pregnant women, veterans and those suffering from a long term conditions, are identified at different parts of the self-referral process.

Some IAPTs have added questions to their self assessment form to identify priority groups. This is to get people into the appropriate team faster and avoids them having 2 assessments.

Others identify priority groups at assessment.

Previous research of self-referral forms with users found:

  • the long term condition question confused users
  • pregnant women expressed concern about divulging this on a form about mental health – fear of social services

Reference - user research report IAPTs discovery December 2018.

Related findings

Role of online self-referral form

Do not attend (DNA) rates and influences

People not turning up for assessment or treatment appointments are a challenge for IAPTs. 
Two services report 20% - 30% DNA rate. Factors thought to influence attendance include:

  • GP referrals are less likely to attend - one service has seen fewer DNAs with increase of self-referral
  • one service has reduced DNAs by use of text reminders
  • service using online self booking system reduced DNAs by 10 to 15%
  • the sooner the appointment the more likely to attend – but one service reports anything less than a week or more than three weeks increases DNA - there is a “golden window” of opportunity
  • hot months and Christmas increase DNAs

“With self-referrals people are already engaging with the service. If there’s that initial drive to get an appointment then they’re more likely to attend that appointment.” 

“We’ve managed to get our DNAs down to 8-10 per cent. Text reminders are a big factor.” 

Related findings

Self booking systems; GP referrals

Software workarounds

Using computer software allows IAPT services to streamline their admin systems.

However, software is often not completely fit for purpose, forcing staff to use time-consuming workarounds.

Some details from online self-referral forms are not automatically pulled through to IAPTUS. Staff have to cut and paste information manually onto IAPTUS pages. This creates an extra admin burden and also raises the risk of human error.

Many IAPTs use several kinds of software to manage patient referrals and records. Information is often transferred manually from one kind of software to another, for example from IAPTUS to Outlook. This again creates a higher admin burden and increases the risk of human error.

“While you’re messing about with this [manually adding information to IAPTUS] you could have put another one on."

Online referrals are growing

The six IAPTs who have online self-referral have seen a growth in online referrals. Only one has statistics on online form completion.

The estimated percentage of online referrals per IAPT varies between 10% to 90%.

One service reported a drop in GP referrals from 50% to 20% since having the online form.

ThinkAction online referrals jumped from 50% - 90% when the form was redesigned and booking system added. Previous to this stats showed 50% were not completing the online form.

ThinkAction also signpost to online when people phone, then guide them through the process. This is usually well received.

“Online form gives patients more choice. You can see by the volume of people using it that it was needed” 

“It’s a lot easier to refer online than to pick up the phone and request an appointment”

“Online referrals have been increasing since we started two years ago – they’re currently 40 per cent of our total referrals - and are showing no signs of letting up.” 

Related findings

Form design, referrals by phone, GP referrals.

Identifying people at risk

Identifying people at risk of harm to self or others is a challenge for all IAPT services.

Assessing risk from online self-referral is difficult and different IAPTs use different approaches. 

Leeds and Norfolk forms ask if person is in crisis and, if so, they are signposted to other services. But people with severe illness still come through self-referral. 

Worthing and ThinkAction signpost to crisis services without asking crisis question on form. ThinkAction sends out measures just before the assessment so any risk can be picked up at assessment.

Berkshire and Norfolk check person on systems – if in treatment within 6-12 months, referral goes to clinicians. 

Doncaster - referrals that suggest risk are referred to senior clinicians.

Related findings

Measures – how and when.

Information required to create patient record

Only minimal information is usually needed to find people on national or local look-up systems and create a patient record. 

Doncaster: Only name and date of birth needed to look people up on local software (System One) and register them.

Norths Staffs: Name, date of birth and phone number needed.

Worthing: Name, date of birth and postcode needed.

Norfolk & Suffolk: Name, address and GP needed.

“To create a profile on the patient record system all we need is their name, date of birth and post code. Then we can look up their NHS number and GP.”

Related findings

Role of online form

How assessments are done

Assessments are usually done by clinicians, either face-to-face or over the phone. Some IAPTs feel strongly that assessments should be face-to-face.

In six of the seven services we visited, clinicians carried out the initial assessment either face-to-face or over the phone.

One IAPT does assessments via their online self-referral form. The forms are analysed by a PWP who then sends out a letter with details of treatment and appointment time. Forms are currently being analysed 8 weeks after they are submitted.

Several IAPTs felt strongly that assessments should be face-to-face so that the first set of measures can be completed and discussed as the part of the therapeutic process. 

“The measures are done at assessment. These are all done by clinicians and have elements of treatment so it’s appropriate.” 

"Clients are asked to fill the measures out 24 hours before their assessment. Then if there’s any risk it can get picked up quickly.” 

Related findings

Measures – how and when.

Referrals by phone

Most phone referrals are taken by admins who collect personal, demographic and eligibility info and then create a patient record. 

Following the phone referral the person is put on a waiting list for assessment. They are then sent a letter, and/or phoned back, and/or sent a text to book an assessment or sent a link to self book online (ThinkAction).

Phone referrals are less structured and can be more time consuming than online referrals.

Sometimes people phoning think they are going to get phone advice. Phone admins have to be trained in appropriate boundaries when taking phone referrals. 

Related findings

Referral channels,information required to create patient record,cat and mouse.

GP/HCP supported referrals

GP referrals are mostly received by fax, phone or email rather than online form. People who face barriers to self-referral, for example, those with social anxiety or ESOL are more likely to be referred by GPs.

One service commented that GP referrals often don’t have enough detail so have to be followed up.

A number of services stressed the need for confirming consent to go ahead with the client as people aren’t always aware they have been referred.

Some services have visited GPs to raise awareness of IAPT services.

“With professional referrals we only expect 50-60% to end up having an assessment, either because they are unsuitable or they don’t engage. With self-referrals generally it’s about 80 per cent of people end up having an assessment.” 

Other findings

Waiting times – waiting times to assessment varied from less than 1 day (this was a service where clients could book appointments online themselves) to more than eight weeks.

Local constraints – some IAPT services are unable to update their software and processes and make innovations because they are constrained by the lengthy, bureaucratic processes required by the wider organisations they site within.

Phone follow-ups - services that don’t send email confirmations after an online form has been submitted often get follow-up phone calls from people checking that their referral has been received or wanting to speed up the referral.

Answerphone messages - some IAPT services offer people lots of information on their answerphone messages, including how to refer yourself online or opening hours. Others have generic messages, such as, “There is no one available to take your call.”

“Sometimes we come up with ideas and it either takes too much time in the trust or they come back with, ‘You can’t do that’.”




Last edited: 9 May 2019 2:42 pm