Police and crime commissioner-commissioned inspection into Norfolk Constabulary

Published on: 24 May 2024

Letter information

Roy Wilsher OBE QFSM
His Majesty’s Inspector of Constabulary
His Majesty’s Inspector of Fire & Rescue Services

Sarah Taylor
Police and Crime Commissioner for Norfolk

Sent on
24 May 2024

I write in response to the letter from your predecessor dated 30 January 2024, which requested us to carry out an inspection of Norfolk Constabulary under section 54(2BA) of the Police Act 1996.

As requested, HMICFRS has reviewed Norfolk Constabulary’s handling of 999 calls from vulnerable people and certain incident types.

We were asked to provide commentary on whether the constabulary would be in an improved position if the Right Care Right Person (RCRP) call script had been used in all the calls.

As part of the review, we have:

  • reviewed 490 emergency 999 calls to Norfolk Constabulary;
  • assessed the computer aided despatch (CAD) records relating to the 999 calls;
  • carried out five interviews and focus groups with personnel from Norfolk Constabulary; and
  • interviewed the then police and crime commissioner.

This letter summarises our main findings. I have enclosed an annex that sets out our findings in more detail.


Initial assessment of 999 calls

The force uses the threat, harm, investigation, vulnerability and engagement (THRIVE) model to assess the initial response to calls for service. This allows the force to judge the risk posed by the call and places the needs of the victim at the centre of that decision.

The THRIVE question set used by the force prior to 6 February 2024 consisted of a limited series of questions to be considered by call handlers. This allowed a degree of discretion and meant there were variations in how these assessments were completed. The use of more questions to gather further information was reliant on the experience and intuition of the call handler.

On 6 February 2024 the force introduced an enhanced THRIVE script which replaced the previous THRIVE question set.

The new process leads the call handler through an extended series of questions. These are designed to maximise information gathering opportunities and identify all associated risks at the call handling stage. The script has further sub-questions that may need to be asked depending on the caller’s previous answer.

During the review we examined calls that were dealt with under both the THRIVE question set and the THRIVE script.

It is worthy of note that the audit that is part of this commission was in addition to the annual PEEL inspection victim service assessment audit. We do this in all forces in England and Wales. The nature of the commission means that this audit was focussed on calls that were more likely to involve high-risk, high-harm incidents, with a greater propensity for vulnerability. By their very nature, these types of calls will ordinarily need a more detailed THRIVE assessment.

We found that overall, the force’s call handlers provide an effective and professional service to the public; often in demanding circumstances. However, there are inconsistencies in when, and if, the THRIVE assessments are completed. And there are inconsistencies in the use of information from force systems and callers. This is important as the assessment process identifies the risks involved and helps the force decide the level of police response. We were concerned that in many cases involving a repeat victim, or where the victim or other persons involved were vulnerable, this wasn’t recorded within the THRIVE assessment.

The introduction of the THRIVE script was an improvement initiative within the force control room. The force continues to refine and establish it as routine working practice. This has resulted in assessments that are more detailed and cover a wider range of issues. While the force has implemented the script relatively recently, the early signs are that the force is getting better at assessments. This is encouraging. But there were some concerns from call handlers that the script restricted a free flow of information. So, the force should review the effectiveness of the new process.

The supervision of incidents within the control room is inconsistent. When the allocated response times for police attendance aren’t met, some incidents are downgraded without a THRIVE re-assessment or any supervisory oversight.

The Right Care Right Person Model

All forces across England and Wales are introducing the Right Care, Right Person model. The model helps forces identify which partner agency is the most appropriate to care for and support vulnerable people. Examples of partner agencies include health and social care services. The model also has guidance on dealing with calls for services relating to mental health, concerns for welfare and missing persons.

In addition to the THRIVE script, the force has developed a process to introduce the RCRP model. This includes a bespoke question script supplemented by training.

The force has commissioned a separate review of its RCRP process and training; led by the National Police Chiefs’ Council and supported by the College of Policing.

The force anticipates that the RCRP model will build on the existing welfare checks policy. This guided the force in responding to incidents where other agencies have requested a visit, or check on the welfare, of a member of the public. The RCRP model was due to be introduced on 6 February alongside the new THRIVE script. But this was delayed.

Personnel broadly see the introduction of the RCRP model as a positive step. The existing policy is established and familiar to them. And they welcomed the RCRP model because it formalises arrangements with partner agencies. This will give all parties a clear understanding of their respective responsibilities and what is expected from them.

As required by section 55(1) and (2) of the Police Act 1996, we will publish this letter and annex in such manner as appears to us to be appropriate.

Annex A: Detailed findings of our police and crime commissioner-commissioned inspection into Norfolk Constabulary


The 999 calls reviewed in this audit were randomly selected across the seven graded response categories used by the force. The calls involved incidents that included vulnerability and related to domestic abuse; missing persons; and concerns for welfare, including mental health. The calls were selected in two categories which covered separate periods of before, and after, personnel had received the Right Care Right Person (RCRP) training.

In the review, we examined the initial triage and risk assessment process. We then considered the use of information about repeat victims and vulnerability. Finally, we considered how the force resolved calls and the appropriateness of the response.

Our terms of reference

The terms of reference agreed with the police and crime commissioner were:

His Majesty’s Chief Inspector of Constabulary (HMCI) will carry out a review of Norfolk Constabulary’s handling of 999 call handling related to calls from vulnerable people and certain incident types. Reference in these terms to HMCI include references to any person acting on his behalf.

The aim is to provide commentary on whether the constabulary would be in an improved position if the RCRP call script had been used in all the calls.

Specifically, HMCI will examine:

  • 490 emergency 999 calls made to Norfolk Constabulary.
  • Calls drawn from a period that covers both before and after RCRP training was provided to personnel. The numbers in each period will be sufficient to allow meaningful comparison.
  • Calls covering a mix of incidents that include vulnerability. Call types will come from domestic abuse, missing persons, concern for welfare including mental health 999 calls.
  • Samples to be taken as follows:
    • 999s with Grade A response (70 calls)
    • 999s with Grade B1 response (70 calls)
    • 999s with Grade B2 response (70 calls)
    • 999s with Grade C Response (70 calls)
    • 999s with Grade D Response (70 calls)
    • 999s with no Computer Aided Despatch (CAD) record created (70 calls)
    • 999s resulting in a diary appointment (70 calls).
  • Interview with the police and crime commissioner at their request.
  • A hot debrief at the conclusion of inspection activity.
  • CADs related to each of the 999 calls. These will be assessed to determine whether they are consistent with the information received and that any risk is correctly recorded.
  • Whether the introduction of the RCRP, and in particular the associated call scripts, would materially alter the service to the public.
  • Focus groups/interviews of personnel in the control room.

The initial THRIVE assessment process

We found that personnel were polite, professional and acted appropriately in their dealings with the public in virtually all calls audited (467/470). Call handlers provide a high level of service to the public and this was particularly evident in some challenging and complex incidents.

For example, in one case the call handler maintained contact with a suicidal missing person for thirty minutes; providing reassurance and support and preventing them from taking their own life. In another case the call handler assessed the initial incident then intuitively asked more questions to establish if the caller and other people involved were being exploited for modern slavery.

Personnel routinely completed a structured triage process. However, on some occasions, relevant information wasn’t recorded on the CAD record.

The force has processes that enable personnel to use force systems to identify if the incident involves a repeat victim, or a vulnerable victim or other person. Call handlers made these checks in most cases.

However, we were concerned to find that information obtained from those checks, or from the conversation with the caller (that identified a repeat victim, or that the victim or other person was vulnerable) wasn’t always recorded.

In too many cases (115/541) that involved a repeat victim and/or where the victim or other persons involved were vulnerable, this wasn’t recorded within the THRIVE assessment. This included children and victims of domestic abuse. So the force may not always give vulnerable people the required safeguarding support or the level of response that is necessary to keep them safe.

We found inconsistency in how THRIVE assessments and checks on force systems were completed. There was also a lack of clarity about when a THRIVE assessment may not be required. One call handler stated that they thought an email had been circulated which stated that for Grade As an entry should be made on the CAD with the THRIVE assessment completed later. Another told us that they believed the THRIVE assessments weren’t required for Grade A or B1 incidents.

We were also told of variations in how checks on force systems were completed. Sometimes checks were completed during the THRIVE process; other times checks were made after the THRIVE process had been completed; or sometimes these checks weren’t completed at all.

For the THRIVE assessment to be effective, personnel should first gather and consider all relevant information; particularly information on force systems. This results in informed decision-making in respect of the needs of the caller, victim or any other person, and the level of response required.

The initial grading and response to incidents was appropriate to the nature of the call on most occasions. However, vulnerability wasn’t always recorded during the THRIVE process. So the grading and response may not be based on a thorough assessment of the needs of the victim or other people.

The force should make sure personnel clearly understand when to complete THRIVE assessments; when to make checks on force systems; and what information to record.

The use of the THRIVE script

Our audit included 57 cases where the THRIVE script had been used. These assessments were more detailed than assessments that used the question set. They considered a wider range of issues. Checks on force systems to identify if the incident involved a repeat victim, or a vulnerable victim or other person (48/52), were completed in virtually all cases. However, in some cases where the THRIVE script was used (14 of the 115 cases mentioned earlier) that involved a repeat victim, or where the victim or other persons involved were vulnerable, this wasn’t recorded.

We were pleased to see an overall improvement in how THRIVE assessments have been completed since the force introduced the script. But the force has more work to do to make sure that it correctly records and assesses information about repeat victims and vulnerability.

Personnel were aware of how to use the THRIVE script. But they were divided about its benefits. Some felt that it was an improvement that allows more information to be gathered. And while it takes longer, it means the process is completed correctly on the first occasion. Others felt that the script was time consuming and unnecessary as grading decisions were made on the information gathered during the call. Some personnel felt that the THRIVE script didn’t allow a free flow of conversation.

A common theme was that personnel felt they hadn’t been consulted before its introduction. They told us that consultation would have identified some areas for improvement in advance. In response, the force told us a stakeholder panel included representatives from various roles including control room personnel and supervisors, representation from the training department, the control room systems lead and the project manager.

We recognise that the THRIVE script had been introduced relatively shortly before our inspection. And that changing force processes takes time and is challenging. We were pleased to find that the force has introduced weekly focus groups which give personnel the opportunity to feed back on the THRIVE script. Personnel welcomed this and spoke positively about changes that had been made in a timely response to their feedback.

It was unclear how the force will review the effectiveness of the THRIVE script assessments following the introduction of this process. However, any review should consider the benefits that the new process provides when considered against the time required for completion. This will allow the force to assess whether it has enough call handling resources to meet any increase in workloads that the THRIVE script may bring (because it takes longer).

We found that training was inconsistent. The force provided training for the THRIVE script by various methods. Some personnel had a brief session on the script with the team tutor in the early hours of a night shift. Other personnel had an opportunity to familiarise themselves with the script. And others attended a two-hour training session. Personnel told us training tended to focus on the process of the script, rather than the application of the separate elements of THRIVE. Some personnel had received previous THRIVE training which considered the wider issues; in some instances this had been quite some time ago.

The force should review the training requirements for call handling; and where necessary train personnel to make sure they are consistent in their THRIVE assessments. The training should also give personnel a clear understanding of issues that contribute to vulnerability; particularly those cases that involve domestic abuse.

Supervisory oversight in the control room needs to be improved

Force policy states that where a member of control room personnel deems it suitable to downgrade an incident, they should speak with a supervisor. The supervisor will apply the THRIVE principles and enter their rationale on the CAD record. Personnel told us that a lack of resources and an inability to meet response times placed pressure on control room personnel to downgrade incidents.

Supervisors didn’t always monitor the timeliness of the response to incidents. And they didn’t intervene when response times were outside the defined timescales. We were concerned to find that incidents were often downgraded because the initial response time hadn’t been met. This means that the force downgrades incidents to manage demand; rather than because there has been a reduction of threat, harm and risk.

On too many occasions (61/180) the force downgraded incidents without undertaking or recording another THRIVE assessment or without effective supervisory oversight.

The oversight of performance and quality assurance in call handling varies across teams. It was unclear how many monthly quality assurance checks supervisors completed. Feedback by supervisors tended to consist of regular one-to-one meetings or informal discussions. However, some personnel hadn’t received any feedback for several months. They told us they would welcome consistency in this area; with a focus on learning to help them improve performance.

The introduction of the Right Care Right Person model

The force’s welfare checks policy is established within the force and personnel were familiar with its use. This policy has a number of similarities with the RCRP model. So personnel are already working within the principles of RCRP. But without the formal agreements with partner agencies.

We found that in some cases the force had dealt with incidents that were more suited to other agencies. Examples of this were where callers had contacted other agencies outside normal working hours and were advised to contact the police.

We didn’t find evidence that this had an adverse effect on community safety. But the RCRP model will reinforce the requirements for a timely and effective response by the appropriate agencies.

In our audit, we found that 54 out of 73 incidents that fell within the welfare checks policy or weren’t police-related matters, had been passed to another agency to deal with. This means that the force dealt with 19 incidents when there was no requirement for it to do so. In 14 of these incidents the force would have benefitted under the RCRP model. In these incidents the responsibilities of the relevant lead agencies would have been clear if the RCRP model had been in place. This would have helped personnel determine the most appropriate response.

A further six incidents were passed to another agency when it would have been more appropriate for police to have been the lead agency. But this didn’t affect public safety.

Personnel were positive about the introduction of the RCRP model. They believed it will complement the existing policy, formalise processes with partner agencies, and provide clarity on the types of incidents that the force will and won’t attend and why. They also spoke of the personal benefits it will bring for them when performing their role. They felt the model gives more support and cover for their reasons and rationale for decision-making. And they told us it allows personnel to use their professional judgment to a greater degree.

The force has provided a wide range of training. This included force training days and access to the College of Policing online learning material. While personnel were generally positive about the training, they felt that they hadn’t been consulted before the RCRP model was created.


Control room personnel displayed high levels of professionalism in their dealings with the public.

In our annual PEEL inspections, we include a victim service assessment. This assessment examines a victim’s journey; from reporting a crime through to the crime being finalised and an outcome assigned.

The results of the recent PEEL VSA inspection and the audit that was part of the commission show that the force performs well in most control room functions.

However, the audit under this commission had a greater focus on 999 calls that potentially involve a greater risk to the public. We found some areas that need to be improved. This is particularly relevant to the identification of, and response to, those that are vulnerable.

The force must make sure it provides personnel with relevant training. This would help them to correctly identify issues of vulnerability; record those issues in their assessments of the incidents; and provide the appropriate response.

The force should review how it communicates its requirements and expectations internally. This would make sure all teams are consistent in performing their roles and processes.

The force should make sure it has robust quality assurance measures to ensure processes are performed consistently. And it should ensure that supervisors have a clear understanding of their roles and responsibilities.

The RCRP model will formalise the working arrangement that the force has with other agencies. The result will be that incidents that aren’t police matters, or are better suited to another agency, are dealt with appropriately.

Clarity on what the force and other agencies will have responsibility for will give personnel greater confidence in their decision-making. Those who contact the force about incidents covered by the RCRP model are more likely to get an improved service.

Back to publication

Police and crime commissioner-commissioned inspection into Norfolk Constabulary