Background

Successful recruitment into the RCT component of the VIP trial (step 3) was contingent on the numbers screened by the police at step 1 of the trial. Our feasibility study [6] suggested that out of 120 victims screened as significantly distressed by the police, 26 would be suitable for a RCT trial. Our pre-power analysis indicated 226 people would be required for the VIP RCT [10]. Thus for 226 participants to be recruited, 1043 victims of crime would need be identified as significantly distressed by the police. Despite adapting screening by using three different screening approaches, we only managed to recruit 131 participants out of our target of 226. Changes in police leadership, major incidents (e.g. terrorist attacks, protests) and resource constraints posed challenges to police screening. Translation from screening to recruitment was also dependent on the screening method used. Our reflections aim to help those wishing to conduct research in this population.

Screening methods

Method 1 (conducted between June 2017 and August 2019), police community support officers aimed to complete home visits with all older victims of community crime.

Method 2 (conducted between May 2021 and June 2022), two police administrators screened victims over the telephone and tasked community support officers to complete home visits and consent only those found to be distressed.

Method 3 (conducted between March 17th and 30th June 2023), a police administrator made initial telephone contact and with the victim’s agreement, passed their details on to a VIP researcher embedded within the police. They provided information about the trial and then directed community support officers to undertake the home visits to those identified as distressed on telephone screening.

Screening results

Under method 1: Of the 9391 crimes reported by older victims during this period, 32% (2915/9391) were successfully screened and demonstrated that 40% (1171/2915) scored above the cut-off for distress. There were differences in screening rates across local authority areas due to varying levels of engagement by police teams, victim factors (such as not speaking English) and the interaction between the victims and the police which affected engagement (e.g. altered public confidence in the police).

Under method 2: Of the 7680 crimes reported during this period, 17% (1340/7680) were successfully screened. There was a lower screen positive rate of 25% (334/1340) as the administrators only classed older victims as distressed if they scored above the distress cut-off and also agreed to a home visit from a community support officer.

Under method 3: Of the 1566 crimes reported by older victims in this brief period, only 152 (10%) were successfully screened. Of these, 61 (40%) were screen positive, consistent with method 1. Of these, 27 (45%) agreed to a home visit from the police and were successfully consented and referred to the VIP team.

Our experience

Consistent with our feasibility study [6], home visits (method 1) were more effective at recruiting older victims into research than telephone contacts. The visiting community support officers fed-back that they often detected non-verbal signs of distress and were able to have a wider discussion on the impact of crime with older victims. However, police resource constraints, along with the COVID-19 pandemic, meant it was necessary to employ initial telephone screening to identify distress. This approach was well-received by the community support officers as it targeted their visits to the most vulnerable older victims. However, unlike with face-to-face visits, many victims did not answer the phone. Furthermore, successful screening was dependent on the quality of the interaction between an administrator and older victim. With only two screeners, screening and recruitment is very much determined by the skills of a limited number of screeners.

Method 1 also had the advantage of streamlining consent into a single point, whereas method 2 required older victims to consent over the telephone to home visits, and again to data-sharing during the visit. Method 3 required consent for the VIP researcher to speak to them, as well as to the visits and to data sharing. Each extra consent point and the gap between telephone calls and home visits increased the chances of attrition. This meant that, although methods 2 and 3 were a more effective use of police resources, they were a less effective use of research funding resources. Projections of the numbers telephone screened at step 1 using method 3 suggested that there would be insufficient numbers of victims referred for step 2 rescreening at 3 months. It was therefore jointly agreed with the funders to end screening and analyse those recruited through methods 1 and 2. This meant the trial was underpowered.

What worked well

Although we did not achieve our target sample size, our methods show that the police were able to use a screening tool to identify older victims in distress. We used four brief screening questions from the PHQ-2 [17, 18] and GAD-2 [19] to minimise demands on police time. Embedding screening into routine practice was feasible for officers and some valued this as a return to traditional community policing. Undertaking screening during SNT welfare visits was also consistent with Ministry of Justice guidelines to offer support for at-risk crime victims [20].

The police integrated their crime reporting system with an in-house software package which was used to send tasking requests to community support officers. This was effective at identifying our target population and could easily be adapted to focus on specific crime types and ensure screening was evenly spread across boroughs.

ChallengesStaff turnover

Our trial was overseen by the Commander for Community Policing, a role which is regularly rotated, and which meant we had eight different commanders over the course of the project. Each new commander required new relationships and a fresh agreement for the VIP trial. The commanders who recognised the value of our project ensured successful screening by providing the approvals we needed, delegating senior officers to oversee the screening, facilitating training and expanding study sites. However, commanders who did not recognise the value of our research delayed or suspended screening until a revised methodology was proposed.

Lesson learned 1

If staff across the police are to engage with a project, there needs to be good rapport and engagement from senior members of the police and researchers. We developed relationships with senior members through visits at police headquarters and with community support officers by accompanying them on home visits. To ensure continuity of study procedures across staff turnover, we recommend that a Memorandum of Understanding is agreed and signed-off at the highest level possible to avoid the need to repeatedly reach new agreements.

Major incidents

Several concurrent challenges (terrorist incidents, climate change protests, knife crime) required police community support officers to be diverted away from the project to assist (‘abstraction’), causing delays to screening and obtaining participant consent. The COVID-19 pandemic during screening method 2 presented further challenges. Police administrators needed training on communicating with older victims and experienced absences due to sickness and lockdowns.

Lesson learned 2

Community support officers are under considerable pressure and service demands are routinely prioritised over research. Use of dedicated administrators within the police reduced the impact of abstraction on screening, but community support officers were still needed to obtain consent for the purpose of the research. Community support officer input needs to be ringfenced for the duration of a significant research project unless there are exceptional reasons to abstract the staff.

Policy changes

In the early phases of the trial, there was a general recognition that the police needed to do more to support older victims, and good practice guidelines advised officers to conduct needs assessments to identify those who were most vulnerable [21]. As our screening methods were compatible with policing priorities at this time, it was supported by senior leaders within the police, and screening gained momentum. An increase in knife crime attracted publicity shortly after the project began (e.g. [22]), which placed competing pressures on the police. Consequently, our project was considered less of a priority and screening was suspended. Although the police have recently committed to training officers in compassionate, evidence-based victim care [23], the extent the police should be involved in supporting people with mental health problems remains a topic of public interest [24]. Public pressure and media attention appears to influence policing priorities, their direction of resources and correspondingly their engagement with our research.

Lesson learned 3

Seeking to increase governmental and public awareness of the need for further research into the impact of crime is a long-term approach which may be promoted through supporting organisations for older people such as Age UK, Victim Support, and through articles in the public media [25]. Support from the victims’ commissioner and the Mayors’ Office, who fund some police services and were able to promote the project as a priority, was also welcomed.

Public confidence in the police

Major incidents concerning the police (e.g. the murder of George Floyd in the USA and Sarah Everard in the UK by serving officers) adversely impacted the reputation of the police. Concerns were also being raised about poor police response times and low conviction rates [26]. Our research thus took place at a time when public confidence in the police was especially low [27, 28] and this appeared to make it challenging for the officers involved in our trial to engage with older victims.

We conducted qualitative interviews and found that, although many older victims recognised the police were under considerable pressure, some participants felt there had been a lack of empathy from the officer they had contact with or that they had been dismissive during their initial crime report [29]. Some of the participants interviewed attributed their ongoing distress to feeling let down by initial police communication.

Lesson learned 4

Whilst it may not be realistic to immediately improve police response time and conviction rates, police training on empathic communication when handling crime reports could still help older victims and crime victims in general to feel heard and their distress acknowledged, which may be an important part of helping them to deal with the situation.

Representative samples of victims

We aimed to ensure that our population was representative of older victims. However, 60–70% of all crimes go unreported [30, 31]. Victims from ethnic minorities or with complex care needs may be less likely to report crime [32, 33], and older victims of property crimes may be more likely to report crime for insurance claims [34].

Lesson learned 5

As the integrated software package was effective at identifying our target population, this could be filtered to target groups known to be under-represented. Future researchers may also consider partnering with charities (e.g. Victim Support) alongside the police to identify older victims who have not reported their crime.