Variability in selecting patients to manage in the community: a service evaluation of community matron's case-finding strategies

Abstract

Background. Community-based case management of high-frequency health service users, also termed ‘high-intensity users’ may reduce the demand on secondary care. In the UK, experienced senior nurses ‘community matrons’ have been appointed to provide more care in the community and to case manage high-intensity users and prevent unnecessary hospital admission. Computerized scoring systems have been developed to help with case finding.

Objective. To evaluate how community matrons select their caseload, in particular the impact of a case finding tool, and access to computer-based systems.

Methods. We used direct observation and the think aloud protocol during case selection and a series of workshops. Analysis was based on the Framework Approach (familiarization, coding, charting, mapping and interpretation) using verbatim transcripts analysed by N-Vivo software.

Results. Community matrons within the same locality have different and sub-optimal caseloads. Although common elements exist in case selection, processes are modified by individuals depending on the influence of their interpersonal relationships, accessibility of computer systems and personal experience and expertise. The supporting IT system often produced data too late and while predicting admission did not identify cases amenable to community intervention.

Conclusions. Inter-professional networks, personal experience and training influence the patients identified for case management. The combination of an improved case finding tool and a better defined role for community matrons could lead to more standardized and equitable case selection.

Case management, chronic disease, community health nursing, computerized, medical record systems, patient readmission

Introduction

The number of people living with long-term health conditions has risen and presents health care systems with a significant challenge. People with chronic diseases impose a significant burden on inpatient services, high-intensity users of services account for 10% of the population, but utilize 55% of inpatient days, and very high-intensity users account for 5% of patients using 42% of inpatient days.1 In 2005, the UK Department of Health set out plans for improving the care of patients with long-term conditions.2 Case management was a key component of the plan and involved the identification of suitable patients, assessment of need and co-ordination of services between care providers to prevent worsening of their condition and unplanned admission to hospital. The role of the community matron was developed to fulfil this need, utilizing the full range of skills of a senior community nurse alongside co-ordinating health care. They were anticipated to have a caseload of 50–80 patients (Box 1).2

BOX 1

The aims and role of the community matron

Aims

• Reduce reliance on hospital care.

• Increase range of responsiveness of community services.

• Improve quality of life for people with long-term conditions.

• Plan for, predict and prevent crises in care management.

Role

• Work collaboratively with all professionals, carers and relatives to understand all aspects of the patients condition.

• Develop personalized care plan, including preventative measures and anticipation of future needs.

• Maintain regular contact with the patient.

• Initiate action when required—ordering tests or prescribing.

• Update patients records; communicate with other professionals.

• Liaise with local agencies—social services, community services, voluntary services to mobilize resources.

• Teach carers and relatives to recognize changes in condition.

• Secure additional support as needed.

• Maintain contact with the patient if they are admitted to hospital.

There are a number of computer-based tools developed to identify patients at risk of readmission.3–6 In the UK, the Department of Health proposed the PARR tool (Patient At Risk of Readmission), designed by the Kings Fund7–9 to identify patients at risk of readmission. The PARR tool was populated with data about hospital attendance, ideally refreshed monthly, but poor availability of local hospital data often delays this. Referrals would also be made by GPs, other health care providers or the patients themselves. At this time, there is mixed evidence for the efficacy of the intervention of a community matron in reducing hospital admissions,10 although their intervention is well received by patients.11

A south London locality had invested in health IT, including providing community matron teams with access to PARR data and the social care database (Framework I). There were concerns that the PARR tool was not effective in helping the community matrons to identify their caseload. We were therefore requested to carry out a service evaluation of how the existing technology supported case selection and how the IT could be improved.

Methods

Literature review

We carried out a Medline literature search using the following key words: community matron, case management, chronic diseases, patient readmission and information technology. We identified literature on the work of community matrons and case management of high intensity service users. We also looked at computer models designed to predict admission. In addition, we visited IT leads in the locality to establish what technological resources were available to the community matrons.

Sample

We used a convenience sample consisting of seven community matrons and two GPs who volunteered to be filmed while they selected their cases and to show how the technology helped or hindered them. Our sample represented just over half (58%) of the 12 community matrons employed across the locality. Two of the community matrons worked within the setting of a virtual ward,12 one of whom was interviewed with a GP. A virtual ward is a multidisciplinary team of primary care professionals who care for people in their own homes who might otherwise be admitted to hospital. They have the same range of professions as found in a hospital ward.13

Direct observations

We carried out direct observation of four community matrons selecting their patients using the ALFA toolkit (multi-channel video observation with screen capture)14,15 and using ‘Think aloud protocol.’16,17 Participants were asked to say what they are looking at, thinking, doing and feeling as they go about the task of selecting patients for their caseload. At the midpoint, a workshop was held to discuss the findings during which it emerged that technology was not the limiting factor in caseload selection. A further four community matrons were then recorded during a semi-structured interview where they discussed patient selection and the use of tools at their disposal. This produced sufficient data to reach saturation. Verbatim transcripts were produced and analysed using the Framework Approach18 with QSRNvivo software involving familiarization, indexing, charting, mapping and interpretation of data.

Ethical considerations

This was a service evaluation carried out for the Public Health Commissioning Directorate in Wandsworth. Meetings were held with the Associate Director of Performance and Information and a Consultant in Public Health Medicine. Written consent was taken from all volunteers prior to video recording the ‘think aloud’ sessions with separate consent obtained to retain the video for analysis afterwards.

Results

We report how technology had a role in case selection using the results from the think aloud protocol and more structured approaches. We then report how inter-professional relationships and the community matrons’ experience and expertise emerged as strong themes and we report the factors that all community matrons took into account and those only used by some at each stage in the case selection process.

Four of the community matrons disclosed their caseload of 10, 20, 28 and 40 patients. No community matron had a full caseload, and all expressed difficulty in identifying suitable patients. Interviews revealed that community matrons received referrals from a variety of sources, they then assessed the patient using a variety of criteria and established an outcome for each patient, either acceptance on to the caseload for short-term or long-term management or referral onto another provider or, in a few cases, rejection. Within this pathway, there were areas of consistent practice among matrons, but variability in practice was evident. Mapping of the data revealed that this was modulated by the themes of access to technology, inter-professional relationships and individual experience and expertise. The variations in referral source and assessment criteria are summarized in Table 1.

TABLE 1

Summary of referral sources and heuristics for selection of caseload

	Source of patients	Assessment criteria
Used by all community matrons	PARR (case finding tool) GP	3 or more chronic conditions
		Polypharmacy—4 or more medications
		Multiple admissions to hospital
Used by some community matrons	District nurse	Age
	Specialist nurse	Past medical history
	Social services	Visits to A&E
	Out of hours service	Reduced mobility
	Chronic disease registers	Falls
	Hospital discharge coordinators	Chronic pain
	Multidisciplinary team meetings in secondary care.	Social isolation—lack of family or carer
	Occupational therapists and physiotherapists in A&E.	Cultural isolation—language barriers
	Nursing homes	Family and patient not coping
	Voluntary sector	Frequent home visits
		Inadequate package of care
		Frequent attendance at A&E
		Factors that may exclude patients from some caseloads
		Palliative care needs
		Mental health problems
		Alcohol problems
		Nursing home residence
		Chronic disease cared for by other specialist team

IT systems and access

The use of technology played a key role in the ability of the community matrons to identify suitable patients. The effectiveness of the technology was influenced by multiple factors: the ease of access to computer systems, the robustness of the system and personal effectiveness in using the tools. Seven of eight community matrons received a list of patients identified by the case finding tool PARR from the primary care trust. From this list, they selected patients with a score predicting risk of readmission >70% to assess for suitability. All community matrons found the PARR tool to be useful but felt that they needed to receive the data more frequently. In addition, they all had access in some form to the electronic patient record (EPR) system. Three community matrons also accessed the social services resource utilization database, Framework I.

It was felt that an impartial case-finding tool such as PARR can identify high-intensity users not known to their GP.

CM1: Now that varies from practice to practice how well people are spotting these patients and dealing with them. I mean the good practice is probably spotting them anyway so there is perhaps a reduced value to using the PARR tool anyway … we can do them opportunistically, reasonably well … but that does leave me wondering if there are people out there that we have not spotted.

It was also considered beneficial to identify patients using some sort of case-finding tool before they presented to hospital in order to be pro-active rather than reactive.

CM7: … I do think we need to be pro-active in identifying risk and preventing, we shouldn't wait for patients to go to A&E … I think we should devise another means of identifying these patients before they actually get to the point where they've gone in.

Patients identified by the case-finding tool were often already known to the practice, had moved out of the area or died.

CM6: These are the PARR patients that I received two days ago … And unfortunately with the list that I received I do know several of these patients. One's moved away. One's gone into a nursing home. One is already being seen by the district nurses and I've also seen this patient. At the moment this patient doesn't require my input.

The EPR formed a rich source of information particularly for the assessment of patients identified by the case finding tool.

CM2: I suppose … before I go and see anyone I also always ask for that EMIS summary so I have got all their prescriptions on there and that just gives a sort of background.

Not all community matrons had equal access to the system or equal access at each of the sites they served.

CM6: I find in both (assessment and management) it's much easier if you have everything at your fingertips and you can just tap onto the computer and check patient details, blood results and anything like that. But actually with the other surgeries having to go in physically and have a look I find that quite cumbersome to do.

In addition to difficulty accessing the EPR system at their surgeries, some community matrons felt they did not have enough understanding of the system.

CM2: … I haven't got very good understanding of the EMIS (the brand of EPR).

The EPR was being used innovatively at two sites. At one site, the community matron had searched the chronic disease database for a source of potential patients.

CM6: … yesterday I had a couple of colleagues that came over and we had a look at the chronic disease registers and I printed off all those so I've got about 400 patients that I need to check over to see whether they're relevant for me. So I'm hoping from that I can also generate patients . …

At another site, the number of home visits was used as a factor for identifying patients.

CM1: And that would be a home visit so they should be coded as V for visit. And that is actually coding that episode of care as a home visit. So we would be able to actually search on home visits, search on patients who have had home visits done through …

The social services database—Framework I was referred to by two community matrons who highly valued the platform as a source of patient information and a means of communication with other professionals involved in their care.

CM2: … So if we go to this guy here for example, who I know has got a home care package…so this guy is known to me so I have made it aware that anyone who logs on, so if the social worker logs on, they know I am involved. He is the one that had the ninety nine point nine risk two years ago. … so I can go down here, I can see what care package, his care package details, I can also see what equipment he has got, I can see here he has been in residential respite. So his personal care is current, this is all his equipment . …

Interprofessional relationships

The community matrons had varying positions within the health care network. Some worked closely with a single GP in the setting of a virtual ward, others served several practices. Referrals from GPs were sometimes made directly to the community matron or at monthly practice meetings.

CM4: … the other two surgeries … they don't … really have a meeting … I wish they would have a meeting just like the others … where you can discuss all of their patients. You know, with the GPs when they're there . …

CM6: I occasionally get them from the district nurses. And I get them from the GPs on occasion. But I do tend to have to really beg for patients from them. They don't automatically think of me as a referral person for these things.

Only two community matrons were receiving referrals from the out-of-hours service (Harmoni).

CM4: … I think all the other matrons were saying that it seemed like I'm the only that is getting referrals from Harmoni.

Other sources of referral were variable and depended on personal links that had been built by either the community matron or the referrer. For example, the heart failure specialist nurse referred to several community matrons, but no other specialist nurses were represented in the interviews.

CM6: The heart specialist nurse and I have quite a close working relationship . … She will usually phone me and then fax through details of a patient she feels I need to see.

One CM was particularly pro-active in seeking sources of referrals for her caseload and raising the profile of her role. She was actively building links with other specialists and the local hospital.

CM6: I personally don't tend to get them from very many of the specialists at all. I'm trying to raise our profile as community matrons and we've just got on the intranet and we're just doing an actual referral form and a leaflet for the community matrons … And I'm hoping from this exposure … that we actually get more referrals and closer working relationships with St George's as well because I do go to the multidisciplinary meetings there and try and close the gap between the hospital and the community.

Communication difficulties were expressed frequently. Community matrons felt that their role was not clearly understood and that their input was not always acknowledged.

CM6: Well from my other surgeries I don't tend to get any feedback at all. None at all …Even quite often in my own surgery here I will write letters and I get no feedback from the GPs. It's quite difficult to work like that, even though I do have quite a good working relationship with them. But they still do not seem to understand the difference in my role and the district nursing role, even though they've had my criteria. Every monthly meeting we have I do a, ‘This is my role. This is what I do. Do you have any patients for me?' But they seem to feel that they're managing their patients quite well.

In contrast, the community matrons universally responded to referrals with a high degree of professionalism. Personal referrals were always assessed by personal contact with the patient and followed up with feedback to the referrer.

CM6: If the GP does a referral I do tend to go out. Sometimes I do a single visit. If it's not the sort of patient that's right for me then I'll refer them on to the correct people …I always inform the GP of my decision.

The idea that their role was not clearly understood thus limiting the referrals they received was echoed by most matrons.

CM 7: Yeah I think there is a lot of confusion … role of the Community Matron and the role of the District Nurse …I think it was rolled out without that education and not many people know exactly what the Community Matron does. I mean it, it will contribute part of it to the referrals we get. If people understood more about the role …

Experience and expertise

Assessment of the suitability of the patient for the caseload was heavily influenced by the experience and expertise of the community matron. All referred to ‘criteria’ or ‘guidelines’ but then imposed their own judgement on the individual situation.

CM8: So this lady has Type 2 Diabetes, schizophrenia, hypertension and osteoarthritis, chronic kidney disease stage four . . . Fitting (criteria) already.

… She's over 65 and she's on more than 4 medications …

CM7: It's a guide, it's a guideline, anything is a guideline, but it's you, the professional who has the knowledge and experience to work out exactly what's best for your patient, and with your patient.

There were often strongly held and contradictory opinions on which patients were suitable for the service. Contrasting opinions were expressed with regard to the acceptance of palliative care patients.

CM5: We're not meant to be a palliative, virtual ward but if we're not careful we will turn into it.

Alternatively

CM8: Actually cancer patients is one thing I've discussed with the virtual ward project leader because I thought, as experienced nurses I feel that we can offer a lot more support to District Nurses in terms of palliative care … and I think what I want to see is a virtual ward, getting to know these patients a lot earlier than just to the end of palliative care because at least with centre management we'd be preventing admissions …

Some community matrons accepted patients in nursing homes, whereas others stated they would reject them.

CM3: She's in … a nursing home for elderly mentally ill … So in terms of long term conditions and identifying her for this project, we would, I would exclude her from the caseload … Because she's already in a nursing home.

Some community matrons excluded patients with certain conditions automatically. Others felt that this approach may overlook someone who could benefit from their input.

CM2: So what I did was I got rid of all the, no not got rid, discarded, that's a better word all of the … dialysis patients because they've got 100% risk and … most of the cancers I discarded as well.

In contrast, when asked whether cancer as a diagnosis would be a reason to reject a patient:

CM6: Not necessarily. And also you often find that they have other long-term conditions as well.

Another matron had a similar approach to sickle cell sufferers. These patients are cared for by specialist nurses but this matron felt she still had a role to play:

CM7: … his condition is sickle cell … He has been to A&E three times … what I will do with him, give him a phone call … and see what sort of services … What I would do first is phone the patient and explain that they have been identified and if he is attending clinic, is there anybody supporting him, I find out from the patient first … So clearly something is not working, …

There was also a mixed approach to patients who had problems associated predominantly with alcohol excess.

CM6: I think I wouldn't be too happy just taking someone who has alcohol problems or mental health problems. I do have two patients that have alcohol problems but they have other long-term conditions and there are various multi disciplinary teams that are involved so it's a joint package which we're fine with.

Where community matrons had been on a training course, they implemented their learning in practice.

CM6: I've just been on extra training for that (expert patient programme19) and yes, if my patient was able to do that I would definitely recommend it.

and in reference to agenda setting before the first visit:

CM4: … we did … a study day … on self management and, erm self management in the community … Yes, Co creating health it was for … Yes and so that came out of it. That's one of the agenda setting (before the meeting) …

Universally, community matrons were in agreement that they could not do a complete assessment unless they went to see the patient in their home environment.

CM7: You cannot do a complete assessment just by sitting her and looking at the paperwork. You've got to go and see your patient and understand the difficulties they have, the need they will require, and then you come to a, well an agreement with the patient what needs to be done.

Among the community matrons, there was acknowledgement that there were variations in practice.

RES 8. ‘I think it's quite apparent that we do all work quite differently from each other though, isn't it?’

Discussion

Principal findings

The variability in case selection was such that it was impossible to make generic recommendations as to how improving the technology would improve case selection. Although there were some common elements, community matrons within the same locality employed varying methods to select their caseloads, a variation they readily acknowledged. The use of the case finding tool (PARR) was important but other data was also needed. It was essential to have access to EPR to assess patients, and although using IT was a barrier for some matrons, others were able to utilize the databases in innovative ways developing heuristics based on the volume of service use or complexity of the case. However, they felt they currently provided a reactive rather than a pro-active service.

Although assessment and acceptance of patients was dependent on the application of standardized criteria, these were modulated by the personal experience, interest and expertise. Some matrons accepted patients who would be rejected by others, based on the place of residence, mental health issues, substance abuse or involvement of other specialist teams. Sources of referrals varied depending on the local perception and understanding of the community matron role, whether practice or virtual ward based, and their individual network of contacts. There was evidence that training influenced practice.

Implications of findings

IT plays a significant role in the selection and assessment of patients by identifying cases that would otherwise not have been referred, and the systems used for routine care need to be accessible to community matrons. Better predictive tools that identify patients with whom the community matron can intervene are needed rather that the current systems that identifies risk of readmission—not the likelihood that community intervention will make a difference. Community matrons need improved IT skills and better teamwork and communication with other health and social care professionals. Personal referrals have some strengths over a technologically generated list and are at the very least assessed by a personal contact with the patient. However, non-standardized selection criteria will lead to some patient groups not receiving potentially beneficial intervention. Training and mentorship can develop and support the expertise of the community matron.

Comparison with the literature

Variation in case selection has been reported in other studies, supporting the need for better identification of people with high risk of hospital admission who may not be current service users and the need for sharing of experience and learning.20 Community matrons also tend to lack the confidence and skills shown by their practice-based colleagues.21 Technology is not isolated from complex social interactions within health care and technological tools offering a standard approach may not ‘fit’ every individual patient.22 The community matron works in a personalized way customizing the solutions to the patient based on her own experience and expertise.

As previously reported,23 we found that although some networks were clearly effective, team-working and inter-professional connections were poor for some community matrons. There was a need for greater collaboration, better communication and improved understanding of the role within primary health care to make the best use of the teams that exist and deliver the best health care to the most appropriate patients.

The UK Department of Health has developed an education framework for community matrons and case managers consisting of a specific education and skills training programme, based on a modular approach and practice-based learning.24 This is a valuable and appropriate approach to learning for community matrons.25 This programme has been successful in helping students to meet the nationally defined competences and provides a high level of guidance for the implementation of the community matron role. However, it is dependent on there being organizational infrastructure to support the role and opportunity for practice.26 Beyond this type of structured training, there are needs for mentorship and sharing of best practice within the local network. Additionally, the caseloads of the community matrons in our study were smaller than those suggested by the Department of Health.

Limitations of the method

Although we used a convenience sample of volunteers, we are happy that this study reached saturation. The initial aim was to investigate how the case finding tool was used in case selection and through an iterative process of modification and observation to develop a better case finding tool. At the midpoint, the emphasis of the study was shifted to explore the heterogeneity in case selection but did not substantially explore the role of training, mentorship or personal experience.

Call for further research

The development and validation of an improved case finding tool is required to help standardize the process of case selection, in particular to identify those patients who may not be in contact with general practitioners or other health care providers and whom the target workforce can influence. Incorporation of the heuristics used by community matrons in this study into the case finding tool may be useful. Further research is required to assess the benefit of the role of the community matron as a case manager of high intensity service users, in particular to compare the community matron with other case managers and to compare those acting independently with those working as part of a virtual ward. Audit of case selection against standards defined by the Department of Health is needed to identify inequalities in case selection and reasons for sub-optimal caseloads.

Conclusions

The complex process of case selection cannot be improved by modifying the case finding tool in isolation. Alongside IT inter-professional networks, the experience and expertise of the individual all impact on selection. However, combining an improved case finding tool with a more standardized role for community matrons could lead to more equitable case selection.

What is known on this subject:

• The management of long-term conditions in the community is a priority for health care.

• There is variation in practice among community matrons.

• There are existing computer-based tools for identifying cases.

• The role of the community matron is developing.

• There is an existing education framework for community matrons.

What this study adds.

• Access to computers and skills to use information technology remain a problem for nurses in the community.

• There is need to improve the specificity of a case finding tool that identifies patients who the community matron can intervene with. Useful heuristics have been identified in this study.

• There is need for improved collaboration among health care professionals and increased awareness of the role of the community matron.

• Training programmes need to be enhanced with sharing of best practice and mentoring of community matrons.

• Existing work practices may perpetuate disparities in caseload selection.

Declaration

Funding: Wandsworth Primary Care Trust.

Conflict of interest: none.

Ethical approval: none.

We would like to acknowledge and thank the community matrons who participated in the interviews. We also gratefully acknowledge the contribution made by the management team and the IT department at Wandsworth Primary Care Trust.