Nurse-led implementation of the single assessment process in primary care: a descriptive feasibility study

Abstract

Objective: to determine the resources required to carry out the single assessment process in primary care.

Design: prospective descriptive study.

Setting: one urban primary care practice, Southampton.

Participants: nine hundred and forty-four people aged 70+ years, registered with the practice, not living in a residential/nursing home, or terminally ill.

Intervention: participants were sent the six-item Sherbrooke questionnaire (case-finding tool). Non-responders were re-mailed after 4 weeks. All those scoring 4, 5 or 6 and a randomly selected half of those scoring 2 or 3 were offered overview assessment and comprehensive assessment as indicated by the Minimum Data Set for Home Care protocol. The nurse assessor identified unmet needs and agreed an action plan with participants. Another researcher conducted semi-structured interviews with a purposive sample of 26 participants to elicit their views of the process.

Main outcome measures: response rates/scores of Sherbrooke questionnaire; numbers/characteristics of people requiring overview and comprehensive assessments; nature of resulting recommendations/referrals and impact on other agencies; resources required; views of service users.

Results: eight hundred and sixty-three (91%) participants replied. Five hundred and seven (54%) scored 2+, triggering an overview assessment, which was offered to 307. One hundred and twenty-four participants (40%) accepted; 64 (52%) had unmet needs (median 8 each, range 2–18), resulting in 34 referrals within the practice including four case conferences, and 21 to community/secondary health services. Few participants with a Sherbrooke score of 2 required comprehensive assessment. Users perceived the process as acceptable and useful, but not always relevant to their current needs.

Conclusion: targeting those scoring 3+ on the Sherbrooke questionnaire (28% of sample) may improve the identification of patients who would benefit from further assessment. A contact approach rather than a case-finding one may improve the relevance of this process to older people.

Received 15 June 2005; accepted in revised form 7 March 2006

Introduction

In 2001, the National Service Framework for Older People [1] set out a 10-year programme to improve the health and social care of older people in the United Kingdom. Through Standard Two (person-centred care), it required local health and social services to cooperate to implement a single assessment process (SAP) by April 2002, later extended to April 2004. The aim of the SAP was to ensure that older people receive appropriate, effective and timely responses to their health and social care needs and that professional resources are used effectively without duplication and in proportion to people’s needs [2]. It is based on four levels of assessment: contact (basic information collected on the first contact with health/social services), overview (if a more rounded assessment is required), specialist (where there are specific needs from one professional) and comprehensive (where specialist assessments are required across several areas or where additional support is required).

Local health and social services were required to collaborate to agree/develop their assessment tools, means of sharing data, preparations for implementation and review of the SAP. Much effort has been directed to adapting or developing assessment tools, often without validation, and a substantial change of policy, practice and culture may be required to encourage a genuine collaboration between health and social services [3]. Progress with implementation is variable, with more activity in community and social services and less activity reported from hospitals. In one acute hospital where it has been implemented, the task has fallen entirely on nurses who expressed concerns over the time required to complete assessments [4].

It is likely that primary health care teams will have to perform contact, overview and comprehensive assessments of older people’s physical, mental and social function and decide where further specialist assessment is warranted. The feasibility of carrying out this process in primary care and its impact on secondary care and other agencies are unknown.

Contact assessments can be opportunistic, as older people are reviewed by health or social care professionals, or a case-finding process may be employed. The Sherbrooke postal questionnaire [5] is a case-finding tool recommended by the Department of Health (DoH). Validated in Canada, it identifies patients requiring further assessment as those who score ≥2 or do not respond to the questionnaire. It is unknown how many assessments its six broad questions will trigger when used in the United Kingdom—the only published study had a low response rate of 61% [6].

The Minimum Data Set for Home Care version 2.3 (MDS-HC) is accredited by the DoH for overview and comprehensive assessments [7]. Validated in many countries including the United Kingdom, it enables the assessment of patients’ function, health, social support and service use. People who have specific problems or who are at risk of functional decline are identified and trigger one of 30 client assessment protocols (CAPs). These allow further assessment and individual care planning, leading to service provision or referrals/recommendations as appropriate.

We have evaluated the feasibility of using these SAP tools within a primary care setting in the United Kingdom.

Participants and methods

Participants

Patients aged 70 and over on 1 April 2003 registered with one primary care practice in urban Southampton were eligible. Excluding 42 living in residential or nursing care homes, 944 patients comprised the study population.

Study design and data collection

Study participants were sent the six-item Sherbrooke postal questionnaire with a letter from their own GP and a reply-paid-envelope addressed to the practice. If no response was received within 4 weeks, and the participant was not known to have moved or died, a further questionnaire was sent. Those who had not replied within another 4 weeks were deemed non-responders.

Over the next 12 months, all participants scoring 4, 5 or 6 positive answers were offered an overview assessment at home by a research nurse (experienced in elderly care), leading to a comprehensive and/or specialist assessment as appropriate. Time constraints required that an identical assessment by the same nurse at the surgery was offered to half (randomly selected) of those participants scoring 2 or 3 positive answers.

To ascertain users’ views, we selected a purposive sample of participants who received the postal questionnaire, overview and comprehensive assessments to encompass a range of gender, age and Sherbrooke scores. Qualitative data were gathered from semi-structured interviews on their experience and understanding of the SAP, its acceptability, usefulness and relevance to their perceived needs.

The resources used to enact the SAP were monitored. Time sheets were maintained by practice administrative and clinical staff for the eighth month of the 1-year assessment period. The research nurse recorded over a 3-month period the time spent for travelling and performing assessments, and enacting recommendations. The costs of equipment were recorded.

Ethical approval was granted by the local research ethics committee who refused permission to follow up the non-responders. Written consent was obtained for all assessments and interviews. The study was supported by a specialist registrar start-up grant from the British Geriatrics Society and a research and development training fellowship from Southampton University Hospitals NHS Trust. The funding bodies played no role in the conduct of the study.

Data analysis

The Sherbrooke results were entered onto a spreadsheet and analysed using descriptive statistics (summation, median and percentages). The MDS-HC assessments were entered onto a software programme which generated the CAPs. The ensuing recommendations/referrals, resource use and whether the needs had been previously met were again analysed using summation, median and percentages.

The semi-structured interviews were tape-recorded and transcribed using a simple word-processing package to facilitate the initial categorisation and ordering of the data. Data analysis followed the inductive coding techniques described by Strauss and Corbin, in which concepts were initially identified that were ‘grounded’ in the data [8]. These were then refined into categories with this process being completed to higher levels of conceptualisation until core categories were selected. Adopting such an approach ensures that the respondents studied are given a voice so that interpretations include the perspectives of those being studied [9]. Thus, data were analysed thematically taking account of the framework of topics incorporated into the interview schedule, alongside issues emerging from the respondents’ concerns and priorities expressed during the interview process.

Results

Eight hundred and sixty-three (91.4%) Sherbrooke questionnaires were returned (see Figure Appendix 1 in the supplementary data on the journal website http://www.ageing.oxfordjournals.org). Five hundred and sixty-six (60%) participants triggered further assessment by scoring ≥2 on the Sherbrooke questionnaire (n = 484), having missing data with the potential to score 2 (n = 23), or failing to reply to both mailings (n = 59) (see Table Appendix 2 in the supplementary data on the journal website http://www.ageing.oxfordjournals.org). We were unable to assess the 59 non-responders further because of ethical committee restrictions. The items on living alone and taking medications received most affirmative replies (Table Appendix 2), and those people with higher scores or who failed to respond were older and included more men (see Table Appendix 3 in the supplementary data on the journal website http://www.ageing.oxfordjournals.org).

We invited, for further assessment, everyone (n = 113) who scored 4, 5 or 6 positive answers on the Sherbrooke, 76 of 152 people who scored 3, 107 of 219 people who scored 2 and 11 of 23 people with missing data who could have scored 2. Eighteen participants had recently died and 165 refused further assessment or were not contactable despite several attempts (Table 1). Three hundred and seven patients were offered MDS-HC assessments, and 124 (40%) accepted. Participants with higher Sherbrooke scores required proportionally more comprehensive assessments than those with lower scores. Sixty-four patients (52%) triggered 475 CAPs in 27 domains (Table 2). The most frequently triggered CAPs (>30) were immunisation and screening, medication management, communication disorders and pain. No one triggered alcohol or elder abuse, or palliative care. The participants triggered a median of eight CAPs each (range 2–18).

For each CAP, the research nurse assessed the situation, including measures already in place, and agreed any action with the patient and main carer where appropriate. The 55 referrals thus generated are listed in Table 3.

Interviews with participants

Saturation of the data was achieved after 26 interviews with 14 male (median age 82 years, range 71–91) and 12 female (median age 80.5 years, range 71–100) participants. Seven lived alone, 15 with a spouse or partner and four lived with a son or daughter. Four interviewees had a Sherbrooke score of 6, six scored 5, eight scored 4, five scored 3 and three scored 2. All were of White British ethnic origin (reflecting the local population), and all but one had undergone comprehensive and overview assessments.

The Sherbrooke questionnaire was perceived as easy to complete. The MDS-HC was also seen as straightforward, and the interviewees appreciated the length of time spent with the nurse, her thoroughness and the opportunity for unhurried discussions. Five people felt they had benefited from the SAP: two through referrals to audiology and occupational therapy and three appreciated the interest shown or general support and advice given, for example on spouse’s dementia. However, many perceived it as a project to benefit the older population in general rather than themselves. The majority felt that they were managing but were in favour of a regular check-up, particularly for those living alone. Two people perceived confusion/duplication between routine well man and diabetic clinic visits, and one person expressed disappointment because of routine delays in implementing bathroom adaptations (requested by the occupational therapist as a result of the SAP assessment).

Resources required

The time spent on this assessment process over a typical month was ∼2.7 h for GPs (reviewing patients and their notes on request), 1 h for clerical staff (pulling notes and filing) and 3 h for the secretary (typing letters and sending appointments). The research nurse spent 30 min with the patient for an overview assessment and 1.75 h for a comprehensive assessment, with an additional 1.25 h to prepare, make referrals and follow them up, and record the data onto a software programme. Return travel time to the patient’s home was typically 30 min. The research nurse undertook a training day on the MDS-HC, and a laptop computer and a licence for the software were required.

Discussion

The Sherbrooke postal questionnaire had a high response rate (91%), which may reflect the general practitioners’ involvement with this study. Further assessment was triggered by 60% by scoring 2+ or failing to respond, compared with 56% in the original Canadian study [5] and 63% in the recent London study [6]. Because few people scoring 2 required comprehensive assessment, a cut-off score of 3+ may be more cost-effective in the United Kingdom. This would be supported by the association of higher risk of A&E attendance with progressively higher Sherbrooke scores found in London [6]. In this highly organised general practice, 6 of 40 patients scoring 5 or 6 on the Sherbrooke were not known to have significant problems (which were similar to those of the other ‘high scorers’) and another three were coping well with obvious problems without primary care input. Of these 40 patients, 10 have died/become gravely ill since completing the Sherbrooke 22 months ago. Fifty-two per cent of participants accepting MDS-HC assessments triggered CAPs for specific problems or risk of functional decline as listed in Table 2, and the need for four case conferences suggests that the SAP highlighted complex issues. Fifty-four per cent of participants offered MDS-HC assessment refused, and it is unknown whether they would have had similar results.

This study was designed to form part of an existing primary care service to maximise the relevance of the findings to service development. Because 307 patients of the 566 screening positive were assessed, the staff time commitments might be doubled per 1,000 elderly people initially screened. The GP’s time may have been minimised by the experience of the elderly care nurse practitioner. This approach was also associated with several limitations. The number of referrals resulting from the SAP may have been reduced by the input of the existing elderly care nurse practitioner and influenced by our use of an experienced nurse researcher. Some participants were assessed at home and some at the surgery, which may have differentially affected the participation rate, although the overall refusal rate was similar at 52–57% in both settings. Finally, we were refused permission to follow up the non-responders who were potentially the most frail and in need of assessment.

The high refusal rate for the MDS-HC overview assessment is of note and may reflect the perceived lack of personal benefit in what was known to be a study. The recently published Medical Research Council trial of multidimensional assessment and management of older people also found a high non-response rate to referrals to general practice (30%) and geriatric management (50%) [10]. They were unable to ascertain reasons for refusal, but our interviews with participants suggest the importance of timing in acceptance of a new service/referral, and that older people may take a while to consider new options which they may accept at a later date. This may partly explain the high rate of refusal of measures to address unmet need, where people felt they were ‘coping’ or just ‘old’. The concepts of self-care and self-management are central to chronic disease management [11], and this may be more difficult to achieve with frail older people than with people with specific conditions such as diabetes.

Older people experienced the SAP as acceptable and useful in accessing formal services to meet their health-orientated needs. Although older people valued their access to formal services, most assumed a major role in managing their own health and well-being, managing difficulties stoically and drawing on support from family members and other informal sources. These assessments may not recognise the complexity of balancing formal and informal support in the context of people’s daily lives.

The Sherbrooke questionnaire may be helpful in identifying people for chronic disease management schemes. People scoring up to 3 might benefit from self-management programmes such as the Expert Patient Programme [12]. People scoring ≥4 on the Sherbrooke required most of the comprehensive assessments, and this might be a method of identifying people suitable for a case-management approach [13]. Further research is required.

This study used two SAP tools, and it is unknown whether the other accredited tools would have produced similar results. It is also not known whether these findings could be translated to other primary care practices or social service settings with assessors of different professional backgrounds. It remains unclear at what intervals the SAP should be repeated. The importance of timing was raised by the user interviews and highlights the possibility that an assessment could become out-of-date as vulnerable older people’s circumstances change. Contact assessment rather than a case-finding approach may avoid this issue.

Key points

• A Sherbrooke questionnaire score of ≥3 identified patients likely to benefit from further assessment.

• The MDS-HC tool identified complex issues but generated a modest number of new referrals to community/secondary health services/social services.

• The participants interviewed perceived the SAP to be acceptable and useful.

• Its relevance is dependent on timing in relation to changes in older people’s circumstances, and this may affect the uptake of recommendations.

Conflict of interest statement

All authors declare that there are no conflicts of interest in connection with this article.

Contributors and acknowledgements

All the authors contributed to study design and/or data analysis and were involved with producing this research paper. We acknowledge the support of all the staff of the Old Fire Station surgery, and Mrs Michelle Morron-Ryan who was involved with the study conception.

References