109 studies were found that fulfilled all inclusion criteria. An overview of these studies is presented in Appendix 3. Due to the large number of studies, each tabulated ‘included study’ (IncS) was allocated a number (e.g. IncS 1, 2 etc) for identification within the following text. Full references details for included studies are presented below the main reference list.
There was a diverse geographical spread across 17 countries, indicating a global interest and potentially concern about the issues surrounding pain assessment and management in care homes.
Methods and appraisal of included studies
Research methods used were diverse, including 13 RCTs, 18 non-randomised trials (typically pre/post-test studies), 16 qualitative studies, 30 of cross-sectional design, and 13 systematic reviews. Where studies were longitudinal (mainly RCTs) the longest duration of follow up was 6 months. While the majority of studies clearly detailed the methods used, study limitations largely related to methodological issues (e.g. cross sectional design) or small participant numbers. The difficulties inherent in including care home participants in reseach, who by the nature of their situation are vulnerable, was not widely acknowledged.
Our analysis provided a breakdown of topics of interest across and between studies, and revealed six initial themes:
Staff variables in relation to pain assessment and management (e.g. experience, qualifications etc.)
Beliefs and perceptions relating to pain (staff, residents, relatives)
Pain assessment tools
Pain assessment and management for nursing care home residents with cognitive impairment
Pain intervention efficacy/effects (pharmaceutical/non pharmaceutical)
Pain training interventions and their outcomes
These six themes were further amalgamated into three broader overarching themes or categories: Staff variables and beliefs, Pain assessment, and Interventions, as illustrated in Fig. 2 below. Due to some studies being multifaceted, results from a particular study may appear in more than one category.
Staff variables and beliefs
There were a number of included papers (n = 17) that explored the training level and experience of health care staff, and the impact this had on pain recognition and practices relating to pain management.
For example, Takai et al. (IncS 94) reported a relationship between increased nursing experience and lower pain prevalence in residents, and a relationship between the years of staff experience and the utilization of non-drug methods to treat it, being more frequently used by those with greater pain experience and training (IncS 95).
Experience tended to affect judgements: Alm & Norbergh (IncS 3) found that nurses with less experience had less confidence in making a judgement about pain themselves; however Baeza et al. (IncS 6) reported that care home staff with more experience tended to feel more stressed when not knowing the cause of pain for residents with dementia, and thereby less able to make their own judgement.
One systematic review (IncS 5) reported a correlation between less experienced staff and the verbally disruptive behaviour of residents, which may have been a manifestation of untreated pain.
Findings were also reported relating to staff turnover, type of working contract, and relationships with pain assessment and management, with evidence that pain assessment and management was performed better when there was lower staff turnover, and staff were hired permanently, and had longer tenure time (IncS 5, 16, 29, 39).
Care homes where staff had higher qualifications appeared to perform better in pain assessment and management, and give better quality of care (IncS 14, 15, 100). In addition, while some GPs lacked knowledge about pain assessment themselves, they valued the role of nurses and other caregivers (IncS 53), even although some staff perceived GPs as disinterested in relation to pain (IncS 78).
Care home staffs’ beliefs and perceptions exerted an influence on pain management, as examined in 40 of the included papers. For example, several papers reported that when staff felt more confident about the identification of pain, they were more certain in relation to its management (IncS 10, 44, 62). However, when staff considered pain to be part of ageing, they were less likely to identify pain in a resident (IncS 62). It was also reported that those residents who thought that pain was part of getting older were influenced by staff attitudes towards treatment, potentially resulting in reduced therapy options (IncS 104).
There was evidence that many staff did not agree with the statement that pain is part of aging and cannot be treated (IncS 8); however, with regard to residents with dementia, 50% of staff did consider pain as part of aging in this population (IncS 14).
When exploring nurses’ knowledge and attitudes to pain assessment in people with dementia, Burns & McIfatrick (IncS 13) reported that even amongst those with very recent training in pain assessment, it was still considered to be a “guessing game”; such beliefs had the potential to result in a higher prevalence of untreated pain in residents with dementia (IncS 94). However, when staff were familiar with residents, they found it easier to interpret non-verbal cues of pain (IncS 49, 61, 68, 104), and potentially differentiate it from other forms of distress (e.g. emotional distress).
Barry et al. (IncS 7) found that relatives who visited care homes more often considered that pain was being noted and treated to a greater extent by the staff. Also, relatives visiting their relatives frequently were more likely to interpret (and report) behavior changes as signs of pain (IncS 37), and thereby felt more actively involved in their care (IncS 24, 59).
Staff beliefs and concerns about the safety of using opioid analgesic were highlighted in several studies (IncS 9, 14, 53, 59, 76). These concerns related to either addiction beliefs, or fears about increased confusion and sedation; drug utilization was also reduced when either nurses or residents refused to use or take certain medications, even when their use may have been benefical.
There were some findings relating to staff perceptions about why residents might not report pain: for example in one paper it was suggested that staff were unsure if residents were displaying pill-seeking behavior because they wanted attention, rather than being truly in pain (IncS 30). In other papers, it was felt that residents might not be reporting pain because they were concerned about potential loss of independence (IncS 61, 76) or did not want to be seen as complainers or difficult residents (IncS 59, 61, 104).
Of the included studies, 16 examined the use of pain assessment tools; in general these did not appeared to be widely utilized (IncS 10, 31, 52), with some GPs in particular having little knowledge of specific tools to assess pain in care home residents (IncS 53). In one of these studies, detailing results from 810 participants across 7 European counries, it was reported that 58% of staff did not use any tool to assess pain in their daily practice (IncS 108). In other studies, although staff had knowledge of the tools, they reported problems interpreting them (IncS 97, 109), or felt them to be time consuming (IncS 42, 79, 84), However, when staff were trained in the utilization of tools, their use was more likely (IncS 97), and with increased usage, staff appeared to recognise their value to a greater extent (IncS 72). The impact on residents of tool usage was highlighted in several studies. For example, the proportion of residents with pain appeared to increase, because pain questions were being asked (and recorded) more frequently, and also pain was being treated to a greater extent (IncS 74, 83). The use of tools was considered to increase awareness, and render staff more sensitive and responsive to residents’ pain-related behaviour (IncS 69).
The most commonly researched tool in the included studies was the Pain Assessment in Advanced Dementia scale (PAINAD). Where used, pain assessment tools that were found to be helpful included the Pain Assessment Checklist for seniors with Limited Ability to Communicate (PACSLAC), DOLOPLUS-2, Pain Assessment in Non-Communicative Elderly Persons (PAINE), and the Pain Assessment for the Dementing Elderly (PADE). In contrast, the Abbey Pain Scale, while commonly used in practice, was found to be inaccurate when assessing pain intensity compared with staff estimated pain intensity, and it was also considered to be lacking in direction about precision of usage (IncS 94).
Of the 22 studies examining pain treatment interventions, most were shown to produce effective results. Where pain management practice improvements were made, analgesic and non-pharmaceutical treatment use increased and pain scores were shown to reduce (e.g. IncS 19, 36, 87). However, a systematic review (IncS 89) reported that pain scores deteriorated (increased) where there was a high turnover of staff, and lack of physician support.
Over half of the pain intervention studies (n = 13) were related to non-pharmaceutical treatments, and many indicated that alternative therapies had the potential to be effective (IncS 20, 28, 36, 62, 101). Non-pharmaceutical treatments, which included massage, aromatherapy, exercise, psychological support, and humour therapy, were viewed favourably by staff (IncS 1), but not necessarily by relatives [IncS 1, 79). Managerial support and staff training for non-pharmaceutical treatments were seen as necessary to facilitate usage (IncS 1, 42, 60). However, a large study by Lukas et al. (IncS 71; participant n = 1900) indicated that moderate to severe pain was more likely to be treated by pharmaceutical means, and less likely to be treated with non-pharmaceutical or alternative therapies.
Among the 21 included papers that considered pain management for care home residents with cognitive and/or communication impairment, there was general acknowledgement that the recognition of pain was challenging (IncS 2, 6, 14, 18). There was also recognition that these residents’ pain may be less well treated (IncS 35, 44, 75), especially where cognitive impairment was more severe (IncS 35). A lack of staff knowledge (IncS 9), inadequate assessment (IncS 27, 46, 75, 77), or availability and appropriateness of pain education programmes were influencing factors (IncS 14, 24). However, training programmes, as further detailed below, could help to increase awareness (IncS 40, 43, 84), as well as improving recognition of behavioural and non-specific manifestations of pain for people with cognitive impairment (IncS 99).
Pain training interventions were examined in 21 papers, and were generally shown to be effective in improving pain management and treatment use/appropriateness (IncS 19, 32, 33, 39, 40, 43, 63, 79, 84, 87, 97). As well as improved pain management, resident quality of care and quality of life also had the potential to be enhanced (IncS 11), which could impact on their longer term wellbeing
There was some suggestion that increasing staff knowledge regarding pain management could reduce anti-psychotic drug use (IncS 17). Training was also helpful in the use of non-pharmaceutical treatments (IncS 19, 79), helping to improve staff awareness and confidence (IncS 102). There were also indications that training for non-registered care staff might be beneficial (IncS 51, 96).
Consultation with stakeholders
Evidence from the field research carried out at 16 care homes is summarized in Table 3, and confirms many of the above points, including staff views regarding lack of eduation and training, lack of systematic use of pain assessment tools, and nearly all mangers expressing the need to improve pain assessment and management in their care home.
Summary of evidence from stakeholders