Expert opinion on the management of pain in hospitalised older patients with cognitive impairment: a mixed methods analysis of a national survey

Older patients in hospital are complex with co-morbidity, polypharmacy and frailty being the norm. Pain is highly prevalent. Although a subjective and personal experience, pain is unanimously unpleasant with sensory and emotional factors impacting on functional and mental capacity, social interaction and quality of life [1]. Multiple physiological and cognitive factors such as damaged nociceptive pathways, lower pain thresholds and mood contribute to the prevalence of pain in the elderly population. With more disease (especially osteoarthritis), immobility, muscle weakness and frequent falls persistent pain is common, particularly chronic musculoskeletal pain. When pain is coupled with cognitive impairment the impact and disability is far greater [2].

Pain is no less frequent or intense in people with dementia, with around 50% of these individuals experiencing pain regularly [3]. However studies have shown that pain is often under recognised and under treated in the cognitively impaired [2,4]. People with dementia are prescribed and given less analgesia than other older people [5]. When analgesia is prescribed to someone with dementia 83% do not receive their medication [6] and 76% of people with dementia did not receive regular analgesia post operatively for hip repair, despite 42% expected to be in severe pain [7].

People with dementia may have difficulty in communicating their pain and often care staff do not recognise the non-verbal signs they may exhibit when in pain, hence missing diagnostic clues [8]. Furthermore, reduction of semantic memory for pain seen in people with dementia and reduced ability to identify pain is associated with a decline in pain reporting [9].

Pain may manifest itself differently in the cognitively impaired and associated changes in behaviour may be attributed to something other than pain leading to late or misdiagnosis of pain [10,11]. The behavioural and psychiatric symptoms of dementia are typically treated with antipsychotics or sedatives which may be unnecessary and harmful if the underlying stimulus such as pain is recognised and treated accordingly. Ultimately untreated pain is devastating in these individuals unable to explain their pain [12]. Using a standardised protocol to treat pain in nursing home residents with moderate to severe dementia has been demonstrated to significantly improve pain, agitation and aggression and could help reduce the unnecessary use of antipsychotics in people with dementia [13].

Management of pain in older people and particularly the cognitively impaired is challenging. Adverse drug reactions are two to three times more common in older people than younger individuals [14]. The risk of delirium is higher and further augmented in those with existing cognitive impairment whether due to vascular or degenerative brain disease. Additionally, lack of recognition of side effects in individuals with dementia may occur [15].

Despite recent publications regarding pain management in older people such as that in Age and Ageing [16], there is little specifically targeted at those with cognitive impairment. It is accepted that clinical guidance should be based on the best available evidence and that from randomised controlled trials is gold standard. Due to the complexity and heterogeneity of this elderly population with cognitive impairment randomised controlled trials are methodologically difficult, hence the evidence base in this group is lacking. However we recognise the wealth of expert opinion in this field. The aim of this paper is to collate this opinion in order for it to be useful in clinical practice.

We obtained a list of consultant Geriatricians/Dementia Leads identified in the National Dementia Audit from the British Geriatrics Society. This amounted to one hundred contacts. We contacted them electronically with an invitation to respond to a questionnaire relating to their role, experience, confidence and opinions on pain management in elderly patients with cognitive impairment. A second and third electronic message was sent after two weeks and one month respectively as a reminder. Their responses were anonymised. Ethical approval was sought but the Local Research Ethics Committee advised review was not necessary.

Those completing the questionnaire were presented with three case scenarios (Additional file 1). They were asked to provide an account of how they would manage the case and what factors needed to be taken into consideration in the management of such cases.

Analysis was undertaken in two parts. Responses to closed questions (individuals’ confidence with the management of each case scenario and their use of specific analgesic medications) were analysed quantitatively. Thematic analysis was undertaken of text responses. Comments were grouped under common themes and reported findings based on the common responses provided. The coding framework used was established from the responses and was not pre-determined.

This work was supported by NHS Fife Endowments-Geriatric Medicine Training Fund. This specifically funded analysis and interpretation of data.

The management of pain is a fundamental part of the systematic and effective delivery of compassionate and dignified care to older people. The experience of pain can be disabling both physically and psychologically. This is heightened in those with cognitive impairment and/or dementia, yet high quality evidence to guide management in this area is lacking. The synthesis of expert opinion in this study highlights the multi-faceted complexity of pain management in this population and significant potential for harm as a result of the analgesic smorgasbord currently available to the clinician. The preferences of respondents in this study do reflect the principles outlined in the BGS standard [16] although it only considered the management of pain in older people in general not specifically those with cognitive impairment. It also recognised the lack of well controlled studies in this population from which to base the evidence.

This study has many strengths; opinion was gathered from experts across the UK. There was no pressure or incentive for them to respond other than assisting the common goal of improving patient care. The response rate to the questionnaire was 48% and the majority of respondents were very confident in the management of these scenarios when asked to self-rate their confidence. A mixed methods analysis was employed to capture the nuances of individual clinicians approach to pain management. Limitations of the study however do include the fact that results are based on expert opinion i.e. the lowest grade of evidence, but as stated previously there is an evidence gap in the management of pain in this population. Despite the good response rate the total number of respondents (37) is small.

The lack of consensus that has resulted from this study highlights the complexity of the clinical field when managing pain in this population. The need for evaluation of the whole individual (a Comprehensive Geriatric Assessment approach), including their family and caregivers and the environment are promoted rather than an approach that would just simply treat the pain. The opinion that the potential to do significant harm has come out loud and clear in all responses: some medications are more likely to cause harm than others and therefore evident that some are a more favourable option than others, so despite no definite consensus the appearance of higher risk and lower risk medications has become apparent. Steps 2 and 3 on the WHO analgesic ladder are blurred in this high risk population. The principles guiding the management of pain set out in the WHO pain ladder are applicable but as severity of pain increases, the choice of analgesic becomes more difficult because of the facility to do harm. A number of results are however unequivocal; paracetamol is safe and should be the first choice analgesic in this situation. Newer opiates should be used in preference in renal impairment. There appeared to be caution noted with prescription of codeine and tramadol. Nefopam appears unsafe for use in this population. These points are immediately transferrable to clinical practice but further unpicking and comparison of the risks associated with other specific analgesics in this population is necessary before a comprehensive clinical guideline can be produced. The principles guiding pain management in this population however are that individuals should be comprehensively assessed and prescription of analgesia should be done with a view to minimising resulting harm; using the principle of “start low and go slow”.