The assessment and management of peri-operative pain in older adults
Correspondence to: P. A. Schofield
A number of recent reports have highlighted the inadequate provision of pain relief for older inpatients. Despite the availability of numerous validated pain measures, pain remains poorly assessed in some cases and, particularly, in the cognitively impaired. Without proper assessment, patients may receive inadequate or inappropriate analgesia, both of which can worsen outcome. Most drugs and techniques that are used for analgesia in younger patients are also suitable for older patients, although dosages may have to be adjusted to avoid the side-effects that are consequent upon age-related changes in drug pharmacokinetics and pharmacodynamics, co-morbidity, frailty, cognitive impairment and polypharmacy. This paper reviews current guidelines and methods of assessing pain in the older adult, and describes the use of, and problems with, mild, moderate, strong, adjuvant and local anaesthetic drugs in the older population for analgesia, advocating multimodal intervention to reduce dose-related side-effects, particularly of opioids.
Pain in older people is under-recognised and undertreated [1-6]. Even when treatment is prescribed, it is often limited to basic medication, and seldom tailored to the individual [7-10]. In addition, professionals often fail to consider alternative pain relief options, such as exercise, rest and thermal strategies . There is undoubtedly more that needs to be done with regards to pain management for the older adult. National guidance is overdue for both acute and chronic pain management, bearing in mind that, as long ago as 1997, Desbiens et al.  demonstrated that 46% of older people admitted to hospital reported pain, 19% of whom had moderate or extremely severe pain, and 13% of whom were dissatisfied with their pain management. Pain relief is as effective for older as for younger patients, but professionals tend to underestimate pain needs, underprescribe and undermedicate .
Admissions to medical and surgical wards are three times more common for over 65-year-olds than for younger patients, yet staff lack the experience in pain assessment and management that might be expected in dealing with the specific problems associated with this population. Attitudinal barriers to pain assessment and analgesia persist among health professionals, related to prescribing potentially harmful drugs for patients with co-morbidities, consequent polypharmacy, cognitive impairment, frailty, reduced physiological reserve, and age-related changes in pharmaco-kinetics and pharmacodynamics .
Both the peripheral and central nervous systems are affected by ageing, with reduction in β-endorphin content and γ-aminobutyric acid (GABA) synthesis in the lateral thalamus, a decline in the concentration of central GABA and serotonin receptors, a decrease in the speed of nociceptive processing, and a decline in C and Aδ fibre function .
The threshold for pain perception appears to be increased in older people when nociceptive stimuli are shorter, of lower spatial extent, are presented at peripheral cutaneous or visceral sites, or are thermally or electrically induced , but it remains uncertain whether or not pain perception threshold generally increases with age .
The assessment of pain in the older population
Pain is a subjective, personal experience, known only by the person who suffers it. The experience of pain is multidimensional and may be described variously according to sensory (intensity, location and character), affective (emotional perception) and functional (impairment of functional ability) components. The process of assessment may be further complicated if the patient has severe cognitive impairment, or there are communication difficulties or language and cultural barriers.
Pain is ‘the fifth vital sign’ and any peri-operative assessment of older surgical patients should include a question about whether the patient has pain , using an intensity rating scale, which also enables assessment of the response to treatment, for example, the Numeric Rating Scale (0 no pain, 1–3 mild pain, 4–6 moderate pain, 7–10 severe pain), Verbal Descriptor Scale none, mild, moderate, severe) or Visual Analogue Scale . Intensity scales can be used for older people with all but severe cognitive impairment, but vary in how easy they are to complete according to age and cognitive function [19-22]. Follow-up questions about the nature of pain (‘does it ache?’, ‘is it sore?) can be useful if a patient denies pain , as older people may be stoic, or reticent about reporting pain for fear of complaining . Whichever scale is selected, attention should be paid to the presentation to ensure that it is in large, clear letters/numbers and presented in good lighting.
Observing the patient can also provide some very useful information regarding pain, particularly when there are communication difficulties. Behavioural indicators of pain, such as facial expressions, physical reactions (e.g. guarding, bracing, rubbing the painful area) and negative reactions (e.g. agitation) , can vary between individuals and within the same individual, and can occur simultaneously. Facial expressions associated with pain, including brow raising and lowering, cheek raising, eyelid tightening, nose wrinkling, lip corner pulling, chin raising and lip puckering, are correlated with the experience of pain, particularly in cognitively impaired patients . However, some indicators, for example, social withdrawal, can be subtle, and associated with causes other than pain . Behavioural changes should prompt carers or healthcare professionals to exclude pain as a cause, through more detailed clinical assessment. It is important to include carers in any pain assessment process as they are often more familiar with the subtle changes in the patient's behaviour that indicate pain, although they may overestimate the presence and degree of pain . Physiological cues, such as pallor, tachycardia and hypertension, can indicate pain, but may be absent if the pain is chronic.
In the peri-operative setting, acute pain is most likely to be related to the surgery performed, but may be superimposed upon chronic pain. Fuller assessment of the patient's background pain should therefore be undertaken as part of general pre-operative assessment, including information about onset, time, course, radiation, aggravating and relieving factors, quality, associated symptoms and medication. Location can be mapped graphically with reasonable test–retest reliability in the elderly . Multidimensional assessment is ideal, because it includes assessment of mood and function, both of which can affect postoperative rehabilitation [30, 31], and can be performed using scales such as the McGill Pain Questionnaire , brief pain inventory  or geriatric pain measure , although further research is needed to support the validity of using these in some older patient groups .
The assessment of pain in older adults with cognitive impairment
A number of behavioural scales have been developed to assess pain in older adults with cognitive impairment [36-42], and consistently include seven main indicators: physiological observation; facial expressions; body movements; verbalisations; and changes in interpersonal interactions, activity/routines and/or mental status. No single instrument can be recommended currently for general use .
The 2007 Royal College of Physicians/British Pain Society/British Geriatrics Society guidelines The Assessment of Pain in Older People  distinguish two different approaches to pain assessment according to an older patient's ability to communicate, providing an algorithm for use in clinical practice (see Appendix 2, reference ). Updated guidelines will be published early in 2014.
Approximately 56% of men and 65% of women aged over 75 years are normally in pain or discomfort, a proportion that rises with institutionalisation and hospitalisation, yet the elderly are consistently less likely than younger patients to receive good pain management [1, 45, 46].
Few studies have specifically investigated the effects of analgesic drugs in older people. Generally, results have been extrapolated across the age spectrum from primary studies involving younger participants , which is problematic in many ways . For many analgesic medicines, for example, a lower initial dose may be required compared with that administered to younger adults, with subsequent doses titrated according to response.
Both the American  and British  Geriatrics Societies advocate a ‘pain ladder’ approach to pharmacological pain management in the older population, using the safest drugs administered by the least invasive route, and only escalating treatment if analgesia remains ineffective. The American guidelines of 2002 , which recommended cyclo-oxygenase-2 (COX-2) inhibitors, were revised in 2009 , after the withdrawal of rofecoxib and valdecoxib due to concerns about cardiovascular safety in patients with heart disease or cerebrovascular accident . Paracetamol is recommended as first-choice oral analgesia [47, 49], due to its efficacy and safety. At the lowest effective dose, non-steroidal anti-inflammatory drugs (NSAIDs) can be added or substituted in patients who do not respond adequately to paracetamol. In patients with an increased risk of gastrointestinal problems, either a COX-2 inhibitor or an NSAID with a gastro-protective agent should be used. Opioid analgesics, with or without paracetamol, can be useful when NSAIDs or COX-2 inhibitors are ineffective, contraindicated or poorly tolerated [47, 49].
Multimodal therapy, using combinations of weaker and stronger analgesics in combination with adjuvant agents, local or regional analgesia and non-pharmacological therapies (e.g. cooling, elevation, splinting, surgery) may provide analgesia whilst sparing the patient side-effects associated with stronger opioid analgesia . Like any other age group, older patients will respond better to treatment if they are given the opportunity to make an informed choice about analgesia. Information should be provided in an age-friendly format (e.g. large print, simple drug information and dosage scheduling) and in accessible packaging after hospital discharge.
Paracetamol is readily available over the counter, and in many over-the-counter cold and flu preparations. It is the preferred analgesic for older adults with musculoskeletal problems and can be used for some mild forms of neuropathic pain. No dosage reduction is necessary for older adults, although care should be taken not to exceed a 4-g limit in 24 h. Liver damage is more likely when the patient is fasted, dehydrated, poorly nourished or has high alcohol consumption.
Non-steroidal anti-inflammatory drugs
Non-selective cyclo-oxygenase inhibition, together with age-related pharmacokinetic changes and co-morbidities, renders the older patient at relatively higher risk of adverse effects from NSAID consumption, including gastrointestinal toxicity and bleeding, renal dysfunction, hyponatraemia and impaired hepatic function . Caution should be used when prescribing in older adults with pre-existing peptic ulceration, cardiac failure, hypertension or renal impairment. NSAIDs with short half-lives, such as ibuprofen and diclofenac, appear to have fewer side-effects. Patients should take the lowest effective dose of NSAIDs or COX-2 selective inhibitors for the shortest time necessary to control symptoms . Co-administration with food, or with a proton pump inhibitor or misoprostil, reduces the risk of gastrointestinal complications.
Despite concerns about the side-effects and adverse reactions associated with NSAIDs, they are highly effective in most cases for reducing mild to moderate pain . NSAIDs alone produce as good analgesia as single or multiple doses of weak opioids alone or in combination with non-opioid analgesics .
Codeine is often used for short-lasting predictable incident pain and can be used alone or, more effectively, in combination with paracetamol. There is a degree of variability in effectiveness, with up to 30% of patients being poor debrisoquine hydroxylators/O-demethylators of codeine to form morphine, such that analgesia is limited or not achieved . Conversely, 1–3% of British patients are ultrafast metabolisers of codeine and tramadol (UM CYP2D6 genotype), converting these prodrugs to their active forms rapidly, leading to their accumulation in renal failure, and resulting in respiratory depression or apnoea . Conversion to active metabolites is inhibited by common medications, including cimetidine, haloperidol, amitriptylline and many selective serotonin re-uptake inhibitors (e.g. fluoxetine). Use is further limited by side-effects, including constipation, confusion and nausea.
In the older population, tramadol may cause fewer respiratory and gastrointestinal side-effects than other opioids, but is associated with delirium , and cannot be tolerated by a third of patients, due to nausea, vomiting, sweating, dizziness, tremors and headaches.
In spite of there being little evidence for their specific use in the elderly [6, 59], opioids are well established for the treatment of cancer pain, and increasingly used for the treatment of chronic pain. Similarly to prescribing other analgesics, the general rule for prescribing opioids in older adults is to ‘start low and go slow’, anticipating side-effects and treating them accordingly  rather than stopping treatment, and introducing different medications sequentially. For example, co-administration of stool softeners and anti-emetics helps to prevent the common side-effects constipation and nausea. Drowsiness is common in the first few days of prescription and patients should be warned about this. From the limited number of studies available in the older population, opioids do not appear to increase postoperative delirium, although this effect is difficult to determine as comparative patients prescribed insufficient non-opioid analgesia are at greater risk of delirium . The validity of a ‘start low and go slow’ approach has been called into question for this reason, i.e. that it leads to inadequate analgesia in patients who are already at higher risk of confusion and agitation . However, pethidine is consistently associated with delirium, and should be avoided, as its metabolite norpethidine can cause excitement, agitation, twitching and tremors. Frail and cognitively impaired patients do not appear to have reduced pain perception and hence appear to require similar doses of analgesia to ablate pain, although more research is needed to confirm this  and analgesia should be prescribed with particular caution to these groups . Buprenorphine may be a more appropriate choice of opioid for patients with renal disease, as its pharmacokinetics are unaltered by renal function .
Older people experience neuropathic pain with a disproportionately higher incidence than younger people, related to increased co-morbidity. Patients may present for surgery already taking antidepressant, anticonvulsant or anti-epileptic drugs, which cause numerous side-effects and potential for drug interactions .
Newer postoperative analgesia strategies involving, amongst others, gabapentin, pregabalin, ketamine and dexmetomidine are promising, but may be limited by side-effects in older patients, in whom more specific research needs to be carried out. Topical therapies involving lidocaine, capsaicin and NSAIDs are not without side-effects, but may be more appropriate for localised pain distant to the site of operation in patients who cannot tolerate systemic analgesia .
Regional techniques, either as the sole method of anaesthesia or as adjuvant analgesia, are well tolerated by the older population, attenuating the surgical stress responses, limiting opioid co-administration and accelerating patient rehabilitation and recovery . Their use may benefit the elderly more than younger patients, due to the relatively greater risk of side-effects in the older adult if opioids were otherwise used [13, 66-68], although, as ever, more research is required comparing postoperative opioid with regional analgesia in the elderly. Age-related alterations in both anatomy and neural micro-anatomy may make correct siting of any block technically more difficult, and drug effects less predictable , blockade being achieved with lower doses of local anaesthetic  and lasting longer .
In conclusion, older adults represent the most rapidly expanding demographic in society and present disproportionately more frequently for surgery than younger patients. They are more likely to experience chronic pain as a result of their increased prevalence of co-morbidities, and, although currently more stoic and less likely to report pain, consequently take greater numbers of analgesic medications long-term, increasing the likelihood of peri-operative drug interactions. Postoperative analgesia is often inadequate, and poorly assessed in patients with communication difficulties or cognitive deficit , despite recent advances in both provision and validation of evaluation tools. Multimodal pain intervention reduces the likelihood of opioid use and side-effects, to which the older population may be particularly prone.
However, the evidence base for specific analgesic interventions in the older population is based on experience and derivation from studies involving younger adults, placing older patients at higher risk of avoidably poor outcome, resulting from both inadequate analgesia and side-effects of medication. Further advances are not simply a case of ‘more research is needed’ so much as identifying what specific questions need asking, and how research might best address these, with the goal of providing universally adequate analgesia for the older population, with minimal side-effects.
No external funding and no competing interests declared.