Patterns of analgesic use, pain and self-efficacy: a cross-sectional study of patients attending a hospital rheumatology clinic

Arthritis is a highly prevalent condition and can lead to physical impairments and psychological morbidity, including depression, anxiety and feelings of helplessness [1]. A long term need for analgesics and non-steroidal anti-inflammatory drugs (NSAIDs) is common, with many people also using non-pharmacological measures and complementary therapies [2, 3]. However, previous research suggests that patients with arthritis commonly do not take NSAIDs as prescribed; taking pain medication in lower doses and less frequently than recommended [4]. Analgesics may be more effective for people with chronic pain when taken regularly and pre-emptively, thus decreasing the peak-and-trough effect of intermittent dosing. This, at least theoretically, enables more continuous pain relief [5] and reduces the likelihood of movement related pain. Research into arthritis patients' use of pain killers has traditionally focused on adherence to NSAIDs [6]. Methods of measuring compliance have included recording proportion of medication taken [7], assaying drug levels [8], self-reported adherence [9–11] and recording non-adherent behaviours such as missing doses [7]. In chronic conditions such as arthritis analgesics may be prescribed to be taken 'as needed' especially where the severity of pain fluctuates, so we have chosen to investigate the pattern of patients' use of pain killers rather than their actual adherence to prescribed regimens. In this study 'patterns' refers to the way in which people take analgesics on a daily basis.

Self-efficacy is defined as an individual's judgement of their capability to carry out a course of action required to achieve a desired goal [12] and in this study refers to a patient's perception of their ability to manage arthritis and its symptoms. Higher self-efficacy scores have been associated with better adherence to non-analgesic medication [13, 14], lower pain intensity [15, 16] and less functional impairment [17–19]. The relationship between self efficacy, pain and patterns of analgesic use is unclear. The aims of this study were to explore patterns of analgesic use among patients attending rheumatology clinics and the relationship between analgesic use, self-efficacy, pain and demographic characteristics.

In this study we have explored the way in which patients attending rheumatology clinics manage their analgesics, and the relationships between pain, self-efficacy and how regularly analgesics are used.

Previous studies indicate that patients with RA [26] and OA [4] may be taking analgesics less frequently than prescribed: a finding that was replicated in our study. We found, however, that people in more pain tended to take analgesics more regularly and pre-emptively (particularly before sleep). (The size of the difference in WOMAC pain scores between those agreeing and disagreeing with statements regarding regularity of use is similar to that obtained by analgesic medication in patients with arthritis [27]). However, patterns of pre-emptive analgesic use before exercise were more complex. Those with moderate pain more frequently took medication before exercise than those with slight pain, but those in severe pain were less likely to do so, perhaps because exercise capacity was limited in those with severe pain or our study may have been underpowered to detect a difference. By contrast a smaller but similar study of RA patients found no relationship between pain and regularity of analgesic use [28].

In our study, a majority of patients (18, 64%) agreed with the statement 'I always take my tablets everyday'. A large proportion of patients, however, varied their use of analgesics depending on their current level of pain; including over a third of people reporting severe pain, who only used their medication on demand, or at times when pain was unbearable. This approach is at odds with recommendations in guidelines for health professionals, in which pre-emptive and regular use of analgesics is recommended in order to prevent breakthrough and movement related pain [29]. Recent draft guidance from the National Institute for Health and Clinical Excellence recognised the surprising lack of data on how patients use analgesics in OA [28, 30]. Also surprisingly, publications for patients by respected arthritis organisations provide no guidance on how analgesics should be used [31], although it is relevant that disease knowledge is poorly related to patients' beliefs about medication [32]. Education and advice about analgesic use is likely to be provided by clinicians during normal clinical contacts. The clinic from which these patients were drawn also ran education classes for patients with newly diagnosed inflammatory arthritis. However, it is likely that few of the patients in this study had recently attended such classes: the median duration of arthritis in our population was 10 years.

Analgesic medication appears to be seen differently to other medications in that purposeful dose restraint is common and often seen as acceptable [4]. Patients may have many reasons for limiting their intake of analgesics in this way. A dislike of tablet taking despite, paradoxically, relatively common use of complementary medications has been previously found and is supported by our study [4]. Fears about dependency have also been well reported [4] and, in our experience, patients often express concerns about developing tolerance to analgesics. Use of regular medication is perceived by many people to indicate a loss of autonomy and to imply weakness [33]. Limiting analgesic consumption may, like the widespread use of complementary therapies, reflect an affirmation of autonomy or personal empowerment [34]. Beliefs and concerns about medicines are reported to influence adherence to medication. Indeed, medication beliefs have been found to have a greater influence on adherence in chronic illness than clinical and sociodemographic factors [35].

A variety of psychosocial factors are known to influence pain severity, which in turn may influence patterns of analgesic use. We found, like others [36], that patients from ethnic minorities reported more pain. However, only 35 (16%) patients from ethnic minorities were included in our population. This also prevented us from exploring the differences in pattern of analgesic use by severity of pain in each ethnicity, as the number of patients in each category was very small.

A measure of self efficacy was included in our analysis. This is a task specific trait [15] which seeks to ascertain how capable an individual believes they are of carrying out a course of action or behaviour needed to achieve a goal [12]; here, the management of their pain. The concept of self-efficacy is widely accepted and acknowledged to be important in the study of pain. We found, like others [37], that patients with higher perceived self efficacy reported less pain. Cross-sectional measurements of self-efficacy are believed to be a meaningful indicator of daily pain and coping, with greater self-efficacy often leading to more effective pain coping mechanisms [28]. Indeed, people with high levels of self efficacy (for pain) have reported less pain then those with low self-efficacy when subjected to painful stimuli in a laboratory [37]. Self efficacy was not, however, associated with regularity of analgesic use reported in our study, after the influence of pain was taken into account. No other measures were included in our study and whilst many have overlapping features, other psychological measures may have provided more insight into the reasons behind patterns of analgesic use. The low variance explained by the multivariate models for severity of pain (R²13.5%) and for regularity of analgesic use (R²3.7%) highlight the fact that that other psychological factors may be more influential. Beliefs about pain and analgesics [38, 39], coping mechanisms [28], depression [40, 41], anxiety [41] and psychological stress [42] may all be of importance in understanding patients' experience of arthritis, pain and their analgesic use.

Our study was cross-sectional and therefore has important limitations particularly because we do not know whether the patterns of analgesic use and perceived self efficacy, remain stable and whether the correlations reported hold true over time. To achieve a high response rate and generalisability we focused on the factors relating to pain that we were most interested in, such as self-efficacy. However we acknowledge that the investigation of other factors such as depression is important and may be worth further study in the future. We did not assess the psychometric properties of the patterns of analgesic use scale prior to its use in this study. It was not our intention to develop a new scale and to report the properties of this scale. We chose our questions to gain an insight into how patients managed their analgesics on a daily basis. A majority of questionnaires were self-completed (81%) but in 19% of cases (41 participants) a researcher recorded the answers. The most common reason for this was difficulty with writing due to arthritis; only 3 of the 41 participants who were assisted were from an ethnic minority. Considering that 16% of the total study population were from ethnic minorities, problems in participants' ability to write English were not a significant factor in this study and are unlikely to result in any bias. Important strengths of our study were an excellent response rate, use of well validated measures, a heterogeneous population, inclusion of consecutive patients and availability of a researcher to help with clarification when patients were completing questionnaires.

We found that self-efficacy, regularity of analgesic use and ethnicity influenced the degree of pain reported by patients. The majority of patients in this hospital outpatient sample (85%) reported moderate or more severe pain. People with greater pain reported more regular analgesic use but a large proportion of patients, including some in severe pain, varied their consumption of analgesics greatly. This may be because of patients' beliefs and concerns about medications. Health professionals who recognise such concerns during consultations may be better able to guide patients and encourage more optimal patterns of analgesic use and promote better function.