Abstract
Objectives:
The association between the presence of pain and mental disorder is well known. The extent of pain treatment in psychiatric patients is estimated to be high, but there is a lack of recent studies focusing on analgesic treatment in patients with mental disorders. The use of analgesics can be associated with side effects, and it is possible that analgesics are not the correct treatment for chronic pain among patients with mental disorders.
Methods:
Data were obtained among inpatients in a geriatric psychiatry department at Aarhus University Hospital, Risskov, Denmark, between 1 April 2013 and 1 October 2013. The study examines the extent of analgesic use by patients at admission with a follow up at discharge to note any change during hospitalization.
Results:
A total of 89 patients aged 68 years or older were included (56 women, 33 men). At admission, 51.7% used analgesics, and this number did not change significantly from admission to discharge. A statistically significant increased risk of analgesic use was found in females (odds ratio 4.0). The indications for analgesic use were not present in 34.5% of the pain-treated patients at admission. At discharge, this number had been reduced to 23.1%. Paracetamol was the drug most frequently used, followed by opioids.
Conclusions:
The use of analgesics among aged psychiatric inpatients is high. An increased focus on this topic is recommended.
Introduction
In Denmark, the use of analgesics is increasing [Statens-Serum-Institut, 2014]. In 2012, a total of 130,770,000 defined daily doses (DDD) of paracetamol with the Anatomical Therapeutic Chemical Classification System number N02BE01 [Sketris et al. 2004; World Health Organization, 2014], were sold in Danish primary care [Statens-Serum-Institut, 2014]. In 2012, the total Danish population included 5,588,006 people [Danmarks Statistik, 2014]; therefore, every citizen consumed 1 DDD of paracetamol for 23 days per year. Paracetamol is categorized as the most frequently used analgesic in primary care in Denmark [Statens-Serum-Institut, 2014]. Statistics performed by The National Serum Institute for the period between 1999 and 2012 describe an increase in the sale of the three main drugs used as analgesics, including paracetamol, nonsteroidal anti-inflammatory drugs (NSAIDs) and opioids. The use of prescribed analgesics is particularly high among the aged population compared with younger age groups [Statens-Serum-Institut, 2014]. In 2013, in Danish primary care, 63 % of those identifiable purchasing paracetamol was made by individuals aged 65 or above [Statens-Serum-Institut, 2014].
Long-term use of analgesics is associated with a high number of side effects. Side effects such as gastrointestinal bleeding, renal failure and cardiovascular events have caused NSAIDs to be contraindicated in most cases when treating chronic pain in aged patients [Barkin et al. 2010; O’Neil et al. 2012; Rastogi and Meek, 2013]. It has been suggested that NSAIDs prevent the occurrence of Alzheimer’s disease but an exact relationship has not yet been established [Wang et al. 2015]. Paracetamol, in contrast to NSAIDs, is known to have a mild profile of adverse effects. Thus, when treating pain, paracetamol is commonly the first drug of choice [Fine, 2012]. Even though paracetamol is known as a mild-acting analgesic, there have been concerns raised about the liver toxicity of paracetamol, especially in frail, aged people [O’Neil et al. 2012]. In addition to the side effects of the treatment with analgesics, polypharmacy can be a severe problem when prescribing analgesics to aged people [Andrade et al. 2011; Rastogi and Meek, 2013].
Increasing age and the presence of a range of mental disorders are independent risk factors for developing chronic pain [Gureje, 2007; Tunks et al. 2008; Andrade et al. 2011]. In particular, the association between chronic pain and depression has been investigated in detail, and it is recognized that each predisposes to the other [Gureje, 2007; Tunks et al. 2008; Goesling et al. 2013; Chopra and Arora, 2014]. It is suggested that the association is due to neurophysiologic overlap [Chopra and Arora, 2014] and to the dysfunction of emotional regulation in patients with depression [Van Hecke et al. 2013]. To emphasize the significance of age and mental disorder when evaluating the risk of chronic pain, a study found that 1 in 4 people above the age of 75 years used analgesics and psychotropics concomitantly [Hartikainen et al. 2005].
Compared with nonpsychiatric patients, the treatment of pain among patients with a mental disorder can be complicated because of a decreased response to pain therapy [Gureje, 2007; Tunks et al. 2008]. In addition, the rehabilitation of patients with a mental disorder can be more difficult because of greater disability and decreased functional recovery related to pain complaints [Gureje, 2007; Tunks et al. 2008]. There is evidence that treating underlying depression among aged patients with pain complaints improves the functional outcome associated with those complaints, as well as the self-reported experience of pain and overall quality of life [Lin et al. 2003].
Bearing in mind that the use of analgesics among aged people can be associated with side effects, and that it is possible that analgesics are not the correct treatment for chronic pain among patients with mental disorders, it is important to reduce any needless use of analgesics in the treatment of aged psychiatric patients. Evidence is lacking of the extent of analgesic use in aged psychiatric patients. Reasons for patients needing analgesic treatment are also unknown. The literature suggests an increased correlation between pain and depression [Gureje, 2007; Tunks et al. 2008; Goesling et al. 2013; Chopra and Arora, 2014], but indication for analgesic use needs investigating. A study on the extent of analgesic use and its indication(s) is important if treatment for pain in aged psychiatric patients is to be improved in the future. It is clinically observed that a great number of inpatients in geriatric psychiatry departments are prescribed analgesics; however, the indications for the prescription are not always clear.
Aims of the study
To investigate the type and extent of analgesic use and the patients’ pain complaints in a geriatric psychiatry department at Aarhus University Hospital, Risskov, Denmark, and to analyse if the total use of analgesics changed during the patients’ hospitalization.
Methods
Design
A systematic screening of analgesic treatment at admission and discharge in patients at a psychiatric department was conducted over six months. The screening was performed by means of all available information: case records, patient interviews, information from relatives and caregivers, as well as from department staff.
Sample
Inclusion criteria
Patients admitted to the Department of Geriatric Psychiatry, Aarhus University Hospital, Risskov, Denmark, between 1 April 2013 and 1 October 2013, who at the time of admission were 68 years or older, were included. Younger patients diagnosed with dementia when leaving the hospital were also included because they were admitted to the department despite their age. The department has the responsibility for all adults concerning the diagnosing process of organic mental disorders, especially dementia diseases, from the hospital’s catchment area, mainly in its outpatient clinic. Some of the patients requiring an inpatient stay are included in this study. Further, it has the responsibility for inpatient treatment of individuals aged 68 years or above with any mental disorder. The catchment area of the hospital is the county of the city Aarhus, Denmark. Aarhus is a city with many educational institutions; therefore, the average age is the lowest for the cities of Denmark and the prevalence of aged people is therefore relatively low.
Exclusion criteria
Patients hospitalized for one day in connection with maintenance electroconvulsive treatment and patients admitted for compulsory treatment for a few hours were excluded as they can be classified as outpatients. If a patient was admitted to the department more than once during the investigation period, we recorded only the patient’s first hospitalization.
Measure
The data were recorded systematically via a questionnaire. At admission, the patient’s daily regular (taken daily at regular intervals) and irregular [only taken when needed (pro necessitate)] use of analgesics was investigated along with the indications for use, as well as the age and sex of the patient. At discharge, we followed up and again recorded current daily regular and irregular use of analgesics along with the patient’s current main International Classification of Diseases (ICD)-10 diagnosis [World Health Organization, 2010] and duration of hospitalization. At both admission and discharge, we queried whether the indications for the daily regular use of analgesics for a certain patient could be confirmed.
Procedure
Two nurses employed by the department were responsible for completing the questionnaire when a patient was admitted. These nurses where specifically trained in psychiatry and thoroughly instructed about the project. The information was initially obtained from the patient, and afterwards from relatives or caregivers if a patient was unable to answer the questions because of his or her psychiatric condition. The medication information provided from the questionnaire was checked in the patient’s electronic case record (ECR) with regard to type and doses. All types of analgesic medication prescribed were recorded. We used the ECR to confirm the indications for the prescription of analgesics. If the patient did not have a diagnosis compatible with daily pain complaints, or if the patient had a diagnosis compatible with pain complaints but in the questionnaire denied having pain complaints as a reason for analgesic treatment, the indication for the current daily regular use was grouped as nonconfirmed. All questions asked in the questionnaire are included in this study.
Irregular analgesic treatment at admission was recorded by asking the patient about the use of analgesics in the month leading up to the admission. Some patients were unable to answer, and some had not been asked because they had been discharged before being asked by the nurses. At admission, if the patient had a prescription of irregularly used analgesics, he or she and his or her relatives or caregiver were also asked if the prescribed analgesics were used. At discharge we used the ECR to investigate if irregularly prescribed analgesics had been used during hospitalization, because if any had, they needed to be recorded.
Classification
The medication was recorded in four groups according to the Anatomical Therapeutic Chemical Classification System [Sketris et al. 2004; World Health Organization, 2014], and we only recorded the drug if it had been prescribed as an analgesic. The four groups were: (1) N02B: other analgesics and antipyretics (the group includes paracetamol), (2) M01A: anti-inflammatory and antirheumatic products (nonsteroids), (3) N02A: opioids; (4) N03A: antiepileptics.
The doses of medication were recorded as DDDs [Sketris et al. 2004; World Health Organization, 2014]. In this way, it was possible to compare doses for two different medical drugs belonging to the same medicine group.
The pain complaints were grouped into four categories [Andrade et al. 2011]: (1) do not know why there is a prescription for analgesics; (2) diffuse pain complaints, including headache; (3) musculoskeletal system pain; and (4) neuropathic pain.
According to age, the patients were divided into two groups: (1) aged less than 75 years; (2) aged 75 years or above.
The main diagnosis was recorded according to the ICD-10 classification system: F00–F09: organic, including symptomatic, mental disorders; F10–F19: mental and behavioural disorders due to psychoactive substance use; F20–F29: schizophrenia, schizotypal and delusional disorders; F30–F39: mood disorders; F40–F49: neurotic, stress-related and somatoform disorders; F50–F59: behavioural syndromes associated with physiological disturbances and physical factors; F60–F69: disorders of adult personality and behaviour; F70–F79: mental retardation; F80–F89: disorders of psychological development; F90–F98: behavioural and emotional disorders with onset usually occurring in childhood and adolescence; F99: unspecified mental disorders [World Health Organization, 2010].
Statistics
Alpha level was determined to be 0.05 and a p value of less than or equal to 0.05 was significant. We performed multiple logistic regression analysis on three groups respectively: (1) regularly pain-treated patients at admission, (2) regularly pain-treated patients at admission with a valid indication for the treatment, (3) regularly or irregularly pain-treated patients at admission. Associations between use of analgesics at admission and the variables of sex, age, diagnoses and duration of hospital stay were investigated using multiple logistic regression analysis and some of the variables where controlled for in the final analysis. Logistic regression analysis was also used to investigate the association between the mentioned variables and the justification for use of analgesics at admission.
To decide whether diagnoses and duration of hospital stay should be included in the logistic regression analysis we performed a Chi-square test or a simple logistic regression analysis, and included the variables if they had a p value of less than 0.25. Age and sex were included regardless of p value, as previous studies have found correlations between these variables and the risk of chronic pain [Tunks et al. 2008; Kurita et al. 2012]. Furthermore, we tested the variables of diagnoses and duration of hospital stay by a likelihood ratio test (LRT) to make a final decision as to whether or not to include the variables in the final multiple logistic regression analysis. The given variables were tested one at a time and were interpreted as being significant for the model if the LRT was significant to a level of 0.05. The significant variables found by the LRT were included in the final multiple logistic regression analysis. The LRT was used to find the significant variables for the final multiple logistic regression analysis and was thereby correct for potential confounders.
To detect any sex difference in age among the admitted patients we performed a t test. The Wilcoxon Mann–Whitney U test was used to investigate sex difference in duration of hospital stay. The McNemar test was used to investigate changes in analgesic use during hospitalization.
Results
A total of 89 newly admitted patients were included during the investigation period (56 women, 33 men; 100% of the eligible patients). The average duration of hospital stay was 26 days (31 days for women, 19 days for men; p = 0.02). The average age of the study population was 76.9 ± 6.7 years with no significant difference between the sexes. The main ICD-10 diagnoses at discharge were distributed as follows: F0: organic, including symptomatic, mental disorders (n = 22), F2: schizophrenia, schizotypal and delusional disorders (n = 15), F3: mood disorders (n = 46) and the ‘rest group’ (n = 6).
Figure 1 presents the frequency of analgesic use at baseline and at follow up. At admission, 29 (32.6%) patients used analgesics daily at regular intervals. The number of patients with a use of analgesics at admission increased to 46 (51.7%) when including patients who used analgesics irregularly. The rest of the 89 patients did not have a use of analgesics. At follow up, the number of patients using analgesics regularly was reduced to 26 (29.2%, nonsignificant); and when including patients using analgesics irregularly, 45 (50.6%, nonsignificant) were still using analgesics.
The number of patients with analgesic use.
Table 1 includes regular use of analgesics; Table 2 shows validation of indications for analgesic treatment at admission; and Table 3 includes regular and irregular use of analgesics. The LRT suggested including the diagnoses given in Table 2 (p = 0.005), but the subgroups included too few patients to obtain reasonable estimates. The duration of hospital stay is included in Table 3 because of significance (p = 0.02) found by the LRT.
Distribution of regular analgesic treatment at admission.
OR, odds ratio; CI, confidence interval. OR 1.0 indicates the reference variable.
Distribution of patients with a valid indication for regular analgesic treatment at admission.
OR, odds ratio; CI, confidence interval. OR 1.0 indicates the reference variable.
Distribution of regular or irregular analgesic treatment at admission.
OR, odds ratio; CI, confidence interval. OR 1.0 indicates the reference variable.
The most frequent drugs among the 29 patients regularly treated with analgesics at admission were the N02B drugs [27 patients (93.1%) had a prescription]. In this study, all the medications included in the N02B group were identified as paracetamol. The next most frequently used analgesic was the N02A group, opioids (31%), followed by N03A group, antiepileptics (3.4%). No patient had a prescription of drugs from the M01A group (anti-inflammatory and antirheumatic products, and nonsteroids). It should be noted that some patients had a prescription of analgesics from more than one group.
The indications for the regularly prescribed analgesic treatment at admission were, in most cases: musculoskeletal pains (65.5%) followed by diffuse pain complaints (17.2%) and neuropathic pain complaints (3.4%). The remaining patients did not know why they had a prescription (10.3%), and one patient (3.4%) had not been interviewed about analgesic treatment. At admission, we could not confirm the indication of regular pain treatment in 34.5% of the regularly pain-treated patients. At discharge, this number was still high (23.1%). No difference with regard to confirmation was observed in patients using paracetamol or opioids.
As paracetamol and opioids were the most frequently used analgesics at admission, we further investigated the use of these drugs. The recommended maximum dose of paracetamol of 1.33 DDD [Dansk Lægemiddel Information A/S, 2014] was used in most cases in the regularly paracetamol-treated patients at admission (55.6%). No prescription exceeded this dose. The average dose of paracetamol regularly prescribed to a patient at admission was 1.1 DDD and was still the same at discharge. The number of patients treated with paracetamol was reduced from 27 of the admitted patients (30.3%) to 22 of the discharged patients (24.7%, nonsignificant). All patients regularly treated with opioids at admission were still using them at discharge, and the DDD remained unchanged. Eight patients had concomitant regular use of paracetamol and opioids at admission. This number had been reduced to six at discharge, which was nonsignificant.
Discussion
Our study is the first to investigate the use of analgesics among inpatients admitted to a geriatric psychiatry department. At admission, one third of the patients in our study regularly used analgesics daily, 93.1% of whom used paracetamol. When including the irregular use of analgesics, half of the admitted patients were treated for pain complaints. The most common pain complaint was related to the musculoskeletal system, followed by diffuse pain complaints.
Chronic pain is a common condition affecting approximately one fifth of the European population [Van Hecke et al. 2013]. Some factors associated with chronic pain are female sex, increasing age, and the presence of physical or mental comorbidity [Brnabic et al. 2012; Kurita et al. 2012; Van Hecke et al. 2013]. The temporal association between mental disorder and pain is found to be bidirectional [Hotopf et al. 1998]. In our study population, we found an increased incidence of using analgesics on a regular or irregular basis when being a woman [odds ratio (OR) 4.0]. The association between the use of analgesics and being a woman is four times greater than expected, if assumed that there was no association between the variables. The association is expected, as women have a greater sensitivity to pain than men and are at greater risk of suffering from a number of common pain conditions, possibly as a result of the differences in sex hormone levels [Bartley and Fillingim, 2013].
We did not find statistically significant correlations between the use of analgesics and the variables of age and psychiatric diagnoses. In accordance with a previous Danish study, we would expect increased use of analgesics in older patients compared with younger ones because of an increased prevalence of chronic pain [Kurita et al. 2012]. We would also expect a difference in the distribution of analgesic treatment compared with diagnosis, as some mental disorders such as depression and anxiety are more frequently described in correlation with chronic pain than other mental disorders [Gureje, 2007; Twillman, 2007; Barš et al. 2010; Chopra and Arora, 2014]. The nonsignificant results in our study concerning age might be a consequence of small sample size. Tables 1 and 3 show a tendency towards increased analgesic use in older patients [OR 1.2 and 1.5, respectively (nonsignificant)]. A larger study population would also allow us to divide diagnoses into smaller groups which, in future studies, could possibly explain a connection between the extent of analgesic use and specific diagnoses.
Women showed a tendency (OR 2.0) towards having the indication for regular analgesic treatment confirmed more often than men, whereas the oldest people (OR 0.3) lacked confirmation of indication more frequently than younger people. The LRT found a statistically significant difference in the ability to confirm the indication of the analgesic treatment compared with the diagnoses. The small number of patients according to the subgroups did not allow us to make reasonable estimates, and we therefore excluded diagnoses from the logistic regression. However, the finding from the LRT emphasizes the previously mentioned need of future studies concentrating on the correlation between analgesics and diagnoses.
Confirmed indications for analgesic use was absent in 34.5% of the patients treated for pain at admission. At discharge, this number was reduced to 23.1%. We reviewed the patients’ ECRs with regard to current diagnoses compatible with pain complaints in order to identify whether the indication of analgesic treatment could be confirmed. It is possible that a patient could have had a diagnosis compatible with pain complaints without having it recorded in the ECR. In such a case, the analgesic treatment would be sufficiently indicated, even though the recording of indication has failed and the patient would be included in the nonconfirmed group.
The analgesic most frequently used regularly among the study population was paracetamol, followed by opioids. NSAIDs were only prescribed as analgesics to be taken when needed. This is in line with the recommendations for pain treatment in the aged [Barkin et al. 2010; O’Neil et al. 2012; Rastogi and Meek, 2013]. No similar cross-sectional studies on the extent of analgesic use in the aged Danish population have been made. Therefore, we were unable to compare the amount of analgesics used in the study group with that of the general population.
A total of 10% of the admitted patients in our study regularly used opioids, which exceeds the 4.5% found in a Danish cross-sectional study on the adult population aged 16 years or above [Kurita et al. 2012]. Bearing in mind that several studies have shown an association between use of opioids, age and mental disorders [Hartikainen et al. 2005; Sullivan et al. 2006; Parsells Kelly et al. 2008], this finding was to be expected.
More than half of the patients in our study with a regular prescription for paracetamol used the recommended maximum dose of 1.33 DDD each day [Dansk Lægemiddel Information A/S, 2014]. There is ongoing research investigating whether doses of paracetamol in aged people should be reduced, as aged people are at increased risk of hepatotoxicity because of age-dependent changes in metabolism [Twycross et al. 2013].
The most common pain complaint at admission in patients treated regularly for pain was related to the musculoskeletal system. Data obtained from two nationwide Danish cross-sectional studies carried out in 2005 and 2010, respectively, also described musculoskeletal pain as the most common type of pain among the adult Danish population [Sjogren et al. 2009; Kurita et al. 2012]. More than a quarter of patients in our study population received daily pain treatment for diffuse pain complaints or without knowing why they had been prescribed analgesics. Fibromyalgia is a common, underdiagnosed pain syndrome that causes diffuse pain complaints [Rahman et al. 2014]. The syndrome is associated with mental disorders like depression and anxiety [Barš et al. 2010]. Fibromyalgia could be of interest in future clinical work, in exploring if some pain-treated psychiatric patients, beside their mental disorder, meet the criteria for a diagnosis of fibromyalgia.
The number of patients requiring analgesic treatment did not change significantly between admission and discharge. Some of the patients changed from regular use of analgesics to irregular use. Prior to the study, we expected a decrease in the use of analgesics when treating the mental disorder, by eliminating mental disorder as a reason for analgesic use. When treating a patient’s depression, we expected the increased requirement for analgesics described in patients with pain and concomitant depression [Chopra and Arora, 2014] to be eliminated. Some antidepressants commonly used are recorded as both analgesics and antidepressants [Goesling et al. 2013; Chopra and Arora, 2014], and this would, per se, indicate a decrease in the pain complaints. The doses used when treating the individual indications of pain and depression differ, and the antidepressants are not used for all kinds of pain [Goesling et al. 2013], which, to some extent, can explain unchanged analgesic use during hospitalization. The unchanged use of analgesics may also be because some patients did not recover from the mental disorder at discharge, or they had a mental disorder that was not correlated with pain. At discharge, some of the patients in the study still had a somatic diagnosis compatible with pain. However, the lack of confirmed indication for analgesic treatment in 34.5% of admitted patients and in 23.1% of discharged patients could indicate a lack of evaluation of the analgesic-prescription effect.
The reliability of the collected data was considered adequate because we used the ECRs and statements from the patients themselves and from relatives or caregivers in order to investigate the use of analgesics. To be sure to record as much irregular analgesic use as possible, we asked patients about their use of analgesics not described in the ECR for the one month prior to admission. Some patients were unable to answer and therefore we may be underestimating irregular analgesic use in the run up to admission.
All patients meeting the inclusion criteria were included; that is, the composition of the study group is not biased. Data from 89 patients were included during the investigation period. In future studies on the correlation between mental disorders and the use of analgesics, a greater number of included patients would be preferable as it would increase the possibility of supporting the suggested correlation between analgesic use and specific diagnoses. Our study supports previous findings that mental disorders and chronic pain are correlated [Hotopf et al. 1998; Gureje, 2007; Tunks et al. 2008; Goesling et al. 2013; Chopra and Arora, 2014]. In future studies, it would be of interest to perform a similar study including an additional comparison with a control group of people without mental disorders.
Strengths and limitations
Strengths
All patients meeting the inclusion criteria during the 6 months of data collection were screened, and we reached 100% inclusion. Data were collected from all possible sources. In contrast to many other studies, this study examined the regular use of analgesics alongside irregular use.
Limitations
Despite a 100% inclusion rate, the sample remains statistically small. As all patients included in our study derive from the same department and the severity of their psychiatric conditions is not known, the sample is of limited representation of the overall group of geriatric psychiatric patients, and it can be difficult to generalize the findings to other departments.
Conclusion
We have found that the prevalence of analgesic use among aged psychiatric inpatients was high. Of specific interest is the large amount of regularly pain-treated patients without a confirmed indication to be found in the ECR, or where the patient did not know why there was a prescription. We found a significantly increased incidence of the use of analgesics at admission among women. The potential of confirming analgesic use is probably associated with different diagnoses, but further studies including more patients are needed.
The distribution of types of regularly prescribed analgesics was as expected. Paracetamol was used to the greatest extent, followed by opioids, while NSAIDs were only prescribed as analgesics to be taken when needed. The indication for analgesic treatment was most frequently musculoskeletal pain, followed by diffuse pain complaints. No major changes in the use of analgesics were observed between admission and discharge. We suggest that reviewing analgesic treatment in patients with a nonconfirmed indication will make it possible to record some of the nonconfirmed indications as confirmed or reduce the use of analgesics in the group studied.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement
The authors declare that there is no conflict of interest.