If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence to Torstein F. Habiger, MD, Department of Global Public Health and Primary Care, Centre for Elderly and Nursing Home Medicine, University of Bergen, Alrek helseklynge, Årstadveien 17, Block D, 5018 Bergen, Norway.
Department of Global Public Health and Primary Care, Centre for Elderly and Nursing Home Medicine, University of Bergen, Bergen, NorwayDepartment of Public Health and Primary Care, Centre for Elderly and Nursing Home Medicine, Leiden University Medical Centre, Netherlands
Department of Global Public Health and Primary Care, Centre for Elderly and Nursing Home Medicine, University of Bergen, Bergen, NorwayDepartment of Clinical Psychology, Faculty of Psychology, University of Bergen, Bergen, Norway
Department of Global Public Health and Primary Care, Centre for Elderly and Nursing Home Medicine, University of Bergen, Bergen, NorwayMunicipality of Bergen, Bergen, Norway
In nursing homes (NHs), 30% to 60% of patients experience daily pain and >80% have dementia. This can lead to neuropsychiatric symptoms, including psychosis symptoms such as delusion. We investigated if there was a relationship between pain and psychosis symptoms over time. We also aimed to investigate the effect of a multicomponent intervention (COSMOS) on pain, psychosis symptoms, and analgesic prescription.
Design
COSMOS is a cluster-randomized, single blinded, controlled trial. Each NH unit was defined as a cluster and randomized to either the COSMOS intervention or care as usual. The COSMOS intervention is a multicomponent intervention, consisting of staff training in communication, pain treatment, medication review, organization of activities, and safety. The intervention lasted for 4 months with a follow-up at month 9.
Setting and Participants
Sixty-seven units from 33 Norwegian NHs in 8 municipalities. The study included 723 patients aged ≥65 years, residing at the NH ≥2 weeks before inclusion. Patients with a life expectancy <6 months were excluded.
Measures
Pain was measured using the Mobilization-Observation-Behavior-Intensity-Dementia Pain Scale. Psychosis symptoms were measured using the Neuropsychiatric Inventory–NH version. Measurements were performed at baseline, and months 4 and 9.
Results
Multilevel Mixed-Effect statistical analysis found that psychosis symptoms as a group (odds ratio [OR] 2.03, P = .009), and delusion (OR 2.12, P = .007) were associated with pain over time. No significant intervention effect on psychosis symptoms was observed. Compared with the control group, people with dementia in the intervention group experienced less musculoskeletal pain (β: −0.47, P = .047). Analgesic prescription was not affected by the intervention.
Conclusion and Implications
Pain is associated with psychosis symptoms, and pain assessment should be done when making treatment decisions on psychosis symptoms in NH patients. The COSMOS intervention improved musculoskeletal pain in people with dementia, but not psychosis symptoms, and there is need for further studies on treatment of psychosis symptoms in NH patients.
People with dementia are at risk of having untreated pain due to difficulties in reporting their own pain location and pain intensity, and this can in turn lead to reduced quality of life (QoL) and increased neuropsychiatric symptoms (NPS).
NPS can be grouped together in clusters according to coexistent symptoms, such as agitation, mood, and psychosis, the latter consisting of delusion and hallucination.
Neuropsychiatric symptom clusters of Alzheimer's disease in Hong Kong Chinese: Prevalence and confirmatory factor analysis of the Neuropsychiatric Inventory.
Stability of the factor structure of the Neuropsychiatric Inventory in a 31-month follow-up study of a large sample of nursing-home patients with dementia.
Resourse use and disease course in dementia - nursing Home (REDIC-NH), a longitudinal cohort study; design and patient characteristics at admission to Norwegian nursing homes.
Further, polypharmacy is common in an NH population, and studies have shown that regular medication reviews are necessary to decrease the risk of unnecessary drug prescriptions, as well as unwanted side effects.
Intervention with the screening tool of older persons potentially inappropriate prescriptions/screening tool to alert doctors to right treatment criteria in elderly residents of a chronic geriatric facility: A randomized clinical trial.
Guidelines on the treatment of psychosis symptoms recommend nonpharmacological measures as the first-line approach and highlight the importance of treating possible underlying causes.
Previous studies have found that other NPSs, such as agitation, can benefit from nonpharmacological measures; however, the effect on psychosis symptoms is uncertain.
The value of personalized psychosocial interventions to address behavioral and psychological symptoms in people with dementia living in care home settings: A systematic review.
The negative impact of psychotropic drug use on quality of life in nursing home patients at different stages of dementia: Cross-sectional analyses from the COSMOS Trial.
investigated the effect of systematic pain treatment on agitation in people with dementia, where agitation was reduced in response to pain treatment. Secondary analyses from the same study also show positive effects on mood symptoms as well as psychosis symptoms.
The COSMOS trial was designed to improve the QoL of NH patients through better COmmunication, Systematic assessment and treatment of pain, Medication review, Organization of activities and Safety, thus the acronym COSMOS.
As the intervention included elements that previous studies have found to improve both pain and psychosis symptoms, we aimed to analyze whether the multicomponent intervention could improve pain and psychosis symptoms in people with and without dementia. We hypothesized that the intervention would have a positive effect on both pain and psychosis symptoms. We also wanted to determine if the use of analgesics changed in response to the intervention, as well as the characteristics, such as QoL, of patients with pain using analgesics.
Further, baseline data from the COSMOS study found an association between pain and psychosis symptoms
; we aimed to investigate if this association persisted over time by analyzing the control group patients who received their usual care.
Methods
This study was based on secondary analyses from the COSMOS trial. The study was a multicenter cluster-randomized, single blinded controlled trial performed from 2014 to 2015, aimed at improving patients’ QoL through the implementation of a multicomponent intervention. The study enrolled 723 patients from 33 NHs and 67 different NH units in Norway. The entire study protocol and a description of the COSMOS intervention have previously been published in full elsewhere,
Patients ≥65 years who had stayed at the NH for at least 2 weeks were included. Patients with a <6 months’ life expectancy were excluded from the study.
Randomization and Intervention
Each NH unit was defined as a cluster and randomized to receive either the COSMOS intervention or care as usual. The COSMOS intervention components were based on current state-of-the-art evidence,
The value of personalized psychosocial interventions to address behavioral and psychological symptoms in people with dementia living in care home settings: A systematic review.
and was implemented through a 2-day education seminar for NH staff, as well as a medication review for all units during the intervention period. All NHs participated with at least 2 staff members who were put in charge of implementing the COSMOS intervention at their respective NH units. The intervention period lasted for 4 months, with a follow-up at month 9. Data collection was performed at baseline, month 4, and month 9. All assessments were performed by NH staff who knew the patient well.
Outcome Measures
Pain was assessed using the Mobilization-Observation-Behavior-Intensity-Dementia-2 (MOBID-2) Pain Scale. The scale has been thoroughly tested for validity, reliability, and responsiveness.
Mobilization-Observation-Behavior-Intensity-Dementia Pain Scale (MOBID): Development and validation of a nurse-administered pain assessment tool for use in dementia.
MOBID-2 consists of 2 parts, where part 1 assesses musculoskeletal pain through 5 actively guided movements during which the raters are encouraged to look for pain behavior. Part 2 consist of 5 items and assesses pain coming from head, skin, and internal organs. For each item, raters assess the patients' pain on a Numerical Rating Scale (NRS) from 0 to 10, where 0 represents no pain and 10 represents the worst pain possible. Finally, raters take all assessments into account and rate the patient's total pain score on an NRS from 0 to 10. A total pain score ≥3 is viewed as clinically significant pain.
Psychosis symptoms were measured using the Neuropsychiatric Inventory–Nursing Home Version (NPI-NH).
The NPI-NH measures the frequency and severity of 12 different NPSs (eg, agitation, delusion, and depression in the last week before assessment). Frequency (F) is measured on a scale from 0 to 4, where 0 represents not present, and 4 represents present daily. Severity (S) is measured on a scale from 1 to 3, where 1 represents mild symptom severity and 3 represents a severe symptom with high stress on the patient. The scores for frequency and severity are multiplied to generate a score for each symptom ranging from 0 to 12. A score ≥4 is considered a clinically significant symptom.
Previous factor analyses of NPS have found different symptom clusters, among others the psychosis symptom cluster, which consists of delusion and hallucinations.
Neuropsychiatric symptom clusters of Alzheimer's disease in Hong Kong Chinese: Prevalence and confirmatory factor analysis of the Neuropsychiatric Inventory.
Stability of the factor structure of the Neuropsychiatric Inventory in a 31-month follow-up study of a large sample of nursing-home patients with dementia.
Information concerning medication and diagnoses were obtained from the patients’ medical records. Analgesics were defined as the group N02 in the Anatomical Therapeutic Chemical (ATC) classification system, which was further subdivided into opioid analgesics (N02A) and nonopioid analgesics (N02B and N02C). In addition, cognitive function was assessed using the Mini Mental State Examination (MMSE).
Analyses were performed by TFH in collaboration with a statistician (JM). The intervention effect on pain and psychosis were analyzed using Multilevel Mixed-Effect Linear Regression, with random intercept for clusters and time as a categorical variable. The association between pain and psychosis over time in the control group was investigated using Multilevel Mixed-Effect Logistic regression with maximum likelihood estimation and random intercept for clusters. A clinically significant symptom of psychosis, representing the presence of 1 or more symptoms of psychosis, was defined as the dependent variable, and a clinically significant MOBID-2 score (≥3) was established as an independent variable. The same analysis was conducted for the individual symptoms of psychosis. Associations were adjusted for the effect of time, defined as a categorical variable, age, dementia severity, and use of opioids. Model fit was evaluated using Akaikes Information criterion.
Information about the COSMOS study and its implications was provided for all patients. Consent was obtained in written and verbal form from patients with the cognitive ability to provide it. For patients lacking this ability, presumed consent was obtained from the patients’ next of kin or legal guardian after explaining the study procedure. The trial was approved by the Regional Committee for Medical and Health Research Ethics, West Norway (REK 2013/1765) and registered at clinicaltrials.gov (NCT02238652).
Results
A total of 723 patients were enrolled; 178 were excluded, leaving 545 patients to be included in the study. A total of 297 patients were randomized to the intervention group, with 248 allocated to the control group. A total of 73.8% of patients were women, with an average age of 86.7 years (Table 1).
There was no significant intervention effect on the total score of the MOBID-2 Pain Scale from baseline to month 9 (β −0.23; 95% confidence interval [CI] −0.88 to 0.42; P = .49) (Table 2). A significant positive intervention effect was found for MOBID-2 part 1 for people with dementia from baseline to month 9 (β −0.45; 95% CI –0.90 to −0.01; P = .047), but not for MOBID-2 part 2 (Figure 1). The number of patients using opioids increased nonsignificantly from baseline to month 9 in both groups, from 31.5% to 38.0% in the control group (odds ratio [OR] 1.31; P = .20) and from 30% to 35% in the intervention group (OR 1.26; P = .27). There was no significant intervention effect on the use of opioid analgesics (OR 0.95; 95% CI 0.53–1.70; P = .86).
Table 2Effect of Intervention Compared With Control on Pain and Psychosis Symptoms
Fig. 1Progression of pain scores (MOBID-2) in people with dementia. MOBID-2 part 1: Musculoskeletal pain (Significant invention effect), MOBID-2 part 2: Pain from internal organs, head, and skin.
The number of people with pain in the intervention group who did not use analgesics decreased from 24.3% to 19.4% from baseline to month 4, and remained stable at month 9 (Table 3). The number of people in the control group with a MOBID-2 score ≥3 who did not use analgesics steadily increased from baseline (19.1%) to month 9 (26.3%) (Table 3). The difference between the control and intervention groups, and changes within groups were not significant at either time point (Table 3). Patients in the intervention group with a MOBID-2 score ≥3 who used analgesics experienced more NPS (F 5.7; P = .001) compared with patients with a MOBID-2 score <3 at baseline and month 4 (Table 4). In the control group, the same was found at month 9 (F: 4.5; P = .005), but no such difference was observed at baseline and month 4 (Table 4). Patients with a MOBID-2 score ≥3 using analgesics had lower QoL than other patients at all time-points in both the control and intervention groups (Table 4).
Table 3Use of Analgesics in People With and Without Clinically Significant Pain (MOBID-2 ≥3)
Baseline (χ2 Control vs. Intervention: χ2 = 4.08, P = .253)
Control (n = 202)
Intervention (n = 251)
MOBID-2 ≥3
MOBID2 <3
MOBID-2 ≥3
MOBID2 <3
Uses analgesics (% of pain group)
76 (80.9)
57 (52.8)
78 (75.7)
65 (43.9)
No analgesics (% of pain group)
18 (19.1)
51 (47.2)
25 (24.3)
83 (56.1)
Month 4 (χ2 Control vs. Intervention: χ2 = 5.19, P = .158)
Control (n = 185)
Intervention (n = 208)
MOBID-2 ≥3
MOBID2 <3
MOBID-2 ≥3
MOBID2 <3
Uses analgesics (% of pain group)
64 (77.1)
51 (50.0)
58 (80.6)
61 (44.9)
No analgesics (% of pain group)
19 (22.9)
51 (50.0)
14 (19.4)
75 (55.1)
Month 9 (χ2 Control vs. Intervention: χ2 = 4.20, P = .241)
Pain and psychosis symptoms as a group (OR 2.03; 95% CI 1.19–3.45; P = .009), and delusion individually (OR 2.12; 95% CI 1.23–3.63; P = .007) were significantly associated over time. There was no significant association between pain and hallucinations (OR 1.47; 95% CI 0.66–3.29; P = .35). Patients who used antipsychotic medication were more likely to experience pain than patients not using antipsychotic medication (OR 1.78; 95% CI 1.02–3.10; P = .043). There was no significant intervention effect on psychosis symptoms – either as a cluster (β 0.23; 95% CI −0.92 to 1.37; P = .70) or for the individual symptoms delusion and hallucinations (Table 2).
Discussion
Pain was significantly associated with psychosis symptoms, and delusion over time, but not with hallucinations. This is important for clinicians, as it suggests that a thorough pain assessment is essential before making treatment decisions concerning psychosis symptoms. This is to our knowledge the first study to investigate the relationship between pain and psychosis symptoms over time. The COSMOS intervention had a positive effect on musculoskeletal pain in people with dementia, highlighting the importance of a thorough pain assessment and treatment strategy in NHs.
The total MOBID-2 pain score was not reduced in response to the COSMOS intervention. Musculoskeletal pain was, however, reduced in people with dementia. The reason that musculoskeletal pain was reduced in people with dementia in the intervention group compared with the control group, and not for the total population, can be explained by the ability of the MOBID-2 pain scale to detect pain in people with dementia. Patients without cognitive impairment are able to report their own pain and the effect of pain treatment, or lack thereof, which assists the physician's decision making.
The assessment of musculoskeletal pain can also be more straightforward compared with the assessment of pain from the internal organs, head, and skin, as musculoskeletal pain can be provoked by active movements, as in part 1 of the MOBID-2 Pain Scale. Our findings are partly in line with a previous study by Sandvik and colleagues,
who found that a Stepwise Protocol for Treating pain reduced pain in NH patients with dementia and behavioral disturbances. However, in contrast to our study, this study focused primarily on the treatment of pain and found an intervention effect on the total pain score, not only on musculoskeletal pain.
In the COSMOS study, no significant intervention effect on pain was found, although the number of patients with pain increased over time in the control group and decreased in the intervention group. This may suggest that the intervention group could have benefited from the COSMOS intervention to a certain degree.
The use of analgesics was high in both groups, especially the number of patients using opioids on a regular basis, which increased from 30% to >35% from baseline to month 9 in both the control group and intervention group. A previous study found a rise in the use of opioids in Norwegian NHs from 2000 to 2011, from 1.9% to 17.9%.
Our findings suggest that this trend has continued, which is particularly worrisome considering the possible side effects from long-term opioid use and risk of polypharmacy in people with dementia.
Pain, complex chronic conditions and potential inappropriate medication in people with dementia. Lessons learnt for pain treatment plans utilizing data from the Veteran Health Administration.
Tolerability of buprenorphine transdermal system in nursing home patients with advanced dementia: A randomized, placebo-controlled trial (DEP.PAIN.DEM).
investigating the effect of analgesic treatment on depression in NH patients with dementia found that patients being prescribed a buprenorphine transdermal patch had a significantly higher chance of dropping out of the study due to adverse events. This highlights the importance of being thorough in evaluating the risk of possible side effects in patients, before and during the prescription of an opioid analgesic.
When investigating the characteristics of patients with and without pain using and not using analgesics, we observed that patients with pain who used analgesics had lower QoL and more NPS than other patients. This supports previous studies, which have found pain to be associated with NPS and poor QoL.
However, it was unexpected that no significant differences were found between patients' QoL scores concerning those with pain using analgesics and those with pain not using analgesics, as the aim with analgesic prescription is to reduce the patient's pain and improve their QoL. In future studies, this could be an important focus point when investigating the effect of long-term analgesic use in NH patients. If long-term use of opioid analgesics causes more harm through unwanted side effects than benefit, more focus on frequent reevaluation of long-term use of opioid analgesics is needed.
As hypothesized, psychosis symptoms and pain were associated over time, as was the individual symptom delusion, but not hallucinations. This is in line with previous cross-sectional findings from the baseline data of the COSMOS study.
who found a relationship between pain and delusion in 2822 NH patients with dementia from 8 countries. As this is the first study that investigates the longitudinal relationship between pain and psychosis symptoms, our finding adds important knowledge regarding psychosis symptoms in NH patients. If psychosis symptoms, and especially delusion, are associated with pain, then a thorough assessment of pain should be a prerequisite when deciding on treatment options for psychosis symptoms, and aid in reducing the use of psychotropic medication to those who benefit the most from them. This is further highlighted by our finding that use of antipsychotic medication was associated with pain. If psychosis symptoms are triggered by underlying pain, then treatment of the underlying factor would be preferred rather than treating only the overt symptoms.
In contrast to our hypothesis, no significant intervention effect on psychosis symptoms was found. This diverges from a previous study from 2016, where systematic pain assessment and treatment in 352 patients with dementia and behavioral disturbances reduced psychosis symptoms.
limiting the potential to discover an intervention effect. The 2 studies also differ regarding the type of intervention. Where the 2016 study only focused on systematic pain assessment and treatment, this was only 1 part of the COSMOS trial, which includes components with the ability to reduce psychosis symptoms, such as organization of activities and medication review. Despite this, psychosis symptoms were not reduced, which underlines the complexity in treating psychosis symptoms in NH patients and that there is no one-size-fits-all treatment. Guidelines state that nonpharmacological options should be the first-line treatment, but knowledge concerning the effect of such treatments is sparse and our study is one of few that investigates the effect of nonpharmacological options.
Strengths and Limitations
This is one of the largest multicomponent intervention studies performed in a NH setting, and it includes a broad NH population both with and without dementia. This increases the generalizability of our findings. There is also a strength in using the MOBID-2 Pain Scale, which has been thoroughly tested for reliability and responsiveness for change. A limitation is that the study was powered with respect to QoL and not for pain and psychosis symptoms. It is a limitation that we only had data on type of pharmacological pain treatment, not dosage, or if nonpharmacological measures had been taken. There was also no assessment concerning the type of psychosis, or if the psychosis symptoms were chronic or acute in nature. A limitation also lies in the lack of knowledge regarding the duration of current pain and pain treatment at baseline, which is important to consider when interpreting the results.
Conclusions and Implications
Pain, psychosis symptoms as a group, and delusion were significantly associated over time, highlighting the importance for clinicians to assess pain when making treatment decisions on psychosis symptoms. The COSMOS intervention had no significant effect on psychosis symptoms. The COSMOS intervention had a significant effect on musculoskeletal pain in patients with dementia, but not on the total pain score, which shows the need for systematic pain assessment and treatment in patients with dementia. The use of opioid analgesics increased in both groups and was not affected by the COSMOS intervention, which shows the importance of frequent reassessment of opioid prescriptions.
Acknowledgments
Tony Elvegaard, Christian Gulla, and Irene Aasmul took part in the data collection. TFH thanks the Medical Students Research Program and the University of Bergen for all their support. We also thank the Norwegian Ministry of Health and Care services, as well as the GC Rieber foundation for their financial support to the Center for Elderly and Nursing Home Medicine.
References
Selbaek G.
Kirkevold O.
Engedal K.
The prevalence of psychiatric symptoms and behavioural disturbances and the use of psychotropic drugs in Norwegian nursing homes.
Neuropsychiatric symptom clusters of Alzheimer's disease in Hong Kong Chinese: Prevalence and confirmatory factor analysis of the Neuropsychiatric Inventory.
Stability of the factor structure of the Neuropsychiatric Inventory in a 31-month follow-up study of a large sample of nursing-home patients with dementia.
Resourse use and disease course in dementia - nursing Home (REDIC-NH), a longitudinal cohort study; design and patient characteristics at admission to Norwegian nursing homes.
Intervention with the screening tool of older persons potentially inappropriate prescriptions/screening tool to alert doctors to right treatment criteria in elderly residents of a chronic geriatric facility: A randomized clinical trial.
The value of personalized psychosocial interventions to address behavioral and psychological symptoms in people with dementia living in care home settings: A systematic review.
The negative impact of psychotropic drug use on quality of life in nursing home patients at different stages of dementia: Cross-sectional analyses from the COSMOS Trial.
Mobilization-Observation-Behavior-Intensity-Dementia Pain Scale (MOBID): Development and validation of a nurse-administered pain assessment tool for use in dementia.
Pain, complex chronic conditions and potential inappropriate medication in people with dementia. Lessons learnt for pain treatment plans utilizing data from the Veteran Health Administration.
Tolerability of buprenorphine transdermal system in nursing home patients with advanced dementia: A randomized, placebo-controlled trial (DEP.PAIN.DEM).
The University of Bergen, the Norwegian Research Council (Protocol code: 222113) and Rebecca Ege Hegermanns Foundation financed the study. The funders had no role in study-design, data collection, data analyses, data interpretation, or writing of the article itself.
00472-2&id=gr1.jpg){kind=link}