Prevalence, natural history and measurement of pain in stroke

Myzoon Ali Marian C. Brady Terence J. Quinn Gillian Mead

Post-stroke pain is common but poorly understood. Prevalence ranges from 10% 1 to 40%2-4 or higher 5. Aetiology includes central post-stroke pain, headaches, and musculoskeletal-issues from post-stroke impairments1. Post-stroke pain is associated with poor outcomes. Pain affects mobility,6 restricts activities of daily living (ADL)7, relates to poor rehabilitation8 and quality of life (QoL)9. Post-stroke pain assessment is complex. Pain may coexist with complications such as fatigue or depression.10 Stroke-related communication problems can hinder the ability to express pain1. Pain is under-diagnosed, inadequately managed10 and often undisclosed unless an active enquiry is made2. Up to 66% of those who identified central pain following stroke had inadequate intervention11. Management is important as improved pain control is associated with improved function and QoL12. To inform treatments for pain, we need a better understanding of assessment and epidemiology. Sample size calculations require robust estimates of post-stroke pain prevalence, patient characteristics associated with pain, and prognostic models of the natural history of pain progression. To guide outcome assessment selection, we need to compare differing methods of quantifying pain. The authors are part of an NIHR and British Association of Stroke Physicians writing group, who are developing a trial of a stroke-pain intervention.13 Findings from this work will inform the design and conduct of this subsequent trial in terms of sample size, trajectory of pain, preferred assessment and optimum time for intervention. Aims: to describe prevalence of, populations affected by, and measurement of post-stroke pain to inform development of a complex intervention trial. Research Questions: 1. How common is post-stroke pain? 2. How do pain scores change over time? 3. What are the characteristics of those who experience pain, versus those who do not? 4. What is the association between pain and every-day function? 5. What are the relationships amongst pain, mobility, fatigue and anxiety/depression?

We will analyse individual participant data (IPD) from clinical trials in the Virtual International Stroke Trials Archives (VISTA: www.virtualtrialsarchives.org). This archive collates and provides access to anonymised data from clinical trials for novel exploratory analyses. Data will be extracted on demography (age, sex, initial stroke severity, medical history, stroke lesion location), Barthel Index, and pain using all available data. Feasibility: Pain measurement on the European QoL Scale (EQ-5D) is available for at least 8000 participants at a single time point; repeated measurements are available for at least n=400. At least 600 IPD on pain are available at a single time-point from the 36 item Short Form survey (SF-36). Visual Analogue Scale (VAS) for pain and corresponding EQ-5D scores are available for at least 300 participants. Data on AE, including pain are routinely available for acute stroke trials within VISTA. Analyses: How common is post-stroke pain and how do scores change over time? We will describe prevalence post-stroke pain using EQ-5D, SF-36 and VAS across all time points, and where appropriate, by synthesising pain data from these assessments into a single outcome measure using an NIHR-funded pre-specified algorithm14 whereby the VAS and SF-36 scores will be transformed to fit the format and scoring of the EQ-5D whilst retaining the original distributions of each scale. We will describe the full range of the transformed ED-5D pain scores at single time-points, and progression of scores over repeated measures, where available, using frequency counts and percentages. Time will be presented as “days since stroke.” Characteristics of those who experience pain, versus those who do not We will describe demography including age, sex, medical history, aphasia (Best Language domain of the NIHSS), mobility (Mobility and Stairs domains of the Barthel Index) and anxiety/depression (EQ-5D) for participants with and without post-stroke pain, using summary statistics. We will examine differences using a Mann-Whitney test, and test for independent association using multivariable models with pain as an ‘outcome’. Association between pain and every-day function We will conduct indirect comparisons of pain scales by examining the strengths of association between each pain scale and the Barthel Index. Pain data from the EQ-5D, SF-36 and VAS will be included in separate regression models to examine the relationships between each assessment and the Barthel Index at follow up, adjusting for relevant covariates. Spearman correlation analyses (partial correlations, adjusting for age, and relevant covariates) will assess relationships between pain on each scale, and the Barthel Index. Relationships amongst pain, mobility, fatigue and anxiety/depression Exploratory analyses of the EQ-5D domains will generate kappa statistics to describe relationships between pain and the other domains (mobility, self-care, usual activities, anxiety/depression), and whether addition of other domain scores to the pain score results in better capture of pain. This will be tested by examining the relationship between combined EQ-5D domains and the VAS for pain using Spearman correlation analyses. We will also examine the relationships between SF-36 fatigue and mood questions (on their own and in combination) and pain, using Spearman correlation analyses.

1. Treister AK, Hatch MN, Cramer SC, & Chang EY. Demystifying post-stroke pain: from etiology to treatment. PM&R , 63-75. 2017. 2. Langhorne P et al. Medical complications after stroke : A multicentre study. Stroke 31[6], 1223-1229. 2000. 3. Ratnasabapathy Y et al. Shoulder pain in people with a stroke: a population-based study. Clin Rehab 17, 304-311. 2003. 4. Gamble GE et al. Poststroke shoulder pain: a prospective study of the association and risk. Eur J Pain 6, 467-474. 2002. 5. Klit H, Finnerup NB, Andersen G, & Jensen TS. Central poststroke pain: a population-based study. Pain 152, 818-824. 2011. 6. Hamzat TK & Osundiya OC. Musculoskeletal pain and its impact on motor performance among stroke survivors. Hong Kong Physiotherapy Journal 28, 11-15. 2010. 7. Lindgren I, Jonsson A-C, Norrving B, & Lindgren A. Shoulder pain after stroke. A prospective population-based study. Stroke 38, 343-348. 2007. 8. Roy CW, Sands MR, Hill LD, Harrison A, & Marshall S. The effect of shoulder pain on outcome of acute hemiplegia. Clinical Rehabilitation 9[21], 27. 1995. 9. Widar M, Ahlstrom G, & Ek AC. Health-related quality of life in persons with long-term pain after a stroke. J Clin Nursing 13[497], 505. 2004. 10. Appelros P. Prevalence and predictors of pain and fatigue after stroke: a population based study. Int J Rehabil Res 29, 329-333. 2006. 11. Widar M, Samuelsson L, Karlsson-Tivenius S, & Ahlstrom G. Long-term pain conditions after a stroke. J Rehabil Med 34, 165-170. 2002. 12. Katz N. The impact of pain management on quality of life. J Pain Symptom Manage 24[Suppl 1], S38-S47. 2002. 13. Ioannou A et al. Incidence, types and nature of post-stroke pain: systematic review of literature and meta-analysis. https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=114940 . 2019. 14. The RELEASE Collaboration. REhabilitation and recovery of peopLE with Aphasia after StrokE (RELEASE): A protocol for a systematic review-based Individual Participant Data (IPD) meta- and network meta-analysis. Aphasiology 34, 137-157. 2020.