Skip to main content

Research Repository

Advanced Search

Musculoskeletal pain and its impact on rates of hospitalisation and mortality in cancer: a linked electronic health record cohort study (MSKCOM)

Mason, Kayleigh; Bailey, James; Heron, Neil; Marshall, Michelle; Achana, Felix; Chen, Ying; Edwards, John J; Frisher, Frisher; Huntely, Alyson L; Mallen, Christian; Mamas, Mamas; Ee Png, May; Tatton, Stephen; White, Simon; Jordan, Kelvin


Felix Achana

Ying Chen

John J Edwards

Alyson L Huntely

May Ee Png

Stephen Tatton


Background: Patients with painful musculoskeletal conditions have an increased risk of developing cancer, but less is known about whether this impacts on the longer-term cancer prognosis. This study aims to determine whether patients with painful musculoskeletal conditions have a greater incidence of hospitalisation and mortality following diagnosis of incident primary cancer. The Approach: Data were obtained from a national primary care records database (Clinical Practice Research Datalink; CPRD Aurum) with linkage to hospitalisation and mortality records. Patients aged >45 years with incident primary breast, colorectal, lung or prostate cancer recorded in primary care were included. Patients were stratified by consultations in primary care for painful musculoskeletal conditions (defined as regional pain, osteoarthritis and inflammatory arthritis) in the 24 months prior to cancer, and also stratified by severity. Severe musculoskeletal pain was defined as prescription of strong/very strong opioids or secondary care referral in the 6 months before cancer. Time-to-event outcomes included hospital admission and all-cause mortality. Incidence rates per 1,000 person-years (IR/1000py) were calculated and multivariable flexible parametric models estimated the adjusted hazard ratio (aHR). Covariates included socio-demographics and comorbidities. Findings: There were 428,866 patients with an incident primary cancer diagnosis (140,335 [33%] breast; 85,580 [20%] colorectal; 84,802 [20%] lung; 118,414 [27%] prostate) with median age 70 years and 50% females in the total cancer cohort. 28% of patients consulted for musculoskeletal pain in the 2 years prior to cancer diagnosis with 17% recorded with regional pain, 7% osteoarthritis and 4% inflammatory arthritis when stratified by condition, and 6% categorised as severe when stratified by severity. Patients with inflammatory arthritis and severe pain were prescribed strong/very strong opioids in the 2 years prior to cancer more commonly than those without pain (46% and 43% vs 19%, respectively). Patients with severe musculoskeletal pain had the highest risk for hospitalisation (IR/1000py 640 vs no pain 419; aHR 1.04, 95% CI 1.02-1.06) and all-cause mortality (IR/1000py 216 vs no pain 145; aHR 1.19, 1.15-1.22). Patients with musculoskeletal conditions had an increased incidence of hospitalisation (IR/1000py regional 477, osteoarthritis 543 and inflammatory 512 vs no pain 419) and mortality (IR/1000py regional 160, osteoarthritis 164 and inflammatory 186 vs no pain 145) compared to those without pain, but after adjustment for socio-demographics and comorbidity there were no increased risks observed by type of musculoskeletal condition. Implications: Patients with severe musculoskeletal pain were at greater risk of poorer long-term outcomes following diagnosis of incident primary cancer. Higher proportions of patients with severe musculoskeletal pain were prescribed strong/very strong opioids. This may have masked cancer pain and impacted the diagnosis of the underlying (and yet to be diagnosed) cancer.

Presentation Conference Type Conference Paper (unpublished)
Conference Name SAPC ASM 2022 - UCLan 50th Annual Scientific Meeting
Conference Location University of Central Lancashire, UK
Start Date Jul 4, 2022
End Date Jul 6, 2022
Acceptance Date Aug 20, 2022
Publication Date Jul 6, 2022
Series Title Society for Academic Primary Care (SAPC) 50th Annual Scientific Meeting
Publisher URL