An exploration of strong opioid utilisation in non-cancer pain patients in UK primary care, 2000-2010

Adan Hag Hersi, Muna (2018) An exploration of strong opioid utilisation in non-cancer pain patients in UK primary care, 2000-2010. PhD thesis, University of Nottingham.

[img] PDF (Thesis - as examined) - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (4MB)



Opioids are the most potent analgesics available and their treatment is well established in cancer and acute pain. However, their long-term use in non-cancer pain is worrisome due to insufficient evidence on long-term effectiveness and safety. Population-based research examining the utilisation of opioids in non-cancer pain patients in UK primary care remains limited and little information exists on patients receiving long-term therapy.


This thesis aimed to examine the demographics, clinical traits, prescribing patterns and healthcare use of non-cancer pain patients prescribed strong opioids in primary care, with a focus on long-term strong opioid use.


The thesis was a retrospective observational study using the Clinical Practice Research Datalink. Prescription data of four strong opioids (morphine, buprenorphine, fentanyl and oxycodone) issued between 2000-2010 to adults (aged ≥18 years) without cancer diagnosis 12 months within the date of first prescription use were included.

The annual prevalence and incidence of non-cancer pain patients prescribed strong opioids over the 11 years were initially assessed using joinpoint regression analysis, and the users’ demographic and clinical characteristics were explored descriptively. The numbers of prescriptions per patients, total number of days of drug supply, and total oral morphine equivalent daily dose (OMED) were subsequently calculated annually and compared between long-term (>90 days/year) and short-term (≤90 days/year) users.

Multivariate linear regression modelling with generalized estimating equations was consecutively used to identify baseline and time-varying covariates linked to long-term strong opioid use. Lastly, long-term strong opioid users’ primary and secondary healthcare use was quantified using CPRDs’ data link with Hospital Episode Statistics (HES), and the nature of hospital admissions and patient-level factors of influence were determined by multivariate regression data analysis.


A total of 135,941 non-cancer pain patients (63.34% female, mean age 66.34±17.98 years) were prescribed strong opioids and were included in the analysis. The incidence of long-term prescribing (>90 days) continued to increase by 16.96% per year (95%CI: 13.70%, 20.30%, p<0.001) from 3.60 (95%CI: 3.38, 3.82) per 10,000 person-years in 2001 to 12.75 (95%CI: 12.41, 13.10) per 10,000 person-years in 2010. The mean number of pain diagnoses amongst long-term users was 3.00±2.16, with back pain, abdominal pain and osteoarthritis presenting the most common diagnostic disorders, and the mean number of co-morbidities was 2.04±1.70. Depression and anxiety were prevalent in 50.02% and 22.13% of long-term users, respectively.

The mean daily oral morphine equivalent dose (OMED) of long-term users (95.88±109.03 mg/day/year) was almost twice that of short-term users (54.80±54.55mg/day/year). Over a quarter (26.01%) of long-term users received high oral morphine equivalent daily dose and their mean annual OMED was 221.19±148.07mg/day and mean annual days’ supply was 277.05±84.55 days.

Forty-one percent of long-term users had one or more all cause in-patient admissions during the study period, and a rising trend of admission spells was noted annually by 28.16% (95% CI: 26.40, 29.90, p<0.001). Musculoskeletal pain disorders were the main cause of hospital admissions and re-admissions. Factors associated with hospitalisation included; greater GP visits per year (>3 consultations/year), >120mg morphine equivalent daily dose (aRR: 1.37; 95%CI: 1.27, 1.49) and co-prescriptions of psychoactive medications, including antidepressants (aRR: 1.07; 95%CI: 1.00, 1.14), benzodiazepines (aRR: 1.17; 95%CI: 1.09, 1.26) and non-benzodiazepines (aRR: 1.15; 95%CI: 1.05, 1.27).


Prescribing of strong opioids over the 11 years was characterized by a shift towards long-term prescribing. Primary care non-cancer pain patients exposed to strong opioids long-term were typified by multiple pain and co-morbid illnesses that included common psychiatric disorders, which suggest complex, and vulnerable, high-risk patients that are susceptible to negative health consequences. The thesis has identified several covariates linked to strong opioid-utilisation and healthcare use, which exemplifies key findings that can be used to inform clinical decision-making, targeted management interventions and monitoring of non-cancer pain patients treated with strong opioids long-term.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Knaggs, Roger
Chen, Li-Chia
Keywords: opioids; primary care; pain relief
Subjects: R Medicine > RB Pathology
Faculties/Schools: UK Campuses > Faculty of Science > School of Pharmacy
Item ID: 53322
Depositing User: Adan Hag Hersi, Muna
Date Deposited: 11 Dec 2018 04:40
Last Modified: 11 Dec 2020 04:30

Actions (Archive Staff Only)

Edit View Edit View