Long-term Opioid Therapy, Misuse and Mortality in Patients With Chronic Non-cancer Pain in Germany

Long-term Opioid Therapy, Misuse and Mortality in Patients With Chronic Non-cancer Pain in Germany

Among patients receiving opioids for noncancer pain, recent research in North America showed a strong association between doses and opioid-related mortality, especially at dosages exceeding thresholds recommended in recent guidelines. However, the focus on over-dosage may underestimate overall opioid-related mortality and data on death associated with opioid use in a population-based cohort of chronic noncancer pain patients in Europe is scarce. Especially comparative studies studying the safety of long-term opioid therapy in a real-world setting relative to non-opioid medication for chronic noncancer pain are needed in a European context.

This is a cross-sectional observational cohort study between 2012 and 2017 of patients with chronic noncancer pain. The primary objective is to investigate the association between mortality among patients with chronic noncancer pain with long-term opioid-therapy compared to non-opioid pain medication. The data will be retrieved from an anonymized German health claims database including 4,00,000 persons insured by 69 German statutory health insurances. The data set includes 5.0% of the population covered by statutory health insurances from January 1, 2012, to December 31, 2017. Only anonymized and aggregated data (no directly or indirectly identifying data) will be extracted.

Patients with headache, diseases of the musculoskeletal system and connective tissue, migraine, trigeminal neuralgia, atypical facial pain, persistent somatoform pain disorder, polyneuropathies, or diabetes mellitus with neurological complications in at least 3 quarters between 01.01.2012 and their first pain medication claim are analyzed. Patients will be stratified in an opioid and a non-opioid group and will be compared with a propensity score matching approach.

Patients' follow-up period include 5 years after start treatment and chronic pain diagnosis between 01.01.2013 and 31.12.2017. Each patient will be censored at death (death date), switching of study group, 12 months without treatment or followed-up 5 years until last known record for the, whichever happens first. The main analyses will be analyzed with a multivariate Cox proportional hazards regression.

Inclusion Criteria:

- Patients are only included if they have been diagnosed in at least three quarters in the study period with one of the following diagnoses: R51, R52*, M00*-M99*, G43*-G44*, G50.0 or G50.1, F45.4*, G62*, or E10.4*-E14.4 plus G63.3. At least one diagnoses must be between 1 January 2012 and index treatment and main and secondary hospital diagnoses (i.e. Haupt- und Nebendiagnosen) will be used to include the patients.

Exclusion Criteria:

• Patients with present opioid and non-opioid pain medication prescriptions in 2012 are excluded from analysis (therapy-naïve patients only).
• Cancer patients will be excluded if the cancer diagnosis is accompanied by at least one of the following treatments in the same quarter: radiation therapy or chemotherapy all defined by the OPS codes in. Diagnoses will be assessed via ICD-10 diagnoses during the inclusion period from 1 January 2012 until index treatment and main and secondary hospital diagnoses (i.e. Haupt- und Nebendiagnosen) will be taken into account.
• Palliative care, coded by ICD-10 code Z51.5 or OPS code 8-982*, 8-98e*, 8-98h* before index date is excluded
• Opioid substitution treatment with ICD-10 code Z51.83 in the study period is excluded.