In this systematic review and meta-analysis, all population base cohort studies evaluated the Association between the use of tramadol Vs. Codeine was included with the risk of all-cause mortality and cardiovascular diseases in patients with osteoarthritis. This systematic review has been registered in PROSPERO (registration number CRD42024512781) and conducted based on the criteria of the PRISMA checklist [20].

The case group included patients with OA who had taken tramadol for at least 1 month. The control group included patients with OA who had taken codeine for at least one month with different doses.

We first specified the keywords related to the research question based on the PICO strategy. The related mesh terms were specified for each keyword. Then, the Overall search strategy of the study was determined. To find relevant studies, we searched four databases: PubMed, Scopus, Embase, Web of sciences, and Google Scholar until January 2024. Each database was searched based on a special search strategy. Two independent researchers carried out the search process. The last search update was on January 12, 2024. The mesh terms were determined by the PICO (population, intervention, comparison, and outcome) format. The search was conducted using the following mesh terms.

((‘’ Mortality’’ OR ‘’Death Rate’’ OR ‘’ Mortality Rate’’ OR ‘’Death’’ OR ‘’ all-cause mortality’’ OR ‘’ all-cause”)) AND ((‘’ Cardiovascular Disease’’ OR ‘’ Major Adverse Cardiac Events’’ OR ‘’ Cardiac Events’’ OR ‘’ Cardiac Event‘’ OR ‘Adverse Cardiac Events’’ OR ‘’ CVD’’)) AND ((‘’Tramundin’’ OR ‘’Biodalgic’’ OR ‘’Jutadol’’ OR ‘’Nobligan’’ OR ‘’Tramadol”) OR (‘’ Morphinan-6-ol, 7,8-didehydro-4,5-epoxy-3-methoxy-17-methyl-, (5alpha,6alpha)-‘’ OR ‘’ N-Methylmorphine’’OR ‘’N Methylmorphine’’OR ‘’ Isocodeine’’ OR ‘’ Codeine Phosphate’’ OR ‘’ Ardinex’’)) AND (‘’ Osteoarthritides’’OR ‘’ Osteoarthrosis ‘’ OR “Degenerative Arthritides’’OR ‘’ Degenerative Arthritis’’ OR ‘’ Arthrosis’’ OR’’ Arthroses’’ OR ‘’ Osteoarthrosis Deformans’’).

Population-based cohort studies by examining the association of tramadol Vs. Codeine use with the risk of all-cause mortality and cardiovascular diseases in osteoarthritis patients, studies with a follow-up of at least 12 months, and the estimation and reporting of the study’s effect size as the criteria for entering this review. A systematic meta-analysis was defined. Exclusion criteria included studies examining the effect of tramadol versus other opioids or stronger compounds derived from codeine, such as Oxycodone or Hydrocodone, examining tramadol versus no-use tramadol, examining the effect of tramadol versus codeine in populations without osteoarthritis, non-English language studies, Letters to the editor articles, case report studies or case series reports, review articles and meta-analyses, laboratory or animal studies, genetic studies and lack of access to the full text of the article.

Searching PubMed, Scopus, Web of Sciences, and Google Scholar databases, 2213 related articles were found. The articles’ titles, authors, and abstracts were extracted with Endnote version 20 software. Using EndNote software, 501 common and duplicate articles were detected and removed after combining the search results of four databases. In the initial screening of the search, 975 articles were selected to check their relationship with the research question. The screening used the title and abstract of studies to find studies related to the research question by two independent researchers. The process of cleaning articles was done to find relevant articles. The full text of 214 studies was read completely to extract the variables. Finally, seven population-based cohort studies were included in this systematic review (Fig. 1).

Flowchart of screening and inclusion of studies

The required variables and data were extracted from the studies using a checklist based on the questions and objectives of this systematic review and entered into an Excel file. The variables examined and extracted in this systematic review include study authors, year of publication, population under study, type of study design, type of statistical analysis, methods for control confounding variables, mean age, gender distribution, setting of the study, the total number of people examined, the number of participants examined in the tramadol and codeine groups, overall mortality, the rate of all-cause mortality and cardiovascular diseases in each group per 1000 people, underlying diseases, history of diabetes, average body mass index, mean follow-up, the risk of all-cause mortality and cardiovascular diseases based on the Hazard ratio (HR) effect size in the 95% confidence Intervals (95% CIs) and the quality of studies in terms of risk of bias. In the data extraction process, if there is a difference between two researchers about a variable, a third researcher was used to resolve the difference.

In this systematic review, the Newcastle-Ottawa Quality Assessment Form for Cohort Studies checklist was used to assess the quality of studies regarding the risk of bias [21]. This checklist evaluates and scores the quality of studies included in regular review studies in the three dimensions: selection, comparability, and outcome/exposure. This checklist’s score range is between 0 and 9; a higher score means higher quality. The entered studies are based on quality scores in one of three categories: good, fair, and poor. They are classified as follows. Good (3 or 4 scores for the selected dimension and one or two stars for the comparability dimension and 2 or 3 stars for the outcome/exposure dimension), Fair (2 scores for the selected dimension and one or two stars for the comparability dimension and 2 or 3 stars for outcome/exposure dimension) and Poor (0 or 1 score for the selected dimension and 0 stars for the comparability dimension and 0 or 1 star for the outcome/exposure dimension).

The extracted data were analyzed using Stata 17 software. The characteristics of the participants included in the study were reported using descriptive and mean statistics. The risk of mortality and cardiovascular diseases in all studies was extracted with the Hazard ratio (HR) in the 95% confidence interval (95% CI) to reduce heterogeneity. To control the effects of the study sample size, the pooled effect size was estimated with a random model. The Cochran Q and I2 tests were used to evaluate the heterogeneity of the studies. Egger’s test was used to evaluate the presence of publication bias. Due to the absence of publication bias for any of the outcomes in different studies, there was no need to use trim and fill analysis to solve publication bias. The risk of tramadol versus codeine use with ACM and CVD was reported using the pooled HR with a 95% CI.