This meta-analysis was conducted according to the guidelines of the PRISMA statement and do not require ethics approval or patient consent as the present study was based on previously published articles.
Search methods
We searched the Embase via Ovid, MEDLINE via Ovid, Web of Science, Cochrane Library and PudMed databases to Feb. 2021. Each database was searched with search strategy that is suitable to them. See supplementary material. Relative studies were also identified through cross-checking of references in the primarily identified studies and review articles. In this period, no publication time or language restriction was used.
Eligibility criteria
Two authors were responsible for the study selection and quality assessment section mentioned below independently. Studies were included in this analysis if they were: 1) studies that evaluated lateral UKA as a primary surgical treatment; 2) studies assessed implant survivorship and clinical outcomes of patients who underwent lateral UKA; 3) had a follow-up of 3 years or longer; 4) reported the confident interval of implant survival rate. Instead, the following studies were excluded due to: 1) missing data or insufficient data; 2) studies reported their conclusion based on the same cohort, and in this circumstance, only the study with the most comprehensive data was included; 3) studies not published in English; 4) study designs, such as, case reports, reviews, commentaries, editorials or conference abstracts were declined; 5) animal experiments or human in vitro work were also excluded. Any discrepancies were resolved by discussion.
Data extraction
The same two authors completed the need for data extraction using a standardized data collection form. Following data were extracted: general information (publication year, study design, first author’s name), sample size (number of participants, number of knee underwent lateral UKA), baseline characteristics (age, sex), survivorship of the implants, mode of failure. When we cannot contact the corresponding author, the survival data were extracted from their Kaplan-Meier curves for its advantage that it accounts for patients who were lost to follow-up or died. For the analysis of survivorship, failure was defined as revision for any reason (defined as operations in which at least one of the components was changed), re-operations, aseptic revision and dislocation of the bearing.[12]
Assessment of methodological quality
As the previous systematic review mentioned, the quality of studies available was low.[13] Therefore, a quality assessment was conducted using the Newcastle–Ottawa Scale (NOS), which was accommodated to the studies included in this meta-analysis.[14] The NOS scores study in the assessment of three domains: (1) selection of study groups; 2) ascertainment of endpoints; 3) follow-up of outcomes; 4) prospective calculation of the study size, respectively.
Statistical analysis
We used random effects models to calculate implant survivorship and 95% confidence intervals. The heterogeneity between studies was assessed using the χ2 test and the inconsistency index (I2). An I2 > 50% with P < 0.05 from the χ2 test is indicative of significant heterogeneity. In this case, a random effect model was chosen to pool the data. Otherwise, a fixed effect model was used. Sensitivity and subgroup analysis were also conducted to identify the source of heterogeneity in the eligible studies. Subgroup analysis was conducted between fix-bearing and mobile bearing prosthesis, lateral parapatellar approach (LPPA) and minimally invasive-surgical approach (MIS), publication year in between 2000–2010 and 2011–2020.
Publication bias was analyzed by using the Deeks funnel plot and an asymmetry test. P < 0.05 is considered the existence of publication bias.
All analyses were performed using Stata version 15.1 (StataCorp LP). Engauge Digitizer version 10.8 was used when only Kaplan-Meier survival curve was reported in the publication.