Abstract Background/Aims Interstitial Lung Disease (ILD) affects approximately 30% of patients with idiopathic inflammatory myopathies (IIM). It is a leading cause of morbidity and mortality in IIM patients. Natural history of ILD varies widely from rapidly-progressive to indolent with minimal symptoms. IIM describes a collection of related autoimmune, inflammatory disorders predominantly affecting combinations of muscles, skin and lungs. Improving understanding of ILD clinical course will aid prognostication and guide sub-categorisation to improve future research in the field. We performed a systematic review and meta-analysis of prognostic factors in IIM-ILD. Methods MEDLINE and EMBASE databases were interrogated on 18/03/2021 using a pre-defined search protocol (PROSPERO ID:CRD42021240206). Studies providing summary data relating to numbers of survivors vs non-survivors according to any baseline criteria were selected. Baseline characteristics reported in ≥ 5 papers were included for meta-analysis. Risk of bias was assessed by Newcastle-Ottawa score. Stata-16 software was used for meta-analyses using a random effects model to report difference as odds ratio for binary variables, and hedge’s g standardised mean difference for continuous variables. A continuity correction was applied for zero effect studies. Heterogeneity was measured by I2, and publication bias assessed with Egger’s test. Results The search returned 4211 articles. 722 were relevant for abstract review, with 454 requiring full text assessment. 78 studies were eligible for inclusion. Overall mortality was 26.1% (+/-0.14 SD). 62 studies were from Asia, two from Mexico and four from Europe, with mortality rates of 26.5%, 13.6% and 25.3% respectively. The strongest risk factor is anti-MDA-5 antibody (OR 6.03). Conversely anti-tRNA-synthetase antibodies (ARS) are protective. When ARS+ was compared to MDA-5-/ARS- patients only, this difference between the groups disappeared. Anti-MDA-5 titre showed a non-significant effect direction towards higher mortality. ANA, anti-SS-A/Ro52 did not significantly impact mortality. Clinical predictors of increased mortality were male gender, acute/sub-acute onset, clinically amyopathic disease (CADM), dyspnoea, ulceration, fever and increasing age. Additional investigations shown to positively predict mortality were CRP, ferritin, LDH, AaO2 gradient, ground glass opacity and fibrosis HRCT scores. Whereas higher SP-D, lymphocytes, %FVC and %DLCO were protective. Conclusion Whilst many factors were associated with increased mortality in IIM-ILD, heterogeneity between studies is high with all having moderate to high risk of bias. A majority of the data come from studies in Japanese and Chinese populations where anti-MDA5 disease appears to be particularly prevalent and associated with deleterious outcomes. Extrapolating to local populations may not be appropriate. The domination of studies by anti-MDA5 disease may be masking risk factors relevant to other IIM subgroups. Due to the rarity of IIMs, these subgroups are often researched together, however this meta-analysis highlights the need to better categorise IIM patients in clinical research. Disclosure J.R. Hannah: None. T. Gordon: None. M. Rooney: None. J. Galloway: None. P. Gordon: None.