Lucy Gardiner, Daniel Lozano-Rojas, Nikki Smith, Jane Espley, Iain D Stewart, Kimon Ntotsis, Raminder Aul, Nawar Dair Bakerly, Paul Beirne, Charlotte E Bolton, Jeremy S Brown, Andrew Briggs, Trudie Chalder, James D Chalmers, Gourab Choudhury, Melanie J Davies, Anthony De Soyza, Annemarie B Docherty, Nicholas Easom, Carlos Echevarria, Claudia Efstathiou, Omer Elneima, Jonathan Fuld, John R Geddes, Anne Florentine Goemans, William Greenhalf, Neil J Greening, Beatriz Guillen-Guio, Victoria C Harris, Ewen M Harrison, Nicholas Hart, Liam G Heaney, Simon Heller, Ling‐Pei Ho, Alex Horsley, Linzy Houchen-Wolloff, Luke Howard, John R Hurst, Muhammad Mamoon Iqbal, Joseph Jacob, R Gisli Jenkins, Caroline J. Jolley, Mark Jones, Steven Kerr, Kamlesh Khunti, Olivia C. Leavy, Keir Lewis, Nazir I Lone, Janet M Lord, Will D.-C. Man, Michael Marks, Hamish McAuley, Gerry P McCann, Stefan Neubauer, Peter Openshaw, Dhruv Parekh, Paul Pfeffer, Krishnah Poinasamy, Joanna C Porter, Jennifer K Quint, Najib M Rahman, Betty Raman, Matthew Richardson, Sarah DM Rowland-Jones, Matthew Rowland, Ruth M Saunders, Janet T Scott, Malcolm G Semple, Marco Sereno, Ajay M Shah, Aziz Sheikh, Aarti Shikotra, Ananga Singapuri, Maxime Taquet, David L. Thomas, A A Roger Thompson, Mathew Thorpe, Mark Toshner, Lin Wang, Dan G Wootton, Bang Zheng, Louise V Wain, Christopher E Brightling, Sally J Singh, Rod Taylor, Rachael A Evans
BACKGROUND: Preexisting multiple (two or more) long-term conditions (MLTCs) may negatively affect recovery after COVID-19. We investigated how preexisting MLTCs, including different categorization and patterns of MLTCs, affect 1-year health outcomes after severe COVID-19. METHODS: Adults post-hospitalization after COVID-19 were recruited during 2020-2021. We compared recovery at 1 year after discharge using adjusted multivariable logistic regression in 1:1 propensity-matched adults (for age, sex, ethnicity, social deprivation, obesity, and smoking history) with and without preexisting MLTCs. In adults with MLTCs, different categorization such as number of conditions, number and types of body systems involved (e.g. respiratory, cardiovascular), and latent class analysis-derived patterns of condition co-occurrence were assessed for their association with recovery at 1 year. RESULTS: A total of 647 adults with MLTCs were matched with 647 adults without MLTCs (n = 1294; 61.9% male, 79.6% of White ethnicity, median age 59 [interquartile range 52-67] years). The presence of MLTCs was associated with lower odds of feeling fully recovered (odds ratio 0.66 [95% confidence interval 0.51-0.85], P = 0.001). In those with MLTCs, recovery was negatively affected by number and type of body systems involved (e.g. respiratory [odds ratio 0.49 (95% confidence interval 0.34-0.69), P <0.001]) but not by the number of conditions (P >0.1). Four latent classes of MLTC co-occurrence were estimated with different risks of recovery (P <0.01). CONCLUSION: Adults with preexisting MLTCs were 34% less likely to feel fully recovered at 1 year after COVID-19 hospitalization than adults without MLTCs. We describe prognostic classifications of MLTCs, with future work needed to understand whether they have prognostication in broader post-acute infection sequalae.