Objective The COVID-19 pandemic has strained healthcare worldwide. Its direct complications, management and prognosis have been described. Downstream effects, including length of hospital stay (LOS), implications on discharge planning, and effect of in-house testing require formal study. Methods A retrospective cohort study of patients suspected of COVID-19 infection admitted at a metropolitan Australian hospital was conducted. Outcomes before and after availability of in-house COVID-19 testing were compared. Results A total of 129 admissions were analysed. Indications for COVID-testing were dyspnoea (61.2%), fever (19.3%) and delirium (10.8%). All tested negative for COVID-19. Prior to in-house testing, mean LOS was 7.17 days (s.d. ± 4.2), and mean isolation of 1.8 days (s.d. ± 0.8). After availability of in-house testing, mean LOS was 4.78 days (s.d. ± 4.3) with mean isolation of 1.3 days (s.d. ± 0.9), both statistically significant differences. There were five inpatient falls, equivalent to 14.8 falls per 1000 patient/days. Twenty-two patients (17%) required subsequent sub-acute admission, 15 before in-house testing and five after (P = 0.058); however, a sub-group analysis for age >65 years was performed, and the results were significant (P < 0.05), showing all patients who required subacute admissions were aged >65. Conclusion In-house COVID-19 testing is suggested to significantly reduce the duration patients spend in isolation and overall LOS in hospital. A shorter period of isolation and hospital LOS may reduce the need for subacute transfer in patients aged greater than 65 years, as well as the rates of inpatient falls. Large scale studies are needed to further elucidate these findings.