BACKGROUND: Readmission to intensive care during the same hospital stay has been associated with a greater risk of in-hospital mortality and has been suggested as a marker of quality of care. There is lack of published research attempting to develop clinical prediction tools that individualise the risk of readmission to the intensive care unit during the same hospital stay. OBJECTIVE: To develop a prediction model using an inception cohort of patients surviving an initial ICU stay. DESIGN, SETTING AND PARTICIPANTS: The study was conducted at Liverpool Hospital, Sydney. An inception cohort of 14 952 patients aged 15 years or more surviving an initial ICU stay and transferred to general wards in the study hospital between 1 January 1997 and 31 December 2007 was used to develop the model. Binary logistic regression was used to develop the prediction model and a nomogram was derived to individualise the risk of readmission to the ICU during the same hospital stay. MAIN OUTCOME MEASURE: Readmission to the ICU during the same hospital stay. RESULTS: Among members of the study cohort there were 987 readmissions to ICU during the study period. Compared with patients not readmitted to the ICU, patients who were readmitted were more likely to have had ICU stays of at least 7 days (odds ratio [OR], 2.2 [95% CI, 1.85- 2.56]); non-elective initial admission to the ICU (OR, 1.7 [95% CI, 1.44-2.08]); and acute renal failure (OR, 1.6 [95% CI, 0.97-2.47]). Patients admitted to the ICU from the operating theatre or recovery ward had a lower risk of readmission to ICU than those admitted from general wards, the emergency department or other hospitals. The maximum error between observed frequencies and predicted probabilities of readmission to ICU was estimated to be 3%. The area under the receiver operating characteristic curve of the final model was 0.66. CONCLUSION: We have developed a practical clinical tool to individualise the risk of readmission to the ICU during the same hospital stay in patients who survive an initial episode of intensive care.
BACKGROUND: Readmission to intensive care during the same hospital stay has been associated with a greater risk of in-hospital mortality and has been suggested as a marker of quality of care. There is lack of published research attempting to develop clinical prediction tools that individualise the risk of readmission to the intensive care unit during the same hospital stay. OBJECTIVE: To develop a prediction model using an inception cohort of patients surviving an initial ICU stay. DESIGN, SETTING AND PARTICIPANTS: The study was conducted at Liverpool Hospital, Sydney. An inception cohort of 14 952 patients aged 15 years or more surviving an initial ICU stay and transferred to general wards in the study hospital between 1 January 1997 and 31 December 2007 was used to develop the model. Binary logistic regression was used to develop the prediction model and a nomogram was derived to individualise the risk of readmission to the ICU during the same hospital stay. MAIN OUTCOME MEASURE: Readmission to the ICU during the same hospital stay. RESULTS: Among members of the study cohort there were 987 readmissions to ICU during the study period. Compared with patients not readmitted to the ICU, patients who were readmitted were more likely to have had ICU stays of at least 7 days (odds ratio [OR], 2.2 [95% CI, 1.85- 2.56]); non-elective initial admission to the ICU (OR, 1.7 [95% CI, 1.44-2.08]); and acute renal failure (OR, 1.6 [95% CI, 0.97-2.47]). Patients admitted to the ICU from the operating theatre or recovery ward had a lower risk of readmission to ICU than those admitted from general wards, the emergency department or other hospitals. The maximum error between observed frequencies and predicted probabilities of readmission to ICU was estimated to be 3%. The area under the receiver operating characteristic curve of the final model was 0.66. CONCLUSION: We have developed a practical clinical tool to individualise the risk of readmission to the ICU during the same hospital stay in patients who survive an initial episode of intensive care.
Authors: Nirav Haribhakti; Pallak Agarwal; Julia Vida; Pamela Panahon; Farsha Rizwan; Sarah Orfanos; Jonathan Stoll; Saqib Baig; Javier Cabrera; John B Kostis; Cande V Ananth; William J Kostis; Anthony T Scardella Journal: J Gen Intern Med Date: 2021-01-22 Impact factor: 5.128
Authors: Rebecca L Hoffman; Jason Saucier; Serena Dasani; Tara Collins; Daniel N Holena; Meghan Fitzpatrick; Boris Tsypenyuk; Niels D Martin Journal: Int J Qual Health Care Date: 2017-06-01 Impact factor: 2.038
Authors: F Shaun Hosein; Niklas Bobrovitz; Simon Berthelot; David Zygun; William A Ghali; Henry T Stelfox Journal: Crit Care Date: 2013-06-29 Impact factor: 9.097