Please use this identifier to cite or link to this item: https://www.arca.fiocruz.br/handle/icict/37634

Type
Article
Copyright
Open access
Embargo date
2020-12-05
Collections
Share

Statistics
Metadata
Show full item record

ICU STAFFING FEATURE PHENOTYPES AND THEIR RELATIONSHIP WITH PATIENTS' OUTCOMES: AN UNSUPERVISED MACHINE LEARNING ANALYSIS
Zampieri, Fernando G. et al. | Date Issued: 2019
Author
Zampieri, Fernando G.
Salluh, Jorge I. F.
Azevedo, Luciano C. P.
Kahn, Jeremy M.
Damiani, Lucas P.
Borges, Lunna P.
Viana, William N.
Costa, Roberto
Corrêa, Thiago D.
Araya, Dieter E. S.
Maia, Marcelo O.
Ferez, Marcus A.
Carvalho, Alexandre G. R.
Knibel, Marcos F.
Melo, Ulisses O.
Santino, Marcelo S.
Lisboa, Thiago
Caser, Eliana B.
Besen, Bruno A. M. P.
Bozza, Fernando A.
Angus, Derek C.
Soares, Marcio
Affilliation
D’Or Institute for Research and Education. Department of Critical Care. Rio de Janeiro, RJ, Brazil / Hospital do Coração. Research Institute. São Paulo, SP, Brazil.
D’Or Institute for Research and Education. Department of Critical Care. Rio de Janeiro, RJ, Brazil / Epimed Solutions. Department of Research and Development. Rio De Janeiro, RJ, Brazil.
Hospital Sírio Libanês. ICU.São Paulo, SP, Brazil.
University of Pittsburgh School of Medicine. Clinical Research, Investigation, and Systems Modeling of Acute Illness Center. Department of Critical Care Medicine. Pittsburgh, PA, USA / University of Pittsburgh Graduate School of Public Health. Department of Health Policy & Management. Pittsburgh, PA, USA.
Hospital do Coração. Research Institute. São Paulo, SP, Brazil.
Epimed Solutions. Department of Research and Development. Rio de Janeiro, RJ, Brazil.
Hospital Copa D’Or. ICU. Rio de Janeiro, RJ, Brazil.
Hospital Quinta D’Or. ICU. Rio de Janeiro, RJ, Brazil.
Hospital Israelita Albert Einstein. Adult ICU. São Paulo, SP, Brazil.
Hospital Santa Paula. ICU. São Paulo, SP, Brazil.
Hospital Santa Luzia Rede D’Or São Luiz. ICU. Brasília, DF, Brazil.
Hospital São Francisco. ICU. Ribeirão Preto, SP, Brazil.
UDI Hospital. ICU. São Luís, MA, Brazil.
Hospital São Lucas. ICU. Rio de Janeiro, RJ, Brazil.
Hospital Estadual Alberto Torres. ICU. São Gonçalo, RJ, Brazil.
Hospital Barra D’Or. ICU. Rio de Janeiro, RJ, Brazil.
Santa Casa de Misericórdia de Porto Alegre. Hospital Santa Rita. ICU. Porto Alegre, RS, Brazil.
Hospital Unimed Vitoria. ICU. Vitoria, ES, Brazil.
Hospital da Luz. ICU. São Paulo, RJ, Brazil.
D’Or Institute for Research and Education. Department of Critical Care. Rio de Janeiro, RJ, Brazil / Fundação Oswaldo Cruz. Instituto Nacional de Infectologia Evandro Chagas. Rio de Janeiro, RJ, Brasil.
University of Pittsburgh School of Medicine. Clinical Research, Investigation, and Systems Modeling of Acute Illness Center. Department of Critical Care Medicine. Pittsburgh, PA, USA / University of Pittsburgh Graduate School of Public Health. Department of Health Policy & Management. Pittsburgh, PA, USA.
D’Or Institute for Research and Education. Department of Critical Care. Rio de Janeiro, RJ, Brazil.
Abstract
Purpose: To study whether ICU stafng features are associated with improved hospital mortality, ICU length of stay (LOS) and duration of mechanical ventilation (MV) using cluster analysis directed by machine learning. Methods: The following variables were included in the analysis: average bed to nurse, physiotherapist and physician ratios, presence of 24/7 board-certifed intensivists and dedicated pharmacists in the ICU, and nurse and physiotherapist autonomy scores. Clusters were defned using the partition around medoids method. We assessed the association between clusters and hospital mortality using logistic regression and with ICU LOS and MV duration using competing risk regression. Results: Analysis included data from 129,680 patients admitted to 93 ICUs (2014–2015). Three clusters were identifed. The features distinguishing between the clusters were: the presence of board-certifed intensivists in the ICU 24/7 (present in Cluster 3), dedicated pharmacists (present in Clusters 2 and 3) and the extent of nurse autonomy (which increased from Clusters 1 to 3). The patients in Cluster 3 exhibited the best outcomes, with lower adjusted hospital mortality [odds ratio 0.92 (95% confdence interval (CI), 0.87–0.98)], shorter ICU LOS [subhazard ratio (SHR) for patients surviving to ICU discharge 1.24 (95% CI 1.22–1.26)] and shorter durations of MV [SHR for undergoing extubation 1.61(95% CI 1.54–1.69)]. Cluster 1 had the worst outcomes. Conclusion: Patients treated in ICUs combining 24/7 expert intensivist coverage, a dedicated pharmacist and nurses with greater autonomy had the best outcomes. All of these features represent achievable targets that should be considered by policy makers with an interest in promoting equal and optimal ICU care.
Keywords
Intensive care unit
Outcomes
Cluster analysis
Nurse autonomy
Stafng features
ICU organization
 
Publisher
Springer Open Choice
Citation
ZAMPIERI, Fernando G. et al. ICU staffing feature phenotypes and their relationship with patients' outcomes: an unsupervised machine learning analysis. Intensive Care Medicine, v. 45, p. 1599-1607, 2019.
DOI
10.1007/s00134-019-05790-z
ISSN
0342-4642