TY - JOUR
T1 - Is Magnet® recognition associated with improved outcomes among critically ill children treated at freestanding children's hospitals?
AU - Rettiganti, Mallikarjuna
AU - Shah, Kavisha M.
AU - Gossett, Jeffrey M.
AU - Daily, Joshua A.
AU - Seib, Paul M.
AU - Gupta, Punkaj
PY - 2018/2
Y1 - 2018/2
N2 - Purpose With increasing emphasis on high-quality care, we designed this study to evaluate the relationship between Magnet® recognition and patient outcomes in pediatric critical care. Materials and methods Post hoc analysis of data from an existing administrative national database. We used inverse probability of treatment weighting and multivariate models to compare outcomes between two study groups after adjusting for confounding variables. Results A total of 823,634 pediatric patients from 41 centers were included. Of these, 454,616 patients (55.2%) were treated in 23 Magnet hospitals. The majority of baseline characteristics did not vary significantly among the two study groups. In adjusted models, there was no difference in mortality between the two groups (Magnet vs. non-Magnet; odds ratio: 0.92, 95% confidence interval: 0.77–1.11). When stratified by various subgroups, such as cardiac, non-cardiac, ECMO, cardiac arrest, respiratory failure, use of nitric oxide, genetic abnormality etc., Magnet status of the hospital did not confer a survival advantage. In a sensitivity analysis on patients from crossover hospitals only, attainment of magnet status was associated with increased hospital charges. Conclusions This large observational study calls into question the utility of the Magnet Recognition Program among children with critical illness, at least among the freestanding children's hospitals.
AB - Purpose With increasing emphasis on high-quality care, we designed this study to evaluate the relationship between Magnet® recognition and patient outcomes in pediatric critical care. Materials and methods Post hoc analysis of data from an existing administrative national database. We used inverse probability of treatment weighting and multivariate models to compare outcomes between two study groups after adjusting for confounding variables. Results A total of 823,634 pediatric patients from 41 centers were included. Of these, 454,616 patients (55.2%) were treated in 23 Magnet hospitals. The majority of baseline characteristics did not vary significantly among the two study groups. In adjusted models, there was no difference in mortality between the two groups (Magnet vs. non-Magnet; odds ratio: 0.92, 95% confidence interval: 0.77–1.11). When stratified by various subgroups, such as cardiac, non-cardiac, ECMO, cardiac arrest, respiratory failure, use of nitric oxide, genetic abnormality etc., Magnet status of the hospital did not confer a survival advantage. In a sensitivity analysis on patients from crossover hospitals only, attainment of magnet status was associated with increased hospital charges. Conclusions This large observational study calls into question the utility of the Magnet Recognition Program among children with critical illness, at least among the freestanding children's hospitals.
KW - Cardiac surgery
KW - Children
KW - Critical illness
KW - In-hospital mortality
KW - Magnet recognition
KW - Outcomes
UR - http://www.scopus.com/inward/record.url?scp=85037617815&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85037617815&partnerID=8YFLogxK
U2 - 10.1016/j.jcrc.2017.09.004
DO - 10.1016/j.jcrc.2017.09.004
M3 - Article
C2 - 28917160
AN - SCOPUS:85037617815
SN - 0883-9441
VL - 43
SP - 207
EP - 213
JO - Journal of Critical Care
JF - Journal of Critical Care
ER -