Ghent University Academic Bibliography

Advanced

On the practice of ignoring center-patient interactions in evaluating hospital performance

Machteld Varewyck, Stijn Vansteelandt UGent, Marie Eriksson and Els Goetghebeur UGent (2016) STATISTICS IN MEDICINE. 30(2). p.227-238
abstract
We evaluate the performance of medical centers based on a continuous or binary patient outcome (e.g., 30-day mortality). Common practice adjusts for differences in patient mix through outcome regression models, which include patient-specific baseline covariates (e.g., age and disease stage) besides center effects. Because a large number of centers may need to be evaluated, the typical model postulates that the effect of a center on outcome is constant over patient characteristics. This may be violated, for example, when some centers are specialized in children or geriatric patients. Including interactions between certain patient characteristics and the many fixed center effects in the model increases the risk for overfitting, however, and could imply a loss of power for detecting centers with deviating mortality. Therefore, we assess how the common practice of ignoring such interactions impacts the bias and precision of directly and indirectly standardized risks. The reassuring conclusion is that the common practice of working with the main effects of a center has minor impact on hospital evaluation, unless some centers actually perform substantially better on a specific group of patients and there is strong confounding through the corresponding patient characteristic. The bias is then driven by an interplay of the relative center size, the overlap between covariate distributions, and the magnitude of the interaction effect. Interestingly, the bias on indirectly standardized risks is smaller than on directly standardized risks. We illustrate our findings by simulation and in an analysis of 30-day mortality on Riksstroke.
Please use this url to cite or link to this publication:
author
organization
year
type
journalArticle (original)
publication status
published
subject
keyword
quality of care, direct and indirect standardization, Firth correction, STROKE, misspecified model, causal effects, MODELS
journal title
STATISTICS IN MEDICINE
Stat. Med.
volume
30
issue
2
pages
227 - 238
Web of Science type
Article
Web of Science id
000367972400005
JCR category
STATISTICS & PROBABILITY
JCR impact factor
1.861 (2016)
JCR rank
20/124 (2016)
JCR quartile
1 (2016)
ISSN
0277-6715
DOI
10.1002/sim.6634
language
English
UGent publication?
yes
classification
A1
copyright statement
I have retained and own the full copyright for this publication
id
7093651
handle
http://hdl.handle.net/1854/LU-7093651
date created
2016-02-21 16:39:30
date last changed
2016-12-19 15:39:20
@article{7093651,
  abstract     = {We evaluate the performance of medical centers based on a continuous or binary patient outcome (e.g., 30-day mortality). Common practice adjusts for differences in patient mix through outcome regression models, which include patient-specific baseline covariates (e.g., age and disease stage) besides center effects. Because a large number of centers may need to be evaluated, the typical model postulates that the effect of a center on outcome is constant over patient characteristics. This may be violated, for example, when some centers are specialized in children or geriatric patients. Including interactions between certain patient characteristics and the many fixed center effects in the model increases the risk for overfitting, however, and could imply a loss of power for detecting centers with deviating mortality. Therefore, we assess how the common practice of ignoring such interactions impacts the bias and precision of directly and indirectly standardized risks. The reassuring conclusion is that the common practice of working with the main effects of a center has minor impact on hospital evaluation, unless some centers actually perform substantially better on a specific group of patients and there is strong confounding through the corresponding patient characteristic. The bias is then driven by an interplay of the relative center size, the overlap between covariate distributions, and the magnitude of the interaction effect. Interestingly, the bias on indirectly standardized risks is smaller than on directly standardized risks. We illustrate our findings by simulation and in an analysis of 30-day mortality on Riksstroke.},
  author       = {Varewyck, Machteld and Vansteelandt, Stijn and Eriksson, Marie and Goetghebeur, Els},
  issn         = {0277-6715},
  journal      = {STATISTICS IN MEDICINE},
  keyword      = {quality of care,direct and indirect standardization,Firth correction,STROKE,misspecified model,causal effects,MODELS},
  language     = {eng},
  number       = {2},
  pages        = {227--238},
  title        = {On the practice of ignoring center-patient interactions in evaluating hospital performance},
  url          = {http://dx.doi.org/10.1002/sim.6634},
  volume       = {30},
  year         = {2016},
}

Chicago
Varewyck, Machteld, Stijn Vansteelandt, Marie Eriksson, and Els Goetghebeur. 2016. “On the Practice of Ignoring Center-patient Interactions in Evaluating Hospital Performance.” Statistics in Medicine 30 (2): 227–238.
APA
Varewyck, Machteld, Vansteelandt, S., Eriksson, M., & Goetghebeur, E. (2016). On the practice of ignoring center-patient interactions in evaluating hospital performance. STATISTICS IN MEDICINE, 30(2), 227–238.
Vancouver
1.
Varewyck M, Vansteelandt S, Eriksson M, Goetghebeur E. On the practice of ignoring center-patient interactions in evaluating hospital performance. STATISTICS IN MEDICINE. 2016;30(2):227–38.
MLA
Varewyck, Machteld, Stijn Vansteelandt, Marie Eriksson, et al. “On the Practice of Ignoring Center-patient Interactions in Evaluating Hospital Performance.” STATISTICS IN MEDICINE 30.2 (2016): 227–238. Print.