We welcome the thoughtful comments by Martin et al. in reply to our study1  in which we aimed to assess the relationship between hospital-level neuraxial anesthesia utilization and outcomes. While we included various outcomes, our main finding was related to cost as we demonstrated that increased hospital-level use of neuraxial anesthesia is associated with lower hospitalization cost for lower extremity joint replacements. Martin et al. note correctly that almost none of the clinical outcomes under study appear significantly associated with hospital-level use of neuraxial anesthesia, and while they provide three potential mechanisms to be responsible they mainly focus on “an accounting error” as the likely culprit. Indeed, accurate cost data are notoriously hard to come by,2  and the authors are right to state that costs captured in the Premier Healthcare (Charlotte, North Carolina) database are dependent on each hospital’s accounting methodology, while a smaller number of hospitals submit charges that then have to be converted using Medicare cost-to-charge ratios. Importantly, however, this is all independent of hospital-level neuraxial anesthesia use and thus should not affect the relative effect estimates provided in our study. We therefore respectfully disagree with Martin et al. on the role of accounting errors on our study results. To further evaluate the proposed mechanisms mentioned by Martin et al., we performed a large number of analyses in our study for which we applied multiplicity (Bonferroni) adjustments, which is not without controversy as it results in wider CIs and increases the likelihood of type II errors3 ; this may have affected our results and thus could have been a potential mechanism behind our findings.1  Moreover, the complication outcomes included in our study were selected based on strengths of association as well as prevalence of outcomes found in our previously published individual-level models.4  This is by no means a complete selection of all complication outcomes, and it could very well be that some of the association between hospital-level neuraxial anesthesia use and cost is driven by unmeasured (and more subtle) complications. Finally, it is important to keep in mind that our study focuses on hospital-level and not individual-level use of neuraxial anesthesia. This crucial distinction may very well imply that hospitals with (a high) neuraxial anesthesia utilization in lower extremity joint arthroplasty are different in other aspects as well, including increased cost-effectiveness levels through other pathways. For example, neuraxial anesthesia is commonly mentioned in enhanced recovery pathways, which have been linked to superior outcomes. Indeed, hospitals with higher volumes of neuraxial anesthesia may therefore be more likely to have adopted these pathways, which could also be one of the drivers of the effects found in the current study.

In summary, we welcome the academic discourse by Martin et al., but have to disagree on their assessment. While we welcome studies that would aim to validate our findings using alternative data sources, we feel that our results are robust, particularly because the association between cost and neuraxial volume persisted across the multitude of analyses performed in our study.

The authors declare no competing interests.

1.
Memtsoudis
SG
,
Poeran
J
,
Zubizarreta
N
,
Olson
A
,
Cozowicz
C
,
Mörwald
EE
,
Mariano
ER
,
Mazumdar
M
:
Do hospitals performing frequent neuraxial anesthesia for hip and knee replacements have better outcomes?
Anesthesiology
2018
;
129
:
428
39
.
2.
Nwachukwu
BU
,
Hamid
KS
,
Bozic
KJ
:
Measuring value in orthopaedic surgery.
JBJS Rev
2013
;
1:01874474-201311000-00003
.
3.
Bland
JM
,
Altman
DG
:
Multiple significance tests: The Bonferroni method.
BMJ
1995
;
310
:
170
.
4.
Memtsoudis
SG
,
Sun
X
,
Chiu
YL
,
Stundner
O
,
Liu
SS
,
Banerjee
S
,
Mazumdar
M
,
Sharrock
NE
:
Perioperative comparative effectiveness of anesthetic technique in orthopedic patients.
Anesthesiology
2013
;
118
:
1046
58
.