If Charles Dickens were alive today, he might be convinced to pen an epic about the current state of perioperative medicine instead of a nineteenth-century 7-yr-old orphan. But just as Pip painfully learned how readily Great Expectations can evaporate, we believe anesthesiology must avoid a similarly painful downfall by immediately adopting vigorous testing procedures to the great expectations of enhanced recovery after surgery (ERAS) programs. ERAS requires the same scientific inquiry, systematic collection and analysis of quality outcome data, and statistical validation—quite simply EVIDENCE—that are required of pharmaceutical medications and medical devices.

Thus, we applaud Dr. Sessler’s editorial “Implications of Practice Variability” in the February issue of Anesthesiology,1  in which he opines that the rampant proliferation of perioperative protocols, pathways, and ERAS guidelines vastly exceeds the sustaining scientific evidence. But we humbly suggest that Dr. Sessler’s editorial may not have gone far enough in raising the alarm about instituting new, obligatory clinical pathways as “best practices” when the published evidence for such interventions is either absent, contradictory, or statistically inadequate. Moreover, of greatest concern is the common practice whereby medical centers and ERAS committees invoke new clinical pathways, but omit the commensurate requirement for concurrent, parallel, and meaningful (i.e., more than just hospital discharge) patient data to provide definitive answers to whether these protocols indeed improve patient value as well as hospital efficiency. We posit that the future of ERAS must embrace three key concepts.

Let’s be clear—the general concept for the standardization of patient care for particular perioperative interventions is prudent, efficient, cost-effective, and—most critically—proven to work with selective and usually simple practice modifications. Thus, interventions such as practicing minimally invasive surgery, avoidance of nasogastric tubes, adoption of updated non per os and antibiotic guidelines, early perioperative oral intake, and early ambulation are prime examples that are proven to improve patient outcomes. But in many institutions, current ERAS protocols have quickly morphed from relatively simple interventions to multipage, detailed clinical mandates that dictate everything from type (and doses) of “approved” anesthetic drugs (inhalation anesthesia vs. total intravenous anesthesia), content and rate of fluid infusions, types of monitoring (BIS, FloTrac, etc.), and perhaps even the exact vasopressor for the treatment of hypotension. Such mandates clearly increase the cost, complexity, and preparation time for intraoperative care. But there is an alarming paucity of data for the majority of these trendy components of recent ERAS pathways. For example, although specific drugs such as infusions of lidocaine, magnesium, and ketamine are often promoted during major spine surgery, they have never, to our knowledge, been proven to be more efficacious than traditional anesthetic regimes. Indeed, a current randomized, controlled trial in patients having multilevel spine surgery with a protocol consisting of preoperative oral gabapentin and acetaminophen with intraoperative infusions of lidocaine and ketamine was stopped early because of futility—lacking evidence of efficacy!2 

Originally, outcome reports from the early ERAS literature focused on the reduction of postoperative complications. More recently, length of stay has emerged as the “gold standard” for tracking protocol efficacy. Although there is no question that length of stay equates with reduction of required hospital resources—pleasing hospital administrators with cost savings—it is not clear that length of stay as a stand-alone metric is necessarily in the patient’s best interest. As pointed out by Memtsoudis and Kehlet in their 2019 editorial,3  a decreased length of stay may be associated with increased postdischarge spending. Furthermore, many of these postdischarge costs are de facto shifted to an alternative facility, such as a rehabilitation hospital, or the invisible but very real costs to the patient’s caregivers (friends, family, neighbors) at home.

Moreover, although ERAS pathways are clearly linked to a reduced length of stay, we challenge clinicians and administrators to avoid the statistical trap of associating events (easy to do and easy to analyze) compared with designing and implementing valid inferential outcome studies proving cause-and-effect (hard to do and more complex to analyze).4  Lastly, it is our impression that some of the length of stay improvement with ERAS could be due to setting new patient expectations during the preoperative anesthesia and surgical visits. Indeed, one must consider either a placebo or even a Hawthorne effect as confounders.

The bulk of the ERAS literature has come out of major academic medical centers, and we challenge investigators to expand and test whether and how these protocols are applicable to community and even small critical-access rural hospitals. In that process, we must recognize a number of barriers to generalized adoption of ERAS, including the following:

  • Cost. Additional resources are needed to initiate and maintain protocols, rewrite standard order sets in your electronic medical record, provide sufficient monitors for protocol-driven algorithms, and hire data collection and analytic personnel to continually monitor quality outcomes within the institution.

  • Robust informatics. Measuring pathway improvements is vital to both intelligently modify the protocols for maximum benefit as well as sustain support from hospital leadership. This requires both a sophisticated functional data system as well as skilled informatics practitioners to analyze the outcomes. At the University of Minnesota, more than $150,000 per year is spent tracking these quality metrics (personal communication, Richard C. Prielipp, M.D., Department of Anesthesiology, University of Minnesota, Minneapolis, Minnesota; verbal communication as of June 2020).

  • A new team culture. Although a local champion is key to initiating an ERAS pathway, the long-term success requires sustained collaboration of anesthesia, surgery, nursing, and hospital administration. Any breakdown in this network exposes impediments to sustained compliance.

We applaud Sessler and Memtsoudis for sounding the alarm about the unbridled enthusiasm for ERAS protocols.1,3,4  As Sessler opined: “There is no basis for giving clinical pathways a ‘free pass’ on evidence.”1  Indeed, we applaud Anesthesiology for increasing the volume of the alarm bell with publications like the randomized, controlled trial by Maheshwari et al.2  Publication of such “negative” trials4  is vital to separating valid ERAS elements from unnecessary or perhaps even detrimental components of proposed pathways. “Great expectations” for ERAS may indeed prove to be true, but in 2020 we still don’t know whether the reality equates to the hype. Regardless, it is time to put the evidence in ERAS.1,5 

Competing Interests

Dr. Prielipp is a member of the Board of Directors of the Anesthesia Patient Safety Foundation (APSF; Rochester, Minnesota) and serves on the speakers’ bureau for Merck Co., Inc. (Kenilworth, New Jersey). Dr. Rice declares no competing interests.

1.
Sessler
DI
.
Implications of practice variability.
Anesthesiology
.
2020
;
132
:
606
8
2.
Maheshwari
K
,
Avitsian
R
,
Sessler
DI
,
Makarova
N
,
Tanios
M
,
Raza
S
,
Traul
D
,
Rajan
S
,
Manlapaz
M
,
Machado
S
,
Krishnaney
A
,
Machado
A
,
Rosenquist
R
,
Kurz
A
.
Multimodal analgesic regimen for spine surgery: A randomized placebo-controlled trial.
Anesthesiology
.
2020
;
132
:
992
1002
3.
Memtsoudis
SG
,
Poeran
J
,
Kehlet
H
.
Enhanced recovery after surgery in the United States: From evidence-based practice to uncertain science?
JAMA
.
2019
;
321
:
1049
50
4.
Sessler
DI
.
Negative trials, and what to do with them?: First, stop calling them “negative”.
Anesthesiology
.
2020
;
132
:
221
4
5.
Joshi
GP
,
Kehlet
H
.
Enhanced recovery pathways: Looking into the future.
Anesth Analg
.
2019
;
128
:
5
7