Title: Cardiac surgery and healthcare quality: Is
the right question being asked?
Authors : Abdullah Nasif, MD1/ Saqib Masroor, MD1
1Division of Cardiothoracic Surgery, Department of Surgery, University
of Toledo Medical Center
Toledo, OH
USA
Manuscript: Minimally Invasive Mitral Valve
Surgery After Previous Sternotomy: A Propensity-Matched Analysis.Disclosure : NoneWord Count : 1381
Even though by 2003, Casselman (and many others) had concluded that
totally endoscopic mitral valve repair can be performed safely with
excellent results and a high degree of patient
satisfaction1, less than a quarter of all isolated
mitral valve procedures were performed using minimally invasive approach
(MIS) by 20162. Conventional sternotomy (ST) remains
the approach of choice in the majority of cardiac surgery centers. Since
2011, partial sternotomy has fallen out of favor and right mini
thoracotomy (RMT) approach has been the major MIS approach (with or
without robotics) for both primary as well as re-operative mitral valve
surgery. At experienced centers, the indications for MIS surgery have
been expanded to include complex pathologies, reoperative surgery,
endocarditis, as well as a hybrid open approach for severely calcified
mitral annuli using an open deployment of transcatheter aortic
valve3-5.
One reason for the slow adoption of MIS has been the lack of randomized
prospective trials comparing the conventional sternotomy approach with
MIS. Most literature supporting the use of MIS has consisted of
retrospective review of series of individual surgeons or centers, which
have shown a shorter length of stay, reduced need for transfusions and a
quicker recovery2,3. Since these reports came from
centers with extensive experience and the fact that initial cohorts of
patients undergoing MIS were relatively lower risk patients, these
retrospective observational studies were not as convincing in their
conclusions, because the two groups of patients were not similar. Only a
few propensity-matched analyses comparing MIS vs sternotomy have so far
been reported in patients undergoing primary
surgery4-6.
For re-operative mitral valve surgery, there has been one
propensity-matched comparison of 42 pairs of patients undergoing right
mini-thoracotomy MIS vs sternotomy from China7. MIS
patients had lower transfusions, shorter length of stay and lower costs,
while having similar mortality. However, the study had a mean length of
stay of 22 days vs 16 days and mortality of 11% vs 7 % for sternotomy
and MIS patients, respectively and thus the results cannot be reliably
generalized.
In this issue of the Journal , Hamandi et al8,
reviewed 305 isolated MV reoperations that were performed in a single
institution between 2007-2018. Patients who underwent MIS MV reoperation
totaled 199, while sternotomy operations were 106. The primary endpoints
were operative mortality and 1-year survival with operative
complications and length of stay being secondary endpoints. Median age
of patients was 69 years with an equal gender distribution. The team
performed propensity-matched analysis to compare the two groups.
There were 88 well-balanced matched pairs. There was no statistically
significant difference in mortality among the matched groups at 30 days
(3.4% vs 8.0%, p=0.19) or at 1-year (15.9% vs. 16.5%, p=0.9).
Comparing long-term survival rates, no statistically significant
difference was found up to 5 years postoperatively. Also, the incidence
of post-operative complications such as atrial fibrillation, valve
dysfunction or renal failure didn’t show any statistically significant
difference. However, intraoperative blood utilization was significantly
lower among the MIS cohort (p<0.01). Patient satisfaction was
not evaluated as is not possible in a retrospective analysis. Neither
was readmission rates and other similar measures which would be
important in a value-based care system.
The 30-day mortality difference (3.4% vs 8%), while not statistically
significant, tended to be lower in MIS patients. 4 patients in the MIS
group converted to sternotomy due to adhesions. It is not clear from the
manuscript, if the mortality in the MIS group was in some way related to
the conversions or not. But based on our experience over the years and
from the analysis of this manuscript, we recommend an early conversion
to sternotomy if one is dealing with difficult adhesions, rather than
risking a long tedious operation and possibly emergently converting to
sternotomy. It is also important to note that 75% of patients were
discharged home, however readmission rate is unknown. With the advent of
value-based purchasing, readmission rates should also be looked at.
Overall, the authors should be congratulated on their excellent
management of this subset of patients and for taking the time share
their experience with us.
Propensity score matching is commonly used in evaluation research to
estimate average treatment effects.9 The main benefit
in using this statistical method is to remove confounding bias from
observational cohorts. It attempts to reduce the effects of confounders
by matching already treated subjects with control subjects who exhibit a
similar propensity for treatment based on preexisting covariates that
influence treatment selection. However, it is limited in that it
requires the removal of data and works primarily on binary treatments.
In this study, by including standardized mean difference (SMD), the
authors were able to balance the covariates in this propensity-matched
analysis.
Other than being a single-center retrospective study, this study
suffered from other short-comings of a propensity match study, such as
the loss of study power due to the decreased sample size after
performing propensity matching. Also, “the surgeon effect” was noted.
Since the MIS MV reoperative surgeries were performed by the same
surgeons who performed the sternotomy cases, the results may not be
generalizable.
The question being addressed by this manuscript (and by most other
similar comparisons of one therapy vs another) is, “Is MIS better than
sternotomy?”
Unfortunately, that question cannot be satisfactorily addressed with
this or similar studies. Healthcare quality has evolved since its
inception in 1999 with the Institute of Medicine report, titled “To Err
is human”. In the subsequent report “Crossing the Quality
Chasm”11, a high-quality care is defined as beingsafe, effective, patient-centered, timely, efficient and
equitable. Our healthcare delivery system is changing, and so should
our research methodologies. Our analyses should go deeper than
scratching the surface with mortality and morbidity data. Most studies,
including this one by Hamandi et al, do not even address
“effectiveness” adequately in the context of healthcare quality.
Having similar mortality and morbidity means that both approaches are
equally ‘safe ’. We have little information about other measures
of safety, such as readmissions, central line associated blood borne
infections. We have not evaluated whether the two approaches were
patient centered (Did the patient participate in choosing the
approach?), efficient (Cost of care) or equitable.
As cardiac surgeons dealing with life and death from up close, we are
not used to viewing healthcare from the rather distant 6-pronged quality
viewpoint mentioned above. But this is important for a very important
reason which I explain below.
Individual surgeons and patients may not have the power to bring about a
meaningful change in the way we do business everyday. But just like
state pension funds pressured oil companies into facing climate
change10, big stakeholders like insurance companies
and other payers may be able to convince the cardiac surgeons to face
the future. For that to happen, quality metrics such as readmission
rates, cost of care and patient satisfaction must be looked at and
reported, because that is how these stakeholders assess quality.
According to some studies7 MIS approach is better in
terms of cost and patient satisfaction. Such comprehensive analyses of
quality will go a long way in answering a slightly different question
than the one posed earlier; “Does MIS offer better quality than
sternotomy?”
If we want to influence healthcare delivery and have a passion for
quality, then our research methodology must reflect the high standards,
that we have set for our clinical work. We should also develop new
measures of quality besides morbidity and mortality. We have to look at
those metrics that have traditionally been ignored by surgeons, but are
important for the payers and the hospitals that rely on these payers for
their success. As far a minimally invasive vs sternotomy approach is
concerned, that question is not going to last for long. Not because one
side would have won or the other lost, but because for those that have
not yet boarded the train of minimally invasive mitral valve surgery,
that train may have already left the station, moving at full speed ahead
towards the “percutaneous station”. It is not a matter of if ,
but when , sternotomy would not be the standard of care for mitral
valve surgery. Today’s vascular surgeons save open repair of abdominal
aortic aneurysm for a very small subset of patients. There is no reason
to believe that tomorrow’s mitral valve surgeons will consider open
sternotomy any differently for mitral valve surgery.